Improvement of intestinal absorption of peptide drugs by glycosylation: Transport of tetrapeptide by the sodium ion-dependent D-glucose transporter

Article abstract of DOI:10.1021/js970269p

A tetrapeptide (Gly-Gly-Tyr-Arg, GGYR), which is not transported by di- or tripeptide transporters, was glycosylated with p-(succinylamido)phenyl α- or β-D-glucopyranoside (α,β-SAPG) to investigate whether these glycosylated molecules are transported by t

Full text of DOI:10.1021/js970269p

Improvement of Intestinal Absorption of Peptide Drugs by
Glycosylation: Transport of Tetrapeptide by the Sodium Ion-Dependent
D-Glucose Transporter
†
‡
MASAHIRO NOMOTO , KAZUHITO YAMADA , MAKOTO HAGA*, AND MASAHIRO HAYASHI
Contribution from Department of Pharmaceutics, Faculty of Pharmaceutical Sciences, Science University of Tokyo, 12 Ichigaya,
Funagawara-machi, Shinjuku-ku, Tokyo 162, Japan
Received July 14, 1997.
Final revised manuscript received November 20, 1997.
Accepted for publication November 21, 1997.
the systemic availability of peptides, it is necessary to
enhance their absorption and to increase their resistance
to enzymatic degradation. Many approaches have been
suggested to circumvent these barriers, such as coadmin-
istration of peptidase inhibitors,1-4 coadministration of
absorption enhancers,5,6 and modifications of the peptide
structure.7-9
It has been shown recently that there are specific carrier-
mediated transport systems in the intestine which facilitate
absorption of amino acids, oligopeptides, monosaccharides,
monocarboxylic acids, phosphate, bile acids, and several
water-soluble vitamins.10 Strategies to improve the bio-
availability of poorly absorbed drugs have been designed
by modifying drugs so that they are absorbed by specialized
intestinal transporters.
Abstract 0 A tetrapeptide (Gly-Gly-Tyr-Arg, GGYR), which is not
transported by di- or tripeptide transporters, was glycosylated with
p-(succinylamido)phenyl R- or â-D-glucopyranoside (R,â-SAPG) to
investigate whether these glycosylated molecules are transported by
the Na+-dependent D-glucose transporter. Their uptake into brush-
border membrane vesicles (BBMVs) and transport through the intestinal
membrane were examined using the rapid filtration technique and
the everted sac method. It was observed that glycosylation at the
R-amino position of GGYR increased resistance to aminopeptidase
activity and inhibited its degradation. When R- and â-SAPG-GGYR
were incubated with BBMVs, overshoot uptake was observed about
2 min after the start of incubation in the presence of an inward Na+
gradient. This uptake remained unaffected by the addition of GGYR
+
while it was significantly inhibited when Na+ was replaced with K or
The role and function of glucose transporters in the
general control of glucose homeostasis are currently being
elucidated. Five human facilitative-diffusion hexose trans-
porters (GLUT-1 to GLUT-5) in different tissues have been
characterized by molecular cloning.11,12 A low-affinity
transporter GLUT-2 is known to be present in the baso-
lateral membrane of intestinal epithelial cells. There are
Na+-dependent glucose transporters, referred to as SGLT-
1, present at a high level in human intestine and also at a
somewhat lower level in rat, which actively transport
glucose across the apical brush-border of epithelial cells.13
In the previous study we synthesized glycosylated insu-
lin, p-(succinylamido)phenyl-R-D-glucopyranoside (SAPG)-
substituted insulin (SAPG-INS) and utilized SGLT-1 to
improve the intestinal absorption of insulin.14 It was
observed that SAPG-INS was less susceptible to enzymatic
degradation than native insulin and had an increased
affinity for brush-border membrane vesicles (BBMVs)
prepared from rat small intestine. It remained doubtful,
however, whether insulin was taken up into BBMVs by this
transporter because of its large molecular size.
In the present study we synthesized glycosylated tet-
rapeptides, R- and â-SAPG-Gly-Gly-Tyr-Arg (R,â-SAPG-
GGYR), to investigate whether glycosylated peptide is
transported by the Na+-dependent D-glucose transporter.
A tetrapeptide is considered to be the smallest peptide
which is not transported by di- or tripeptide transporters.
We selected GGYR because of its ease of radioiodination
of tyrosine and commercial availability. We also examined
their uptake into BBMVs and transport through the
intestinal membrane by the rapid filtration technique and
the everted sac method.
R- and â-SAPG-GGYR were incubated with BBMVs at 4 °C. Uptake
was also markedly inhibited either with 1 mM phloridzin or 10 mM
D-glucose. These findings suggested that the Na+-dependent glucose
transporter (SGLT-1) played an important role in the uptake of both
R- and â-SAPG-GGYR into BBMVs. A comparison of R- with
â-SAPG-GGYR revealed that the amount of â-SAPG-GGYR taken
up was greater than that of R-SAPG-GGYR. From the everted sac
method data, it was shown that the elimination clearance from the
mucosal side, CLel, and permeation clearance to the serosal side,
CLp, were 15.82 ± 6.83 and 0.83 ± 0.06 µL/min/cm for R-SAPG-
GGYR and 44.52 ± 3.61 and 3.50 ± 0.81 µL/min/cm for â-SAPG-
GGYR, respectively, and that R-SAPG-GGYR was more resistant to
enzymatic degradation than â-SAPG-GGYR. Permeation of both R-
and â-SAPG-GGYR was inhibited in the presence of D-glucose and
in the absence of a Na+ gradient, suggesting that both R- and
â-SAPG-GGYR were transported by the Na+-dependent D-glucose
transporter. The permeation clearance transported by the Na+-
dependent D-glucose transporter, (CLp)Na+, of â-SAPG-GGYR was
about 5 times greater than that for R-SAPG-GGYR. This result may
be ascribable to the fact that the â-form of glucose has higher affinity
to SGLT-1 than the R-form. The results of the present study
encourage further investigations on improvements in intestinal absorp-
tion of peptide drugs by glycosylation.
1. Introduction
Development of peptide drugs as therapeutic agents has
proved very difficult because of their poor membrane
penetrability and extreme susceptibility to enzymatic
degradation in the gut lumen and enterocytes. To improve
2. Materials and Methods
* Corresponding author. Tel: +81-3-3260-6725, ex. 5033. Fax: +81-
3-3260-8597. E-mail: haga@ps.kagu.sut.ac.jp.
