Serine side chain-linked peptidomimetic conjugates of cyclic HPMPC and HPMPA: Synthesis and interaction with hPEPT1

Article abstract of DOI:10.1021/mp100186b

Cidofovir (HPMPC), a broad spectrum antiviral agent, cannot be administered orally due to ionization of its phosphonic acid group at physiological pH. One prodrug approach involves conversion to the cyclic form (cHPMPC, 1) and esterification by the side chain hydroxyl group of a peptidomimetic serine. Transport studies in a rat model have shown enhanced levels of total cidofovir species in the plasma after oral dosing with l-Val-l-Ser-OMe cHPMPC, 2a. To explore the possibility that 2a and its three l/d stereoisomers 2b-d undergo active transport mediated by the peptide-specific intestinal transporter PEPT1, we performed radiotracer uptake and electrophysiology experiments applying the two-electrode voltage clamp technique in Xenopus laevis oocytes overexpressing human PEPT1 (hPEPT1, SLC15A1). 2a-d did not induce inward currents, indicating that they are not transported, but the stereoisomers with an l-configuration at the N-terminal valine (2a and 2b) potently inhibited transport of the hPEPT1 substrate glycylsarcosine (Gly-Sar). A "reversed" dipeptide conjugate, l-Ser-l-Ala-OiPr cHPMPC (4), also did not exhibit detectable transport, but completely abolished the Gly-Sar signal, suggesting that affinity of the transporter for these prodrugs is not impaired by a proximate linkage to the drug in the N-terminal amino acid of the dipeptide. Single amino acid conjugates of cHPMPC (3a and 3b) or cHPMPA (5, 6a and 6b) were not transported and only weakly inhibited Gly-Sar transport. The known hPEPT1 prodrug substrate valacyclovir (7) and its l-Val-l-Val dipeptide analogue (8) were used to verify coupled transport by the oocyte model. The results indicate that the previously observed enhanced oral bioavailability of 2a relative to the parent drug is unlikely to be due to active transport by hPEPT1. Syntheses of the novel compounds 2b-d and 3-6 are described, including a convenient solid-phase method to prepare 5, 6a and 6b.

Full text of DOI:10.1021/mp100186b

brief articles
Serine Side Chain-Linked Peptidomimetic Conjugates of
Cyclic HPMPC and HPMPA: Synthesis and Interaction with
hPEPT1†
Larryn W. Peterson,‡,§ Monica Sala-Rabanal,§,|,⊥ Ivan S. Krylov,‡
Michaela Serpi,‡ Boris A. Kashemirov,‡ and Charles E. McKenna*,‡
Department of Chemistry, UniVersity of Southern California, Los Angeles,
California 90089-0744, and Department of Physiology, DaVid Geffen School of Medicine at
UCLA, Los Angeles, California 90095
Received June 1, 2010; Revised Manuscript Received August 9, 2010; Accepted August 24, 2010
Abstract: Cidofovir (HPMPC), a broad spectrum antiviral agent, cannot be administered orally due
to ionization of its phosphonic acid group at physiological pH. One prodrug approach involves conversion
to the cyclic form (cHPMPC, 1) and esterification by the side chain hydroxyl group of a peptidomimetic
serine. Transport studies in a rat model have shown enhanced levels of total cidofovir species in the
plasma after oral dosing with L-Val-L-Ser-OMe cHPMPC, 2a. To explore the possibility that 2a and its
three L/D stereoisomers 2b-d undergo active transport mediated by the peptide-specific intestinal
transporter PEPT1, we performed radiotracer uptake and electrophysiology experiments applying the
two-electrode voltage clamp technique in Xenopus laevis oocytes overexpressing human PEPT1
(hPEPT1, SLC15A1). 2a-d did not induce inward currents, indicating that they are not transported,
but the stereoisomers with an L-configuration at the N-terminal valine (2a and 2b) potently inhibited
transport of the hPEPT1 substrate glycylsarcosine (Gly-Sar). A “reversed” dipeptide conjugate, L-Ser-
L-Ala-OiPr cHPMPC (4), also did not exhibit detectable transport, but completely abolished the Gly-
Sar signal, suggesting that affinity of the transporter for these prodrugs is not impaired by a proximate
linkage to the drug in the N-terminal amino acid of the dipeptide. Single amino acid conjugates of
cHPMPC (3a and 3b) or cHPMPA (5, 6a and 6b) were not transported and only weakly inhibited
Gly-Sar transport. The known hPEPT1 prodrug substrate valacyclovir (7) and its L-Val-L-Val dipeptide
analogue (8) were used to verify coupled transport by the oocyte model. The results indicate that the
previously observed enhanced oral bioavailability of 2a relative to the parent drug is unlikely to be due
to active transport by hPEPT1. Syntheses of the novel compounds 2b-d and 3-6 are described,
including a convenient solid-phase method to prepare 5, 6a and 6b.
Keywords: hPEPT1 transporter; SLC15A1; oral bioavailability; intestinal transport; Xenopus
laevis oocyte; prodrugs
used as a treatment for human cytomegalovirus (HCMV)
retinitis in immune-suppressed individuals.1 HPMPC, and other
Introduction
Cidofovir (HPMPC) is a broad spectrum antiviral agent active
against herpes-, adeno-, polyoma- and poxviruses.1 It is currently
transporter, hPEPT1 and serine side chain-linked cidofovir prodrugs;
Poster-119, 22nd International Conference on AntiViral Research,
Miami Beach, FL, May 3-7, 2009. Krylov, I. S.; Peterson, L. W.;
Kashemirov, B. A.; Breitenbach, J.; Borysco, K.; Drach, J. C.; Kim,
J. S.; Hilfinger, J. M.; McKenna, C. E. In Vitro transport, activation
and antiviral evaluation of new HPMPA prodrugs synthesized on a
solid support; Poster-190, 22nd International Conference on AntiViral
Research, Miami Beach, FL, May 3-7, 2009.
* To whom correspondence should be addressed. Mailing address:
University of Southern California, Department of Chemistry, Uni-
versity Park, Los Angeles, CA 90089-0744. E-mail: mckenna@
usc.edu. Tel: +1 213 740 7007. Fax: +1 213 740 0930.
