Oligoarginine-based prodrugs with self-cleavable spacers for caco-2 cell permeation

Article abstract of DOI:10.1248/cpb.56.1515

In the development of oligoarginine-based prodrugs with self-cleavable spacers for intestinal absorption, we previously reported a series of spacers with variable half-lives of parent compound release based on a neighboring group participation mechanism f

Full text of DOI:10.1248/cpb.56.1515

November 2008
Chem. Pharm. Bull. 56(11) 1515—1520 (2008)
1515
Oligoarginine-Based Prodrugs with Self-Cleavable Spacers for Caco-2 Cell
Permeation
Kentaro TAKAYAMA,a,b Yuka SUEHISA,b Takuya FUJITA,b Jeffrey-Tri NGUYEN,c Shiroh FUTAKI,a,d
Akira YAMAMOTO,b Yoshiaki KISO,c and Yoshio HAYASHI*,c,e
a Institute for Chemical Research, Kyoto University; Uji, Kyoto 611–0011, Japan: b Department of Biopharmaceutics, 21st
Century COE Program, Kyoto Pharmaceutical University; c Department of Medicinal Chemistry, Center for Frontier
Research in Medicinal Science, 21st Century COE Program, Kyoto Pharmaceutical University; Kyoto 607–8412, Japan:
d SORST, JST; Kawaguchi, Saitama 332–0012, Japan: and e Department of Medicinal Chemistry, Tokyo University of
Pharmacy and Life Sciences; Hachioji, Tokyo 192–0392, Japan.
Received November 23, 2007; accepted August 26, 2008; published online September 2, 2008
In the development of oligoarginine-based prodrugs with self-cleavable spacers for intestinal absorption, we
previously reported a series of spacers with variable half-lives of parent compound release based on a neighbor-
ing group participation mechanism from an amino acid side-chain structure next to the succinyl moiety. In the
present study, to diversify the half-life of the spacer, we first synthesized several additional fluorescein isothio-
cyanate ethanolamine (FE)-heptaarginine conjugates (4d—g) and evaluated their conversion time. To investigate
the overall cellular uptake of FE-heptaarginine conjugates, the cellular uptakes of FE-heptaarginines 4a and 4b
possessing the longest and shortest half-lives, respectively, were evaluated using HeLa cells by confocal mi-
croscopy and flow cytometry. Conjugate 4a with a longer half-life was more efficiently taken up by the cells than
conjugate 4b. However, in term of the transport rate of parent FE 1 in in vitro Caco-2 cell permeation assay, con-
jugate 4b with a short half-life could function more efficiently that conjugate 4a. To understand the reason for
this discrepant finding, fluorescence on the basal side medium after treatment with conjugate 4b in the perme-
ation assay was determined. It became apparent that the fluorescence was mostly from the parent FE 1 itself, and
not conjugate 4b, suggesting that the conjugate was cleaved inside the cells. Moreover, the conversion time of con-
jugate 4b (t1/2ꢀ9.4 min at pH 7.4) was significantly extended in slightly acidic media. These results suggest that
the conversion rate was slowed in the relatively acidic endosomal environment where the conjugate was trans-
ferred after endocytosis, and resulted in a favorable migration time across the cells. The other conjugates, in-
cluding conjugate 4a, were more stable inside of the cell, resulting in very long conversion times that were inef-
fective in increasing the permeation rate. Therefore, spacers with shorter half lives, in order to produce a larger
amount of the parent compound inside the cells are promising development for effective oligoarginine-based
cargo-transporter systems to enhance intestinal absorption of parent drugs with low permeability.
Key words cell penetrating peptide; oligoarginine; self-cleavable spacer
Many kinds of cell penetrating peptides (CPPs)1—9) includ- oligoarginine-drug model conjugates 4a—c, which were con-
ing oligoarginine peptides have been widely used as attrac- nected by this chemical-triggered self-cleavable spacers,
tive tools for intracellular delivery of various substances with were synthesized, and their conversion times to parent FE 1
low membrane permeability.2,10—18) Highly cationic clusters via an intramolecular succinimide formation was investi-
composed of constitutive basic guanidino groups in oli- gated.25) The conversion times of these compounds (4a—c)
goarginine peptides are known to interact with negative- were different with t1/2 values ranging from 9 to 100 min.
charged proteoglycans on the surface of cells and internalize Moreover, the conversion time was well controlled by the
by macropinocytosis.19—22) However, the precise internaliza- neighboring-group participation of an adjacent amino acid
tion mechanism of these CPPs has yet been elaborated.
side-chain structure to the succinyl moiety within the spacer
Only recently drug conjugates with oligoarginine peptides
have been expected to increase the penetration of low perme-
able drugs through the intestinal epithelial cell layer into
blood.23—25) In our previous study, we proposed a new self-
cleavable spacer in the oligoarginine-based cargo-transporter
(OACT) system toward effective intestinal absorption.25)
This spacer was designed to release the parent drug from a
hydrophilic oligoarginine moiety inside the cells, in order
to transport the parent drug into blood through the hydropho-
bic basal-side cell membrane, after the conjugate had pene-
trated from the apical-side of cells with the assistance of oli-
goarginine (Fig. 1). Fluorescein isothiocyanate (FITC)-
ethanolamine (FE 1) and H-GABA-D-Arg7-NH2 were chosen
as a drug-model with low intestinal permeability and an
oligoarginine-based CPP, respectively. Using on Fmoc-
based solid phase peptide synthesis (SPPS), three kinds of
Fig. 1. A New Oligoarginine Prodrug Strategy Based on Chemically Trig-
gered Self-Cleavage for Effective Intestinal Absorption
∗ To whom correspondence should be addressed. e-mail: yhayashi@ps.toyaku.ac.jp
© 2008 Pharmaceutical Society of Japan

1516
Vol. 56, No. 11
mass spectrometer equipped with the PerSeptive Biosystems.