† Present address: Meiji Seika Kaisha, Ltd., Pharmacokinetics
Research Lab., Pharmaceutical Research Center, 760 Morooka-cho,
Kohoku-ku, Yokohama 222, J apan.
2.1. Ma ter ia lssGly-Gly-Tyr-Arg (GGYR), amastatin,
chymostatin, and phosphoramidon were purchased from
Peptide Institute Inc. (Osaka, J apan). Na125I was pur-
chased from Amersham J apan (Tokyo, J apan). Lactoper-
‡ Present address: Santen Pharmaceutical Co., Ltd. Ophthalmic
Laboratories, 8916-16 Takayama-chou, Ikoma-shi 630-01, J apan.
326 / Journal of Pharmaceutical Sciences
S0022-3549(97)00269-4 CCC: $15.00
Published on Web 01/27/1998
© 1998, American Chemical Society and
Vol. 87, No. 3, March 1998
American Pharmaceutical Association

oxidase and phloridzin were purchased from Sigma Chemi-
cal Co. (St. Louis, MO). Dimethyphosphinothioyl chloride
(Mpt-Cl) was purchased from Tokyo Kasei Co. (Tokyo,
J apan). p-Nitrophenyl R-D-glucopyranoside (R-NPG), p-
nitrophenyl â-D-glucopyranoside (â-NPG), palladium car-
bon, and dansyl chloride were purchased from Nakalai
Tesque Inc. (Kyoto, J apan). All other chemicals were
analytical grade and were used without further purifica-
tion. Deionized redistilled water was used to prepare all
solutions.
2.2. Syn t h esis of SAP G a n d SAP G-GGYR sThe
synthesis of SAPG was performed according to the method
of J eong et al.15 Briefly, the nitrophenyl group on R-NPG
was reduced to an aminophenyl group with ammonium
formate in the presence of palladium carbon yielding
p-aminophenyl R-D-glucopyranoside (R-APG). APG was
then allowed to react with succinic anhydride at room
temperature for 4 h to produce SAPG. The coupling of
SAPG to GGYR was performed using a Mpt (dimeth-
ylphosphinothionyl) mixed anhydride method according to
the work of Ueki et al.16 Briefly, GGYR (170 mg, 321 µmol)
was dissolved in a 1:1 (v/v) mixture of water and dimethyl
formamide (DMF, 36 mL total volume) which was cooled
in an ice bath after adjustment of the pH to 9.5 with 0.1 N
NaOH. One hundred and twenty milligrams of SAPG was
dissolved in 1.08 mL of DMF containing 80 µL (320 µmol)
of tri-n-butylamine, then 34 µL (320 µmol) of Mpt-Cl was
added at 0 °C and the solution was stirred for 15 min. Next
160 µL (640 µmol) of tri-n-butylamine was added. This
SAPG solution was added dropwise to the GGYR solution
in an ice bath and the pH of mixed solution was adjusted
to 9.5 with 0.1 N NaOH. After the solution was allowed
to stand at room temperature for 1 h, it was stirred
overnight while the pH was maintained at 9.5.
2.3. P u r ifica tion of SAP G-GGYRsThe organic re-
agents and solvents were separated from glycosylated
GGYR using a Sephadex LH-20 gel column (1.0 × 15 cm).
The crude mixture (5 mL) was applied to the column and
eluted with 100 mM phosphate buffer at a flow rate of 18
mL/h. The UV absorption of each fraction (3 mL) was
measured at 254 nm, and the fractions showing increased
absorption, indicating the presence of peptide, were col-
lected and pooled. The glycosylated GGYR was further
purified by a DEAE anion-exchange chromatography using
a Con Sep LC 100 with a Mem Sep 1010 cartridge
(Millipore, Bedford, MA), which was eluted with a solution
of 20 mM Tris-HCl buffer (pH 8.5) at a rate of 3 mL/min
using a linear gradient to 1 M NaCl. Fractions comprising
the first peptide peak that eluted were pooled and dialyzed
using a Microasilyser S1 with a Acipex Cartridge AC 110-
10 (Asahi Chemical Industry Co. Ltd., Tokyo, J apan). The
solution obtained was lyophilized to give a white-yellow
powder.
2.4. Deter m in a tion of SAP G-GGYRsHPLC Analy-
siss Samples were analyzed using a Waters 600E system
(484 UV/Vis detector, 741 Data Module, Millipore Corp.,
Waters Chromatography Division, Milford, MA) on an
ASAHI PAK GS-320 column (7.6 × 500 mm, Asahi Chemi-
cal Industry, Co. Ltd., Tokyo, J apan) or on a Carbonex
TCAS column (4.6 × 100 mm, Tonen Co. Ltd., Tokyo,
J apan) at a flow rate 1.0 or 0.4 mL/min. The mobile phase
consisted of 50 mM ammonium acetate/acetonitile (80:20%
v/v).
dansyl chloride solution (2.5 mg/mL in acetone) was added
and the solution was incubated at 37 °C. After 1 h the
solution was dried in a vacuum and 100 µL of 6 N HCl
was added. The tube was sealed by heating in a fine
oxygen-gas flame. Hydrolysis was carried out at 105 °C
for 18 h. After this period, the tube was centrifuged at
3000 rpm for 5 min. The tube was then opened and HCl
was removed under vacuum. After dissolving of the
hydrolyzed products in ammonium acetate-acetonitile
solution, they were analyzed on the HPLC system described
above. As a reference, a sample of the peptide was
hydrolyzed before dansylation and the composition of
amino acids was analyzed. Four dansyl amino acids (O-
DNS-L-Tyr, DNS-Gly, DNS-L-Arg, and N,O-diDNS-L-Tyr)
were detected when the lyophilized product was dansylated
after hydrolysis while no amino acid was detected when it
was hydrolyzed after dansylation. The retention times of
O-DNS-L-Tyr, DNS-Gly, DNS-L-Arg, and N,O-diDNS-L-Tyr
were 20.1, 26.5, 33.6, and 34.6 min, respectively (mean
value of two experiments). These findings suggested that
the lyophilized product was the compound substituted at
the Gly terminal position.
