Pyrimidine and nucleoside γ-esters of L-Glu-Sar: Synthesis, stability and interaction with hPEPT1

Article abstract of DOI:10.1016/j.ejps.2005.02.007

The aim of the present study was to improve the synthetic pathway of bioreversible dipeptide derivatives as well as evaluate the potential of using l-Glu-Sar as a pro-moiety for delivering three newly synthesised nucleoside and pyrimidine l-Glu-Sar deriva

Full text of DOI:10.1016/j.ejps.2005.02.007

European Journal of Pharmaceutical Sciences 25 (2005) 145–154
Pyrimidine and nucleoside ␥-esters of l-Glu-Sar: Synthesis,
stability and interaction with hPEPT1
a
b
b
a
´
Andre H. Eriksson , Peter L. Elm , Mikael Begtrup , Birger Brodin ,
Robert Nielsenc, Bente Steffansena,∗
a
Department of Pharmaceutics, The Danish University of Pharmaceutical Sciences, Universitetsparken 2, DK-2100 Copenhagen, Denmark
b
Department of Medicinal Chemistry, The Danish University of Pharmaceutical Sciences, Universitetsparken 2, DK-2100 Copenhagen, Denmark
c
Department of Biochemistry, August Krogh Institute, Universitetsparken 13, DK-2100 Copenhagen, Denmark
Received 21 July 2004; received in revised form 1 February 2005; accepted 10 February 2005
Available online 23 March 2005
Abstract
The aim of the present study was to improve the synthetic pathway of bioreversible dipeptide derivatives as well as evaluate the potential
of using l-Glu-Sar as a pro-moiety for delivering three newly synthesised nucleoside and pyrimidine l-Glu-Sar derivatives. l-Glu(trans-2-
thymine-1-yl-tetrahydrofuran-3-yl ester)-Sar (I), l-Glu(thymine-1-yl-methyl ester)-Sar (II) and l-Glu(acyclothymidine)-Sar (III) were syn-
thesised and in vitro stability was studied in various aqueous and biological media. Affinity to and translocation via hPEPT1 was investigated
in mature Caco-2 cell monolayers, grown on permeable supports. Affinity was estimated in a competition assay, using [14C] labelled Gly-
Sar (glycylsarcosine). Translocation was measured as pHi-changes induced by the substrates using the fluorescent probe BCECF and an
epifluorescence microscope setup. All dipeptide derivatives released the model drugs quantitatively by specific base-catalysed hydrolysis at
pH > 6.0. II was labile in aqueous buffer solution, whereas I and III showed appropriate stability for oral administration. In 10% porcine
intestinal homogenate, the half-lives of the dipeptide derivatives indicated limited enzyme catalyzed degradation. All compounds showed
good affinity to hPEPT1, but the Compounds I and III showed not to be translocated by hPEPT1. The translocation of the l-Glu-Sar derivative
of acyclovir, l-Glu(acyclovir)-Sar was also investigated and showed not to take place. Consequently, l-Glu-Sar seems to be a poor pro-moiety
for hPEPT1-mediated transport.
© 2005 Elsevier B.V. All rights reserved.
Keywords: Bioreversible derivatives; Dipeptide; Aqueous stability; hPEPT1; Caco-2; Intracellular pH
1. Introduction
low oral bioavailability (Morse et al., 1993), which may be as-
cribed to their relatively hydrophilic character, and hence lim-
ited passive permeability across intestinal mucosa. However,
targeting the peptide transporter PEPT1 and intracellular es-
terases has proven to efficiently increase the oral bioavail-
ability of aciclovir and ganciclovir when administered as
the bioreversible valine ester pro-drugs (Beauchamp et al.,
1992; Burnette and de Miranda, 1994; Crooks and Murray,
1994; Ganapathy et al., 1998; Han et al., 1998; Sugawara et
al., 2000).
Several chemotherapeutic agents share structural similar-
ities to nucleosides or nucleobases and act by interrupting
either the various steps of replication or various steps of
the biosynthesis of nucleosides. Examples are the antiviral
agents AZT, ganciclovir and acyclovir (Brigden et al., 1981;
Matthews and Boehme, 1988; Mitsuya and Broder, 1987).
Another example is 5-fluorouracil which is used in the treat-
ment of cancer (Heidelberger et al., 1957; Longley et al.,
2003). These drugs are often characterised by variable and
Another class of bioreversible model pro-drugs targeting
hPEPT1 is based on metabolically stable dipeptidomimetic
pro-moieties, such as l-Glu-Sar and d-Glu-Ala (Friedrichsen
et al., 2001; Lepist et al., 2000; Nielsen et al., 2001c;
∗
Corresponding author. Tel.: +45 35 30 62 21; fax: +45 35 30 60 30.
E-mail address: bds@dfuni.dk (B. Steffansen).
0928-0987/$ – see front matter © 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.ejps.2005.02.007

146
A.H. Eriksson et al. / European Journal of Pharmaceutical Sciences 25 (2005) 145–154
([14C]Gly-Sar) with a specific activity of 49.94 mCi/mmol
was purchased from New England Nuclear (Boston, MA,
USA).
Steffansen et al., 1999; Taub et al., 1997, 1998). These com-
pounds showed promising drug delivery properties in vitro
in terms of aqueous and enzymatic stability and affinity to
hPEPT1. The parent drug release from ␥-esters of l-Glu-
Sar and d-Glu-Ala, is thus generally controlled by a specific
base-catalysed hydrolysis in vitro, consequently suggesting
that the pro-drugs will remain relatively stable during pas-
sage of the upper small intestine at pH 6.0, yet release parent
drug within intracellular or blood environments at pH 7.4
(Thomsen et al., 2003). On the other hand, it has recently
been shown that acyclovir when linked to l-Glu-Sar has lim-
ited bioavailability after oral administration to rats (Thomsen
et al., 2004). The general application of the pro-moiety l-Glu-
Sar in targeting of nucleoside analogues to PEPT1; however,
still remains to be investigated.
2-[N-Morpholino]ethanesulfonic acid (MES), 2[4-(2-
hydroxyethyl)-1-piperazine]-ethanesulfonic acid (Hepes)
and glycylsarcosine were purchased from Sigma–Aldrich
(St. Louis, MO, USA). 2ꢀ,7ꢀ-bis-(2-carboxyethyl)-5-(and-
6)-carboxyfluorescein, acetoxymethyl ester (BCECF/AM),
2ꢀ,7ꢀ-bis-(2-carboxyethyl)-5-(and-6)-carboxyfluorescein
(BCECF acid) and Pluronic F-127 were purchased from
Molecular Probes (Eugene, OR, USA). Cremophor EL
was purchased from BASF AG (Ludwigshafen, Germany).
Lys[Z(NO2)]-Pro was kindly donated by Professor Klaus
Neubert, Department of Biochemistry, Marthin-Luther-
¨
Universitat, Halle-Wittenberg, Germany.
The aim of the present study was to improve the synthetic
pathway of bioreversible dipeptide derivatives as well as eval-
uate the potential of using l-Glu-Sar as a pro-moiety for de-
livering three newly synthesised nucleoside and pyrimidine
l-Glu-Sar derivatives.
l-Glu(acyclovir)-Sar was synthesised as described by
Steffansen and co-workers (Thomsen et al., 2003).
2.2. Apparatus
1H NMR and 13C spectra were recorded on a 300 MHz
Varian instrument. Chemical shifts are given in parts per mil-
lion (δ-values) relative to tetramethylsilane (TMS) as an in-
ternal standard. Elemental analyses were performed by Mi-
croanalytical Laboratory, Department of Physical Chemistry,
University of Vienna, Austria.
