6A-O-[(4-biphenylyl)acetyl]-alpha-, -beta-, and -gamma-cyclodextrins and 6A-deoxy-6A-[[(4-biphenylyl)acetyl]amino]-alpha-, -beta-, and -gamma-cyclodextrins: potential prodrugs for colon-specific delivery.

Article abstract of DOI:10.1021/jm970130r

Cyclodextrins (CyDs) are known to be fermented to small saccharides by colonic microflora, whereas they are only slightly hydrolyzable and thus are not easily absorbed in the stomach and small intestine. This property of CyDs is particularly useful for colon-specific delivery of drugs. In this study, an antiinflammatory 4-biphenylylacetic acid (BPAA) was selectively conjugated onto one of the primary hydroxyl groups of alpha-, beta-, and gamma-CyDs through an ester or amide linkage, 6A-O-[(4-biphenylyl)acetyl[-alpha-, -beta-, and -gamma-CyDs (1-3) and 6A-deoxy-6A-[[(4-biphenylyl)acetyl]amino]-alpha-, -beta-, and -gamma-CyDs (4-6). In rat cecal and colonic contents (10%, w/v), 1 and 3 released more than 95% of BPAA within 1-2 h, and 2 released about 50% of the drug within 12 h. The amide prodrugs, 4-6, did not release BPAA in the cecal contents, but gave BPAA/maltose or BPAA/triose conjugates linked through an amide bond. On the other hand, these prodrugs were found to be stable in the contents of rat stomachs and small intestines, in intestinal or liver homogenates, and in rat blood. The serum levels of BPAA increased about 3 h after oral administration of 1 and 3 to rats, accompanying a marked increase in the serum levels, whereas 2 and 4-6 resulted in little increase of the serum levels. These facts suggest that BPAA is released after the ring opening of CyDs followed by the ester hydrolysis, and the BPAA activation takes place site-specifically in the cecum and colon. Therefore, the present CyD prodrug approach provides a versatile means of constructing a novel colon-specific drug delivery system.

Full text of DOI:10.1021/jm970130r

J . Med. Chem. 1997, 40, 2755-2761
2755
6A-O-[(4-Bip h en ylyl)a cetyl]-r-, -â-, a n d -γ-cyclod extr in s a n d
6A-Deoxy-6A-[[(4-bip h en ylyl)a cetyl]a m in o]-r-, -â-, a n d -γ-cyclod extr in s: P oten tia l
P r od r u gs for Colon -Sp ecific Deliver y
Kaneto Uekama,* Kunihiro Minami, and Fumitoshi Hirayama
Faculty of Pharmaceutical Sciences, Kumamoto University, 5-1 Oe-honmachi, Kumamoto 862, J apan
Received March 3, 1997X
Cyclodextrins (CyDs) are known to be fermented to small saccharides by colonic microflora,
whereas they are only slightly hydrolyzable and thus are not easily absorbed in the stomach
and small intestine. This property of CyDs is particularly useful for colon-specific delivery of
drugs. In this study, an antiinflammatory 4-biphenylylacetic acid (BPAA) was selectively
conjugated onto one of the primary hydroxyl groups of R-, â-, and γ-CyDs through an ester or
amide linkage, 6A-O-[(4-biphenylyl)acetyl]-R-, -â-, and -γ-CyDs (1-3) and 6A-deoxy-6A-[[(4-
biphenylyl)acetyl]amino]-R-, -â-, and -γ-CyDs (4-6). In rat cecal and colonic contents (10%,
w/v), 1 and 3 released more than 95% of BPAA within 1-2 h, and 2 released about 50% of the
drug within 12 h. The amide prodrugs, 4-6, did not release BPAA in the cecal contents, but
gave BPAA/maltose or BPAA/triose conjugates linked through an amide bond. On the other
hand, these prodrugs were found to be stable in the contents of rat stomachs and small
intestines, in intestinal or liver homogenates, and in rat blood. The serum levels of BPAA
increased about 3 h after oral administration of 1 and 3 to rats, accompanying a marked increase
in the serum levels, whereas 2 and 4-6 resulted in little increase of the serum levels. These
facts suggest that BPAA is released after the ring opening of CyDs followed by the ester
hydrolysis, and the BPAA activation takes place site-specifically in the cecum and colon.
Therefore, the present CyD prodrug approach provides a versatile means of constructing a
novel colon-specific drug delivery system.
In tr od u ction
thus absorbed as small saccharides in the large intes-
tine.10 Most bacteroid strains isolated from the human
colon are capable of degrading CyDs, as evidenced by
their ability to grow on CyDs using them as the sole
carbon source and by the stimulation of cyclodextranase
activity by exposure to CyDs.11 This biodegradation
property of CyDs may be useful as a colon-targeting
carrier, and thus CyD prodrugs may serve as a source
of site-specific delivery of drugs to colon. In a previous
paper,12 we reported that the ester-type prodrug, rather
than the amide-type prodrug, of the antiinflammatory
drug, 4-biphenylylacetic acid (BPAA), with â-CyD may
serve as a colon-targeting prodrug, on the basis of the
preliminary in vitro hydrolysis experiment. In this
study, we prepared six BPAA prodrugs coupled to R-,
â-, and γ-CyDs through an ester or amide linkage, 6A-
O-[(4-biphenylyl)acetyl]-R-, -â-, and -γ-CyDs (1-3) and
6A-deoxy-6A-[[(4-biphenylyl)acetyl]amino]-R-, -â-, and -γ-
CyDs (4-6), and their in vitro drug release and in vivo
oral absorption behavior in the rat model were inves-
tigated.
The potential use of natural cyclodextrins (CyDs) and
their synthetic derivatives has been extensively studied
in order to improve certain properties of drugs such as
solubility, stability and bioavailability.1 CyD complexes
are in equilibrium with guest and host molecules in
water, the degree of the dissociation being dependent
on the magnitude of the stability constant.2 This
property of CyD complexes is a desirable quality,
because the complex dissociates to free CyD and drug
at the absorption site, and only the drug in free form
enters into systemic circulation. However, such an
equilibrium is disadvantageous when drug targeting is
to be attempted because the complex dissociates before
it reaches the organs or tissues to which it is to be
delivered. One of the method to circumvent the dis-
sociation is to bind a drug covalently to CyDs.