Determination of the Sugar/ Peptide RatiosThe number
of sugar groups on the glycosylated peptide was determined
using the phenol/sulfuric acid method18 and a Micro BCA
protein assay reagent kit (Pierce, Rockford, IL). Briefly,
the glycosylated peptide (1 mg) was dissolved in 1 mL of a
5% aqueous solution of phenol. To this solution was added
5 mL of concentrated H2SO4, followed by mixing. After this
mixture was allowed to stand for 30 min its absorbance
was measured at 490 nm. The concentration of sugar was
determined by comparison with a standard curve for SAPG
over the range from 0 to 0.1 mg/mL. The peptide concen-
tration was determined according to the standard protocol
of the Micro BCA protein assay reagent. From the ob-
served concentrations of sugar and peptide, the sugar/
peptide ratio of the glycosylated peptide was shown to be
1.
Measurement of Molecular Weight and PolaritysOne
milligram of the purified glycosylated peptide was dissolved
in glycerol and a secondary ion mass spectrogram (SIMS)
was obtained using a Hitachi M-83A mass spectrometer.
SAPG-substituted GGYR was identified by a molecular ion
peak, MH+, at 805. The polarity of R- or â-SAPG-GGYR
was measured using a digital polarimeter (DIP-360, J ASCO
Corporation, Tokyo J apan) at room temperature. The [R]D
values were 140.74° and 24.99° for R-SAPG and R-SAPG-
GGYR, respectively, and -48.00° and -8.33° for â-SAPG
and â-SAPG-GGYR, respectively. Samples were also ana-
lyzed by HPLC on a Carbonex TCAS column at 0 °C. The
mobile phase consisted of 50 mM ammonium acetate
/acetonitile (80:20) at a flow rate of 0.4 mL/min. Two main
peaks were observed: one at a retention time of 3.07 min
for R-SAPG-GGYR and one at a retention time of 5.31 min
for â-SAPG-GGYR. The purity of R- and â-SAPG-GGYR
were 96.5 and 96.0%, respectively, as determined from
HPLC chromatograms (mean value of two experiments).
2.5. In Vitr o Exp er im en tsPreparation of Brush Border
Membrane Vesicles (BBMVs)sBBMVs were prepared by
Mg2+ precipitation according to the methods of Tiruppathi
et al.19 and Takuwa et al.20 Briefly, male Wistar rats (body
weight 200 ( 20 g) which had been fasted for 18-24 h were
intraperitoneally anesthetized with sodium pentobarbital
(50 mg/kg) and then sacrificed by exsanguination through
the carotid arteries. The abdomen was opened by a median
incision and the small intestine (from the duodenum to
ileum) was excised. The portion of the intestine was cut
into small pieces (7-8 cm) and rinsed with ice-cold 0.9%
saline and then cut open longitudinally. The mucosa was
softly scraped off with a glass slide and was homogenized
Analysis of the Position of SAPG SubstitutionsThe
position of SAPG-substitution on GGYR was estimated
according to the method of Gray17 with some modifications.
Briefly, peptide (1 mg) was dissolved in 10 µL of 0.2 M
sodium hydrogen carbonate solution and dried in a vacuum.
This was redissolved in 10 µL of distilled water and the
pH of the solution was adjusted to 8.5-9.0. Then 10 µL of
Journal of Pharmaceutical Sciences / 327
Vol. 87, No. 3, March 1998

with an Excel Auto Homogenizer (Model OX-7, Nihonseiki
Co. Ltd., Tokyo J apan) at 12000 rpm for 90 s in 20 vol (v/
w) of 6 mM Tris-NaOH buffer solution (pH 7.5) containing
150 mM mannitol and 2.5 mM EGTA. To this homogenate
was added 1 M MgCl2 solution to a final concentration of
10 mM, and the mixture was allowed to stand for 15 min.
After centrifugation at 3000g for 10 min, the supernatant
fraction was centrifuged at 42000g for 30 min. The
resultant pellet was suspended in 10 mL of uptake buffer
(16 mM Tris-10 mM HEPES-HCl containing 15 mM KCl
and 270 mM mannitol, pH 7.5) and the mixture was
centrifuged again at 42000g for 30 min. The second pellet
was resuspended in 1 mL of uptake buffer and the solution
was passed slowly through a syringe fitted with a 25-gauge
needle to make the vesicle size uniform. The final protein
concentration of the brush-border membrane preparation,
which was determined by a Micro BCA protein assay, was
adjusted to 10 mg/mL. The purity of BBMVs was assessed
by assaying the activity of the marker enzyme alkaline
phosphatase (ALP) using a Alkaline Phospha-K Test Wako
(Wako Pure Chemical Industries, Ltd., Osaka J apan). The
ALP activity was about 10 times higher in BBMVs com-
pared to the homogenate.
Uptake of D-Glucose into BBMVssA radiolabeled D-
glucose solution containing 185 kBq/mL [3H]-D-glucose, 1
mM unlabeled D-glucose, and 100 mM NaCl in uptake
buffer solution was prepared. A 100 µL portion of this
solution and 20 µL of BBMVs suspension were incubated
in each test tube at 25 °C for 5 min. The reaction was
started by adding the labeled D-glucose solution to the
BBMVs suspension. After an appropriate incubation time,
3 mL of ice-cold stop buffer solution (10 mM Tris-11 mM
HEPES-HCl containing 153 mM KCl, pH7.5) was added.
BBMVs were collected on nitrocellulose filters (0.45 µm)
and washed three times with 4 mL of ice-cold stop buffer
solution. The filter was added to 10 mL of Scintisol EX-H
(Dojindo Laboratories Co. Ltd., Kumamoto J apan), and
vesicle-associated radioactivity was determined after solu-
bilization of the filter in a liquid scintillation counter (ASC-
3600, Aloka Co. Ltd., Tokyo J apan). The blank value was
determined by mixing the labeled D-glucose solution with
the BBMVs suspension followed by immediate filtration
and washing with the stop buffer solution. This value was
subtracted from the data.
Enzymatic Degradation of R- and â-SAPG-GGYR and
GGYRsA 50 µL portion of BBMVs suspension (5 mg of
protein/mL) was mixed with 150 µL of 5 mM R- or â-SAPG-
GGYR or GGYR solution and the mixture was incubated
at 37 °C. At set time, 250 µL of acetonitrile and 50 µL of
internal standard (200 µM tryptophan) were added, and
the mixture was centrifuged at 12000 rpm for 15 min. The
supernatant was filtered through a membrane filter (0.45
µm). An aliquot of this supernatant was analyzed using a
Waters 600E HPLC system on a TSK-gel ODS-80TM
column (4.5 × 150 mm, Tosoh, Co. Ltd., Tokyo, J apan) at
a flow rate of 1 mL/min. The mobile phase consisted of 50
mM ammonium acetate/methanol (90:10% v/v).