† Preliminary accounts of some of this work were presented: Peterson,
L. W.; Kashemirov, B. A.; Blazewska, K. M.; Breitenbach, J. M.;
Borysko, K.; Drach, J. C.; Kim, J. S.; Kijek, P.; Hilfinger, J. M.;
McKenna, C. E. Synthesis and structure-activity aspects of some
cyclic cidofovir peptidomimetic prodrugs; Oral-15, 20th International
Conference on AntiViral Research, Palm Springs, CA, April 29-May
3, 2007. Sala-Rabanal, M.; Peterson, L. W.; Serpi, M.; Krylov, I. S.;
Kashemirov, B. A.; Kim, J. S.; Mitchell, S.; Hilfinger, J. M.;
McKenna, C. E. Interactions between the human oligopeptide
‡ University of Southern California.
§ Contributed equally.
| David Geffen School of Medicine at UCLA.
⊥ Present address: Department of Cell Biology & Physiology, Wash-
ington University in St. Louis School of Medicine, St. Louis, MO
63110.
10.1021/mp100186b  2010 American Chemical Society
Published on Web 10/07/2010
VOL. 7, NO. 6, 2349–2361 MOLECULAR PHARMACEUTICS 2349

brief articles
Peterson et al.
phosphate/phosphonate drugs, contain an ionizable P(O)(OH)2
group, which exists as a dianion at physiological pH resulting
in low membrane permeability and low oral bioavailability. To
address this problem, a number of prodrug strategies have
recently been developed.2,3 In one approach,2,4,5 the first
negative charge is masked by intramolecular cyclization to form
cyclic cidofovir (cHPMPC, 1), while the second P-OH is
esterified by the hydroxy side chain of an appropriate dipeptide,
e.g. X-Ser where X is an amino acid with a hydrophobic side
chain and the carboxyl group of the Ser is esterified to mask
its ionizable OH and increase lipophilicity. The use of cHPMPC,
which is converted to HPMPC by intracellular cyclic CMP
phosphodiesterases, offers the advantage of decreased nephro-
toxicity, while maintaining potency.6 The amino acid and
dipeptide promoieties offer the potential to “fine-tune” the
properties of the prodrug and are expected to be toxically
benign.
esterify the carboxylic acid group of the valine to form the
conjugated prodrug, which exhibits enhanced oral bioavail-
ability. This has been attributed to uptake and transport by
the human oligopeptide transporter, hPEPT1 (SLC15A1),
based on direct evidence obtained using radiolabeled 7 in
Xenopus laeVis oocytes expressing hPEPT1.9,12 hPEPT1 is
a proton-coupled di- and tripeptide transporter located in the
brush border membrane of enterocytes of the small intestine.
Its broad substrate specificity allows it to also transport
nonpeptidic compounds, such as ꢀ-lactam antibiotics, making
it attractive as a drug delivery target.13 While approaches to
modeling SAR and STR properties of hPEPT1 have been
recently reported,14-17 predictive rules for affinity and
transport have been elusive.18
We have reported the synthesis of several L-X-L-Ser-OMe
cHPMPC conjugates where X ) Ala, Val, and Leu and the
dipeptide promoiety is attached to the drug via esterification
of its phosphonic acid group by the serine side chain
hydroxyl function.4,5 When investigated for permeability in
an in situ single pass perfusion rat model, these conjugates
showed enhanced transport properties compared to the parent
compound, 1.5 The L-X-L-Ser-OMe cHPMPC derivatives
were also evaluated for transport in a murine model, and
after oral gavage dosing, increased levels of 1 species were
observed in the plasma compared to dosing with 1 itself.4,5
To investigate the possibility that hPEPT1 was implicated
The application of prodrug strategies has led to numerous
examples of polar drugs being made accessible orally.2,7,8
Typically, these approaches involve creating a drug precursor
having improved membrane permeability due to enhanced
lipophilicity3 or to targeting and exploiting the function of
specific transporters located in the gastrointestinal tract. One
example of the latter is valacyclovir (7), the 5′-valyl ester
of acyclovir (ACV).9-11 This drug does not contain a
phosphonate group, but has a terminal hydroxyl that can
(10) Guo, A.; Hu, P.; Balimane, P. V.; Leibach, F. H.; Sinko, P. J.
Interactions of a nonpeptidic drug, valacyclovir, with the human
intestinal peptide transporter (hPEPT1) expressed in a mammalian
cell line. J. Pharmacol. Exp. Ther. 1999, 289 (1), 448–54.
(11) Han, H.-K.; De Vrueh, R. L. A.; Rhie, J. K.; Covitz, K.-M. Y.;
Smith, P. L.; Lee, C.-P.; Oh, D.-M.; Sadee, W.; Amidon, G. L.
5′-Amino acid esters of antiviral nucleosides, acyclovir, and AZT
are absorbed by the intestinal PEPT1 peptide transporter. Pharm.
Res. 1998, 15 (8), 1154–1159.
(12) Beauchamp, L. M.; Orr, G. F.; De Miranda, P.; Burnette, T.;
Krenitsky, T. A. Amino acid ester prodrugs of acyclovir. AntiViral
Chem. Chemother. 1992, 3 (3), 157–164.
(13) Sala-Rabanal, M.; Loo, D. D. F.; Hirayama, B. A.; Turk, E.;
Wright, E. M. Molecular interactions between dipeptides, drugs
and the human intestinal H+-oligopeptide cotransporter hPEPT1.
J. Physiol. 2006, 574 (1), 149–166.
(1) De Clercq, E. Clinical potential of the acyclic nucleoside
phosphonates cidofovir, adefovir, and tenofovir in treatment of
DNA virus and retrovirus infections. Clin. Microbiol. ReV. 2003,
16 (4), 569–596.
(2) Peterson, L. W.; McKenna, C. E. Prodrug approaches to improving
the oral absorption of antiviral nucleotide analogues. Expert Opin.
Drug DeliVery 2009, 6 (4), 405–420.
(3) Hostetler, K. Y. Alkoxyalkyl prodrugs of acyclic nucleoside
phosphonates enhance oral antiviral activity and reduce toxicity:
Current state of the art. AntiViral Res. 2009, 82 (2), A84–A98.
(4) Eriksson, U.; Peterson, L. W.; Kashemirov, B. A.; Hilfinger, J. M.;
Drach, J. C.; Borysko, K. Z.; Breitenbach, J. M.; Kim, J. S.;
Mitchell, S.; Kijek, P.; McKenna, C. E. Serine peptide phospho-
ester prodrugs of cyclic cidofovir: Synthesis, transport, and
antiviral activity. Mol. Pharmaceutics 2008, 5 (4), 598–609.