HPLC-UV used to quantify the compounds consisted of
a Shimadzu Model LC-6A pump and a Rheodyne 7125
injection valve with a 20-␮l loop connected to a Milton
Roy Spectromonitor 3000 UV detector and a Merck D-
2000 GPC integrator. Waters spherisorb S5OdS2 columns
4.6 mm × 250 mm connected to S5ODS2 precolumns were
applied in all HPLC analyses.
The following three l-Glu-Sar derivatives were syn-
thesised: an ester linked cyclic nucleoside analogue
l-Glu(trans-2-thymine-1-yl-tetrahydrofuran-3-yl ester)-Sar
(I), an N-acyloxymethyl linked nucleobase l-Glu(thymine-
1-yl-methyl ester)-Sar (II) and an ester linked acyclic nu-
cleoside analogue l-Glu(acyclothymidine)-Sar (III). The po-
tential of l-Glu-Sar as pro-moiety was evaluated in part by
studying the rate and mechanism of the in vitro model drug
release from the bioreversible derivatives in various aqueous
and biological media. Furthermore, affinity to hPEPT1 was
investigated in Caco-2 cells by inhibition studies. Transloca-
tion of compounds via hPEPT1 was estimated from studies
of intracellular acidification.
Transepithelial resistance (TEER) was measured in tis-
sue resistance measurement chambers (Endohm) with a
voltohmmeter from World Precision Instruments. The Swip
KS 10 Digi shaking plate used for cell culture experi-
2. Materials and methods
2.1. Materials
¨
ments was from Edmund Buhler. Radioactivity was counted
with a Tri-Carp 2110TR Liquid Scintillation Analyzer from
Packard.
Chemicals used in the synthesis were purchased from
Sigma–Aldrich (St. Louis, MO, USA) or BACHEM AG
(Bubendorf, Switzerland) and used without further purifi-
cation with the following exceptions: dimethyl formamide
(DMF) and dichloromethane (DCM) were dried and stored
Intracellular pH was monitored using an upright Nikon
Optiphot microscope with a Leitz (50 × W.I., NA. 1) objec-
tive. The microscope was equipped with a PTI Delta-RAM
monochromator, coupled to the light port via fibre-optics and
a PTI fluorescence detection system.
˚
over 3-A molecular sieves.
HPLC solvents were of analytical grade and chemicals
used in buffer preparations for stability studies were of lab-
oratory grade. Flash chromatography was performed with
Matrex silica, 60A/35-70 ␮m purchased from Millipore (Bil-
lerica, MA, USA).
Porcine intestine was kindly donated by The Danish
Meat Trade College (Roskilde, Denmark). Caco-2 cells
were obtained from American Type Culture Collection
(Manassas, VA, USA). Cell culture media and Hanks bal-
anced salt solution (HBSS) were obtained from Life Tech-
nologies (Høje Ta˚strup, Denmark). [14C]glycylsarcosine
2.3. Synthesis
All reactions involving air-sensitive reagents were per-
formedundernitrogen. Allglasswarewasflame-driedpriorto
use. N-(3-Dimethylaminopropyl)-Nꢀ-ethylcarbodiimide hy-
drochloride (EDAC) was handled under a nitrogen atmo-
sphere as previously described (Friedrichsen et al., 2001).
The crude products were purified by flash chromatography
(Still et al., 1978). Intermediates (a–f) and final products
(I–III) are shown in Fig. 1.

A.H. Eriksson et al. / European Journal of Pharmaceutical Sciences 25 (2005) 145–154
147
(278 mg, 1.45 mmol) and DMAP (0.66 mmol, 81 mg) in
DCM (2 ml). Trans (2-thymine-1-yl-tetrahydrofuran-3-ol)
(0.6 g, 2.83 mmol) in DMF (2.00 ml) was added under ni-
trogen followed by stirring at rt for 16 h. Evaporation to
dryness in vacuo gave oil, which was dissolved in boil-
ing ethyl acetate (50 ml). Washing with saturated aqueous
NaCl (2× 30 ml), saturated aqueous NaHCO3 (2× 30 ml),
10% aqueous citric acid (2× 30 ml), saturated aqueous NaCl
(30 ml), drying over MgSO4, removal of the ethyl acetate and
flash chromatography (ethyl acetate–heptane 1:1 → ethyl ac-
etate) afforded 0.31 g (94%) of Boc-Glu(trans-2-thymine-1-
yl-tetrahydrofuran-3-yl ester)-OBn (a), rf 0.12 (ethyl acetate-
heptane 1:1). δH (CDCl3) 7.37 (5H, s, phenyl CH), 7.06/7.03
(1H, s, C C H in thymine), 5.69 (1H, d, J = 2.9 Hz, H-2ꢀ),
5.37 (1H, m, H-3ꢀ) 5.21 (1H, m, NH in Glu), 5.17 (2H, s, CH2
in benzyl), 4.40 (1H, m, CH in Glu), 4.32 (1H, m, H-5ꢀ), 4.09
(1H, m, H-5ꢀ), 2.44 (2H, m, CH2 in Glu), 2.39 (1H, m, H-4ꢀ),
2.20 (1H, m, CHA in CH2 in Glu), 2.12 (1H, m, H-4ꢀ), 1.93
(1H, m, CHB in CH2 in Glu), 1.91 (3H, s, CH3 in thymine),
1.42 (9H, s, Boc).
Boc-Glu(trans-2-thymine-1-yl-tetrahydrofuran-3-yl
ester)-Sar-Ot-Bu (b) was prepared by first dissolving Boc-
Glu(trans-2-thymine-1-yl-tetrahydrofuran-3-yl ester)-OBn
(a) (0.16 g, 0.301 mmol) in ethyl acetate (10 ml). Ten percent
palladium on carbon (300 mg) was added and the solution
was stirred under hydrogen for 16 h. Filtration, extraction
with additional 100 ml of methanol and evaporation to
dryness gave 0.10 g (75%) of Boc-Glu[1-(cis-3-hydroxy-
2-tetrahydrofuranyl)-thymine]-OH as a white solid, rf 0.10
(ethyl acetate).
Fig. 1. The chemical structure of the three products l-Glu-(trans-2-thymine-
1-yl-tetrahydrofuran-3-yl ester)-Sar (I), l-Glu-(thymine-1-yl-methyl ester)-
Sar (II) and l-Glu-(acyclothymidine)-Sar (III) as their trifluoroacetates.
The compounds (a–f) are intermediates in the synthesis as described in
Section 2.
2.3.1. l-Glu(trans-2-thymine-1-yl-tetrahydrofuran-3-yl
ester)-Sar-OH, trifluoroacetate (I)
The ester linked cyclic nucleoside analogue was syn-
¨
thesised by a Vorbruggen coupling between 2,4-bis(trime-
Sar was attached to Boc-Glu(trans-2-thymine-1-yl-
tetrahydrofuran-3-yl ester)-OH by the following procedure:
Boc-Glu(trans-2-thymine-1-yl-tetrahydrofuran-3-yl ester)-
OH (100 mg, 0.23 mmol), 1-hydroxybenzotriazole, hydrate
(33 mg, 0.22 mmol), diisopropylethylamine (0.2 ml) and
EDAC (66 mg, 0.34 mmol) were dissolved in DMF (1.0 ml).