Iodination of R- and â-SAPG-GGYR and GGYRsR- and
â-SAPG-GGYR and GGYR were iodinated according to the
method of Hamlin et at.21 Briefly, 2.5 mg of R- or â-SAPG-
GGYR or GGYR was dissolved in 600 µL of 0.1 M sodium
citrate buffer (pH 5.2). To this solution were added 2 µL
of 0.05 M H2O2 , 2 µL of KI/Na125I mixture solution (0.05
M KI:3.7 GBq/mL Na125I ) 3:1 by vol), and 2 µL of
lactoperoxidase solution in 125 mM EDTA (final concentra-
tion of lactoperoxidase 9.25 µM) successively with stirring
every 10 min. The mixture was applied to a Sephadex A-25
column (1.6 × 5 cm), eluted with 0.1 M sodium citrate
buffer (pH 5.2), and then fractionated into labeled and
unlabeled GGYR. The radioactivity of each 0.6 mL fraction
was determined in an autowell γ-counter (Aloka, ARC-300).
Concentration measurements of peptide solutions were
performed using a Hitachi model U-3000 recording spec-
trophotometer, and the specific radioactivity was calcu-
lated.
Uptake of R- and â-SAPG-GGYR and GGYR into
BBMVssUptake or adsorption of R- and â-SAPG-GGYR
and GGYR into BBMVs was measured at 37 °C by a rapid
filtration technique according to the method of McCarthy
et al.22 The 125I-labeled peptide solution (195 µL) was
mixed with an equal volume of 32 mM Tris-20 mM Hepes-
HCl buffer solution containing 30 mM KCl, 200 mM NaCl,
and 540 mM mannitol (pH 7.5) in the uptake experiment
with an inward Na+ gradient or 32 mM Tris-20 mM
Hepes-HCl buffer solution containing 230 mM KCl and 540
mM mannitol (pH 7.5) in the uptake experiment with an
inward K+ gradient. This mixture (390 µL) and the
BBMVs suspension (130 µL) were incubated separately at
37 °C (or 4 °C) for 5 min then mixed and incubated at 37
°C (or 4 °C). The osmolarity of this mixed solution was
502 mOsm, which was measured using a osmometer
(Osmette, Precision Systems, Inc.). Hyperosmolar and
hypoosmolar buffers (350, 635, and 773 mOsm) were
prepared by adjusting the amount of D-mannitol added to
the solution. At set time intervals after mixing, the vesicles
(65 µL) were sampled from a plastic test tube and sucked
through a prewashed nitrocellulose filter with a pore size
of 0.45 µm (Advantec Toyo, Tokyo, J apan). The filter was
washed three times with 4 mL of ice-cold stop buffer
solution. The filter was transferred to an RIA tube and
radioactivity was measured in an auto-well γ-counter
(Aloka, ARC-300). The blank value was determined by
mixing the peptide with the vesicle suspension followed by
immediate filtration and washing with the stop buffer
solution. This value was subtracted from the uptake data.
Everted Gut-Sac TechniquesThe permeability of R- and
â-SAPG-GGYR and GGYR was studied using the everted
gut-sac technique.23 Male Wistar rats (weighing 200 ( 20
g), which had been fasted for 18-24 h, were anesthetized
by an ip injection of 50 mg/kg sodium pentobarbital. After
making a median incision in the abdomen, the small
intestine was cut at two positions i.e., 2 and 9 cm from
Treitz’s ligaments. The entire length of the small intestine
was carefully removed and placed in Krebs Ringer buffer
solution (KRBS, pH 7.4) which was bubbled with 95% O2:
5% CO2 and then rinsed. After 15 min, the end of the
segment was ligated and the segment was everted on a
glass rod and rinsed. The other end of the segment was
cannulated with a Teflon tube to form a sampling port. A
2.0 g glass weight was tied to the end of the sac to prevent
changes in its internal volume. The sac was filled with 1
mL of KRBS and placed in a test tube containing 10 mL of
test solution (or 10 mL of KRBS containing 3 mM fluores-
cein isothiocyanate-dextrans 4000 as a leakage marker)
that was continuously bubbled with 95% O2:5% CO2. Every
20 min samples (100 µL) were drawn from the serosal or
mucosal side of the sac, and 110 µL acetonitrile and 10 µL
of internal standard were added. The volumes removed
were replaced with fresh KRBS. The mixture was then
centrifuged at 12000 rpm for 15 min and the supernatant
was filtered through a membrane filter (0.45 µm). An
aliquot of this supernatant was analyzed in the same way
as those described in the paragraph “Enzymatic Degrada-
tion of R- and â-SAPG-GGYR and GGYR”.
Estimation of Elimination Clearance (CLel) and Perme-
ation Clearance (CLp)sThe concentration of glycosylated
GGYR on the mucosal side decreased so rapidly that we
could not consider it constant. The elimination clearance
(CLel) of intact glycosylated GGYR or GGYR per unit length
of small intestinal gut was calculated, therefore, according
328 / Journal of Pharmaceutical Sciences
Vol. 87, No. 3, March 1998

Figure 3sDegradation of R- and â-SAPG-GGYR in the presence of BBMVs.
A 50 µL sample of BBMVs suspension (5 mg of protein/mL) was mixed with
150 µLof 5mMR- or â-SAPG-GGYR solution and the mixture was incubated
at 37 °C. At set time, 250 µL of acetonitrile and 50 µL of internal standard
(tryptophan) were added, and the mixture was centrifuged at 12000 rpm for
15 min. The supernatant was analyzed by HPLC after the filtration with a
membrane filter (0.45 µm). The results are given as means ± SE (n ) 4).
Key: (light gray) R-SAPG-GGYR; (dark gray) â-SAPG-GGYR.
Figure 1sStructures of R-SAPG-GGYR (a) and â-SAPG-GGYR (b).
potent aminopeptidase inhibitor, about 80% of intact GGYR
remained 10 min after the start of incubation. It was
reported1 that 20 µM amastatin inhibited hydrolysis of Leu-
enkephalin, YGGFL, by over 90%. When GGYR was
incubated with 33.3 µM phosphoramidon or 33.3 µM
chymostatin, about 60% and 20% of the initial amount
remained 10 min after the start of incubation, respectively.