(5) McKenna, C. E.; Kashemirov, B. A.; Eriksson, U.; Amidon, G. L.;
Kish, P. E.; Mitchell, S.; Kim, J.-S.; Hilfinger, J. M. Cidofovir
peptide conjugates as prodrugs. J. Organomet. Chem. 2005, 690
(10), 2673–2678.
(14) Kottra, G.; Frey, I.; Daniel, H. Inhibition of intracellular dipeptide
hydrolysis uncovers large outward transport currents of the peptide
transporter PEPT1 in Xenopus oocytes. Pfluegers Arch.: Eur.
J. Physiol. 2009, 457 (4), 809–820.
(6) Mendel, D. B.; Cihlar, T.; Moon, K.; Chen, M. S. Conversion of
1-[((S)-2-hydroxy-2-oxo-1,4,2-dioxaphosphorinan-5-yl)methyl]cy-
tosine to cidofovir by an intracellular cyclic CMP phosphodi-
esterase. Antimicrob. Agents Chemother. 1997, 41 (3), 641–646.
(7) Anastasi, C.; Quelever, G.; Burlet, S.; Garino, C.; Souard, F.;
Kraus, J.-L. New antiviral nucleoside prodrugs await application.
Curr. Med. Chem. 2003, 10 (18), 1825–1843.
(8) Ariza, M. E. Current prodrug strategies for the delivery of
nucleotides into cells. Drug Des. ReV.sOnline 2005, 2 (5), 373–
387.
(9) Balimane, P. V.; Tamai, I.; Guo, A.; Nakanishi, T.; Kitada, H.;
Leibach, F. H.; Tsuji, A.; Sinko, P. J. Direct evidence for peptide
transporter (PepT1)-mediated uptake of a nonpeptide prodrug,
valacyclovir. Biochem. Biophys. Res. Commun. 1998, 250 (2),
246–51.
(15) Larsen, S. B.; Jorgensen, F. S.; Olsen, L. QSAR models for the
human H+/peptide symporter, hPEPT1: Affinity prediction using
alignment-independent descriptors. J. Chem. Inf. Model. 2008, 48
(1), 233–241.
(16) Winiwarter, S.; Hilgendorf, C. Modeling of drug-transporter
interactions using structural information. Curr. Opin. Drug
DiscoVery 2008, 11 (1), 95–103.
(17) Pedretti, A.; De Luca, L.; Marconi, C.; Negrisoli, G.; Aldini, G.;
Vistoli, G. Modeling of the intestinal peptide transporter hPepT1
and analysis of its transport capacities by docking and pharma-
cophore mapping. ChemMedChem 2008, 3 (12), 1913–1921.
(18) Foley, D.; Pieri, M.; Pettecrew, R.; Price, R.; Miles, S.; Lam,
H. K.; Bailey, P.; Meredith, D. The in vitro transport of model
thiodipeptide prodrugs designed to target the intestinal oligopeptide
transporter, PepT1. Org. Biomol. Chem. 2009, 7 (18), 3652–3656.
2350 MOLECULAR PHARMACEUTICS VOL. 7, NO. 6

Peptidomimetic Conjugates of Cyclic HPMPC and HPMPA
brief articles
Figure 1. Compounds tested in the hPEPT1 oocyte expression system. Compounds 2-6 were isolated and evaluated
as the diastereomeric mixtures obtained in synthesis due to generation of a new chiral center at phosphorus.
in the observed transport, these derivatives were codosed with
glycylsarcosine (Gly-Sar), a well-known substrate for hPEPT1,
in DC5 cells overexpressing hPEPT1.5 All three compounds
showed affinity for hPEPT1, with Ki values of 2.7 to 5.4
mM.5
unless specified otherwise. Cidofovir and also HPMPA were
prepared as previously described,4 and L-Val acyclovir was
a kind gift from GlaxoSmithKline, Brentford, Middlesex,
U.K. [Glycyl-2-3H]-Gly-Sar (specific activity 60 Ci/mmol)
was from American Radiolabeled Chemicals, St. Louis, MO.
In the present study, we report the synthesis of the three
L/D stereoisomers of 2a (2b-2d), the L/D stereoisomeric
single amino acid serine side chain-linked conjugates of
cHPMPC (3a,b) and its adenine analogue, cHPMPA (5, 6a
and 6b) and a “reversed” dipeptide conjugate of cHPMPC
(4) (Figure 1). To investigate the interactions between these
prodrugs and hPEPT1, we employed a two-electrode voltage
clamp (TEVC) technique and tracer uptakes in Xenopus
laeVis oocytes expressing the transporter.13 In addition to
Gly-Sar, valacyclovir (7, Figure 1) and its dipeptide analogue,
L-Val-L-Val acyclovir (8, Figure 1), were included in the
study as positive controls.
General Method A. Synthesis of Amino Acid Isopropyl
Esters. Isopropanol (1.5 mL/mmol amino acid) was added
to a flask containing serine (500 mg, 4.8 mmol) or alanine
(1.00 g, 11.2 mmol). TMSCl (2.7 equiv) was added to the
reaction vessel, and the slurry was heated to reflux while
stirring. Additional portions of TMSCl were added as
1
necessary, while the reaction was monitored by H NMR
(after ∼5 h, the solution became transparent, and the reaction
was complete.) The sample was cooled to room temperature,
and the volatiles were removed under reduced pressure
yielding a solid. The excess isopropanol was coevaporated
with acetonitrile, leaving a white solid.
Experimental Section
Nomenclature. IUPAC-style names were generated with
the assistance of ACD/Name (ACD Labs, Inc.). The Au-
tonom nomenclature program, which gave somewhat dif-
ferent “IUPAC name” results for compounds such as 2a as
reported in our previous publication,4 is no longer available.
Isopropyl L-Serinate Hydrochloride (9a). 9a was obtained,
using general method A, as a white solid (759 mg, 87%).
1H NMR (400 MHz, D2O): δ ) 1.17 (d, J ) 6.3 Hz, 6H),
3.87 (dd, J ) 12.5 and 3.5 Hz, 1H), 3.96 (dd, J ) 12.5 and
4.4 Hz, 1H), 4.08 (dd, J ) 3.5 and 4.4 Hz, 1H), 5.00 (m,
1H).