H-Sar-Ot-Bu·HCl (36 mg, 0.20 mmol) was added and the so-
lution stirred under nitrogen for 20 h. The solution was evap-
orated to dryness in vacuo and the residue dissolved in boil-
ing ethyl acetate (25 ml). Washing with saturated aqueous
NaCl (2× 15 ml), saturated aqueous NaHCO3 (2× 15 ml),
10% aqueous citric acid (2× 20 ml), saturated aqueous NaCl
(1× 15 ml), drying over MgSO4, filtration, removal of the
ethyl acetate in vacuo and flash chromatography (ethyl ac-
etate) yielded 80 mg (70%) of semi-solid Boc-Glu(trans-
2-thymine-1-yl-tetrahydrofuran-3-yl ester)-Sar-Ot-Bu (b), rf
0.70 (ethyl acetate). δH (CD3OD) 7.45/7.38 (1H, s, thymine
H C C), 5.73 (1H, m, H-2ꢀ), 5.43 (1H, m, H-3ꢀ), 4.70/4.51*
(1H, m, CH in Glu), 4.35 (1H, m, H-5ꢀ), 4.20 (1H, d,
J = 17.2 Hz, HA in CH2 in Sar), 4.08 (1H, m, H5ꢀ), 3.83 (1H,
d, J = 17.2 Hz, HB in CH2 in Sar), 3.22/2.93* (3H, s, CH3 in
Sar), 2.50 (2H, m, CH2 in Glu), 2.50 (beneath CH2 signals
from Glu) (1H, m, H-4ꢀ), 2.12 (1H, m, HA in CH2 in Glu),
2.12 (beneath CH2 signals from Glu) (1H, m, H-4ꢀ), 1.88
(3H, s, CH3 in thymine), 1.46 (9H, s, COO-t-Bu), 1.43 (9H,
s, Boc). *Two conformers.
thylsilyl)thymine and acetic acid (2-methoxytetrahydro-
furan-3-yl) ester catalyzed by SnCl4 to give almost selec-
tively the acetic acid (trans-2-thymin-1-yl-tetrahydrofuran-
3-yl) ester (Lin et al., 1979). The key intermediate and acetic
acid (2-methoxytetrahydrofuran-3-yl) ester were synthesised
by treatment of 2-methoxytetrahydrofuran-3-ol with acetyl
chloride/pyridine. 2,3-Dihydrofuran was treated with per-
acetic acid in methanol to give almost selectively the trans-2-
methoxytetrahydrofuran-3-ol (Craig et al., 1999; Sweet and
Brown, 1966). The acetyl group was removed using 1 M
LiOH in THF.
By this procedure, 0.67 g of yellow–white, solid trans
(2-thymine-1-yl-tetrahydrofuran-3-ol) was obtained, rf 0.25
(ethyl acetate). δH (CDCl3) 10.30 (1H, s, thymine NH), 7.22
(1H, s, thymine H C C), 5.72 (1H, s, H-2ꢀ), 4.93 (1H, s,
OH), 4.47 (1H, d, J = 4.6 Hz, H-3ꢀ), 4.32 (2H, m, H5ꢀ), 2.12
(1H, m, H-4ꢀ), 1.98 (1H, m, H-4ꢀ), 1.92 (3H, s, thymine
CH3). δC (CDCl3) 164.6, 151.2, 134.6, 110.9, 94.3, 76.4,
70.5, 31.3 and 13.0. Anal. calc. for C9H12N2O4:C, 50.94;
H, 5.70 and N, 13.20. Found: C, 50.74; H, 5.64 and N,
12.91. The 3-hydroxy group was then coupled to l-Glu-Sar
using the following sequential procedure. Boc-Glu(trans-2-
thymine-1-yl-tetrahydrofuran-3-yl ester)-OBn (a) was pre-
pared by dissolving Boc-Glu-OBn (334 mg, 0.99 mmol), 1-
hydroxybenzotriazole hydrate (152 mg, 0.99 mmol), EDAC

148
A.H. Eriksson et al. / European Journal of Pharmaceutical Sciences 25 (2005) 145–154
l-Glu(trans-2-thymin-1-yl-tetrahydrofuran-3-yl ester)-
(3H, s, CH3 in thymine), 1.73 (1H, m, HB in CH2 in Glu),
1.49/1.45* (9H, s, COO-t-Bu), 1.41 (9H, s, Boc). *Two con-
formers.
Sar-OH, trifluoroacetate (I) was achieved using the fol-
lowing deprotection procedure: Boc-Glu(trans-2-thymine-
yl-tetrahydrofuran-3-yl ester)-Sar-Ot-Bu (b) (80 mg,
0.14 mmol) was dissolved in dichloromethane (3 ml). After
addition of TFA (1.5 ml), the solution was stirred for 3 h.
The solvent was removed in vacuo. Co-evaporation with
chloroform (3× 3 ml) yielded 74 mg (100%) of semi-solid
l-Glu(trans-2-thymine-1-yl-tetrahydrofuran-3-yl-ester)-
Sar-OH, trifluoroacetate (I). δH (CD3OD) 7.38 (1H, s,
thymine H C C), 5.79 (1H, m, H-2ꢀ), 5.42 (1H, m, H-3ꢀ),
4.57 (1H, m, CH in Glu), 4.38 (1H, m, H-5ꢀ), 4.30 (1H, d,
J = 17.4 Hz, HA CH2 in Sar), 4.07 (1H, m, H5ꢀ), 4.03 (1H,
d, J = 17.2 Hz, HB in CH2 in Sar), 3.19/3.00* (3H, s, CH3
in Sar), 2.65 (2H, m, CH2 in Glu), 2.40 (1H, m, HA in CH2
in Glu), 2.22 (1H, m, HB in CH2 in Glu), 2.17 (beneath
CH2 signals from Glu) (2H, m, H-4), 1.89 (3H, s, CH3 in
thymine). *Two conformers.
l-Glu(thymin-1-yl-methyl ester)-Sar, trifluoroacetate (II)
was prepared from Boc-Glu(thymine-1-yl-methyl ester)-Sar-
Ot-Bu (d) (50 mg) using the same deprotection procedure as
described in Section 2.3.1 to yield 46 mg (100%) of Boc-
Glu(thymine-1-yl-methyl ester)-Sar, trifluoroacetate (II) as
a semi-solid. δH (CDCl3) 7.54 (1H, s, thymine H C C),
5.71 (2H, s, CH2 in methyl-linker), 5.30 (1H, d, J = 4.1 Hz,
NH in Glu), 4.58/4.40* (1H, m, CH in Glu), 4.29 (1H, d,
J = 17.5 Hz, HA in CH2N), 4.01 (1H, d, J = 17.4 Hz, HB in
CH2), 3.18/3.00* (3H, s, NCH3), 2.66 (2H, m, CH2 in Glu),
2.23 (1H, m, HA in CH2 in Glu), 2.15 (1H, m, HB in CH2 in
Glu), 1.88 (3H, s, CH3 in thymine). *Two conformers.
2.3.3. l-Glu(acyclothymidine)-Sar, trifluoroacetate (III)
The 1-(2-hydroethoxy)methyl linker was attached to N-1
of thymine by reacting 2,4-bis(trimethylsilyl)thymine with
1,3-dioxolane, potassium iodide and chlorotrimethylsilane
in acetonitrile at room temperature (Ubasawa et al., 1995).
A 3.4 g of (thymine-1-yl-methyl)-ethan-2-ol was obtained
with rf 0.33 (DCM-MeOH 9:1), which was crystallised from
96% ethanol (18 ml) to give melting point of 175–177 ◦C.
δH (CD3OD) 7.51 (1H, s, thymine H C C), 5.18 (2H, s,
NCH2O), 3.64 (4H, m, CH2O in linker), 1.89 (3H, s, CH3 in
thymine).
2.3.2. l-Glu(thymine-1-yl-methyl ester)-Sar,
trifluoroacetate (II)
Boc-Glu(thymine-1-yl-methyl ester)-OBn (c) was pre-
pared by adding thymine to formaldehyde.