Phosphoramidon is a potent inhibitor of brush-border
endopeptidase, and chymostatin is an inhibitor of chymot-
rypsin which will very likely cleave GGYR between Y and
R. In addition, more than 90% of the initial amount
remained 20 min after the start of incubation when GGYR
was incubated with BBMVs in the presence of 33.3 µM each
of amastatin, phosphoramidon, and chymostatin. It was
reported that the brush-border membrane of intestinal
mucosal cells contains various endo- and exopeptidases
such as aminopeptidases, dipeptidylaminopeptidase, car-
boxypeptidase, and angiotensin-converting enzyme (dipep-
tidyl carboxypeptidase).25 The effective inhibition of GGYR
degradation by amastatin and phosphoramidon indicated
that GGYR was mainly degraded by aminopeptidase and
partially degraded by endopeptidase. The degradation of
R- and â-SAPG-GGYR is shown in Figure 3. The remain-
ing amounts of R- and â-SAPG-GGYR were 90.8 and 93.5%,
respectively, 10 min after incubation started, while the
remaining amount of GGYR in the presence of 33.3 µM
amastatin was 77.5%. These findings indicated that gly-
cosylation at the R-amino position of GGYR increased its
resistance to aminopeptidase degradation. These results
also indicated that glycosylation might increase resistance
not only to the degradation of aminopeptidase but also, to
some extent, to the degradation of endopeptidase.
Figure 2sDegradation of GGYR in the presence of BBMVs. A 50 µL sample
of BBMVs suspension (5 mg of proten/mL) was mixed with 150 µL of 5 mM
GGYR solution and the mixture was incubated at 37 °C. At set time, 250 µL
of acetonitrile and 50 µL of internal standard (tryptophan) were added, and
the mixture was centrifuged at 12000 rpm for 15 min. The supernatant was
analyzed by HPLC after the filtration with a membrane filter (0.45 µm). The
results are given as means ± SE (n ) 4). Key: (4) control; (0) amastatin
(33.3 µM); (]) phosphoramidon (33.3 µM); (O) chymostatin (33.3 µM); (3)
amastatin + phosphoramidon + chymostatin (33.3 µM each).
to the method of Taki et al.,24 using an area under the curve
(AUC) of mucosal concentration vs time plot from time 0
to 80 min and the eliminated amount as follows:
CLel ) (eliminated amount in the mucosal side
solution)/AUC0-80 (1)
Similarly, the permeation clearance CLp of intact glyco-
sylated GGYR or GGYR was calculated using eq 2.
CLp ) (amount that permeated to the serosal side
solution)/AUC0-80 (2)
3.2. Up ta k e of D-Glu cose a n d r- a n d â-SAP G-
GGYRsWe studied the uptake of D-glucose into BBMVs
by the rapid filtration method as a control experiment to
check the function of a Na+-dependent D-glucose trans-
porter. Overshoot uptake was observed about 1 min after
incubation started in the presence of an inward gradient
of Na+ at 37 °C (data not shown). We considered from this
finding that BBMVs with Na+-dependent D-glucose trans-
porters were properly prepared. The mechanism of trans-
port of glycosylated peptide into BBMVs was then studied.
Since the uptake values represent the sum of transport into
the intravesicular space and the nonspecific binding of the
peptide to the external membrane surface, the amount of
3. Results and Discussion
3.1. En zym a tic Degr a d a tion of r- a n d â-SAP G-
GGYRsThe glycosylated GGYRs used in this study were
confirmed as R- and â-SAPG-GGYR by the analytical
methods described in Materials and Methods. The struc-
tures of these compounds are shown in Figure 1.
The degradation of GGYR in the presence of BBMVs is
shown in Figure 2. GGYR was degraded so rapidly that
only 7% of the initial amount remained 10 min after the
start of incubation. When GGYR was incubated with
BBMVs in the presence of 33.3 µM amastatin, a known
Journal of Pharmaceutical Sciences / 329
Vol. 87, No. 3, March 1998

Figure 4sEffect of osmolarity on the uptake of R- and â-SAPG-GGYR by
BBMVs. The 125I-labeled peptide solution (195 µL) was mixed with an equal
volume of 32 mM Tris−20 mM Hepes-HCl buffer solution containing 30 mM
KCl, 200 mM NaCl, and 540 mM mannitol (pH 7.5). This mixture (390 µL)
and the BBMVs suspension (130 µL) were mixed and incubated at 37 °C.
The osmolarity of this mixed solution was 502 mOsm. Hyperosmolar and
hypoosmolar buffers (350, 635, and 773 mOsm) were prepared by adjusting
the amount of D-mannitol added to the solution. The sampling and
measurement of radioactivity were described in the experimental section. The
results are given as means ± SE (n ) 4). Key: (O) R-SAPG-GGYR; (9)
â-SAPG-GGYR.
Figure 6sTime-course changes in â-SAPG-GGYR uptake by BBMVs. The
experimental conditions were the same as those described in the legend to
Figure 5 except that no experiments using GGYR and glucose were performed.
The results are given as means ± SE (n ) 4). Key: (0) in the presence of
a Na+ gradient at 37 °C; ([) in the presence of a Na+ gradient at 4 °C; (b)
+
in the presence of a K gradient at 37 °C; (2) in the presence of a Na+
gradient and 1mM phloridzin at 37 °C.
tors. We therefore incubated R- or â-SAPG-GGYR with
BBMVs without these inhibitors, and overshoot uptake was
observed after about 2 min in the presence of an inward
Na+ gradient. This uptake remained unaffected by the
addition of 7.5 mM GGYR. Conversely, uptake was
significantly inhibited when Na+ was replaced with K+ or
when BBMVs were incubated with R-SAPG-GGYR at 4 °C.
The transport of R-SAPG-GGYR was markedly inhibited
either with 1 mM phloridzin or with 10 mM D-glucose.
These findings indicated that the Na+-dependent glucose
transporter played an important role in the uptake of
R-SAPG-GGYR into BBMVs.