Synthesis As shown in Chart 1, similar to a previously reported proce-
dure for preparing FE-heptaarginines,25) to a solution of fluorescein isothio-
cyanate (FITC, 320 mg, 0.82 mmol) in DMF was added ethanolamine
(75 ml, 1.23 mmol) and the reaction mixture was stirred for 2 h at room tem-
perature. After removing DMF in vacuo, the residue was dissolved in
EtOAc, washed with 5% citric acid and water for three-times each, dried
over Na2SO4 to obtain title compound 1 as a yellow solid (260 mg, 70%). mp
1
ꢁ300 °C, H-NMR (300 MHz, DMSO-d6) d: 10.10 (br s, 2H), 9.98 (s, 1H),
8.30 (s, 1H), 8.04 (br s, 1H), 7.74 (br d, Jꢂ8.4 Hz, 1H), 7.16 (d, Jꢂ8.1 Hz,
1H), 6.66 (d, Jꢂ1.8 Hz, 2H), 6.62—6.54 (m, 4H), 3.59 (m, 4H, partially
overlapped with water peak). HR-MS (FAB) m/z: 451.0960 for [MꢃH]ꢃ
(Calcd 451.0964 for C23H19O6N2S). Anal. Calcd for C23H18N2O6S·0.5H2O:
C, 60.12; H, 4.17; N, 6.10; Found: C, 60.38; H, 4.29; N, 6.32.
To a solution of FITC-ethanol amine 1 (200 mg, 0.444 mmol) in DMF was
added succinic anhydride (49 mg, 0.490 mmol) and DMAP (18.1 mg,
0.148 mmol), and the reaction mixture was stirred for 16 h at 50 °C. After re-
moving DMF in vacuo, the residue was purified by preparative TLC
(20ꢀ20 cm, eluent; CHCl3 : MeOH : H2Oꢂ8 : 3 : 1 upper phase) to afford
title compound 2 as a yellow solid (137 mg, 56%). mp 262—264 °C, 1H-
NMR (300 MHz, DMSO-d6) d: 9.80 (s, 1H), 8.22 (s, 1H), 7.93 (d, Jꢂ8.4
Hz, 1H), 7.75 (m, 1H), 7.58 (d, Jꢂ1 Hz, 1H), 7.09 (d, Jꢂ8.1 Hz, 1H), 6.75
(br s, 2H), 6.68 (d, Jꢂ8.4 Hz, 2H), 6.59 (d, Jꢂ8.4 Hz, 2H), 4.33 (t, Jꢂ4.8
Hz, 2H), 3.93 (dd, Jꢂ5.1, 10.1 Hz, 2H), 2.63 (m, 4H). HR-MS (FAB) m/z
551.1129 for [MꢃH]ꢃ (Calcd 551.1124 for C27H23O9N2S). Anal. Calcd for
C27H22N2O9S·H2O: C, 57.04; H, 4.25; N, 4.93; Found: C, 56.82; H, 4.38; N,
4.85.
Fig. 2. Oligoarginine-Drug Model Conjugates
Spacers Based on Intramolecular Succinimide Formation
4 with Self-Cleavable
region. Furthermore, an ideal conversion time seemed to be
important to increase parent compound permeability in our
human intestinal cell line (Caco-2) model experiment, be-
To a Rink amide AM resin (340 mg, 0.25 mmol), Fmoc-amino acids
cause only conjugate 4b with a t1/2 value of 9.4 min exhibited (0.75 mmol), including Fmoc-Arg(Pmc)-OH, Fmoc-GABA-OH and then
improved Caco-2 cell permeability.25)
In the present study, at first, to diversify the half-life of the
spacer, we synthesized additional FE-heptaarginine conju-
Fmoc-Xaa-OH (XaaꢂD-Ala, Dap, Dab, D-Asn, D-His, Orn or D-Gln) were
sequentially coupled using a DIPCDI (0.75 mmol)–HOBt (0.75 mmol)
method for 2 h in DMF after removal of each Fmoc group with 20% piperi-
dine–DMF (7 mlꢀ2 times, 2, 20 min) to obtain H-Xaa-GABA-D-Arg(Pmc)7-
gates 4d—g (Fig. 2) and evaluated their conversion time.
Next, we wanted to investigate the overall cellular uptake of
FE-heptaarginine conjugates. The cellular uptakes of FE-
heptaarginines 4a and 4b possessing the longest and shortest
half-lives, respectively, were evaluated using HeLa cells by
confocal microscopy and flow cytometry. Furthermore, the
efficacy of these conjugates in in vitro permeation assay
using Caco-2 cell monolayer was determined. The results in-
dicated that only 4b with a t1/2 value of 9.4 min was the better
conjugate to increase Caco-2 cells permeability. To under-
stand the mechanism under the observed increased transport
of the parent compound in 4b, the conversion time depend-
ency on pH, and fluorescence of FE 1 at the basal side in the
in vitro permeation assay were determined. All synthetic de-
tails of oligoarginine conjugates 4a—g are also described.