A N-
acyloxymethyl linker was introduced selectively in the
1-position by condensing 1,3 bis(hydroxy)thymine with
Boc-Glu-OBn in the presence of dicyclohexylcarbodiimide
(DCC) and catalytic amount of N,N-dimethylaminopyridine
(DMAP) (Ahmad et al., 1987). Boc-Glu(thymine-1-yl-
methyl ester)-OBn (c) appeared as a solid, rf 0.70 (EtOAc).
δH (CDCl3) 8.14 (1H, s, NH in thymine), 7.36 (5H, s, benzyl
CH), 7.28 (1H, s, H C C), 5.63 (2H, s, CH2O), 5.17 (2H, s,
benzyl CH2), 4.40 (1H, m, CH in Glu), 2.47 (2H, m, CH2 in
Glu), 2.23-1.80 (2H, m, CH2 in Glu), 1.92 (3H, s, thymine
CH3), 1.41 (9H, s, Boc).
Sar was attached by the following sequential proce-
dure: Boc-Glu(thymine-1-yl-methyl ester)-OBn (c) (0.24 g,
0.50 mmol) was dissolved in ethyl acetate under nitrogen as
described in Section 2.3.1. Ten percent palladium on carbon
(200 mg) was added and the mixture stirred under hydro-
gen for 2 h. Subsequent filtration through Celite and evapora-
tion to dryness gave 200 mg (100%) of Boc-l-Glu(thymin-1-
yl-methyl ester)-OH as an oil. Boc-Glu(thymin-1-yl-methyl
ester)-OH (0.20 g, 0.52 mmol) and 1-hydroxybenzotriazole
hydrate (75 mg, 0.51 mmol) were dissolved in dry DMF un-
der nitrogen. After 2–3 min, DIPEA (0.2 ml, 1.15 mmol),
H-Sar-Ot-Bu·HCl (82 mg, 0.45 mmol) and EDAC (0.15 g,
0.78 mmol) were added and the solution was stirred under
nitrogen overnight. After the same evaporation and wash-
ing procedures as described in Section 2.3.1, 0.15 g (65%)
of Boc-Glu(thymine-1-yl-methyl ester)-Sar-Ot-Bu (d) ap-
peared as an oil, rf 0.57 (ethyl acetate). δH (CDCl3) 7.56
(1H, s, thymine H C C), 5.71 (2H, s, CH2 in methyl linker),
5.40/5.30* (1H, s, NH in Glu), 4.70/4.54* (1H, m, CH in
Glu), 4.20 (1H, d, J = 17.1 Hz, HA in CH2N), 3.79 (1H, d,
J = 17.2 Hz, HB in CH2N), 3.20/2.94* (3H, s, NCH3), 2.52
(2H, m, CH2 in Glu), 2.06 (1H, m, HA in CH2 in Glu), 1.87
Boc-Glu(acyclothymidine ester)-OBn (e) was pre-
pared from Boc-Glu-OBn (2.53 g, 7.49 mmol), 1-
hydroxybenzotriazole hydrate (1.15 g, 7.48 mmol), DMAP
(0.92 g, 7.53 mmol) and EDAC (2.10 g 10.95 mmol), which
were dissolved in dry dichloromethane (15 ml). After
stirring for 10 min, thymine-1-yl-methyl-ethan-2-ol (1.00,
5.00 mmol) dissolved in 15 ml hot dry DMF, was added.
Upon stirring for 16 h and after evaporation and washing
procedures as described in Section 2.3.1, 1.55 g (60%) of
Boc-Glu(acyclothymidine)-OBn (e), rf 0.59 (ethyl acetate),
in the form of a solidifying oil was obtained. δH (CDCl3)
8.53 (1H, s, thymine NH), 7.36 (s, 5H, phenyl), 7.13 (1H, s,
H C C in thymine), 5.23 (1H, d, J = 8.4 Hz, NH in Glu),
5.17 (2H, s, NCH2O), 5.15 (2H, s, benzyl CH2), 4.46 (1H,
m, CH in Glu), 4.22 (2H, t, J = 3.4 Hz, CH2 in linker), 3.77
(2H, t, J = 4.6 Hz, CH2 in linker), 2.40 (2H, m, CH2 in Glu),
2.18 (1H, m, HA in CH2 in Glu), 1.97 (1H, m, HB in CH2 in
Glu), 1.93 (3H, s, CH3 in thymine), 1.43 (9H, s, Boc).
Boc-Glu(acyclothymidine)-Sar-Ot-Bu (f) was pre-
pared from Boc-Glu(acyclothymidine)-OBn (e) (0.77 g,
1.48 mmol) as described in Sections 2.3.1 and 2.3.2 to yield
0.60 g (87%) of Boc-Glu(acyclothymidine)-Sar-Ot-Bu(f)
as an oil, rf 0.39 (ethyl acetate). δH (CDCl3) 8.85 (1H, s,
thymine NH), 7.16 (1H, s, thymine H C C), 5.46 (1H, d,
J = 8.4 Hz, Glu NH), 4.74/4.56* (1H, m, CH in Glu), 4.23
(2H, m, CH2O in linker), 4.23 (beneath CH2O signal) (1H,
HA in CH2N), 3.79 (2H, t, J = 4.5 Hz, CH2O in linker), 3.79
(beneath CH2O signal) (1H, HB in CH2N), 3.16/2.95* (3H,
s, CH3 in Sar), 2.44 (2H, m, CH2 in Glu), 2.11 (1H, m, HA in

A.H. Eriksson et al. / European Journal of Pharmaceutical Sciences 25 (2005) 145–154
149
clear supernatant was stored for maximally 2 h at 5 ◦C until
analysis.
CH2 in Glu), 1.93 (3H, s, CH3 in thymine), 1.75 (1H, m, HB
in CH2 in Glu), 1.45 (9H, s, Boc), 1.40 (9H, s, COO-t-Bu).
*Two conformers.
2.5. Cell culture
l-Glu(acyclothymidine)-Sar, trifluoroacetate (III) was ob-
tained by deprotection of Boc-Glu(acyclothymidine)-Sar-Ot-
Bu (f) as described in Section 2.3.1 to yield 46 mg (100%) of
l-Glu(acyclothymidine)-Sar, trifluoroacetate (III) as a brown
oil. δH (CD3OD) 7.90 (1H, s, thymine NH), 7.49 (1H, s,
thymine H C C), 5.17 (2H, s, NCH2O), 4.58/4.41* (1H, m,
CH in Glu), 4.29 (beneath CH2O signal) (1H, d, J = 17.5 Hz,
HA in CH2N), 4.26 (2H, m, CH2O in linker), 4.03 (1H, d,
J = 17.4 Hz, HB in CH2N), 3.79 (2H, t, J = 4.6 Hz, CH2O in
linker), 3.19/3.00* (3H, s, NCH3), 2.59 (2H, m, CH2 in Glu),
2.22 (1H, m, HA in CH2 in Glu), 2.09 (1H, m, HB in CH2
in Glu), 1.88 (3H, s, CH3 in thymine). *Two conformers are
observed.
Caco-2 cells were cultured as previously described
(Nielsen et al., 2001a). In short, cells were seeded in culture
flasks and passaged in Dulbecco’s Modified Eagle’s medium
(DMEM) supplemented by 10% fetal bovine serum, peni-
cillin/streptomycin (100 U/ml and 100 ␮g/ml, respectively),
1% l-Glutamine and 1% non-essential amino acids. For
affinity, studies cells were seeded onto tissue culture treated
Transwells® (1.13 cm2, 0.4 ␮m pore size) at a density of
105 cells/cm2 and grown to confluent monolayers. Experi-
ments were performed 25–27 days after seeding at passage
numbers ranging from 32 to 45. TEER was measured at room
temperature just before the experiments and ranged from 300
to 500 ꢀ cm2. Cells for studying hPEPT1-mediated translo-
cation were cultured under similar conditions except that they
were seeded onto 4.52 cm2, 0.4 ␮m pore size Transwells®
with clear PET filters. Translocation studies were performed
18–41 days after seeding at passage numbers ranging from
26 to 49.