A number of CyD derivatives have been synthesized
for many purposes such as the construction of sophis-
ticated enzyme models and reagents or catalysts for
regio- and stereoselective synthesis or chiral recogni-
tion.3 CyD derivatives coupled to drugs such as en-
kephalin,4 doxorubicin,5 methicillin,6 ibuprofen,7 and
nucleobase8 were recently synthesized for modification
of biological potencies. However, potential applications
of drug/CyD derivatives in medical and pharmaceutical
fields are rather scarce, because CyD derivatizations
usually cause a decrease in biological activity of drugs.5,9
CyDs are known to be barely capable of being hydro-
lyzed and only slightly absorbed in passage through the
stomach and small intestine; however they are fer-
mented into small saccharides by colonic microflora and
Resu lts a n d Discu ssion
Ch em istr y. 1, 2, and 3 were prepared in two steps
according to the method of Coates et al. (Scheme 1).13
One of the primary alcohol of R- and â-CyDs was
tosylated using p-toluenesulfonyl chloride and that of
γ-CyD was sulfonylated using 2-naphthalenesulfonyl
chloride in pyridine. The sulfonyl groups of CyDs were
then replaced by BPAA in DMF. 1 and 3 were purified
by ion-exchange chromatography and then lyophilized.
2 was purified by preparative thin layer chromatogra-
phy, because it was less soluble in water/methanol and
ester hydrolysis occurred during evaporation of a large
amount of DMF used as an eluent of the chromatogra-
X Abstract published in Advance ACS Abstracts, J uly 15, 1997.
S0022-2623(97)00130-1 CCC: $14.00 © 1997 American Chemical Society

2756 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 17
Sch em e 1. Preparation of 1-6a
Uekama et al.
a
Reagents: (a) p-toluenesulfonyl chloride; (b) 2-naphthalenesulfonyl chloride; (c) sodium 4-biphenylylacetic acid; (d) sodium azide; (e)
triphenylphosphine + ammonium hydroxide; (f) 4-biphenylylacetic acid + 1,3-dicyclohexylcabodiimide + hydroxysuccinimide.
F igu r e 1. Time courses of BPAA liberation from 1 (O), 2 (4), 3 (0), and ethyl 4-biphenylylacetate (b) (1.0 × 10-5 M) in rat
intestinal contents and biological fluids in isotonic buffer/DMF (1.0%, v/v) solution after incubation at 37 °C: panel A, isotonic
buffer (pH 4.3 and 7.4); panel B, stomach contents (10% (w/v), pH 4.3); panel C, small intestinal contents (10% (w/v), pH 7.4);
panel D, cecal contents (10% (w/v), pH 6.8); panel E, colonic contents (10% (w/v), pH 6.8); panel F, small intestinal homogenates
without contents (10% (w/v), pH 6.7); panel G, large intestinal homogenates without contents (10% (w/v), pH 7.0); panel H, liver
homogenates (4% (w/v), pH 6.6); panel I, blood (50% (v/v), pH 7.2).
phy. 4, 5, and 6 were prepared in four steps according
to the method of Belanger et al. (Scheme 1).14 The
p-toluenesulfonyl-R- and -â-CyDs and 2-naphthalene-
sulfonyl-γ-CyD were converted to mono(6-azido-6-deoxy)-
CyDs by the reaction with sodium azide in water. These
compounds were reduced by treatments with triphenyl-
phosphine in DMF and then with concentrated am-
monium hydroxide. The monoamino-CyDs were coupled
to the BPAA carboxyl group which was first activated
using 1,3-dicyclohexylcarbodiimide-hydroxysuccinimide

Prodrugs for Colon-Specific Delivery
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 17 2757
in ethyl acetate. The final materials were purified by
ion-exchange chromatography. The chemical purity was
1
checked by H NMR and FAB mass spectra and by the
BPAA liberation after alkaline hydrolysis, confirming
that the substitution degree of BPAA is unity.
Hyd r olysis of P r od r u gs in Ra t Biologica l Med ia .
In aqueous solution, the ester prodrugs, 1-3, were
hydrolyzed to BPAA with first-order rate constants of
9.60 × 10-2 h-1 (t1/2 ) 7.2 h, 3), 8.58 × 10-2 h-1 (t1/2
)
8.1 h, 2), 6.36 × 10-2 h-1 (t1/2 ) 10.9 h, 1), 3.11 × 10-2
h-1 (t1/2 ) 22.3 h, ethyl 4-biphenylylacetate) at pH 9.0
at 37 °C and only small amounts (<10%) within 24 h
under the physiological pH condition (pH 7.4). The
amide prodrugs, 4-6, were stable under these pH and
temperature conditions.
Figure 1 shows the amounts of BPAA liberated from
1-3 and ethyl 4-biphenylylacetate (1.0 × 10-5 M) after
the incubations with contents (10%, w/v) of rat intestinal
tracts such as stomach, small intestine, cecum, and
colon or in rat intestinal homogenates (10%, w/v)
without contents, liver homogeneous (4%, w/v), and
blood (50%, v/v) at 37 °C. In the contents of stomach
and small intestine, the ester prodrugs released BPAA
only in small amounts (<10%) which could be accounted
for by spontaneous hydrolysis in aqueous solution. On
the other hand, in the cecal and colonic contents the
ester prodrugs released BPAA significantly in the order
of 3 > 1 . 2 > ethyl 4-biphenylyl acetate: 100% release
of BPAA from 3 and 1 within 1 and 2 h, respectively,
and more than 50% from 2 within 12 h. In the
homogenates and blood, 1-3 liberated BPAA only in
small amounts (<10%) for 24 h, whereas ethyl 4-bi-
phenylylacetate was rapidly hydrolyzed to BPAA be-
cause of high esterase activities of the homogenates and
blood. The resistance of the prodrugs against the ester
hydrolysis in the blood and homogenates may be due
to a large steric effect of the bulky CyD moiety. From
4-6, no release of BPAA was observed within 24 h in
all intestinal contents and biological media employed.