Time-course changes in uptake of â-SAPG-GGYR by
BBMVs are shown in Figure 6. Overshoot uptake was
observed about 2 min after incubation started in the
presence of an inward gradient of Na+ at 37 °C. The
uptake was significantly inhibited when Na+ was replaced
with K+, when BBMVs were incubated with â-SAPG-GGYR
at 4 °C or in the presence of 1 mM phloridzin. These
findings showed that â-SAPG-GGYR was also taken up by
the Na+-dependent glucose transporter. To clarify the
difference between R- and â-SAPG-GGYR, time-course
changes in uptake were compared (Figure 5 and 6) together
with the uptake of GGYR in the presence of 33.3 µM each
of amastatin, chymostatin, and phosphoramidon (data not
shown). The uptake of â-SAPG-GGYR by BBMVs was
significantly greater than that of R-SAPG-GGYR at 1 min
after the incubation, and GGYR was taken up or associated
with BBMVs by about 50% of the value seen for R-SAPG-
GGYR, but no overshoot was observed (data not shown).
Two types of glucose transporter have been reported so far,
namely, facilitated-diffusion transporter and Na+-coupled
active transporter. The former are referred to as GLUT-1
to GLUT-5 and the latter is referred as SGLT-1. SGLT-1
can transport both glucose and galactose and is present at
a high level in the intestine.27 The structual requirements
for substrates to interact with SGLT-1 are considered to
be (1) the substrate must have a D-pyranose ring configu-
ration, (2) it must possess a C1 chair form, and (3) the
hydroxyl group in the glucose molecule at carbon 2 must
be in the equatorial position.28-30 As both R- and â-SAPG-
GGYR satisfy these conditions, it is feasible that they may
interact with SGLT-1 and be actively taken up by BBMVs
or absorbed from rat intestine. It is well-known that
phloridzin, which has a â-configuration of glucose, inhibits
the transport of R-D-glucosides. This means that the â-form
has higher affinity for SGLT-1 than the R-form of glucose.
Mizuma et al.31 studied the intestinal absorption of ano-
Figure 5sTime-course changes in R-SAPG-GGYR uptake by BBMVs. The
125I-labeled peptide solution (195 µL) was mixed with an equal volume of 32
mM Tris−20 mM Hepes-HCl buffer solution containing 30 mM KCl, 200 mM
NaCl, and 540 mM mannitol (pH 7.5) in the uptake experiment with an inward
Na+ gradient or 32 mM Tris−20 mM Hepes-HCl buffer solution containing
230 mM KCl and 540 mM mannitol (pH 7.5) in the uptake experiment with an
+
inward K gradient. This mixture (390 µL) and the BBMVs suspension (130
µL) were mixed (the mixed solution contained 1 mM phloridzin or 7.5 mM
GGYR or 10 mM glucose in some experiments) and incubated at 37 °C (or
4 °C). The sampling and measurement of the radioactivity was described in
the experimental section. The results are given as means ± SE (n ) 4). Key:
(0) in the presence of a Na+ gradient at 37 °C; ([) in the presence of a Na+
+
gradient at 4 °C; (b) in the presence of a K gradient at 37 °C; (2) in the
presence of a Na+ gradient and 1mM phloridzin at 37 °C; (*) in the presence
of a Na+ gradient and 7.5 mM GGYR at 37 °C; (+) in the presence of a Na+
gradient and 10 mM glucose at 37 °C.
nonspecific binding was first analyzed by determining the
uptake of the peptide as a function of the reciprocal of
osmolarity.26 Figure 4 shows the dependence of R- and
â-SAPG-GGYR transport on extravesicular medium osmo-
larity. Extrapolation of the straight lines for R- and
â-SAPG-GGYR to infinite osmolarity gave values of about
15 and 38 pmol/mg of peptide, respectively which represent
the amount of nonspecific surface binding. These values
represented about 3.6 and 7.5% of each uptake value at
502 mOsm. It was shown from these values that the
observed uptake was mainly intravesicular and surface
binding was not significant.
Time-course changes in the uptake of R-SAPG-GGYR by
BBMVs are shown in Figure 5. As described above, the
glycosylation of GGYR inhibited enzymatic degradation as
did a mixture of aminopeptidase and endopeptidase inhibi-
330 / Journal of Pharmaceutical Sciences
Vol. 87, No. 3, March 1998

Figure 7sTime-course changes in the elimination of R- and â-SAPG-GGYR
on the mucosal side (a) and appearance of intact constituents on the serosal
side (b). The sac was filled with 1 mL of KRBS and was placed in a test tube
containing 10 mL of 3 mM GGYR or R- or â-SAPG-GGYR solution. Every 20
min, samples (100 µL) were drawn from the serosal or mucosal side of the
sac, and 110 µL of acetonitrile and 10 µL of internal standard were added.
The volumes removed were replaced with fresh KRBS. The mixture was then
centrifuged at 12000 rpm for 15 min and the supernatant was filtered through
a membrane filter (0.45 µm). The HPLC analysis was described in the
experimental section. The results are given as means ± SE (n ) 4). Key (3)
GGYR; (]) R-SAPG-GGYR; (O) â-SAPG-GGYR.
Figure 8sPermeation clearance (CLp) of R-SAPG-GGYR (a) and â-SAPG-
GGYR (b). The amount of intact R- or â-SAPG-GGYR that permeated to the
serosal side solution during 80 min in the presence of 10 mM of phloridzin,
10 mM of D-glucose and in the absence of Na+ was determined in the same
way as those described in the legen to Figure 7, and the permeation clearance
(CLp) was calculated according to eq 2. The results are given as means ±
SE (n ) 4).
( 0.06 µL/min/cm for R-SAPG-GGYR, and 44.5 ( 3.61 and
3.50 ( 0.81 µL/min/cm for â-SAPG-GGYR, respectively. The
amount of intact constituent eliminated from the mucosal
side is considered to be the sum of the amount that
degraded on the mucosal side and the amount that perme-
ated to the serosal side. Although we did not calculate the
total amount that permeated including metabolites which
were degraded enzymatically during transport through the
intestinal membrane and on the serosa, degradation clear-
ance (CLd) for â-SAPG-GGYR was greater than that for
R-SAPG-GGYR (data not shown). These findings suggested
that R-SAPG-GGYR was more stable to enzymatic degra-
dation than â-SAPG-GGYR. In addition, a considerable
amount of metabolite GGYR was detected on the serosal
side solution both for R- and â-SAPG-GGYR. The amount
of GGYR 80 min after the start of the permeation experi-
ment was about 100 and 200 nmol for R- and â-SAPG-
GGYR (data not shown). Since GGYR itself did not
permeate to the serosal side because of enzymatic degrada-
tion and very low permeability, it was considered that R-
or â-SAPG-GGYR was bioactivated to the parent peptide
during transport through the intestinal membrane of the
serosa. Bioactivation following the absorption step is one
important issue in the use of the prodrug principle for
improving epithelial delivery of peptides.
mers of glucosides and showed that the affinity of the
â-anomer for SGLT-1 was higher than that of the R-anomer
of glucosides. Our finding described above was in ac-
cordance with this result.