General Experimental Methods. 1H, 13C, and 31P NMR
spectra were obtained on a 400 MHz Varian or 500 MHz
Bruker spectrometer. Chemical shifts (δ values) are reported
in ppm, and all coupling constants (J values) are quoted in
Hz. The following NMR abbreviations are used: s (singlet),
d (doublet), t (triplet), m (unresolved multiplet), dd (doublet
of doublets), brs (broad signal). Chemical shifts (δ) are
reported in parts per million (ppm) relative to internal CHCl3
(δ 7.24, 1H) or 85% H3PO4 (δ 0.00, 31P). Data for 31P NMR
are 1H decoupled. LC-MS analyses were done using a
Thermo-Finnegan LCQ DECA XPmax system. HRMS spectra
were recorded at the UCR High Resolution Mass Spectrom-
etry Facility, Riverside, CA. All reagents were purchased
from Sigma Aldrich or Alfa Aesar and used as obtained,
Isopropyl D-Serinate Hydrochloride (9b). 9b was obtained
using general method A, as a white solid (558 mg, 64%).
1H NMR (400 MHz, D2O): δ ) 1.18 (d, J ) 6.3 Hz, 6H),
VOL. 7, NO. 6 MOLECULAR PHARMACEUTICS 2351

brief articles
Peterson et al.
3.88 (dd, J ) 12.5 and 3.5 Hz, 1H), 3.97 (dd, J ) 12.5 and
4.4 Hz, 1H), 4.08 (dd, J ) 3.5 and 4.4 Hz, 1H), 5.01 (m,
1H).
Isopropyl L-Alaninate Hydrochloride (10). 10 was ob-
tained using general method A, as a white solid (972 mg,
52%). 1H NMR (400 MHz, D2O): δ ) 1.17 (d, J ) 6.3 Hz,
6H), 1.42 (d, J ) 7.3 Hz, 3H), 4.01 (q, J ) 7.3 Hz, 1H),
4.98 (m, 1H).
Isopropyl N-(tert-Butoxycarbonyl)-L-seryl-L-alaninate (13).
13 was obtained using general method B and starting from
250 mg of L-alanine isopropyl ester hydrochloride (10) and
306 mg of Boc-L-serine, as an off-white film (353 mg, 74%).
1H NMR (400 MHz, CDCl3): δ ) 1.17 and 1.19 (2d, J )
6.5 and 6.8 Hz, 6H), 1.33 (d, J ) 7.4 Hz, 3H), 1.37 (s, 9H),
3.62 (m, 2H), 3.90-3.93 (m, 1H), 4.19 (brs, 1H), 4.44 (m,
1H), 4.96 (m, 1H), 5.69 (brs, 1H), 7.30 (brs, 1H).
General Method B. Synthesis of Boc-Protected
Dipeptides. Serine methyl ester hydrochloride (9a,b) or
alanine isopropyl ester hydrochloride (10) (3.2 mmol, 1
equiv) and 1 equiv (3.2 mmol) of Boc-valine or Boc-L-serine,
respectively, were dissolved in 30 mL of dry CH2Cl2. The
reaction mixture was cooled to 0 °C before addition of HOBt
hydrate (4.8 mmol, 1.5 equiv) and Et3N (16 mmol, 5 equiv).
The reaction mixture was kept at 0 °C for 15 min before
EDC · HCl (4.0 mmol, 1.25 equiv) was added. The reaction
mixture was stirred at rt overnight. An additional 100 mL
of CH2Cl2 was added, and the organic layer was washed
successively with a citric acid solution (1.6 M, 25 mL),
saturated NaHCO3 solution (25 mL), and saturated NaCl
solution (20 mL). The organic phase was dried over Na2SO4
and concentrated under reduced pressure to yield the product.
Methyl N-(tert-Butoxycarbonyl)-L-valyl-L-serinate (11a).
11a was obtained from 250 mg of L-serine methyl ester
hydrochloride and 349 mg of Boc-L-valine using general
Boc-Protection of Serine Isopropyl Esters (12a,b).
Serine isopropyl ester 9 (200 mg, 1.08 mmol, 1 equiv) was
dissolved in 4 mL of CH3CN along with 240 mg (1 equiv)
of di-tert-butyl dicarbonate. To the resulting solution was
added 0.25 µL of Et3N at once. The resulting suspension
was stirred at room temperature for 60 min and then diluted
with 50 mL of CH2Cl2. The resulting solution was washed
with 10 mL of H2O, 10 mL of 10% citric acid solution, and
10 mL of saturated NaHCO3 solution. An additional wash
with saturated NaHCO3 solution was done as necessary. The
organic layer was dried over anhydrous Na2SO4 and filtered.
The filtrate was concentrated to afford a clear oil.
Isopropyl N-(tert-Butoxycarbonyl)-L-serinate (12a). 12a
1
was obtained as a transparent oil (269 mg, 99%). H NMR
(400 MHz, CD3OD): δ ) 1.19 and 1.20 (2d, J ) 6.3 Hz,
6H), 1.38 (s, 9H), 3.18 (brs, 1H), 3.80 (dd, J ) 11.4 and 3.5
Hz, 1H), 3.85 (dd, J ) 11.4 and 4.4 Hz, 1H), 4.23 (brs, 1H),
5.00 (m, 1H), 5.57 (brs, 1H).
1
method B, as an off-white film (386 mg, 76%). H NMR
Isopropyl N-(tert-Butoxycarbonyl)-D-serinate (12b). 12b
1
(400 MHz, CDCl3): δ ) 0.89 and 0.92 (2d, J ) 6.6 Hz,
6H), 1.36 (s, 9H), 1.97-2.05 (m, 1H), 3.70 (s, 3H),
3.78-3.96 (m, 3H), 4.61-4.63 (m, 1H), 5.43 (brs, 1H), 7.28
(brs, 1H). The 1H NMR chemical shift values are similar to
those previously reported for the same compound in
CD3OD.4
Methyl N-(tert-Butoxycarbonyl)-L-valyl-L-serinate (11b).
11b was obtained using general method B and starting from
500 mg of D-serine methyl ester hydrochloride and 698 mg
of Boc-L-valine as an off-white film (907 mg, 89%). 1H NMR
(400 MHz, CDCl3): δ ) 0.97 (d, J ) 7.0 Hz, 3H), 1.03 (d,
J ) 7.0 Hz, 3H) 1.48 (s, 9H), 2.25 (m, 1H), 3.82 (s, 3H),
3.98-4.05 (m, 3H), 4.68 (m, 1H), 5.08 (brs, 1H), 7.03 (brs,
1H).
was obtained as a transparent oil (232 mg, 86%). H NMR
(500 MHz, CD3OD): δ ) 1.19 and 1.20 (2d, J ) 6.3 Hz,
6H), 1.38 (s, 9H), 3.17 (brs, 1H), 3.78-3.87 (m, 2H), 4.22
(brs, 1H), 5.00 (m, 1H), 5.57 (brs, 1H).