These results indicate that the BPAA activation took
place site-specifically in the rat cecal and colonic con-
tents.
F igu r e 2. HPLC chromatograms of the hydrolysate after
incubation of 4 (5.7 × 10-3 M) in rat cecal contents (10%, w/v)
in isotonic buffer/DMF (1.0%, v/v) solution at 37 °C. 4 (a),
triose conjugates (b), maltose conjugates (c). Eluent (0.1 M
acetic acid/methanol (54:46, v/v)) on a CAPCEll PAK C18
column. One of R groups in structural formula of the maltose
and triose conjugates is NHCOCH2Ph-Ph, and the others are
OH.
deoxy-6B-[[(4-biphenylyl)acetyl]amino]maltose (7). The
BPAA/linear oligosaccharide conjugates longer than
tetraose could not be detected under the present HPLC
conditions.
Figure 3 shows time profiles for appearance/disap-
pearance of 1-6, BPAA, 7, and BPAA/triose conjugate
(8) in rat cecal contents (10%, w/v). 7 has two positional
isomers described above, and 8 has three isomers of 6A-
deoxy-6A-amino-, 6B-deoxy-6B-amino-, and 6C-deoxy-6C-
[[(4-biphenylyl)acetyl]amino]triose. However, these iso-
mers were determined as a single component, because
the quantitative determination of each isomer was
difficult under the present HPLC conditions. 3, 6, 1,
and 4 disappeared within 1 and 2 h, respectively, in the
cecal contents, and about half of 2 and 5 within 1 h.
The disappearance rate of the ester prodrugs was almost
the same as that of the corresponding amide prodrugs;
the disappearance rate constants in the 10% (w/v) rat
cecal contents were 2.08((0.02) × 10-2 and 1.98((0.01)
× 10-2 min-1 for 1 and 4, 1.50((0.02) × 10-2 and
1.57((0.01) × 10-2 min-1 for 2 and 5, and 4.43((0.03)
× 10-2 and 4.53((0.02) × 10-2 min-1 for 3 and 6,
respectively. 1 and 3 produced quantitatively BPAA,
whereas 4 and 6 formed 7 and 8 without the production
of BPAA. The total amount of 7 and 8 was almost the
same as that of BPAA produced from 1 and 3. If the
ester hydrolysis precedes the ring opening, the disap-
pearance rate of prodrugs can be different between the
ester and amide prodrugs. However, the rate was same
between them, although the products were different.
These facts indicate that 1, 3, 4, and 6 were firstly
subject to the ring opening to give the maltose and triose
conjugates directly or indirectly via BPAA/longer linear
oligosaccharide conjugates which were rapidly hydro-
lyzed to the small saccharide conjugates. After the
fermentation to small saccharide conjugates, they were
rapidly hydrolyzed to BPAA, whereas the amide conju-
To gain insight into the hydrolysis mechanism, the
hydrolysate of the R-CyD amide prodrug, 4, in rat cecal
contents was analyzed in detail, because the amide
prodrugs released no BPAA in rat cecum and colon
contents in spite of the disappearance of the intact
prodrugs with time, as will be discussed later. Figure
2 shows HPLC chromatograms 0, 3, 24, 48, and 72 h
after incubation of 4 (5.7 × 10-3 M) in the cecal contents
(10%, w/v). The peak corresponding to 4 (retention time
) 6.5 min) decreased with time, and two peaks appeared
in the chromatogram during the incubation, which was
detected when a large amount of water (0.1 M acetic
acid/methanol, 54:46 (v/v)) was used as an eluent of
HPLC. The second fraction (retention time ) 7.1 min)
increased and decreased with time and converted finally
to the third fraction (retention time ) 8.4 min). These
fractions were analyzed by mass spectroscopy after the
separation by HPLC and gave molecular ions [M + H]+
of 536 and 698, respectively. (See the Supporting
Information for the mass spectra.) These mass numbers
were identical with those of the BPAA conjugates
coupled to maltose and triose, respectively, indicating
that 4 was hydrolyzed finally to 6A-deoxy-6A- or 6B-

2758 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 17
Uekama et al.
F igu r e 3. Time courses of disappearance of 1-6 (O, 1.0 × 10-5 M) and appearance of BPAA (b) and BPAA conjugates with
maltose (2) and triose (4) in rat cecal contents (10% (w/v), pH 6.7) in isotonic buffer/DMF (1.0%, v/v) solution at 37 °C: panel A,
1; panel B, 4; panel C, 2; panel D, 5; panel E, 3; panel F, 6. Each point represents the mean ( SE of four experiments.
Ta ble 1. Pharmacokinetic Parameters of BPAA after Oral Administrations of 1, 2, 3, BPAA, and â-CyD Complex (Equivalent to 10
mg/kg BPAA) to Rats
MRTc (h)
AUCd (h‚µg/mL)
Fe (%)
a
b
system
tmax (h)
Cmax (µg/mL)
1
2
3
9.0 ( 0.0* f
11.8 ( 6.8
7.8 ( 0.7*
0.5 ( 0.3
15.3 ( 2.5*
1.3 ( 0.4*
23.6 ( 3.2*
6.9 ( 1.2
9.3 ( 0.5*
11.4 ( 1.5*
8.8 ( 0.7*
6.6 ( 0.7
119.6 ( 12.0*
10.9 ( 3.3*
166.1 ( 22.9*
33.6 ( 2.5
64.3 ( 6.4
5.9 ( 1.8
89.3 ( 12.3
18.1 ( 1.3
34.0 ( 0.9
BPAA alone
â-CyD complex
1.5 ( 0.7*
10.5 ( 0.9
5.9 ( 0.2
63.3 ( 1.7*
a
b
d
The time required to reach Cmax
.