2.4. Stability in aqueous solution
The degradation of I–III as well as the appearance of the
respective pyrimidine or nucleoside analogues was studied
in various buffer solutions at 37 0.5 ◦C. The degradation
of III was investigated in the pH range from 2.0 to 10.0,
whereas the degradation of I and II was investigated at pH
5.0–7.4. The applied buffers were phosphate (pH 2.0, 3.0, 6.0
and 7.4), acetate (pH 4.0 and 5.0) and borate (pH 8.0, 9.0 and
10.0). The total buffer concentrations were 0.02–0.06 M. The
ionic strength of each solution was adjusted with potassium
chloride to 0.15. The reaction was initiated by adding 100 ␮l
stock solution to 10 ml pre-equilibrated buffer solution result-
ing in initial concentrations ranging from 10−5 to 10−4 M.
At various times, samples were taken and either analysed
immediately or kept for maximally 2 h at 5 ◦C until analy-
sis. The mobile phase consisted of acetate buffer, methanol,
acetonitrile and triethylamine (90:5:5:0.01 or 95:0:5:0.01).
The flow rate was 1.0 ml/min and detection wavelengths
were either 268 or 270 nm. The limit of quantification for
the compounds in aqueous solution was generally approx-
imately 5 × 10−6 M, and the retention times were 15 and
16 min for the two isomers of I, and 7 and 8 min for II and
III. For Compound III, a peak appeared at 26 min. This peak
was interpreted as the peak corresponding to the proposed
cyclisation product arising from cyclisation of the dipeptide
backbone.
2.6. [14C]Gly-Sar uptake studies
Ki values for the dipeptide derivatives were determined
by studying the inhibition of [14C]Gly-Sar uptake into fil-
ter grown Caco-2 cell monolayers as described previously
(Nielsen et al., 2001c). In short, the applied buffers were
HBSS supplemented with 0.05% BSA and 10 mM MES (pH
6.0) or 10 mM HEPES (pH 7.4). The cell monolayer was
initially incubated for 15 min at 37 ◦C with buffers at pH
6.0 applied apically and 7.4 applied basolaterally. The ex-
periment was initiated by apically incubating the monolayer
for 5 min at 37 ◦C with a mixture of 0.5 ␮Ci [14C]Gly-Sar
and various concentrations of dipeptide derivative ranging
from 0 to 3.0 mM. After incubation, the cells were washed
with ice-cold HBSS and filters were removed to count the
cell-associated radioactivity by liquid scintillation analysis.
Measurements were carried out in doublet on three different
passages.
2.7. pHi measurements
The degradation of the dipeptide derivatives was also
investigated at 37 ◦C 0.5 ◦C in 10% porcine intestinal
mucosal homogenate prepared in saline 0.05 M phosphate
buffer, pH 7.4 prepared as described previously (Krondahl et
al., 1997). The reaction was initiated by adding a stock so-
lution to 5 ml pre-equilibrated biological media resulting in
initial concentrations ranging from 4 × 10−5 to 8 × 10−4 M.
At various times, samples were taken and analysed as pre-
viously described; however, a protein precipitation step was
included. This was done by adding a saturated solution of
(NH4)2SO4 to the sample at a ratio of 2:1 followed by mix-
ing and centrifugation at 18,000 rpm for 5 min at 4 ◦C. The
hPEPT1-mediated translocation of substrates across the
apical membrane of Caco-2 cells was investigated by moni-
toring the intracellular pH of cells, which were perfused api-
cally with test solutions (Thwaites et al., 1993). The solu-
tions used were Krebs–Ringer’s solution: NaCl (137.3 mM),
KCl (5.1 mM), CaCl2 (2.8 mM), MgSO4 (1.0 mM), KH2PO4
(0.6 mM), Glucose (10.0 mM) and MES (at pH 6.0) or
HEPES (at pH 7.4) (20 mM). All test solutions were prepared
in Krebs–Ringer’s solution, pH 6.0.
The experiments were carried out on filter grown Caco-2
cell monolayers. The cells were loaded in Krebs–Ringer’s

150
A.H. Eriksson et al. / European Journal of Pharmaceutical Sciences 25 (2005) 145–154
of test compound causing 50% inhibition of [14C]-Gly-Sar
uptake. IC50-values were transformed to a Ki-value using
the Cheng–Prusoff equation (Cheng and Prusoff, 1973).
Data from pHi measurements were compiled by deter-
mining the slope of the pH versus time curve for each com-
pound (app 150 data points) and testing it against the slope
resulting from perfusion of a blank buffer (pH 6.0) using
paired t-tests of the slopes from the respective blank and test
solutions.
solution pH 7.4 with a cocktail of BCECF/AM (2.6 mM),
Pluronic F-127 (0.03%, w/v) and Cremophor EL (0.03%,
w/v) for 60 min. The filters were mounted in an Ussing
chamber on the stage of an upright microscope with
the apical side facing upwards. The basolateral side was
bathed in Krebs–Ringer’s solution (pH 7.4) throughout the
experiments. The apical side was exposed to a constant flow
of solution (blank buffers as well as test solutions) by a
peristaltic pump (6 ml/min). The objective was immersed in
the apical solution.
The tissue was allowed to equilibrate with Krebs–Ringer’s
solution, pH 7.4 for 5 min after loading and was then excited
with light at 490 and 432 nm alternately (HBW 2–4 nm).
The emitted light was collected at 530 10 nm.
3. Results
3.1. Stability in aqueous solution
The ratio between light emitted when the probe was
excited at 490 and 432 nm, respectively, was used to
approximate pHi by means of a calibration curve. The
calibration curve was obtained by solutions of BCECF in
Krebs–Ringer-like media (NaCl (5.1 mM), KCl (137.3 mM),
CaCl2 (2.8 mM), MgSO4 (1.0 mM), KH2PO4 (0.6 mM),
Glucose (10.0 mM) and MES (at pH 6.0 and 6.5) or HEPES
(at pH 7.0, 7.5 and 8.0) (20 mM)). The viscosity of the solu-
tions for calibration was increased by adding PVP (25 g/l).
Compounds were tested in concentrations ranging from one
to three times the corresponding Ki value determined in
inhibition studies. Based on Michaelis–Menten kinetics, a
positive response would be expected to be between 50 and
75% of maximal response, and thus would be detectable.
Experiments were performed at least in duplicate. Positive
(Gly-Sar) and negative(Lys[Z(NO2)]-Pro) controls were
At fixed temperature, buffer concentration and ionic
strength, the overall dipeptide derivative degradation as well
as the model drug release followed first-order kinetics. The
pH–rate profile of III was determined in the range between pH
2.0 and 10.0 (Table 1). The overall degradation was catalysed
by specific acid and base and the overall stability of the dipep-
tide derivative was highest at pH ∼ 5. At pH < 5, the predomi-
nant route of degradation was through a proposed cyclisation
of the dipeptide backbone similar to previous observations
by Steffansen and co-workers (Thomsen et al., 2003).
The stability of the dipeptide derivatives as well as the
model drug release from I and II was investigated in the
pH range between 5.0 and 7.4 under the same conditions
as for III. In this pH range, ester hydrolysis catalysed by
specific base was predominant for both compounds. The
overall half-lives of the three compounds at pH 6.0 and 7.4 in
aqueous zero buffer solution are shown in Table 1. Standard
error for half-lives in zero buffer concentration was obtained
from the regression analysis of half-lives versus total buffer
concentration.
¨
included for validation of the method (Knutter et al.,
2001). Furthermore, the l-Glu-Sar derivative of acyclovir
l-Glu(acyclovir)-Sar was included (Thomsen et al., 2003).