General Method C. Synthesis of the Peptidomimetic
cHPMPC Conjugates. To HPMPC (0.30 mmol, 1 equiv),
5 mL of anhydrous DMF and 0.5 mL of dry DIPEA were
added. The reaction flask was warmed by a heat gun to
facilitate the dissolution of the HPMPC-DIPEA salt. The
solvent was then removed under vacuum. To the residue, 4
mL of anhydrous DMF, DIPEA (5 equiv), the relevant Boc-
protected amino acid or dipeptide (1.5 equiv) in dry DMF
and PyBOP (2.1 equiv) were added, and the reaction mixture
was stirred under N2 at 35 °C for 2 h. The reaction was
monitored by 31P NMR, and additional portions of PyBOP
were added as necessary. Solvent was removed under
vacuum, and the brownish-red residue was washed with Et2O.
The product was purified by preparative TLC [2 mm silica
gel on 20 × 20 glass support, Analtech]. The Boc-protected
derivative was recovered after extraction in CH3OH for 3 h.
TFA (4 mL, 99%) was added to a solution of the Boc-
protected derivative in 4 mL of dry CH2Cl2. After 5 h at
room temperature, the solvent was removed under vacuum.
Purification of the target compounds was conducted by
preparative TLC (2 mm silica gel on 20 × 20 cm glass
support). The desired product was found in the lower band,
cut out, and extracted using 100 mL of methanol. The silica
was removed by filtration, the solutions of the bis(trifluo-
roacetic acid) salts in methanol were concentrated, and the
products were precipitated by the addition of Et2O and
Methyl N-(tert-Butoxycarbonyl)-D-valyl-L-serinate (11c).
11c was obtained using general method B and starting from
250 mg of L-serine methyl ester hydrochloride and 345 mg
1
of Boc-D-valine, as an off-white film (352 mg, 69%). H
NMR (400 MHz, CDCl3): δ ) 0.87 (d, J ) 6.8 Hz, 3H),
0.94 (d, J ) 6.8 Hz, 3H) 1.38 (s, 9H), 2.15 (m, 1H), 3.72 (s,
3H), 3.87-3.96 (m, 3H), 4.58 (m, 1H), 4.96 (brs, 1H), 6.86
(brs, 1H).
Methyl N-(tert-Butoxycarbonyl)-D-valyl-D-serinate (11d).
11d was obtained using general method B and starting from
1.250 g of D-serine methyl ester hydrochloride and 1.746 g
1
of Boc-D-valine, as an off-white film (1.933 g, 76%). H
NMR (400 MHz, CDCl3): δ ) 0.92 (d, J ) 6.5 Hz, 3H),
0.97 (d, J ) 6.5 Hz, 3H) 1.41 (s, 9H), 2.13 (m, 1H), 3.68 (s,
3H), 3.73-4.18 (m, 3H), 4.56 (m, 1H), 5.93 (brs, 1H), 7.49
(brs, 1H).
2352 MOLECULAR PHARMACEUTICS VOL. 7, NO. 6

Peptidomimetic Conjugates of Cyclic HPMPC and HPMPA
brief articles
collected by filtration. The white crystals obtained were
carefully dried in vacuo. The final products were analyzed
37%). Mobile phase for first TLC purification: CH2Cl2/
CH3OH (6:1). Mobile phase for second TLC purification:
CH2Cl2/CH3OH (4:1). H NMR (400 MHz, CD3OD): δ )
1
by H NMR and 31P NMR, HRMS, and LC-MS.
1
1.32-1.37 (m, 6H), 3.77-4.63 (m, 11H), 5.11-5.20 (m,
1H), 5.90 (d, J ) 6.7 Hz, 1H), 7.55 (d, J ) 6.7 Hz, 0.8H),
7.60 (d, J ) 6.7 Hz, 0.2H). 31P NMR (202 MHz, CD3OD):
δ ) 14.63 (s, 0.78P), 15.51 (s, 0.22P). HRMS-FAB m/z [M
+ H]+ calcd for C14H23N4O7P: 391.1377, found 391.1380.
LC-MS: tR ) 9.22 min; [M + H]+ 391.1.
Methyl L-Valyl-O-{(5S)-5-[(4-amino-2-oxopyrimidin-1(2H)-
yl)methyl]-2-oxido-1,4,2-dioxaphosphinan-2-yl}-L-serinate (2a).
2a was obtained from 100 mg of HPMPC and 131 mg of
11a using general method C, as a white powder (123 mg,
56%). Mobile phase for first TLC purification: CH2Cl2/
CH3OH (6:1). Mobile phase for second TLC purification:
CH2Cl2/CH3OH (10:3). 1H NMR (400 MHz, CD3OD): δ )
1.09-1.16 (m, 6H), 2.22-2.34 (m, 1H), 3.78-4.60 (m,
14H), 5.91-5.94 (brs, 1H), 7.60 (d, J ) 7.4 Hz, 0.7H), 7.65
(d, J ) 7.4 Hz, 0.3H). 31P NMR (202 MHz, CD3OD): δ )
Isopropyl O-{(5S)-5-[(4-Amino-2-oxopyrimidin-1(2H)-yl)-
methyl]-2-oxido-1,4,2-dioxaphosphinan-2-yl}-D-serinate (3b).
3b was obtained from 150 mg of HPMPC and 176 mg of
12b using general method C, as a white powder (62 mg,
21%). Mobile phase for first TLC purification: CH2Cl2/
CH3OH (9:1). Mobile phase for second TLC purification:
1
14.41 (s, 0.65P), 15.53 (s, 0.35P). The H and 31P NMR
chemical shift values are similar to those previously reported
for the same compound.4 LC-MS: tR ) 9.83 and 11.17 min;
[M + H]+ 462.1.