The maximum levels in serum. c The mean residence time. The area under the serum BPAA
concentration-time curves, determined up to 24 h postadministration. e Bioavailability compared with the AUC value of BPAA administered
intravenously. f An asterisk (*) indicates p < 0.05 versus BPAA alone.
gates resisted to the hydrolysis and gave 7 as a final
product. In rat small intestinal contents (10%, w/v),
homogenates (10%, w/v), and blood (50%, v/v), neither
appreciable disappearance of the prodrugs nor appear-
ances of BPAA, 7, and 8 were observed. In the case of
2 and 5, the ring opening was slower than those of 1, 3,
4, and 6. This order of the ring-opening rates was
different from those expected from acid-catalyzed15 and
most CyD-hydrolyzing enzymatic ring openings,16 i.e.,
the larger the cavity size, the faster the rate. Several
CyD-hydrolyzing enzymes may take part in the ring
opening of the prodrugs by colonic microflora, particu-
larly at the low concentration of the substrates (1.0 ×
10-5 M), because it is reported that enzyme preparations
from a number of sources show different rates of attack
on different CyDs and each CyD-hydrolyzing enzyme
has its own preferred CyD substrate.17 Further, the
presence of various biological components such as
cholesterol that fit the â-CyD cavity may suppress the
ring opening of 2 and 5, because the inclusion of guests
slows down the rate.18 2 released BPAA in only small
amounts, in spite of the significant disappearance of the
prodrugs. The reason for this inconsistency of mass
balance is unclear at present. The ring-opening mech-
anism of 2 and 5 may be different from that of the R-
and γ-CyD prodrugs, which should be further studied.
In Vivo Absor p tion Beh a vior . Figure 4 shows the
plasma levels of BPAA after oral administration of the
F igu r e 4. Serum levels of BPAA after oral administration of
1 (O), 2 (4), 3 (0), BPAA (b), or BPAA-â-CyD complex ([)
(equivalent to 10 mg/kg BPAA) to rats. Each point represents
the mean ( SE of three experiments.
1, 2, 3, BPAA, alone, or BPAA/â-CyD complex to rats,
and Table 1 summarizes their pharmacokinetic param-
eters. The BPAA and BPAA/â-CyD complex (molar
ratio of 1:1) systems showed a rapid increase and
decrease in the serum levels of BPAA, where the
complex showed higher serum levels than the BPAA
alone. On the other hand, the serum BPAA levels of 1
and 3 increased after a lag time of about 3 h and reached
maximum levels after about 9 and 8 h, respectively,
accompanying a significant increase in the serum levels.

Prodrugs for Colon-Specific Delivery
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 17 2759
mmol) was added to the monosulfonylated R- (9.86 g, 8.75
mmol), â- (12.0 g, 9.31 mmol), or γ-CyDs (12.2 g, 8.20 mmol)
in DMF (100 mL), and the mixture was stirred at 100 °C for
30 h. The reaction solution was concentrated under reduced
pressure, a large amount of acetone (1000 mL) was added, and
the precipitate was collected under vacuum. The solid was
dissolved in a small amount (about 10 mL) of DMF and the
solution added to a large amount (about 100 mL) of acetone.
This procedure was repeated several times. 1 and 3 were
purified by ion-exchange column chromatography (DIAION
HP-20, 5 × 20 cm) eluting with methanol/water with increas-
ing methanol content. The eluates were monitored by TLC,
and the prodrugs appeared in the eluates of 20-40% methanol
in water. Methanol in the eluate was removed under reduced
pressure, and the solution was lyophilized to yield 3.5 g (36%)
and 3.1 g (25%) of 1 and 3, respectively. 2 was purified by
preparative thin-layer silica gel chromatography (silica gel
60F, 20 × 20 cm, Merck) with an eluent of acetonitrile/water,
3:7 (v/v). The fraction of TLC containing 2 was collected and
suspended in water, and after filtration the solution was
lyophilized to yield 7.4 g (62%) of 2. 1: mp 255 °C dec; Rf )
0.67; 1H NMR (DMSO-d6) δ 7.17-7.49 (m, 9H, J ) 8 Hz,
biphenyl), 5.26-5.40 (m, 12H, CyD 2,3-OH), 4.59-4.66 (d, 6H,
CyD 1-H), 4.13-4.38 (m, 5H, CyD 6-OH), 3.45-3.74 (m, 26H,
CyD 3,5,6-H, BPAA CH2), 3.12-3.25 (m, overlaps with HOD,
CyD 2,4-H); 13C NMR (DMSO-d6) δ 171.92 (BPAA CdO),
140.93, 139.72, 134.54, 130.96, 129.84, 128.29, 127.59 (BPAA
biphenyl), 40.5 (overlaps with solvent peak, BPAA CH2), 102.97
(CyD C1), 83.32, 83.14, 82.95 (CyD C4), 74.18 (CyD C3), 74.02
(CyD C2), 73.06 (CyD C5), 69.87, 64.49, 60.87 (CyD C6); FAB
mass [M - H]- m/z 1165 (nominal molecular weight 1166). 2:
The lag time phenomenon indicates that BPAA was
released and absorbed in the cecum and colon after the
administration of the prodrugs, since the time required
to reach the colon after oral administration is reported
to be about 1.5-3 h in rats.19 The area under serum
BPAA concentration-time profiles of 1 and 3 were 3.6
and 4.9 times larger than that of the BPAA alone. 2
gave little increase in serum levels, because of the
slower drug liberation. 4, 5, and 6 gave no appearance
of BPAA in the serum, reflecting the in vitro hydrolysis
behavior. 7, 8, and intact ester and amide prodrugs
were not detected in the serum under the HPLC
conditions, suggesting the sugar-substituted BPAA
conjugates are poorly absorbed from intestinal tracts
probably due to high hydrophilicity.