2.8. Data analysis
The stability of the dipeptide derivatives I and III was also
followed in 10% porcine intestinal mucosal homogenate at
pH 7.4 and at 37 ◦C. Compound II was not investigated in this
biological medium due to the high lability in aqueous buffer
solution. The degradation of I and III followed pseudo-first-
order kinetics and their respective half-lives in 10% porcine
intestinal mucosal homogenate is listed in Table 1. The degra-
dation of III was faster in 10% intestinal mucosal homogenate
than in aqueous solution at pH 7.4, whereas the degradation
of I in this medium was in the same order of magnitude as in
aqueous solution at pH 7.4.
Kinetic analysis of model drug release in aqueous
solutions was performed using observed rate constants (kobs
)
obtainedatvariouspHvaluesandbufferconcentrationsextra-
polated to zero buffer concentration (Larsen and Bundgaard,
1977). The degradation pathways of I and III at pH ≤ 7.4
were elucidated by following the release of the respective
model drugs as well as following the disappearance of
dipeptide derivative.
Ki for the dipeptide derivatives and Km for Gly-Sar
were calculated. In short the inhibition of [14C]Gly-Sar
uptake at various concentrations of test compound was
determined. The degree of inhibition was fitted to a
Michaelis–Menten-type equation (Equation (1)).
3.2. [14C]Gly-Sar uptake studies
ꢀ
ꢁ
U
1 − (U/U0)max × [I]
IC50 + [I]
The dipeptide derivatives’ affinity to hPEPT1 was deter-
mined by displacement studies of [14C]-Gly-Sar in Caco-2
cell monolayers as described in Section 2. The inhibition
curves are shown along with the inhibition constants (Ki) in
Fig. 2. Data are given as means S.E. All three compounds
showed an affinity in terms of Ki less than 1 mM.
1 −
=
(1)
U0
where U is the uptake of [14C]Gly-Sar, U0 the uptake of
[14C]Gly-Sar at zero inhibitor concentration, [I] the con-
centration of test compound and IC50 was the concentration

A.H. Eriksson et al. / European Journal of Pharmaceutical Sciences 25 (2005) 145–154
151
Table 1
Half-lives of the dipeptide derivatives at pH 6.0 and 7.4 at zero buffer concentration as well as in the biological medium 10% intestinal porcine mucosal
homogenate, mean S.E. (n = 3)
Compound
t1/2 S.E.
Zero buffer concentration
Biological medium
pH 6.0
pH 7.4
10% intestinal porcine
mucosal homogenate
I
11.3 1.2 h
40 1 min
59 4 min
II
8.7 0.7 min
27 1 s
nd
III
3.74 0.38 days
3.6 0.5 h
39 3 min
nd: not determined. Inset: pH–rate profile with log kobs (min−1) vs. pH for the dipeptide derivatives at 37 ◦C and zero buffer concentration. The curves are the
results of the fits to the theoretical expression for degradation.
Fig. 2. Inhibition data for Compounds I–III. Inhibition of [14C]Gly-Sar uptake (1 − U/U0) as a function of concentration of test compound. Ki S.E. (mM) for
the three compounds were determined as described in Section 2 to be: Compound I: 0.38 0.05, Compound II: 0.37 0.05 and Compound III: 0.075 0.004.
The values are means from at least three different passages and were calculated on the basis of a Km for Gly-Sar of 1.37 0.10 mM (mean S.E., n = 3).
3.3. pHi measurements
The hPEPT1-mediated translocation of Compounds I
and III, Gly-Sar, Lys[Z(NO2)]-Pro and l-Glu(acyclovir)-Sar
were investigated by monitoring the intracellular pH of Caco-
2 cell monolayers (Fig. 3). Slopes of the ratio versus time
curve are depicted in Fig. 4. Only Gly-Sar proved to induce
a drop in intracellular pH significantly different (p < 0.001)
from what is induced by a blank buffer (pH 6.0). The Com-
pounds III and l-Glu(acyclovir)-Sar proved not to induce an
intracellular acidification indicating that they are not translo-
cated via a proton cotransporter such as hPEPT1. The magni-
Fig. 3. Translocation studies on Caco-2 cell monolayers. Filters mounted in
tude of the responses for the compounds I and Lys[Z(NO2)]-
Ussing-like chambers with Krebs–Ringer solution pH 7.4 basolaterally are
perfused apically with (a) buffer pH 7.4, (b) buffer pH 6.0, (1) 2 mM Gly-Sar
in buffer pH 6.0 and (2) 0.23 mM III (l-Glu(acyclothymidine)-Sar).
Pro indicate that these compounds are not translocated
either.

152
A.H. Eriksson et al. / European Journal of Pharmaceutical Sciences 25 (2005) 145–154
A number of dipeptide coupled ester model pro-drugs
have been characterised recently (Nielsen et al., 2001b;
Steffansen et al., 1999; Thomsen et al., 2003). This class
of model pro-drugs releases typically the model drug by spe-
cific base-catalysed ester hydrolysis in the pH range between
5and10. Consequently, themodelpro-drugsareconsiderably
more stable at the luminal pH than at intracellular and plasma
pH with half-lives typically in the range of days or hours at
pH 6.0 and hours or minutes at pH 7.4. The mechanism of
specific base-catalysed ester hydrolysis may be similar to the
mechanism proposed by Johansen and Bundgaard, suggest-
ing that the protonated amine is involved in an intramolecular
intermediate facilitating the ester bond cleavage (Johansen
and Bundgaard, 1981). However, the much shorter half-life
of Compound II might indicate a different, kinetically equiva-
lent, mechanism of release for this particular compound. The
very short half-life of I (9 min) at pH 6.0 indicates that the
derivative would be degraded rapidly in the small intestine
before interaction with hPEPT1. Therefore, the drug deliv-
ery properties of this model drug (thymine) seem to not have
improved by this approach.
Fig. 4. pH (per s) S.D. for blank buffer (pH 6.0, n = 15), Gly-Sar (2 mM,
n = 5), Lys[Z(NO2)]-Pro (20 ␮M, n = 2), I (0.37 mM, n = 2), III (0.16 mM,
n = 3) and l-Glu(acyclovir)-Sar (0.85 mM, n = 3); ***indicates significant
(p < 0.001) difference from blank buffer.
4. Discussion
The dipeptide derivatives I and III proved sufficiently sta-
ble at pH 6.0 in order for the compounds to be available for
hPEPT1-mediated absorption with half-lives of close to 4
days and 9 h, respectively, at pH 6.0. Furthermore, the pre-
dominant mechanism of degradation of the dipeptide deriva-
tives at pH > 5 was specific base-catalysed hydrolysis, which
indicates that the model drugs will be released quantitatively
once absorbed.
It has previously been proposed that dipeptide pro-drugs
with an l-Glu-Sar dipeptide backbone form a cyclic product
in parallel to drug release (Thomsen et al., 2003). A similar
reaction has been reported on dipeptidic structures in ACE
inhibitors (Gu and Strickley, 1987, 1988). The stability stud-
ies in aqueous solution suggest that the same is the case for
Compound III. However, since this proposed mechanism of
degradation is predominant at pH < 5 it could be avoided in a
potential drug formulation by a gastro-resistant formulation.
The in vitro metabolism studies of Compounds I and III
further indicate that model drug release is controlled by pH
and to a lesser extent enzymatic activity in 10% intestinal
mucosal homogenate.