NH-resins 3a—g. One-third of this resin 3a (XaaꢂD-Ala) was reacted with
compound 2 (137 mg, 0.25 mmol) by the same coupling method for 2 h and
resultant protected FE-heptaarginine was deprotected with TFA–m-
cresol–thioanisole (5.5 ml, 20 : 1 : 1) for 150 min at rt, followed by prepara-
tive HPLC purification in a 0.1% aqueous TFA–CH3CN system to obtain
FE-heptaarginine 4a as a TFA salt. Yield: 24%; HR-MS (FAB) m/z
1798.9249 for [MꢃH]ꢃ (Calcd 1798.9261 for C76H120O17N33S), purity was
ꢁ97% (HPLC analysis at 230 nm). Other FE-heptaarginines 4b—g were
synthesized by the same method as 4a.
4b: Yield 16%; HR-MS (FAB) m/z: 1813.9381 for [MꢃH]ꢃ (Calcd
1813.9370 for C76H121O17N34S), purity was ꢁ96% (HPLC analysis at
230 nm).
4c: Yield: 29%; LR-MS (MADLI-TOF) m/z: 1829 for [MꢃH]ꢃ (Calcd
1829 for C77H123O17N34S), purity was ꢁ99% (HPLC analysis at 230 nm).
4d: Yield: 5%; HR-MS (FAB) m/z: 1841.9329 for [MꢃH]ꢃ (Calcd
1841.9319 for C77H121O18N34S), purity was ꢁ97% (HPLC analysis at
230 nm).
4e: Yield: 30%; HR-MS (FAB) m/z: 1864.9493 for [MꢃH]ꢃ (Calcd
1864.9479 for C79H122O17N35S), purity was ꢁ97% (HPLC analysis at
230 nm).
Experimental
Materials Reagents and solvents were purchased from Wako Pure
4f: Yield: 7%; HR-MS (FAB) m/z: 1841.9689 for [MꢃH]ꢃ (Calcd
Chemical Ind., Ltd. (Osaka, Japan), Nakalai Tesque (Kyoto, Japan), and 1841.9683 for C78H125O17N34S), purity was ꢁ97% (HPLC analysis at
Aldrich Chemical Co., Inc. (Milwaukee, WI, U.S.A.) and used without fur- 230 nm).
ther purification. Analytical thin-layer chromatography (TLC) was per-
4g: Yield: 6%; HR-MS (FAB) m/z: 1855.9489 for [MꢃH]ꢃ (Calcd
formed on Merck silica gel 60F254 precoated plates. Analytical high-per- 1855.9475 for C78H123O18N34S), purity was ꢁ96% (HPLC analysis at
formance liquid chromatography (HPLC) was carried out on a C18 reverse- 230 nm).
phase column (4.6ꢀ150 mm; YMC Pack ODS AM302) with a binary sol-
vent system: a linear gradient of CH3CN in 0.1% aqueous trifluoroacetic
Conversion of FE-Heptaarginines to FE 1 In vitro conversion profiles
of FE-heptaarginines 4a—g to FE 1 were determined under physiological
acid (TFA) at a flow rate of 0.9 ml/min, detected at UV 230 nm. Preparative conditions with Krebs–Ringer bicarbonate buffer (KRBB, pH 7.4). A solu-
TLC was performed on Merck silica gel 60F254, 2 mm precoated plates.
tion of FE-heptaarginie (0.5 mM, 10 ml) in DMSO was added to KRBB
Preparative HPLC was performed using a C18 reverse-phase column (990 ml) and the mixture was incubated at 37 °C. At the desired time point,
(19ꢀ100 mm; SunFireTM Prep C18 OBDTM 5 mm) with a binary solvent sys-
whole samples (1 ml) were directly applied to RP-HPLC and their con-
tem: a linear gradient of CH3CN in 0.1% aqueous TFA at a flow rate of version profiles were analyzed using
C18 reverse-phase column
15 ml/min, detected at UV 228 nm and 495 nm. Solvents used for HPLC
(4.6ꢀ150 mm; YMC Pack ODS AM302) with a binary solvent system: a
a
were of HPLC grade. All other chemicals were of analytical grade or better. linear gradient of CH3CN (0—50%, 25 min) in 0.1% aqueous TFA at a flow
1H nuclear magnetic resonance (NMR) spectra were obtained on a JEOL rate of 0.9 ml·minꢄ1, detected at UV 230 nm. Conversion time courses for
300 MHz spectrometer with TMS as an internal standard. Fast atom bom-
bardment mass spectrometry (FAB-MS) was performed on a JEOL JMS-
SX102A spectrometer equipped with the JMA-DA7000 data system. Ma-
trix-assisted laser desorption ionization time-of-flight mass spectrometry
compounds 4a—g are shown in Fig. 3.
Cellular Uptake of FE-Heptaarginines Human cervical cancer-de-
rived HeLa cells were maintained in a-minimum essential medium (a-
MEM) with 10% heat-inactivated calf serum. These cells were grown on
(MALDI-TOF-MS) was performed on a Voyager-DETM RP time-of-flight 100 mm dishes and incubated at 37 °C under 5% CO2 to approximately 70%

November 2008
1517
confluence. A subculture was performed every 3—4 d.