Con clu sion
Colonic drug delivery can be achieved with carriers
by making prodrugs that survive passage through
stomach and small intestine, but active moiety is
released by enzymes specifically produced in colon. A
well-known demonstration of this concept is the delivery
of 5-aminosalicylic acid by the use of azo-linked pro-
drugs.20 Further, Friend et al. demonstrated the colon-
targeting of steroids by the use of glycoside prodrugs.21
The present results indicated that the ester type pro-
drugs of BPAA/CyD, particularly R- and γ-CyDs, re-
leased BPAA preferentially in cecal and colonic contents
of rats after the fermentation of CyDs to small oligo-
saccharides, suggesting that CyDs can serve as a new
class of site-specific drug carriers. A number of CyD
derivatives, where the hydroxyl groups were selectively
mono- or multisubstituted by various functional groups
such as amine, carboxylic acid, aldehyde, and thiol, have
been synthesized mainly for the construction of enzyme
models. These derivatives may serve as promoieties for
colon-specific targeting prodrugs. Thus, the CyD con-
jugation approach may provide a versatile means for
construction of colon-specific delivery systems for drugs
such as steroids, 5-aminosalicylic acid, and anti-cancer
agents, some of which are now under investigation.
1
mp 258-268 °C dec; Rf ) 0.63; H NMR (DMSO-d6) δ 7.34-
7.66 (m, 9H, J ) 8 Hz, biphenyl), 5.66-5.78 (m, 14H, CyD
2,3-OH), 4.80-4.85 (d, 7H, CyD 1-H), 4.20-4.52 (m, 6H, CyD
6-OH), 3.55-3.87 (m, 30H, CyD 3,5,6-H, BPAA CH2), 3.30-
3.39 (m, overlaps with HOD, CyD 2,4-H); 13C NMR (DMSO-
d6) δ 171.03 (BPAA CdO), 139.94, 138.72, 133.57, 130.00,
128.87, 127.29, 126.61 (BPAA biphenyl), 39.5 (overlaps with
solvent peak, BPAA CH2), 102.37, 101.95 (CyD C1), 82.08,
81.92, 81.60, 81.52, 81.15 (CyD C4), 73.04 (CyD C3), 72.39
(CyD C2), 72.00 (CyD C5), 68.87, 63.54, 59.89 (CyD C6); FAB
mass [M - H]- m/z 1327 (nominal molecular weight 1328). 3:
1
mp 277-282 °C dec; Rf ) 0.53; H NMR (DMSO-d6) δ 7.37-
7.66 (m, 9H, J ) 8 Hz, biphenyl), 5.76-5.86 (m, 16H, CyD
2,3-OH), 4.84-4.90 (d, 8H, CyD 1-H), 4.19-4.58 (m, 7H, CyD
6-OH), 3.53-3.81 (m, 34H, CyD 3,5,6-H, BPAA CH2), 3.31-
3.44 (m, overlaps with HOD, CyD 2,4-H); 13C NMR (DMSO-
d6) δ 170.96 (BPAA CdO), 139.86, 138.63, 133.51, 129.95,
128.82, 127.23, 126.49 (BPAA biphenyl), 39.4 (overlaps with
solvent peak, BPAA CH2), 102.33, 101.56 (CyD C1), 81.74,
81.37, 81.00, 80.81, 80.72, 80.25 (CyD C4), 72.77 (CyD C3),
72.61, 72.50 (CyD C2), 72.04 (CyD C5), 68.98, 63.51, 59.91
(CyD C6); FAB mass [M - H]- m/z 1489 (nominal molecular
weight 1491).
Exp er im en ta l Section
CyDs were donated by J apan Maize products Co. Ltd.
(Tokyo, J apan), and BPAA and ethyl biphenylylacetate were
supplied from Ledrele J apan Co. Ltd. (Saitama, J apan). The
inclusion complex of BPAA with â-CyD in a molar ratio of 1:1
was prepared by the kneading method, and the formation of
the solid complex was confirmed by powder X-ray diffractom-
etry and differential scanning calorimetry, as reported previ-
ously.22 1H and 13C NMR spectra were taken on a J EOL GX-
400 or A-500 spectrometer (Tokyo, J apan) at 25 °C, using
DMSO-d6 as a solvent. Fast atom bombardment (FAB) mass
spectra were recorded by a J EOL J MS-DX 303 mass spec-
trometer (Tokyo, J apan); 1-6 in a negative mode using a
matrix of diethanolamine/DMSO and 7 and 8 in a positive
mode using glycerol/H2O. Analytical HPLC was performed on
a Hitachi 655A-11 pump with a L-4000 variable wavelength
UV detector. Thin layer chromatography (TLC) was conducted
using normal-phase plates precoated with silica gel 60 F254
obtained from Merck (Darmstadt, Germany) and an eluent of
ethyl acetate/2-propanol/ammonium hydroxide/water ) 7:7:
5:4 (indicator: p-anisaldehyde).
6A-Deoxy-6A-[[(4-bip h en ylyl)a cetyl]a m in o]-r-, -â-, a n d
-γ-CyDs (4-6). Mono(6-deoxy-6-amino)-R-, -â-, and -γ-CyDs
were obtained in three steps from the parent CyDs and
purified by gel chromatography (CM-Sephadex C-25 (NH4
form), 2 × 25 cm, Pharmacia Biotech, Sweden) with aqueous
NH4HCO3 solution.14,24 To BPAA (0.42 g, 2 mmol)/anhydrous
ethyl acetate (20 mL) were added hydroxysuccinimide (0.23
g, 2 mmol) and 1,3-dicyclohexylcarbodiimide (0.41 g, 2 mmol),
and the mixture was stirred at room temperature (about 25
°C) for 4 h. After filtration, the solution was dried under
reduced pressure. The monoamino-R-, -â-, or -γ-CyD (1.9 g,
2.2 g, 2.6 g, 2 mmol) was added to the residue (activated BPAA
ester)/DMF (5 mL), and the mixture was stirred at room
temperature for 5 h. The reaction solution was concentrated
under reduced pressure, and acetone (50 mL) was added. The
precipitate was collected by filtration, washed with acetone,
and applied on the DIAION column (eluent: methanol/water,
about 1:4 (v/v)). Methanol in the eluates was removed under
reduced pressure, and the solution was lyophilized to yield 1.8
g (89%), 2.3 g (100%), and 1.3 g (67%) of 4, 5, and 6,
respectively. 4: mp 249-256 °C dec; Rf ) 0.66; 1H NMR
(DMSO-d6) δ 7.92 (br s, 1H, BPAA, NH), 7.33-7.64 (m, 9H, J
6A-O-[(4-bip h en ylyl)a cetyl]-r-, -â-, a n d -γ-cyclod extr in s
(1-3). Monotosyl R-CyD and mononaphthalenesulfonyl-γ-
CyD were purified by ion-exchange chromatography (DIAION
HP-20, 5 × 20 cm, Mitsubishi Chemical Co., Tokyo, J apan)
with methanol/water.13,23 Monotosyl-â-CyD was recrystallized
three times from water. Sodium 4-biphenylylacetate (2 g, 8.54

2760 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 17
Uekama et al.