All three Compounds I–III showed high affinity to
hPEPT1, comparable to other known substrates such as
valaciclovir(Ki = 0.51 mM)andcephalexin(Ki = 11 mM)and
to affinities of other similar nucleoside and pyrimidine dipep-
tide pro-drugs such as l-Glu(acyclovir)-Sar (Ki = 0.28 mM)
and l-Glu[1-(2-hydroxyethyl)-thymine]-Sar (Ki = 0.21 mM)
(Nielsen et al., 2001c; Thomsen et al., 2003). On the
other hand, it has recently been demonstrated that l-
Glu(acyclovir)-Sar has limited in vivo bioavailability when
administered orally to rats (Thomsen et al., 2004).
The aim of the present study was to improve the synthetic
pathway of bioreversible dipeptide derivatives as well as eval-
uate the potential of using l-Glu-Sar as a pro-moiety for de-
livering three newly synthesised nucleoside and pyrimidine
l-Glu-Sar derivatives.
Previously, Steffansen and co-workers have prepared Boc-
Glu-Sar-Ot-Bu and Boc-d-Glu-Ala-Ot-Bu and coupled the
dipeptide directly to other compounds such as acyclovir and
1,2 hydroxyethylthymine (Thomsen et al., 2003). This stra-
tegy seems to have some disadvantages due to the very
ineffective synthesis involving the Fm protection of the
␥-carboxylic acid. According to our experiments, the de-
protection of the Fm-group with triethylamine followed by
ion-exchange chromatography was very time consuming. In-
stead, l-Glu was introduced to the pyrimidine/nucleoside
analogues. The benzyl- and Boc-protecting groups in Boc-
l-Glu(model drug)-OBn are orthogonal to deprotection and
various amino acids can subsequently be introduced in either
the C- or N-terminal. This synthetic pathway proved more
flexible giving more opportunities for future new combina-
tions of pro-drugs.
In evaluating the potential of l-Glu-Sar as pro-moiety,
the strategy applied was to investigate the in vitro stability
to gather qualitative as well as quantitative information on
mechanisms and rates of degradation. Furthermore, target-
ing to hPEPT1 was investigated in Caco-2 cells by means of
inhibition studies and hPEPT1-mediated intracellular acidi-
fication.
The pH in the upper small intestine is estimated to range
from 5 to 6 and the transit time to be between 2 and 3 h
(Daugherty and Mrsny, 1999). Since hPEPT1-mediated ab-
sorption is considered to take place in the upper small in-
testine, a drug or pro-drug targeted to hPEPT1 must remain
stable under these conditions.
The present ␥-derivatives I, III and l-Glu(acyclovir)-Sar
were not translocated by hPEPT1 and l-Glu-Sar, therefore
seems to be a poor pro-moiety for hPEPT1-mediated trans-
port. A similar observation has been made for ␧-derivatives

A.H. Eriksson et al. / European Journal of Pharmaceutical Sciences 25 (2005) 145–154
153
¨
of Lys-Pro (Knutter et al., 2001). Based on a series of dipep-
Crooks, R.J., Murray, A., 1994. Valaciclovir—a review of a promising
new antiherpes agent. Antivir. Chem. Chemother. 5 (Suppl. 1), 31–37.
Daugherty, A.L., Mrsny, R.J., 1999. Transcellular uptake mechanisms of
the intestinal epithelial barrier Part one. Pharm. Sci. Technol. Today
2, 144–151.
Friedrichsen, G.M., Nielsen, C.U., Steffansen, B., Begtrup, M., 2001.
Model prodrugs designed for the intestinal peptide transporter. A
synthetic approach for coupling of hydroxy-containing compounds
to dipeptides. Eur. J. Pharm. Sci. 14, 13–19.
Ganapathy, M.E., Huang, W., Wang, H., Ganapathy, V., Leibach, F.H.,
1998. Valacyclovir: a substrate for the intestinal and renal peptide
transporters PEPT1 and PEPT2. Biochem. Biophys. Res. Commun.
246, 470–475.
Gu, L., Strickley, R.G., 1987. Diketopiperazine formation, hydrolysis and
epimerization of the new dipeptide angiotensin-converting enzyme in-
hibitor RS-10085. Pharm. Res. 4, 392–397.
tide derivatives with variations of the N-terminal amino acid,
Brandsch and co-workers proposed that hPEPT1-mediated
transport of these derivatives rely on the size of the side chain
modification along with the distance from the ␣-carbon of
the amino acid. Furthermore, the hydrophobicity of the in-
¨
troduced group was suggested to be of importance (Knutter
et al., 2004).
The compounds presented in this study are less lipophilic
than previously reported inhibitors with side-chain modifi-
cations, thus indicating that size alone may be a restrictive
parameter. A contributing factor to the lack of translocation of
the investigated compounds may also be the N-methylation of
the dipeptide backbone. The potential of hPEPT1-mediated
pro-drug delivery, is however still evident. So far, only a lim-
ited number of dipeptide backbones have been investigated
and there are still possibilities in terms of other side chain
positions available for derivatising. Furthermore, the dipep-
tide backbone is available for other forms of modification.
Promising ketomethylene dipeptidomimetic backbone pro-
moieties have recently been proposed by Luthman and co-
workers (Va˚benø et al., 2004a,b). Viable future strategies for
targeting pro-drugs to hPEPT1 may be to explore combina-
tions of alternative dipeptidomimetic backbone pro-moieties
and drug attachment in various positions.
Gu, L., Strickley, R.G., 1988. Preformulation stability studies of the new
dipeptide angiotensin-converting enzyme inhibitor RS-10029. Pharm.
Res. 5, 765–771.
Han, H., de Vrueh, R.L., Rhie, J.K., Covitz, K.M., Smith, P.L., Lee,
C.P., Oh, D.M., Sadee, W., Amidon, G.L., 1998. 5ꢀ-Amino acid es-
ters of antiviral nucleosides, acyclovir, and AZT are absorbed by the
intestinal PEPT1 peptide transporter. Pharm. Res. 15 (8), 1154–1159.
Heidelberger, C., Chaudhuri, N.K., Danneberg, P., Mooren, D., Griesbach,
L., Duschinsky, R., Schnitzer, R.J., Pleven, E., Scheiner, J., 1957.
Fluorinated pyrimidines, a new class of tumour-inhibitory compounds.
Nature 179 (4561), 663–666.
Johansen, M., Bundgaard, H., 1981. Pro-drugs as drug delivery systems.
Part 16: novel water soluble pro-drug types for chlorzoxazone by
esterification of the N-hydroxymethyl derivative. Arch. Pharm. Chem.
Sci. Ed. 9, 43–54.
Knu¨tter, I., Hartrodt, B., Theis, S., Foltz, M., Rastetter, M., Daniel, H.,
Neubert, K., Brandsch, M., 2004. Analysis of the transport properties
of side chain modified dipeptides at the mammalian peptide trans-
porter PEPT1. Eur. J. Pharm. Sci. 21, 61–67.
Acknowledgments
The project is partly funded by LEO Pharma A/S. The
Danish Medical Research Council supported this work di-
rectly (Project Grant #22-01-0310) and via the Centre for
Drug Design and Transport. Birger Brodin was supported
by Novo Nordisk Fonden. The technical assistance of Stine
Bruun, Bettina Dinitzen, Brian Larsen and Susanne Sørensen
is highly appreciated.
Knu¨tter, I., Theis, S., Hartrodt, B., Born, I., Brandsch, M., Daniel, H.,
Neubert, K., 2001. A novel inhibitor of the mammalian peptide trans-
porter PEPT1. Biochemistry 40, 4454–4458.
Krondahl, E., Orzechowski, A., Ekstrom, G., Lennernas, H., 1997. Rat
jejunal permeability and metabolism of mu-selective tetrapeptides
in gastrointestinal fluids from humans and rats. Pharm. Res. 14,
1780–1785.
Larsen, C., Bundgaard, H., 1977. Kinetics and mechanism of degradation
of mecillinam in aqueous solution. Arch. Pharm. Chem. Sci. Ed. 5,
66–86.