For the flow cytometry study, HeLa cells (7.0ꢀ104) in fresh culture
medium (1 ml) were plated into 24-well microplates (Iwaki) and cultured for
48 h in a-MEM containing 10% heat-inactivated calf serum. After complete
adhesion, the cells were incubated at 37 °C for 30 min with fresh medium
(200 ml) containing peptide conjugates prior to washing with PBS containing
sodium heparin (0.5 mg/ml). The cells were then treated with 0.01% trypsin
in PBS (200 ml) at 37 °C for 10 min prior to the addition of PBS (300 ml).
The cells were centrifuged at 3000 rpm for 5 min. After the supernatant was
removed, the cells were washed with 400 ml of PBS and centrifuged at
3000 rpm for 5 min. After this washing cycle was repeated, the cells were
suspended in PBS (400 ml) and subjected to fluorescence analysis on a FAC-
Scalibur (BD Biosciences) flow cytometer using 488 nm laser excitation and
a 515 to 545 nm emission filter.
For the confocal microscopy study, HeLa cells (2ꢀ105) were plated on
35 mm glass-bottomed dishes (Iwaki) and cultured in a-MEM with 10%
heat-inactivated calf serum for 48 h. After complete adhesion, the culture
medium was exchanged, and then the cells were incubated at 37 °C with
fresh medium (200 ml) containing the peptide conjugates and washed with
fresh medium (ꢀ3). Distribution of the peptide conjugates was then ana-
lyzed using a confocal scanning laser microscope FV300 (Olympus)
equipped with a 40ꢀ objective without fixing the cells to avoid artifactual
localization of the internalized peptides.26)
Reagents and conditions: (a) 20% piperidine/DMF; (b) Fmoc-Arg(Pmc)-OHꢀ7,
Fmoc-GABA-OH, then Fmoc-Xaa-OH (Xaa: D-Ala, Dap(Boc), Dab(Boc), D-Asn(Trt),
D-His(Trt), Orn(Boc) or D-Gln(Trt), respectively) in turn, DIPCD, HOBt, DMF; (c)
FITC-ethanolamine monosuccinate 2, DIPCD, HOBt, DMF; (d) TFA, thioanisole, m-
cresol; (e) preparative HPLC.
Chart 1
Penetration Measurement in Caco-2 Cell Monolayer (in Vitro Perme-
ation Assay) Caco-2 cells cultured on TranswellTM for 14—21 d were used
for the transport assay. Transepithelial electric resistance (TEER) was meas-
ured to ensure cell monolayer integrity. Cell monolayers with TEER values
greater than 500 W·m2 were used in transport experiments. After removal of
Results and Discussion
Synthesis As shown in Chart 1, similar to a previously
reported procedure,25) seven kinds of FE-heptaarginines 4a—
c25) and 4d—g were designed and synthesized. Briefly, FE 1,
the culture medium, the cells were washed once with KRBB (pH 7.4), and prepared by coupling FITC with ethanolamine in DMF, was
preincubated for 10 min in KRBB at 37 °C. To analyze the transport of the
samples from the apical-to-basal side, a stock solution of each sample
(2.5 mM) in 0.01 N HCl was diluted 50 times with KRBB (pH 7.4) and this
solution (500 ml) was immediately applied to the apical side (final concentra-
reacted with succinic anhydride in the presence of 4-di-
methylaminopyridine (DMAP) in DMF to afford FE-mono-
succinate 2. Protected peptide resins 3a—g were synthesized
by Fmoc-based SPPS. After loading Fmoc-D-Arg(Pmc)-OH
(Pmc: 2,2,5,7,8-pentamethylchroman-6-sulfonyl), to a Rink
amide AM resin by DIPCD-HOBt method (DIPCD: 1,3-di-
isopropylcarbodiimide; HOBt: 1-hydroxybenzotriazole), pep-
tide chains were elongated by the same coupling method and
the respective Fmoc groups were deprotected with 20%
piperidine/DMF to obtain H-Xaa-GABA-D-Arg(Pmc)7-NH-
resins 3a—g. Compound 2 was then reacted to peptide-resins
3a—g using the same coupling method. Finally, the peptide
resins were deprotected with a TFA–thioanisole–m-cresol
system and the resultant crudes 4a—g were purified by
HPLC as TFA salts. A control peptide, FITC-GABA-(D-
Arg)7-NH2 (5), was also synthesized by similar SPPS.
tion: 50 mM) after removing the preincubation media from the apical and
basal sides. Fresh KRBB (1.5 ml) was also added to the basal side. The cells
were incubated at 37 °C. An aliquot of each sample was taken from each
basal medium at 3 or 5 time points (15, 30, 45, 60, 120 min) and the fluores-
cence intensity was measured using a microplate reader with an excitation
wavelength at 485 nm and emission wavelength at 535 nm. FE 1 and FITC-
GABA-(D-Arg)7-NH2 5 were used as controls (50 mM in KRBB, pH 7.4). The
results are shown in both Fig. 5A (4b—d, 1, 5) and Fig. 5B (4a, e—g, 1, 5).
Paap Analysis From the in vitro permeation assay, the apparent perme-
ability coefficient (Papp) was calculated using the equation:
Papp (cm/sꢀ106)ꢂFlux/(AꢀC0ꢀ60)
where Flux (nmol/min) is the slope of the plot’s best fitted line, (i.e. pene-
trated quantity per minute); A (cm2) is the area of the membrane; and C0
(mM) is the initial concentration of the applied sample solution (in this case,
calculated as Areaꢂ1.12, and C0ꢂ50 was a constant).