) 8 Hz, biphenyl), 5.43-5.57 (m, 12H, CyD 2,3-OH), 4.80-
4.86 (d, 6H, CyD 1-H), 4.44-4.54 (m, 5H, CyD 6-OH), 3.59-
3.77 (m, 26H, CyD 3,5,6-H, BPAA CH2), 3.25-3.48 (m, overlaps
with HOD, CyD 2,4-H); 13C NMR (DMSO-d6) δ 170.25 (BPAA
CdO), 140.04, 138.19, 135.71, 129.58, 128.82, 127.17, 126.50,
126.38 (BPAA biphenyl), 39.8 (overlaps with solvent peak,
BPAA CH2), 101.89, 101.68 (CyD C1), 83.61, 82.25, 82.03, 81.88
(CyD C4), 73.21 (CyD C3), 73.01 (CyD C2), 72.20, 72.05 (CyD
C5), 69.89, 60.06, 59.94, 59.84 (CyD C6); FAB mass [M - H]-
m/z 1164 (nominal molecular weight 1165). 5: mp 258-263
°C dec; Rf ) 0.60; 1H NMR (DMSO-d6) δ 7.89 (br s, 1H, BPAA
NH), 7.34-7.65 (m, 9H, J ) 8 Hz, biphenyl), 5.69-5.82 (m,
14H, CyD 2,3-OH), 4.83 (d, 7H, CyD 1-H), 4.45-4.57 (m, 6H,
CyD 6-OH), 3.33-3.64 (m, overlaps with HOD, CyD 2,3,4,5,6-
H, BPAA CH2); 13C NMR (DMSO-d6) δ 170.31 (BPAA CdO),
140.08, 138.22, 135.71, 129.66, 128.90, 127.24, 126.57, 126.45
(BPAA biphenyl), 39.4 (overlaps with solvent peak, BPAA
CH2), 102.13, 101.98, 101.86 (CyD C1), 83.55, 81.73, 81.67,
81.46, 81.33 (CyD C4), 73.14, 72.99, 72.86 (CyD C3), 72.41
(CyD C2), 72.20, 72.01 (CyD C5), 69.80, 59.93, 59.84 (CyD C6);
FAB mass [M - H]- m/z 1325 nominal molecular weight 1327).
6: mp 279-280 °C dec; Rf ) 0.48; 1H NMR (DMSO-d6) δ 7.87
(br s, 1H, BPAA NH), 7.33-7.64 (m, 9H, J ) 8 Hz, biphenyl),
5.70-5.85 (m, 14H, CyD 2,3-OH), 4.87-4.91 (d, 7H, CyD 1-H),
4.46-4.67 (m, 7H, CyD 6-OH), 3.45-3.62 (m, 34H, CyD 3,5,6-
H, BPAA CH2), 3.29-3.38 (m, overlaps with HOD, CyD 2,4-
H); 13C NMR (DMSO-d6) δ 170.27 (BPAA CdO), 140.04,
138.22, 135.68, 129.66, 128.87, 127.20, 126.53, 126.41 (BPAA
biphenyl), 39.8 (overlaps with solvent peak, BPAA CH2),
102.37, 101.58 (CyD C1), 83.44, 81.02 (CyD C4), 72.71 (CyD
C3), 72.59 (CyD C2), 72.11 (CyD C5), 70.01, 60.00, (CyD C6);
FAB mass [M - H]- m/z 1488 (nominal molecular weight
1490).
DMF/pH 7.4 isotonic phosphate buffer) were added to the
filtrates (5.0 mL) and incubated at 37 °C. Rat liver (wet weight
about 20 g) was washed thoroughly with saline (about 500 mL)
and homogenized with 5 volumes of cold 1.15 w/v KCl using a
tissue homogenizer (Potter-Elvejem, Corning Glassw) at 0 °C.
The homogeates were filtered through gauze, and the filtrate
was centrifuged at 9000g for 30 min at 4 °C. The supernatant
(2.0 mL) was added to the prodrug solutions (8.0 mL, 1.25 ×
10-5 M in 1.25 v/v DMF/pH 7.4 isotonic phosphate buffer)
containing MgCl2, 101.6 mg/mL, glucose 6-phosphate, 60.8 mg/
mL, nicotinamide, 9.16 mg/mL, and nicotinamide-adenine
dinucleotide phosphate buffer, 3.35 mg/mL at 37 °C. Rat blood
was collected using injection syringes treated with sodium
citrate. The prodrug solution (5.0 mL, 2.0 × 10-5 M in 2.0 v/v
DMF/pH 7.4 isotonic phosphate buffer) was added to 5.0 mL
of blood at 37 °C. At appropriate intervals, the reaction
solutions were analyzed for BPAA by HPLC as described
above.