Lepist, E.I., Kusk, T., Larsen, D.H., Andersen, D., Frokjaer, S., Taub,
M.E., Veski, P., Lennernas, H., Friedrichsen, G., Steffansen, B., 2000.
Stability and in vitro metabolism of dipeptide model prodrugs with
affinity for the oligopeptide transporter. Eur. J. Pharm. Sci. 11, 43–50.
Lin, A.J., Benjamin, R.S., Rao, P.N., Loo, T.L., 1979. Synthesis and
biological activities of ftorafur metabolites: 3ꢀ- and 4ꢀ-hydroxyftorafur.
J. Med. Chem. 22, 1096–1100.
Longley, D.B., Harkin, D.P., Johnston, P.G., 2003. 5-fluorouracil: mecha-
nisms of action and clinical strategies. Nat. Rev. Cancer 3, 330–338.
Matthews, T., Boehme, R., 1988. Antiviral activity and mechanism of
action of ganciclovir. Rev. Infect. Dis. 10, S490–S494.
Mitsuya, H., Broder, S., 1987. Strategies for antiviral therapy in AIDS.
Nature 325, 773–778.
Morse, G.D., Shelton, M.J., O’Donnell, A.M., 1993. Comparative phar-
macokinetics of antiviral nucleoside analogues. Clin. Pharmacokinet.
24, 101–123.
Nielsen, C.U., Amstrup, J., Steffansen, B., Frokjaer, S., Brodin, B., 2001a.
Epidermal growth factor inhibits glycylsarcosine transport and hPepT1
expression in a human intestinal cell line. Am. J. Physiol. Gastrointest.
Liver Physiol. 281, G191–G199.
References
Ahmad, S., Ozaki, S., Nagase, T., Iigo, M., Tokuzen, R., Hoshi, A., 1987.
A facile method for synthesis of N-acyloxymethyl-5-fluorouracils, as
a class of antitumor agents. Chem. Pharm. Bull. 35, 4137–4143.
Beauchamp, L.M., Orr, G.F., de Miranda, P., Burnette, T.C., Krenitsky,
T.A., 1992. Amino acid ester prodrugs of acyclovir. Antivir. Chem.
Chemother. 3, 157–164.
Brigden, D., Fiddian, P., Rosling, A.E., Ravenscroft, T., 1981.
Acyclovir—a review of the preclinical and early clinical data of a
new antiherpes drug. Antiviral Res. 1, 203–212.
Burnette, T.C., de Miranda, P., 1994. Metabolic disposition of the acy-
clovir prodrug valaciclovir in the rat. Drug Metab. Dispos. 22, 60–64.
Cheng, Y., Prusoff, W.H., 1973. Relationship between the inhibition con-
stant (KI) and the concentration of inhibitor, which causes 50 per cent
inhibition (I50) of an enzymatic reaction. Biochem. Pharmacol. 22,
3099–3108.
Craig, D., Munasinghe, V.R., Tierney, J.P., White, A.J.P., Williams,
D.J., Williamson, C., 1999. Template-directed intramolecular C-
glycosidation. Cation-mediated synthesis of ketooxetanes from thio-
glycosides. Tetrahedron 55, 15025–15044.
Nielsen, C.U., Andersen, R., Brodin, B., Frokjaer, S., Steffansen, B.,
2001b. Model prodrugs for the intestinal oligopeptide transporter:

154
A.H. Eriksson et al. / European Journal of Pharmaceutical Sciences 25 (2005) 145–154
model drug release in aqueous solution and in various biological me-
dia. J. Control. Release 73, 21–30.
Thomsen, A.E., Christensen, M.S., Bagger, M.A., Steffansen, B., 2004.
Acyclovir prodrug for the intestinal di-/tri-peptide transporter PEPT1:
comparison of transport in Caco-2 cells and in vivo bioavailability in
rats. Eur. J. Pharm. Sci. 23, 319–325.
Thomsen, A.E., Friedrichsen, G.M., Sorensen, A.H., Andersen, R.,
Nielsen, C.U., Brodin, B., Begtrup, M., Frokjaer, S., Steffansen, B.,
2003. Prodrugs of purine and pyrimidine analogues for the intestinal
di-/tri-peptide transporter PepT1: affinity for hPepT1 in Caco-2 cells,
drug release in aqueous media and in vitro metabolism. J. Control.
Release 86, 279–292.
Nielsen, C.U., Andersen, R., Brodin, B., Frokjaer, S., Taub, M.E.,
Steffansen, B., 2001c. Dipeptide model prodrugs for the intestinal
oligopeptide transporter. Affinity for and transport via hPepT1 in the
human intestinal Caco-2 cell line. J. Control. Release 76, 129–138.
Steffansen, B., Lepist, E.I., Taub, M.E., Larsen, B.D., Frokjaer, S., Lenner-
nas, H., 1999. Stability, metabolism and transport of d-Asp(OBzl)-
Ala—a model prodrug with affinity for the oligopeptide transporter.
Eur. J. Pharm. Sci. 8, 67–73.
Still, W.C., Kahn, M., Mitra, A.K., 1978. Rapid chromatographic tech-
nique for preparative separations with moderate resolution. J. Org.
Chem. 43, 2923–2925.
Sugawara, M., Huang, W., Fei, Y.J., Leibach, F.H., Ganapathy, V., Gana-
pathy, M.E., 2000. Transport of valganciclovir, a ganciclovir pro-
drug, via peptide transporters PEPT1 and PEPT2. J. Pharm. Sci. 89,
781–789.
Thwaites, D.T., Hirst, B.H., Simmons, N.L., 1993. Direct assessment of
dipeptide/H+ symport in intact human intestinal (Caco-2) epithelium:
a novel method utilising continuous intracellular pH measurement.
Biochem. Biophys. Res. Commun. 194, 432–438.
Ubasawa, M., Takashima, H., Sekiya, K., 1995. A convenient one-
pot synthesis of acyclonucleosides. Chem. Pharm. Bull. 43, 142–
143.
Sweet, F., Brown, R.K., 1966. 2-Alkoxy-3-hydroxytetrahydropyrans and
2-alkoxy-3-hydroxytetrahydrofurans. Part I: preparation structure and
some derivatives. Can. J. Chem. 44, 1571.
Taub, M.E., Moss, B.A., Steffansen, B., Frokjaer, S., 1997. Influence of
oligopeptide transporter binding affinity upon uptake and transport of
d-Asp(OBzl)-Ala and Asp(OBzl)-Sar in filter grown Caco-2 mono-
layers. Int. J. Pharm. 156, 219–228.
Taub, M.E., Moss, B.A., Steffansen, B., Frokjaer, S., 1998. Oligopep-
tide transporter-mediated uptake and transport of d-Asp(OBzl)-Ala,
d-Glu(OBzl)-Ala, and d-Ser(Bzl)-Ala in filter-grown Caco-2 mono-
layers. Int. J. Pharm. 174, 223–232.
Va˚benø, J., Lejon, T., Nielsen, C.U., Steffansen, B., Chen, W.Q., Hui,
O.Y., Borchardt, R.T., Luthman, K., 2004a. Phe-Gly dipeptidomimet-
ics designed for the di-/tri-peptide transporters PEPT1 and PEPT2:
synthesis and biological investigations. J. Med. Chem. 47, 1060–
1069.
Va˚benø, J., Nielsen, C.U., Ingebrigtsen, T., Lejon, T., Steffansen, B., Luth-
man, K., 2004b. Dipeptidic ketomethylene isosteres as pro-moieties
for drug transport via the human intestinal di-/tri-peptide transporter
hPEPT1: Design, synthesis, stability and biological investigations. J.
Med. Chem. 47, 4755–4765.