Fluorophotometric Analysis of Basolateral Medium for Compound
4b After the cells were incubated at 37 °C for 15 and 60 min in in vitro
permeation assay, 1 ml of each basolateral medium, from the six Transwells
from which the same 4b was treated, was collected (total, 6 ml), acidified
Conversion of Compounds 4a—g to FE 1 In addition
to the previously reported compounds 4a—c,25) the conver-
sion time (t1/2) for synthesized FE-heptaarginines 4d—g to
FE 1 was evaluated by HPLC under physiological conditions
with 1 N HCl (0.5 ml) to stop the conversion, and then lyophilized. After (Krebs–Ringer bicarbonate buffer, KRBB, pH 7.4, 37 °C). As
lyophilization, each resultant powder was dissolved in 0.1 M HCl–DMSO
(600 ml, 1 : 1) and the solution (50 ml) was injected to fluorophotometric
HPLC to measure the fluorescence intensity using a linear gradient of
shown in Table 1 and Fig. 3, in combination with the previ-
ously reported data,25) conjugate 4a having D-Ala with no
basic side chain functionality next to the succinyl moiety
CH3CN (10—42%) in 0.1% aqueous TFA over 16 min with a flow rate of
(Xaa site), exhibited the longest t1/2 of ca. 106 min, while the
basic side chains of FE-heptaarginines 4b—g in the spacer
significantly shortened the conversion time. Especially, Dap
(L-diaminopropanoic acid) derivative 4b, which is expected
to undergo a nucleophilic neighboring-group participation
through a five-membered ring intermediate, exhibited a quick
conversion with a t1/2 value of 9.4 min and almost no side re-
action was observed from this conversion as previously re-
ported.25) Dab (L-diaminobutanoic acid) and D-Asn deriva-
tives 4c and 4d, respectively, going through a six-membered
ring intermediate, showed a slightly longer conversion time
(t1/2ꢂ18, 28 min, respectively). The difference in conversion
times between 4c and 4d indicates that the more nucleophilic
1.0 ml/min detected at Ex. 485 nm and Em. 535 nm. Standard compounds 1
and 4b were analyzed under the same HPLC conditions for comparison. The
Fluorophotometric HPLC profiles are shown in Fig. 6.
pH-Dependent Conversion of Compound 4b In vitro conversion pro-
files of FE-heptaarginine 4b to the FE 1 were determined under acidic con-
ditions with KRBB (pH 6.5, 5.5). A solution of compound 4b (1 mM, 15 ml)
in DMSO was added to KRBB (985 ml) and the mixture was incubated at
37 °C. At the desired time point, a portion of the sample (200 ml) was ap-
plied to RP-HPLC and their conversion profiles were analyzed using a C18
reverse-phase column (ChromolithTM Performance RP-18 endcapped 100—
4.6 mm) with a binary solvent system: a linear gradient of CH3CN contain-
ing 0.1% TFA (5—37%, 8 min) in 0.1% aqueous TFA at a flow rate of
4 ml/min, detected at UV 215 nm.

1518
Vol. 56, No. 11
Table 1. Biological Evaluations of FE-Heptaarginines
a)
b)
Ring size of
intermediate
t1/2
PappAB
Compd
Xaa
(min)
(cm/sꢀ10ꢄ6
)
4ac)
4bc)
4cc)
4d
4e
4f
D-Ala
Dap
Dab
D-Asn
D-His
Orn
D-Gln
—
—
5
6
6
6
7
7
—
—
—
106ꢅ2.6
9.4ꢅ0.10
18ꢅ0.37
28ꢅ0.69
65ꢅ1.2
85ꢅ2.1
88ꢅ3.2
—
0.41ꢅ0.08
0.75ꢅ0.10
0.48ꢅ0.09
0.50ꢅ0.06
0.44ꢅ0.08
0.41ꢅ0.07
0.33ꢅ0.04
0.26ꢅ0.11
0.38ꢅ0.01
0.18ꢅ0.02
4g
FE 1c)
5c,d)
LYe)
—
—
—
—
a) Values were calculated from each degradation profile (Fig. 3). b) Values were
calculated from the results of in vitro model permeation assays (Fig. 4), with S.E.M.
obtained from 8 (4a—g), 6 (FE 1 and 5) or 4 (6) experiments. c) See ref. 25. d) 5:
FITC-GABA-(D-Arg)7-NH2. e) LY: lucifer yellow. —: not applicable.
Fig. 3. Conversion of FE-Heptaarginines 4a—g
Black diamond: 4a; black square: 4b; black triangle: 4c; white circle: 4d; white trian-
gle: 4e; ꢀ-mark: 4f and white square: 4g. The conjugate concentrations at each time
point are represented as relative percentages to the starting concentrations. Plots are
meansꢅS.E.M. of three experiments in KRBB (pH 7.4 at 37 °C). Data for 4a—c were
obtained from ref. 25.
amine in Dab increased the conversion rate, although this
effect is not as significant as the steric stability of the inter-
mediate participating neighboring-groups (i.e., ring size).