3.3. P er m ea tion of r- a n d â-SAP G-GGYR th r ou gh
In testin a l Ever ted Sa csTo study further the mechanism
of transport of glycosylated GGYR across the intestinal
tract, the everted sac technique was employed. Time-
course changes in elimination of glycosylated GGYR on the
mucosal side and the appearance of intact constituents on
the serosal side are shown in Figure 7a,b. As GGYR was
eliminated quickly from the mucosal side and no intact
GGYR was detected on the serosal side, it is evident that
GGYR was degraded enzymatically on the mucosal side.
Both R- and â-SAPG-GGYR were eliminated relatively
slowly on the mucosal side and intact R- and â-SAPG-
GGYR permeated to the serosal side. These findings
showed that glycosylation of GGYR increased its stability
to enzymatic degradation and facilitated its transport
through the intestinal membrane. The amount of â-SAPG-
GGYR eliminated from the mucosal side was greater than
that of R-SAPG-GGYR and the amount of â-SAPG-GGYR
that permeated to the serosal side was about twice that of
R-SAPG-GGYR. The fractions of R- and â-SAPG-GGYR
across the everted sac as compared to the amount intro-
duced at the mucosal side were about 0.25 and 0.45%,
respectively. To express elimination and permeation more
quantitatively, we estimated the elimination clearance
(CLel) and permeation clearance (CLp) according to eqs 1
and 2. The CLel and CLp values were 15.8 ( 6.83 and 0.83
We then studied the contribution of the Na+-dependent
D-glucose transporter to the absorption of R- and â-SAPG-
GGYR. Permeation clearances of R- and â-SAPG-GGYR
in the presence of 10 mM of phloridzin, 10 mM D-glucose
and in the absence of a Na+ gradient are shown in Figure
8a,b. Permeation of R-SAPG-GGYR was not markedly
inhibited in the presence of phloridzin (CLp ) 0.55 µL/min/
cm), while in the presence of D-glucose and in the absence
Journal of Pharmaceutical Sciences / 331
Vol. 87, No. 3, March 1998

of Na+ it was significantly inhibited (CLp ) 0.28, 0.285 µL/
min/cm). In the case of â-SAPG-GGYR, CLp was signifi-
cantly decreased in the presence of phloridzin and D-glucose
and in the absence of Na+. These findings showed that
both R- and â-SAPG-GGYR were transported by the Na+-
dependent D-glucose transporter. However, the finding
that permeation of R-SAPG-GGYR was slightly inhibited
by phloridzin suggested the possibility that this R-anomer
was actively transported by the Na+-dependent D-glucose
transporter with low affinity for â-D-glucose. Ohnishi et
al. have reported that there are two kinds of transporter
which have high and low affinity for â-D-glucoside.32
To clarify the difference in absorption between R- and
â-SAPG-GGYR, we calculated the permeation clearance
transported by the Na+-dependent D-glucose transporter
(CLp)Na+ by subtracting the CLp value in the absence of
Na+ from that in the presence of Na+. The (CLp)Na+ was
2.8 µL/min/cm for â-SAPG-GGYR and was about 5 times
greater than that for R-SAPG-GGYR (0.5 µL/min/cm).
Since the â-form has higher affinity for SGLT-1 than the
R-form of glucose, as described above, a large amount of
â-SAPG-GGYR might be transported to the serosal side.
of insulin in the small intestine: Relative importance of
insulin association characteristics in aqueous solution. Pharm.
Res. 1994, 11, 1115-1120.
10. Tsuji, A.; Tamai, I. Carrier-mediated intestinal transport of
drugs. Pharm. Res. 1996, 13, 963-977.
11. Bell, G. I.; Burant, C. F.; Takeda, J .; Gould, G. W. Structure
and function of mammalian facilitative sugar transporters.
J . Biol. Chem. 1994, 268, 19161-19164.
12. Gould, G. W.; Holman, G. D. The glucose transporter
family: Structure, function and tissue-specific expression.
Biochem. J . 1993, 295, 329-341.
13. Hediger, M. A.; Coady, M. J .; Ikeda, T. S.; Wright, E. M.
Expression cloning and cDNA sequencing of the Na+/glucose
cotransporter. Nature Lond. 1987, 330, 379-381.
14. Haga, M.; Saito, K.; Shimaya, T.; Maezawa, Y.; Kato, Y.; Kim,
S. W. Hypoglycemic effect of intestinally administered
monosaccharide-modified insulin derivatives in rats. Chem.
Pharm. Bull. 1990, 38, 1983-1986.
15. J eong, S. Y.; Kim, S. W.; Eenink, M. J . D.; Feijen, J . Self-
regulating insulin delivery systems. I. Synthesis and char-
acterization of glycosylated insulin. J . Controlled Release
1984, 1, 57-66.
16. Ueki, M.; Inazu, T. Synthesis of peptide containing hy-
droxyamino acids by the mixed anhydride method without
protecting the hydroxyl functions. Chem. Lett. 1982, 45-48.
17. Gray, W. R. In Method in Enzymology, 25th ed.; Hiers, C.
H. W.; Timasheff, S.N. Eds.; Academic Press: New York,
1972; pp 128.
18. Hodge, J . E.; Hofreiter, B. T. In Methods in Carbohydrate
Chemistry, 1st ed.; Whister, R. L., Wolfrom, M. L., Eds.;
Academic Press: New York, 1962; pp 388.
19. Tiruppathi, C.; Miyamoto, Y.; Ganapathy, V.; Leibach, F. H.
Fatty acid-induced alterations in transport systems of the
small intestinal brush-border membranes. Biochem. Phar-
macol. 1988, 37, 1399-1405.