1
CH2Cl2/CH3OH (10:2.7). H NMR (400 MHz, CD3OD): δ
) 1.35-1.38 (m, 6H), 3.74-4.70 (m, 11H), 5.14-5.22 (m,
1H), 5.90 (d, J ) 7.3 Hz, 1H), 7.52 (d, J ) 7.3 Hz, 0.8H),
7.59 (d, J ) 7.3 Hz, 0.2H). 31P NMR (202 MHz, CD3OD):
δ ) 14.51 (s, 0.80P), 15.73 (s, 0.20P). HRMS-FAB m/z [M
+ H]+ calcd for C14H23N4O7P: 391.1377, found 391.1381.
Methyl L-Valyl-O-{(5S)-5-[(4-amino-2-oxopyrimidin-1(2H)-
yl)methyl]-2-oxido-1,4,2-dioxaphosphinan-2-yl}-D-serinate (2b).
2b was obtained from 100 mg of HPMPC and 152 mg of
11b using general method C, as a white powder (47 mg,
22%). Mobile phase for first TLC purification: CH2Cl2/
CH3OH (6:1). Mobile phase for second TLC purification:
Isopropyl O-{(5S)-5-[(4-Amino-2-oxopyrimidin-1(2H)-yl)-
methyl]-2-oxido-1,4,2-dioxaphosphinan-2-yl}-L-seryl-L-alani-
nate (4). 4 was obtained from 150 mg of HPMPC and 257
mg of 13 using general method C, as a white powder (172
mg, 53%). Mobile phase for first TLC purification: CH2Cl2/
CH3OH (8:1). Mobile phase for second TLC purification:
1
CH2Cl2/CH3OH (4:1). H NMR (400 MHz, CD3OD): δ )
1.06-1.14 (m, 6H), 2.26 (m, 1H), 3.79-4.58 (m, 14H), 5.88
(d, J ) 7.3 Hz, 1H), 7.53 (d, J ) 7.3 Hz, 0.7H), 7.57 (d, J
) 7.5 Hz, 0.3H). 31P NMR (202 MHz, CD3OD): δ ) 14.40
(s, 0.7P), 15.68 (s, 0.3P).
1
CH2Cl2/CH3OH (4:1). H NMR (400 MHz, CD3OD): δ )
Methyl D-Valyl-O-{(5S)-5-[(4-amino-2-oxopyrimidin-1(2H)-
yl)methyl]-2-oxido-1,4,2-dioxaphosphinan-2-yl}-L-serinate (2c).
2c was obtained from 300 mg of HPMPC and 455 mg of
11c using general method C, as a white solid (378 mg, 58%).
Mobile phase for first TLC purification: CH2Cl2/CH3OH (6:
1). Mobile phase for second TLC purification: CH2Cl2/
CH3OH (10:3). 1H NMR (400 MHz, CD3OD): δ ) 1.07-1.13
(m, 6H), 2.25 (m, 1H), 3.75-4.58 (m, 14H), 5.92 (d, J )
7.3 Hz, 0.7H), 5.92 (d, J ) 7.5 Hz, 0.3H), 7.59 (d, J ) 7.3
Hz, 0.7H), 7.64 (d, J ) 7.5 Hz, 0.3H). 31P NMR (202 MHz,
CD3OD): δ ) 14.21 (s, 0.8P), 15.26 (s, 0.2P).
Methyl D-Valyl-O-{(5S)-5-[(4-amino-2-oxopyrimidin-1(2H)-
yl)methyl]-2-oxido-1,4,2-dioxaphosphinan-2-yl}-D-serinate (2d).
2d was obtained from 200 mg of HPMPC and 321 mg of
11d using general method C, as a white powder (199 mg,
45%). Mobile phase for first TLC purification: CH2Cl2/
CH3OH (6:1). Mobile phase for second TLC purification:
CH2Cl2/CH3OH (10:3). 1H NMR (400 MHz, CD3OD): δ )
1.09-1.16 (m, 6H), 2.29 (m, 1H), 3.78-4.57 (m, 14H), 5.89
(d, J ) 7.5 Hz, 0.3H), 5.89 (d, J ) 7.3 Hz, 0.7H), 7.53 (d,
J ) 7.3 Hz, 0.7H), 7.57 (d, J ) 7.5 Hz, 0.3H). 31P NMR
(202 MHz, CD3OD): δ ) 14.77 (s, 0.7P), 15.73 (s, 0.3P).
MS-ESI m/z [M + H]+ calcd for C17H28N5O8P: 462.18, found
462.18. LC-MS: tR ) 13.34 and 14.88 min; [M + H]+
462.1.
1.27-1.31 (m, 6H), 1.44 and 1.47 (2d, J ) 7.4 and 7.4 Hz,
3H), 3.75-4.64 (m, 11H), 5.01-5.07 (m, 1H), 6.00 (d, J )
7.5 Hz, 1H), 7.72 (d, J ) 7.5 Hz, 0.6H), 7.77 (d, J ) 7.5
Hz, 0.4H). 31P NMR (202 MHz, CD3OD): δ ) 14.98 (s,
0.57P), 16.26 (s, 0.43P). HRMS-FAB m/z [M + H]+ calcd
for C17H28N5O8P: 462.1748, found 462.1752. LC-MS: tR
) 12.99 and 13.18 min; [M + H]+ 462.2.
General Method D. TBDMS Protection of Serine
Methyl and Isopropyl Esters. Serine methyl ester hydro-
chloride or serine isopropyl ester hydrochloride (9a) (2.3
mmol, 1 equiv) and tert-butylchlorodimethylsilane (4.8
mmol, 2.1 equiv) were dissolved in 20 mL of dry CH2Cl2.
The reaction mixture was cooled to 0 °C before addition of
imidazole (7.1 mmol, 3.1 equiv). The reaction mixture was
stirred overnight at rt. An additional 100 mL of CH2Cl2 was
added to the reaction mixture, and the organic layer was
washed with a saturated citric acid solution (25 mL) and
saturated NaHCO3 solution (25 mL). Combined water phases
were extracted twice with 30 mL of CH2Cl2 (to increase the
yield, product as a salt is soluble in water). All organic phases
were combined and dried over Na2SO4 and concentrated
under reduced pressure yielding a white solid. The solid was
washed on a filter with 20 mL of hexanes to remove the
TBDMSOH, giving a clean product.