FE-heptaarginines 4f and 4g with respective Orn and D-Gln,
undergoing a seven-membered ring intermediate, exhibit-
ed much longer conversion times (t1/2ꢂ85, 88 min, respec-
tively). These results suggest a correlation between the
chemical structure of Xaa and conversion time. Interestingly,
a conversion time of 65 min was observed for D-His deriva-
Fig. 4. Cellular Uptake of FE-Heptaarginines
(A) Confocal microscopic images; cell line: HeLa; incubation: 50 mM, 120 min,
37 °C, a-MEM(ꢃ); cells were washed three times with a-MEM(ꢃ) after incubation
and then observed by confocal microscopy. (B, C) Flow cytometric analysis; cell line:
HeLa; incubation: 50 mM, 30 min (B, black), 120 min (B, white) 30 min after cell-free
pre-incubation for 30 min (C), 37 °C, a-MEM(ꢃ); meansꢅS.D. of three experiments
are shown in flow cytometric analysis.
tive 4e. Thus, by introducing amino acids of different basic- bated at 37 °C in a-MEM containing 10% serum, washed,
ity to the Xaa position in the novel spacers, we successfully and trypsinized. The amount of the respective peptide taken
developed FE-heptaarginines conjugates with variable con- up by these cells was then analyzed by FACS. The majority
version times ranging from 9 to 106 min, resulting in control- of the cell-surface adsorbed peptides were removed by he-
lable release of FE 1.
parin-containing (0.5 mg/ml) PBS wash and trypsin treat-
Cellular Uptake of FE-Heptaarginines We investi- ment, and thus the data reflected the total cellular uptake of
gated the overall cellular uptake of FE-heptaarginines using the peptides (see Experimental).27) As shown in Fig. 4A, a
HeLa cells. The cellular uptakes of FE-heptaarginines 4a and brighter intracellular fluorescence in the 120 min treatment of
4b possessing the longest and shortest half-lives, respec- FE-heptaarginines 4a and 4b than that of FE 1 was observed
tively, were evaluated by confocal microscopy and flow cy- by microscopy. Corresponding results that both FE-hep-
tometry. In confocal microscopy, the cells were incubated at taarginines 4a and 4b could increase intracellular fluores-
37 °C with fresh medium (200 ml) containing the peptide cence were obtained by flow cytometric analysis (Fig. 4B).
conjugates and observed after washing with serum-contain- Although FE-heptaarginine 4a or 4b in the 120 min incuba-
ing medium (ꢀ3). In flow cytometry, FE-heptaarginines 4a, tion increased or slightly increased the total amount of fluo-
4b and FE 1 (50 mM each) were added to HeLa cells, incu- rescence delivered into cells than that in the 30 min incuba-

November 2008
1519
tion, respectively (Fig. 4B), FE-heptaarginine 4a more effec-
tively increased intracellular fluorescence than conjugate 4b.
Furthermore, under the conditions of cell-free pre-incubation
for 30 min before adding the conjugates to the cells (Fig.
4C), only FE-heptaarginine 4a enhanced intracellular fluores-
cence relative to FE 1 alone. Consequently, we concluded
that FE-heptaarginine 4a possessing a longer conversion time
was continuously taken up into the cells for 120 min, while
the cellular uptake of FE-heptaarginine 4b having a t1/2 value
of 9.4 min was mostly completed within 30 min, probably
due to the self-cleavage of the heptaarginine moiety outside
the cells.
In Vitro Permeation Assay To investigate the efficacy
of FE-heptaarginine conjugates on parent drug intestinal cell
permeation after cellular uptake, in vitro permeation assays
using Caco-2 cell monolayer were performed.28,29) The quan-
tity of compounds permeated to the basal side was evaluated
as fluorescence intensity, derived from both FE-heptaarginine
and produced FE 1, in the basal medium. Paap values were
calculated based on this basal fluorescence. Among all FE-
heptaarginine including previously reported 4a—c,25) conju-
gate 4b having a t1/2 value of 9.4 min increased the transport
rate of FE 1 by at least two times higher than FE 1 alone.
Other FE-heptaarginine with longer conversion times exhib-
ited similar or slight increase in permeation over FE 1 alone
(Table 1, Fig. 5). To investigate the mechanism that increased
permeation in conjugate 4b, compounds with fluorescence
in the basal medium were analyzed by fluorophotometric
HPLC. A tiny amount of conjugate 4b was detected only in
the 15 min incubation. The major fluorescence detected was
mostly from FE 1 in both 15 and 120 min incubations (Fig.
6). The Papp value for lucifer yellow (LY), a paracellular
transport marker, was only 0.18ꢅ0.02, suggesting that para-
cellular transport is not involved in increasing the transport
Fig. 5. In Vitro Permeation Assay
(A) Compounds 4b—d (possessing shorter half life), FE 1 and 5. Black diamond:
4b; black triangle: 4c; white square: 4d; ꢀ-mark: FE 1 and white triangle: 5. Plots are
meansꢅS.E.M. of 8 (4b—d) or 6 (FE 1, 5) experiments. (B) Compounds 4a, e—g
(possessing longer half-lives), FE 1 and 5. White diamond: 4a; black diamond: 4e;
white square: 4f; black triangle: 4g; ꢀ-mark: FE 1 and black triangle: 5. Plots are
meansꢅS.E.M. of 6 (FE 1, 5) or 8 (4a, e—g) experiments.
rate in 4b. These results suggest that compound 4b is trans- in FE-heptaarginine 4b, up to 120 min despite its very short
ported intracellularly, mostly converted to its FE 1, and then t1/2 value at pH 7.4, that resulted in at least two-times higher
migrates to the basal medium.