In Vivo Absor p tion Stu d ies. Male Wistar rats weighing
about 200 g were fasted for 18 h prior to drug administration,
while water was allowed ad libitum. BPAA, BPAA/â-CyD
complex, and 1-6 (equivalent to 10 mg of BPAA/5 mL of water
per kg of rat body weight) were orally administered. Blood
samples (about 0.6 mL) were taken periodically from the
jugular vein, and centrifuged at 10 000 rpm for 5 min. The
serum (0.1 mL) was assayed for BPAA, intact prodrugs, and
the sugar conjugates by HPLC under the same condition as
those described above. The bioavailability parameters were
determined using a MULTI program.25
Ack n ow led gm en t. The authors are grateful to Dr.
T. Kondo and Miss Y. Tanaka for their helpful discus-
sion and technical assistance.
Hyd r olysis in Ra t Ga str oin testin a l Con ten ts. Male
Wistar rats, 400-500 g, were fed a standard diet (CE-2, CLEA
J apan Inc., Tokyo). The rats were killed by decapitation, and
stomach, small intestine, cecum, and colon were removed.
Stomach and other contents were diluted to 20% w/v with
isotonic acetate buffer (pH 4.4) and phosphate buffer (pH 7.4),
respectively, and the dispersions of contents were filtered
through a gauze to remove large particles. The prodrug
solution (5.0 mL, 2.0 × 10-5 M in 2.0% v/v DMF/isotonic buffer)
was added to the filtrate (5.0 mL) in air-tight vessels and
incubated at 37 °C. The pH of incubation solutions was
adjusted to 4.4 (stomach) or 7.4 (other contents) by the addition
of small amounts of 0.1 M NaOH. No attempt was made to
replace the atmosphere with nitrogen, because of the lack of
large difference in the hydrolysis rate of the â-CyD prodrugs
between aerobic and anaerobic conditions as reported previ-
ously.12 At appropriate intervals, the reaction solutions were
determined for BPAA, intact prodrugs, and 7 and 8 by HPLC.
BPAA: an aliquot (0.5 mL) of the reaction solution was added
to 0.1 M HCl (0.2 mL), and BPAA was extracted with
cyclohexane/diethyl ether (3:1, 6.0 mL) containing flurbiprofen
as an internal standard (1 µg/mL, 0.5 mL). The organic phase
(5.0 mL) was evaporated under reduced pressure and redis-
solved in methanol (0.1 mL), 0.02 mL of which was subjected
to HPLC analysis under the following conditions: a CAPCEll
PAK C18 column (6 × 150 mm, Shiseido Co., Tokyo, J apan),
a mobile phase of 0.1 M acetic acid/methanol (7:13), a flow rate
of 1.5 mL/min, and a detection of 255 nm. Prodrugs (1-6)
and 7 and 8; an aliquot (0.2 mL) of the reaction solution was
ultrafiltered using a membrane filter (Amicon Kit (Tokyo,
J apan), a centrifugation of 3500 rpm for 15 min at 4 °C), and
the filtrate (0.02 mL) was subjected to HPLC analysis using
the same column and the following mobile phases (0.1 M acetic
acid/methanol): 54:46 v/v for 1 and 4 (retention time, 10.8 and
6.9 min, respectively), 65:35 for 2 and 5 (7.1 and 7.8 min), 62:
38 for 3 and 6 (9.3 and 5.5 min), and 54:46 or 62:38 for 7 and
8 (8.5 and 7 min or 20 and 16 min).
Su p p or tin g In for m a tion Ava ila ble: 1H and 13C NMR
and FAB mass spectra of 1-6, FAB mass spectra of the
maltose and triose conjugates with BPAA, and the first-order
plots for the hydrolysis of 1-3 at pH 9.0, 37 °C (9 pages).
Ordering information is given on any current masthead page.
Refer en ces
(1) (a) Szejtli, J . Cyclodextrin technology; Kluwer: Dordrecht, 1988.
(b) Loftsson, T.; Brewster, M. E.; Derendorf, H.; Bodor, N.
2-Hydroxypropyl â-cyclodextrin: properties and usage in phar-
maceutical formulations. Pharm. Ztg. Wiss. 1991, 4, 5-10. (c)
Ducheˆne, D., Ed. New trends in cyclodextrins and derivatives.
Editions de Sante´: Paris, 1991. (d) Uekama, K.; Hirayama, F.
Improvements of drug properties by cyclodextrins. In The
Practice of Medicinal Chemistry; Wermuth, C. G., Ed.; Academic
Press: London, 1996; pp 793-825.
(2) (a) Bender, M. L.; Komiyama, M. Cyclodextrin chemistry;
Springer-Verlag: Berlin, 1978. (b) Saenger, W. Cyclodextrin
inclusion compounds in research and industry. Angew. Chem.,
Int. Ed. Engl. 1980, 19, 344-362.
(3) (a) Ko¨nig, W. A.; Lutz, S.; Wenz, G. Modified cyclodextrins-novel,
highly enantioselective stationary phases for gas chromatogra-
phy. Angew. Chem., Int. Ed. Engl. 1988, 27, 979-980. (b) Ueno,
A.; Kuwabara, T.; Nakamura, A.; Toda, F. A modified cyclodex-
trin as a guest responsive colour-change indicator. Nature 1992,
356, 136-137. (c) Breslow, R. Biomimetic chemistry and
artificial enzymes: catalysis by design. Acc. Chem. Res. 1995,
28, 146-153.
(4) Parrot-Lopez, H.; Djedaini, F.; Perly, B.; Coleman, A. W.; Galons,
H.; Miocque, M. An approach to vectorisation of pharmacologi-
cally active molecules: the covalent binding of Leu-enkephalin
to a modified â-cyclodextrins. Tetrahedron Lett. 1990, 31, 1999-
2002.
(5) Tanaka, H.; Kominato, K.; Yamamoto, R.; Yoshioka, T.; Nishida,
H.; Tone, H.; Okamoto, R. Synthesis of doxorubicin-cyclodextrin
conjugates. J . Antibiot. 1994, 47, 1025-1029.
(6) Nedra Karunaratne, D.; Farmer, S.; Hancock, R. E. W. Synthesis
of bulky â-lactams for inhibition of cell surface â-lactamase
activity. Bioconjugate Chem. 1993, 4, 434-439.