Isopropyl O-[tert-Butyl(dimethyl)silyl]-L-serinate (14). 14
was obtained using general method D (521 mg, 76%). H
NMR (400 MHz, CDCl3): δ ) 0.08 (s, 3H), 0.12 (s, 3H),
0.88 (s, 9H), 1.29 (d, J ) 6.4 Hz, 6H), 4.13 (dd, J ) 2.6
Isopropyl O-{(5S)-5-[(4-Amino-2-oxopyrimidin-1(2H)-yl)-
methyl]-2-oxido-1,4,2-dioxaphosphinan-2-yl}-L-serinate (3a).
3a was obtained from 150 mg of HPMPC and 176 mg of
12a using general method C, as a white powder (110 mg,
1
VOL. 7, NO. 6 MOLECULAR PHARMACEUTICS 2353

brief articles
Peterson et al.
and 10.7 Hz, 1H), 4.20 (t, J ) 2.6 Hz, 1H), 4.25 (dd, J )
2.6 and 10.7 Hz, 1H), 5.11 (sept, J ) 6.7 Hz, 1H), 8.72
(brs, 3H).
Methyl O-{(5S)-5-[(6-Amino-9H-purin-9-yl)methyl]-2-
oxido-1,4,2-dioxaphosphinan-2-yl}-L-serinate (17a). 17a was
obtained using general method E, as a white powder (117
mg, 29%). 1H NMR (400 MHz, CD3OD): δ ) 3.82 (s, 3H),
4.01-4.12 (m, 1H), 4.25-4.37 (m, 3H), 4.42-4.65 (m, 6H),
8.28 (s, 0.4H), 8.31 (s, 0.6H), 8.39 (s, 1H). 31P NMR (202
MHz, CD3OD): δ ) 14.32 (s, 0.31P), 15.45 (s, 0.69P).
HRMS-FAB m/z [M + H]+ calcd for C13H19N6O6P: 387.1176,
found 387.1181.
Methyl O-[tert-Butyl(dimethyl)silyl]-L-serinate (15a). 15a
1
was obtained using general method D (434 mg, 69%). H
NMR (400 MHz, CDCl3): δ ) 0.08 (s, 3H), 0.11 (s, 3H),
0.88 (s, 9H), 3.83 (s, 3H), 4.15 (dd, J ) 3.0 and 10.7 Hz,
1H), 4.26 (dd, J ) 3.0 and 10.7 Hz, 1H), 4.32 (m, 1H), 8.72
(brs, 3H).
Methyl O-[tert-Butyl(dimethyl)silyl]-D-serinate (15b). 15b
1
Methyl O-{(5S)-5-[(6-Amino-9H-purin-9-yl)methyl]-2-
oxido-1,4,2-dioxaphosphinan-2-yl}-D-serinate (17b). 17b was
obtained, using general method E as a white powder (106
mg, 26%). 1H NMR (400 MHz, CD3OD): δ ) 3.90 and 3.92
(2s, 3H), 4.01-4.13 (m, 1H), 4.29-4.75 (m, 9H), 8.23 (s,
0.4H), 8.28 (s, 0.6H), 8.37 and 8.38 (2s, 1H). 31P NMR (202
MHz, CD3OD): δ ) 14.15 (s, 0.33P), 15.54 (s, 0.67P).
HRMS-FAB m/z [M + H]+ calcd for C13H19N6O6P: 387.1176,
found 387.1181.
was obtained using general method D (452 mg, 72%). H
NMR (400 MHz, CDCl3): δ ) 0.07 (s, 3H), 0.10 (s, 3H),
0.87 (s, 9H), 3.82 (s, 3H), 4.14 (dd, J ) 3.0 and 10.7 Hz,
1H), 4.25 (dd, J ) 3.0 and 10.7 Hz, 1H), 4.32 (t, J ) 3.0
Hz, 1H), 8.78 (brs, 3H).
General Method E. Solid-Phase Synthesis of the Pepti-
domimetic cHPMPA Conjugates. Addition of Amino Acid
to the Trityl Chloride Polystyrene Resin (TCP). The TB-
DMS-protected serine alkyl ester (14, 15a or 15b, 4 equiv)
was reacted with the trityl chloride polystyrene (TCP) resin
(1 equiv) in CH2Cl2 in the presence of DIPEA (15 equiv)
overnight. The resin was then filtered and washed using 150
mL of CH2Cl2. Removal of the TBDMS protecting group
was accomplished with TBAF (4 equiv) in THF for 5 h at
rt. The resin was then filtered and washed with 50 mL of
THF followed by 100 mL of CH2Cl2. The resin (16) was
dried under reduced pressure in a desiccator before being
used for the coupling reaction.
General Method F. Conversion of TFA Salts to HCl
Salts. The peptidomimetic cHPMPA conjugates as trifluo-
roacetic acid salts (17a or 17b, 0.08 mmol) were dissolved
in 3 mL of 3 N methanolic HCl and evaporated under
reduced pressure. This process was repeated 4 times. Finally,
the product was precipitated from a methanolic solution by
the addition of ethyl ether. The solid hydrochloride salts were
collected by filtration.
Coupling of HPMPA with Amino Acid Promoiety.
PyBOP (1.2 equiv) was added to a solution of HPMPA (1
equiv) and DIPEA (15 equiv) in DMF (10 mL). The
cyclization reaction proceeded at rt. After 1 h an additional
portion of PyBOP (3-4 equiv) and the TCP resin containing
serine alkyl ester (16) were added to the reaction mixture.
The reaction mixture was shaken overnight at 38 °C. The
resin was then filtered and washed with 60 mL of DMF
followed by 200 mL of CH2Cl2. The product was cleaved
from the resin with TFA (2 mL) in CH2Cl2 (10 mL) in the
case of serine methyl ester cHPMPA conjugates (17a and
17b) and with 1.4 M HCl in dioxane/CH2Cl2 (10 mL/5 mL)
in the case of serine isopropyl ester conjugate 5. Finally,
TFA salts of the peptidomimetic cHPMPA conjugates were
purified using silica gel TLC plates (2 mm silica gel on 20
× 20 cm glass support, Analtech) using CH2Cl2/CH3OH (6:
1) as the mobile phase. The products were precipitated as
trifluoroacetic acid salts from methanol by addition of diethyl
ether and collected by filtration. Conjugate 5 was purified
by reprecipitation from a methanolic solution by the addition
of diethyl ether.