transport than FE 1 alone. The other oligoarginine conjugates
In the cellular uptake of fluorescence in HeLa cells and in with longer conversion times are expected to be more stable
vitro permeation of parent FE 1 in Caco-2 cells experiments, under endosomal conditions and might result in slower trans-
the observed discrepancy between FE-heptaarginines 4a and portation rates in the Caco-2 permeation assay. Additionally,
4b with different FE 1 formation rates might be explained as the quantum efficiency of fluorescein is known to be reduced
follows; FE-heptaarginine 4a with a longer half-life could be with environmental acidification.31) Endosomal fluorescence
taken up intracellularly better than conjugate 4b with a of conjugates and/or FE 1 that was taken up by the cells via
shorter half-life. However, the intracellular free FE 1 concen- endocytosis might fade, and the amount of internalized fluo-
tration would be increased for conjugate 4b more quickly rescence might be underestimated in our flow cytometric
than that of conjugate 4a. The reason is that self-cleavable analysis. Taking into account all of these considerations, a
elimination of the highly hydrophilic heptaarginine moiety shorter conversion time at pH 7.4 is effective for improving
and the formation of relatively hydrophobic and small molec- permeability, and thereby suggesting the importance of con-
ular weight FE 1, would lead to enhanced transport of FE 1 version time controlled by the self-cleavable spacer.
to the basal medium.
Because 1) the conversion time from the conjugate to FE 1
pH-Dependent Conversion of Compound 4b We eval- is dependent on the chemical structure of the spacer and the
uated the conversion of the conjugates to FE 1 under differ- pH of the environment, 2) FE 1 itself, which was added to
ent pH conditions such as 7.4, 6.5 and 5.5. In FE-heptaargi- the apical medium, did not enhance cellular transport as
nine 4b with enhanced transport of FE 1, in spite of being at shown in Table 1, 3) the detected major fluorescence in the
t1/2ꢂ9.4 min at pH 7.4, the conversion time extended to 15 basal medium was mostly derived from FE 1 and not from
and 59 min at pH 6.5 and 5.5, respectively. A similar pH-de- conjugate 4b that was added to the apical medium in Caco-2
pendent succinimide formation has been described in water- permeation assay (Fig. 6), and 4) a conjugate can internalize
soluble prodrug studies based on a chemical mechanism.30) into the cell using the oligoarginine moiety, it is reasonable
In cellular assay, the prolonged conversion to FE 1 under to consider that the conjugate was delivered into the cells and
slightly acidic conditions in endosomes (pH ꢁ5) is presum- most FE 1 was released from FE-heptaarginine 4b intracellu-
ably associated with the observed gradual transport of FE 1 larly, then transferred to the basal medium.

1520
Vol. 56, No. 11
495—504 (2005).
5) Wright L. R., Rothbard J. B., Wender P. A., Curr. Prot. Pept. Sci., 4,
105—124 (2003).
6) Vivès E., Brodin P., Lebleu B., J. Biol. Chem., 272, 16010—16017
(1997).
7) Derossi D., Joliot A. H., Chassaing G., Prochiantz A., J. Biol. Chem.,
269, 10444—10450 (1994).
8) Futaki S., Suzuki T., Ohashi W., Yagami T., Tanaka S., Ueda K., Sug-
iura Y., J. Biol. Chem., 276, 5836—5840 (2001).
9) Wender P. A., Mitchell D. J., Pattabiraman K., Pelkey E. T., Steinman
L., Rothbard J. B., Proc. Natl. Acad. Sci. U.S.A., 97, 6043—6050
(2000).
Fig. 6. Fluorophotometric HPLC Profiles
(A) 1 alone, (B) 4b alone, (C) a sample taken from the basal medium at 15 min after
4b was added to the apical medium, and (D) a sample taken from the basal medium at
120 min after 4b was added to the apical medium.
10) Astriab-Fisher A., Sergueev D. S., Fisher M., Shaw B. R., Juliano R.
L., Biochem. Pharmacol., 60, 1253—1257 (2000).
11) Torchilin V. P., Rammohan R., Weissig V., Levchenko T. S., Proc. Natl.
Acad. Sci. U.S.A., 98, 8786—8791 (2001).
12) Torchilin V. P., Levchenko T. S., Rammohan R., Volodina N., Papahad-
jopoulos-Sternberg B., D’Souza G. G., Proc. Natl. Acad. Sci. U.S.A.,
100, 1972—1977 (2003).
13) Torchilin V. P., Levchenko T. S., Curr. Prot. Pept. Sci., 4, 141—150
(2003).
14) Khalil I. A., Kogure K., Futaki S., Harashima H., J. Biol. Chem., 281,
3544—3551 (2006).
15) Rothbard J. B., Garlington S., Lin Q., Kirshberg T., Kreider E., Mc-
Grane P. L., Wender P. A., Khavari P. A., Nat. Med., 6, 1253—1257
(2000).
16) Kirshberg T. A., VanDeusen C. L., Rothbard J. B., Yang M., Wender P.
A., Org. Lett., 5, 3459—3462 (2003).