(7) Coates, J . H.; Easton, C. J .; van Eyk, S. J .; May, B. L.; Singh,
P.; Lincoln, S. F. Chiral differentiation in the deacylation of 6A-
O-{2-[4-(2-methylpropyl)phenyl]propanoyl}-â-cyclodextrin. J .
Chem. Soc., Chem. Commun. 1991, 759-760.
(8) Djedaini-Pilard, F.; Perly, B.; Dupas, S.; Miocque, M.; Galons,
H. Specific interaction and stabilization between host and
guest: complexation of ellipticine in a nucleobase functionalized
cyclodextrin. Tetrahedron Lett. 1993, 34, 1145-1148.
Hyd r olysis in Ra t Biologica l F lu id s. Rat intestine
segments were washed fully with isotonic buffer, cut into small
pieces, and homogenized with 5 volume of cold 1.15 w/v KCl
using a tissue homogenizer (Physcotron NS-50, Nichion,
Tokyo) at 0 °C. The homogenates were filtered through a
gauze. The solutions of 1-6 (5.0 mL, 2.0 × 10-5 M in 2.0 v/v

Prodrugs for Colon-Specific Delivery
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 17 2761
(9) Habus, I.; Zhao, Q.; Agrawal, S. Synthesis, hybridization proper-
ties nuclease stability, and cellular uptake of the oligonucleotide-
amino-â-cyclodextrins and adamantane conjugates. Bioconju-
gate Chem. 1995, 6, 327-331.
(10) (a) Andersen, G. H.; Robbins, F. M.; Domingues, F. J .; Moores,
R. G.; Long, C. L. The utilization of schardinger dextrins by the
rat. Toxicol. Appl. Pharmacol. 1983, 5, 257-266. (b) Gerlo´czy,
A.; Fo´nagy, A.; Keresztes, P.; Perlaky, L.; Szejtli, J . Absorption,
distribution, excretion and metabolism of orally administered
14C-â-cyclodextrin in rat. Arzneim. Forsch. Drug Res. 1985, 35,
1042-4017. (c) Flourie´, B.; Molis, C.; Achour, L.; Dupas, H.;
Hatat, C.; Rambaud, J . C. Fate of â-cyclodextrin in the human
intestine. J . Nutr. 1993, 123, 676-680.
(11) Antenucci, R.; Palmer, J . K. Enzymatic degradation of R- and
â-cyclodextrins by bacteroides of the human colon. J . Agric.
Food Chem. 1984, 32, 1316-1321.
(12) Hirayama, F.; Minami, K.; Uekama, K. In-vitro evaluation of
biphenylyl acetic acid-â-cyclodextrin conjugates as colon-target-
ing prodrugs: drug release behaviour in rat biological media.
J . Pharm. Pharmacol. 1996, 48, 27-31.
(13) Coates, J . H.; Easton, C. L.; Fryer, N. L.; Lincoln, S. F.
Complementary diastereoselectivity in the synthesis and hy-
drolysis of acylated cyclodextrins. Chem. Lett. 1994, 1153-1156.
(14) Bellanger, N.; Perly, B. NMR investigations of the conformation
of a new cyclodextrin-based amphiphilic transporter for hydro-
phobic drugs: molecular lollipops. J . Mol. Struct. 1992, 273,
215-226.
(15) Uekama, K.; Otagiri, M. Cyclodextrins in drug carrier system.
CRC Crit. Rev. Ther. Drug Carrier Syst. 1978, 3, 1-40.
(16) Suetsugu, N.; Koyama, S.; Takeo, K.; Kuge, T. Kinetic studies
on the hydrolyses of R-, â-, and γ-cyclodextrins by taka-amylase
A. J . Biochem. 1974, 76, 57-63.
(17) Bender, H. A bacterial glucoamylase degrading cyclodextrins.
Eur. J . Biochem. 1981, 115, 278-291.
(18) Hirayama, F.; Kurihara, M.; Utsuki, T.; Uekama, K. Inhibitory
effect of guest molecules on acid-catalyzed ring-opening of
â-cyclodextrin. J . Chem. Soc., Chem. Commun. 1993, 1578-
1590.
(19) Saffran, M.; Kumar, G. S.; Savariar, C.; Burnham, J . C.;
Williams, F.; Neckers, D. C. A new approach to the oral
administration of insulin and other peptide drugs. Science 1986,
233, 1081-1084.
(20) Kopecˇek, J .; Kopecˇkova´, P. N-(2-Hydroxypropyl)methacrylamide
copolymers for colon-specific drug delivery. In Oral colon-specific
drug delivery; Friend, D. R., Ed.; CRC Press: Boca Raton, 1992;
pp 189-211.
(21) Friend, D. R. Glycosides in colonic drug delivery. In Oral colon-
specific drug delivery; Friend, D. R., Ed.; CRC Press: Boca
Raton, 1992; pp 153-187.
(22) Arima, H.; Adachi, H.; Irie, T.; Uekama, K.; Pitha, J . Enhance-
ment of the antiinflammatory effect of ethyl 4-biphenylyl acetate
in ointment by â-cyclodextrin derivatives: increased absorption
and localized activation of the prodrug in rats. Pharm. Res.
1990, 7, 1152-1156.
(23) Wang, Y.; Ikeda, T.; Ikeda, H.; Ueno, A.; Toda, F. Dansyl-â-
cyclodextrins as fluorescent sensors responsive to organic com-
pounds. Bull. Chem. Soc. J pn. 1994, 67, 1598-1607.
(24) Croft, A. P.; Bartsch, R. A. Synthesis of chemically modified
cyclodextrins. Tetrahedron 1983, 39, 1417-1474.
(25) Yamaoka, K.; Tanigawara, Y.; Nakagawa, T.; Uno, T. A phar-
macokinetic analysis program (MULTI) for microcomputer. J .
Pharmacobio-Dyn. 1981, 4, 879-885.
J M970130R