20. Takuwa, N.; Shimada, T.; Matsumoto, H.; Hoshi, T. Proton-
coupled transport of glycylglycine in rabbit renal brush-
border membrane vesicles. Biochim. Biophys. Acta 1985, 814,
186-190.
21. Hamlin, J . L.; Arquilla, E. R. Preparation, purification, and
characterization of a biologically active derivative substituted
predominantly on tyrosine A14. J . Biol. Chem. 1974, 10, 21-
32.
4. Conclusion
It was shown from the study using BBMVs and the
everted sac method that glycosylation at the R-amino
position of GGYR increased its resistance to aminopepti-
dase and inhibited its degradation and that the Na+-
dependent glucose transporter played an important role in
the intestinal absorption of both R- and â-SAPG-GGYR. It
was also indicated that the amount of â-SAPG-GGYR
transported was greater than that of R-SAPG-GGYR. The
finding that a tetrapeptide was transported by the Na+-
dependent glucose transporter may be useful in the im-
provement of intestinal absorption of peptide drugs which
show poor membrane penetrability.
22. McCarthy, C. F.; Borland. J r. J . L.; Lynch, H. L., J r.; Owen,
E. E.; Tyor, M. P. Defective uptake of basic amino acids and
L-cystine by intestinal mucosa of patients with cystinuria.
J . Clin. Invest. 1964, 43, 1518-1524.
23. Schilling, R. J .; Mitra, A. K. Intestinal mucosal transport of
References and Notes
insulin, Int. J . Pharm. 1990, 62, 53-64.
1. Friedman, D. I.; Amidon, G. L. Oral absorption: Influence
of pH and inhibitors on the intestinal hydrolysis of leu-
enkephalin and analogues. Pharm. Res. 1991, 8, 93-96.
2. Yamamoto, A.; Hayakawa, E.; Lee, V.H. L. Insulin and
proinsulin proteolysis in mucosal homogentes of the albino
rabbit: Implication in peptide delivery from nonoral routes.
Life Sci. 1990, 47, 2465-2474.
3. Morishita, M.; Morishita, I.; Takayama, K.; Machida, Y.;
Nagai, T. Site-dependent effect of aprotinin, sodium caprate,
Na2EDTA and sodium glycocholate on intestinal absorption
of insulin. Biol. Pharm. Bull. 1993, 16, 68-72.
4. Yamamoto, A.; Taniguchi, T.; Rikyuu, K.; Tsuji, T.; Fujita,
T.; Murakami, M.; Muranishi, S. Effects of various protease
inhibitors on the intestinal absorption and degradation of
insulin in rats. Pharm. Res. 1994, 11, 1496-1499.
5. Burcham, D. L.; Angust, B. A.; Hussain, M.; Gorko, M. A.;
Quon, C. Y.; Huang, S.M. The effect of absorption enhancers
on the oral absorption of the GP IIB/IIIA receptor antagonist,
DMP 728, in rats and dogs. Pharm. Res. 1995, 12, 2065-
2070.
6. Hoogstraate, A. J .; Coos Verhoef, J .; Pijpers, A.; van Leen-
goed, L. A.; Verheijden, J . H.; J unginger, H. E.; Bodde, H.
E. In vivo buccal delivery of the peptide drug buserelin with
glycodeoxycholate as an absorption enhancer in pigs. Pharm.
Res. 1996, 8, 1233-1237.
24. Taki, Y.; Sakane, T.; Nadai, T.; Sezaki, H.; Amidon, G. L.;
Languth, P.; Yamashita, S. Gastrointestinal absorption of
peptide drug: Quantitative evaluation of the permeation of
metkephamide in rat small intestine. J . Pharmacol. Exp.
Ther. 1995, 274, 373-377.
25. Bai, J . P. F.; Amidon, G. L. Structural specificity of mucosal-
cell transport and metabolism of peptide drugs: Implication
for oral peptide drug delivery. Pharm. Res. 1992, 9, 969-
978.
26. Langguth, P.; Bohner, V.; Biber, J .; Merkle, H. P. Metabolism
and transport of the pentapeptide metkephamid by brush-
border membrane vesicles of rat intestine. J . Pharm. Phar-
macol. 1994, 46, 34-40.
27. Thorens, B. Glucose transporters in the regulation of intes-
tinal, renal, and liver glucose fluxes. Am. J . Physiol. 1996,
270 (Gastrointest. Liver Physiol. 33), G541-553.
28. Crane, R. K. Na+-dependent transport in the intestine and
other animal tissues. Fed. Proc. 1965, 24, 1000-1005.
29. Wilson, T. H.; Landeu, B. R. Specificity of sugar transport
by the intestine of the hamster. Am. J . Physiol. 1960, 198,
99-102.
30. Landeu, B. R.; Bernstein, L.; Wilson, T. H. Hexose transport
by hamster intestine in vitro. Am. J . Physiol. 1962, 203, 237-
240.
7. Samanen, J .; Wilson, G.; Smith, P. L. Lee, C. P.; Bondinell,
W.; Ku, T.; Rhodes, G.; Nicols, A. Chemical approaches to
improve the oral bioavailability of peptidergic molecules. J .
Pharm. Pharmacol. 1996, 2, 119-135.
8. Ribadeneira, M. D.; Aungst, B. J .; Eyermann, C. J .; Huang,
S. M. Effects of structural modifications on the intestinal
permeability of angiotensin II receptor antagonists and the
correlation of in vitro, in situ, and in vivo absorption. Pharm.
Res. 1996, 2, 227-233.
9. Asada, H.; Douen, T.; Mizokoshi, Y.; Fujita, T.; Murakami,
M.; Yamamoto, A.; Muranishi, S. Stability of acyl derivatives
31. Mizuma, T.; Ohta, K.; Awazu, S. The â-anomeric and glucose
preferences of glucose transport carrier for intestinal active
absorption of monosaccharide conjugates. Biochim. Biophys.
Acta 1994, 1200, 117-122.
32. Ohnishi, T.; Mizuma, T.; Awazu, S. Elucidation of intestinal
transport mechanism of glycoside using intestinal brush-
border membrane vesicles: I. Elucidation of transport mech-
anism of glucoside. Xenobio. Metabol. Dispos. 1996, 11
(Suppl.) S194.
J S970269P
332 / Journal of Pharmaceutical Sciences
Vol. 87, No. 3, March 1998