Although the sterically unhindered ester in the present
self-cleavable model drug-oligoarginine conjugates might be
hydrolyzed by esterases in in vivo intestinal fluids, a more
sterically hindered ester, to be developed in future studies,
could be effective in avoiding such enzymatic cleavage, when
taking into consideration that our previously developed
chemically cleavable water-soluble paclitaxel prodrug (iso-
taxel), which has a bulky secondary ester bond, was stable
against esterases.30) This design could realize into a practical
OACT system based on chemical cleavage for effective intes-
tinal absorption of a parent drug. Consequently, our system
might be applicable to orally non-bioavailable or relatively
low-bioavailable drugs such as taxoids and aspartic protease 17) Futaki S., Niwa M., Nakase I., Tadokoro A., Zhang Y., Nagaoka M.,
inhibitors30,32,33) that contain bulky secondary hydroxyl moi-
eties. Furthermore, we believe that this system is useful for
the design of prodrugs, because the parent drugs are quanti-
tatively reproduced.
Wakako N., Sugiura Y., Bioconjug. Chem., 15, 475—481 (2004).
18) Futaki S., Nakase I., Suzuki T., Nameki D., Kodama E., Matsuoka M.,
Sugiura Y., J. Mol. Recog., 18, 169—174 (2005).
19) Console S., Marty C., Garcia-Echeverria C., Schwendener R.,
Ballmer-Hofer K., J. Biol. Chem., 278, 35109—36114 (2003).
20) Goncalves E., Kitas E., Seelig J., Biochemistry, 44, 2692—2702
(2005).
21) Wadia J. S., Stan R. V., Dowdy S. F., Nat. Med., 10, 310—315 (2004).
22) Nakase I., Niwa M., Takeuchi T., Sonomura K., Kawabata N., Koike
Y., Takehashi M., Tanaka S., Ueda K., Simpson J. C., Jones A. T., Sug-
iura Y., Futaki S., Mol. Ther., 10, 1011—1022 (2004).
23) Morishita M., Kamei N., Ehara J., Isowa K., Takayama K., J. Con-
trolled Release, 118, 177—184 (2007).
24) Liang J. F., Yang V. C., Biochem. Biophys. Res. Commun., 335, 734—
738 (2005).
25) Hayashi Y., Takayama K., Suehisa Y., Fujita T., Nguyen J.-T., Futaki
S., Yamamoto A., Kiso Y., Bioorg. Med. Chem. Lett., 17, 5129—5132
(2007).
In conclusion, oligoarginine-model drug conjugates pos-
sessing a series of peptidic self-cleavable spacer, converting
to their parent drugs with different half-lives via succinimide
formation, were developed. The conversion time is controlled
by a basic amino acid side-chain structure next to the suc-
cinyl moiety on the spacer. Caco-2 permeation assays indi-
cated that an ideal time-dependent self-cleavage of the pep-
tide spacer seems to be important for improving permeability
of drugs. These novel peptidic self-cleavable spacers with
shorter conversion times are promising for the development
of safe and effective oligoarginine-based cargo-transporter
(OACT) systems, to enhance Caco-2 cell permeability of par-
ent drugs with low permeability, as well as intracellular de-
livery of substances with low membrane permeability.
26) Richard J. P., Melikov K., Vives E., Ramos C., Verbeure B., Gait M. J.,
Chernomordik L. V., Lebleu B., J. Biol. Chem., 278, 585—590 (2003).
27) Nakase I., Tadokoro A., Kawabata N., Takeuchi T., Katoh H., Hi-
ramoto K., Negishi M., Nomizu M., Sugiura Y., Futaki S., Biochem-
istry, 46, 492—501 (2007).
Acknowledgements This research was supported by various grants
from MEXT (Ministry of Education, Culture, Sports, Science and Technol-
ogy-Japan, including the 21st Century COE Program and Japan Society for
the Promotion of Science’s Post-Doctoral Fellowship for Foreign Re-
searchers. K. T. is grateful for JSPS Research Fellowship for Young Scien-
tists. We are grateful to Ms. K. Oda and Mr. T. Hamada for mass spectra
measurements.
28) Fujita T., Majikawa Y., Umehisa S., Okada N., Yamamoto A., Ganapa-
thy V., Leibach F. H., Biochem. Biophys. Res. Commun., 261, 242—
246 (1999).
29) Fujita T., Kawahara I., Quan Y.-S., Hattori K., Takenaka K., Muranishi
S., Yamamoto A., Pharm. Res., 15, 1387—1392 (1998).
30) Hayashi Y., Skwarczynski M., Hamada Y., Sohma Y., Kimura T., Kiso
Y., J. Med. Chem., 46, 3782—3784 (2003).
31) Urano Y., Kamiya M., Kanda K., Ueno T., Hirose K., Nagano T., J.
Am. Chem. Soc., 127, 4888—4894 (2005).
References
1) Futaki S., Biopolymers, 84, 241—249 (2006).
2) Snyder E. L., Dowdy S. F., Expert Opin. Drug Deliv., 2, 43—51
(2005).
3) Zorko M., Langel Ü., Adv. Drug Deliv. Rev., 57, 529—545 (2005).
4) Rothbbard J. B., Jessop T. C., Wender P. A., Adv. Drug Deliv. Rev., 57,
32) Matsumoto H., Sohma Y., Kimura T., Hayashi Y., Kiso Y., Bioorg.
Med. Chem. Lett., 11, 605—609 (2001).
33) Sohma Y., Hayashi Y., Ito T., Matsumoto H., Kimura T., Kiso Y., J.
Med. Chem., 46, 4124—4135 (2003).