Methyl O-{(5S)-5-[(6-Amino-9H-purin-9-yl)methyl]-2-
oxido-1,4,2-dioxaphosphinan-2-yl}-L-serinate Hydrochloride
(6a). 6a was obtained using general method F, as a white
1
powder (32 mg, 82%). H NMR (400 MHz, CD3OD): δ )
3.89 (s, 2H), 3.93 (s, 1H), 4.05-4.19 (m, 1H), 4.33-4.77
(m, 9H), 8.35 (s, 0.4H), 8.40 (s, 0.6H), 8.46 and 8.47 (s,
1H). 31P NMR (202 MHz, CD3OD): δ ) 14.28 (s, 0.31P),
15.41 (s, 0.69P).
Methyl O-{(5S)-5-[(6-Amino-9H-purin-9-yl)methyl]-2-
oxido-1,4,2-dioxaphosphinan-2-yl}-D-serinate Hydrochloride
(6b). 6b was obtained using general method F, as an off-
white solid (32 mg, 87%). 1H NMR (400 MHz, CD3OD): δ
) 3.91 (s, 2H), 3.93 (s, 1H), 4.02-4.18 (m, 1H), 4.34-4.76
(m, 9H), 8.35 (s, 0.4H), 8.39 (s, 0.6H), 8.46 (s, 1H). 31P
NMR (202 MHz, CD3OD): δ ) 14.20 (s, 0.33P), 15.56 (s,
0.67P).
Synthesis of L-Val-L-Val Acyclovir (8). The product was
obtained as a white solid as described in the literature19 (140
mg, 39%).
2-[(2-Amino-6-oxo-1,6-dihydro-9H-purin-9-yl)methoxy]-
Isopropyl O-{(5S)-5-[(6-Amino-9H-purin-9-yl)methyl]-2-
oxido-1,4,2-dioxaphosphinan-2-yl}-L-serinate (5). 5 was ob-
tained using general method E (96 mg, 30%). 1H NMR (400
MHz, CD3OD): δ ) 1.30-1.39 (m, 6H), 4.04-4.20 (m, 1H),
4.32-4.76 (m, 9H), 5.13-5.20 (m, 1H), 8.36 (s, 0.4H), 8.41
(s, 0.6H), 8.47 (s, 1H). 31P NMR (202 MHz, CD3OD): δ )
14.18 (s, 0.44P), 15.13 (s, 0.56P). HRMS-FAB m/z [M +
H]+ calcd for C15H23N6O6P: 415.1489, found 415.1495.
1
ethyl L-Valyl-L-valinate (8). H NMR (400 MHz, D2O): δ
) 0.59 (d, J ) 6.9, 6H), 0.72 (d, J ) 7.0, 6H), 1.74-1.82
(m, 1H), 1.86-1.95 (m, 1H), 3.57 (d, J ) 6.1, 1H), 3.67
(m, 2H), 3.89 (d, J ) 5.7, 1H), 4.03-4.07 (m, 2H), 5.43 (s,
2H), 8.85 (s, 1H). 13C NMR (100 MHz, CD3OD): δ ) 16.27,
17.01, 17.44, 17.85, 29.99, 30.23, 57.98, 58.13, 63.38, 67.03,
72.13, 115.77, 138.34, 151.95, 154.29, 157.82, 168.44,
2354 MOLECULAR PHARMACEUTICS VOL. 7, NO. 6

Peptidomimetic Conjugates of Cyclic HPMPC and HPMPA
brief articles
170.94. The NMR chemical shift values are in agreement
with the literature values:19 LC-MS tR ) 19.15 min; [M +
H]+ 424.3.
6, and in the presence or in the absence of 10 mM Gly-Sar,
cHPMPC (1), L-Val-L-Ser-OMe cHPMPC (2a), D-Val-D-Ser-
OMe cHPMPC (2d), L-Ser-OiPr cHPMPC (3a), D-Ser-OiPr
cHPMPC (3b), L-Ser-OMe cHPMPA (6a), and D-Ser-OMe
cHPMPA (6b), L-Val ACV (7), or L-Val-L-Val ACV (8).
After incubation, oocytes were rinsed thoroughly with ice
cold Na+ 7.5 buffer, individually lysed with 5% sodium
dodecyl sulfate, and assayed for radioactivity in commercial
scintillation cocktail (Ultima Gold, PerkinElmer, Waltham,
MA). Unless otherwise noted, data are shown as mean (
SEM of at least 3 experiments from different frogs, with
5-8 oocytes each. At the concentrations used, none of the
compounds had an effect on Gly-Sar uptake in noninjected
oocytes (data not shown).
Functional Expression of hPEPT1 in Oocytes. Mature
female Xenopus laeVis were purchased from Nasco (Fort
Atkinson, WI). All animal protocols followed guidelines
approved by the University of California Chancellor’s
Committee on Animal Research and the National Institutes
of Health. Frogs were anesthetized with 0.1% tricaine (Sigma
Chemical Company, St. Louis, MO) buffered with 0.1%
NaHCO3, a portion of the ovary was surgically removed,
and the frogs were sacrificed by an overdose of nembutal
(60 mg for 60 min). Stage V-VI oocytes were selected and
maintained at 18 °C in modified Barth’s solution, supple-
mented with antibiotics.13 hPEPT1 plasmids were linearized
with BamHI (New England Biolabs, Ipswich, MA) and
transcribed in vitro using the T7MEGAScript kit and RNA
cap analogue (Ambion, Austin, TX), and the cRNAs were
prepared as described.20 Oocytes were injected with 50 ng
of hPEPT1 cRNA one day after isolation, and incubated at
18 °C for 4-7 days postinjection. Experiments were
performed at 20-22 °C. Noninjected oocytes served as
controls.
Electrophysiology. A two-microelectrode voltage clamp
system was used to determine the interactions between
hPEPT1, glycylsarcosine (Gly-Sar) and the prodrugs in
oocytes expressing the transporter.13 Oocytes were super-
fused in a buffer containing (in mM) 100 NaCl, 2 KCl, 1
MgCl2, 1 CaCl2, and 10 Hepes/Tris (pH 7.5) or 10 2-(N-
morpholino)ethanesulfonic acid (Mes)/Tris (pH 6.0), and the
membrane potential was held at -50 mV. Continuous current
data were acquired with a chart recorder, and the current
responses to the addition of 0.5-10 mM of the hPEPT1
substrate glycylsarcosine (Gly-Sar) and/or 10 mM of the
following compounds were recorded: L-valine (L-Val), L-
serine (L-Ser), cHPMPC (1), L-Val-L-Ser-OMe cHPMPC
(2a), L-Val-D-Ser-OMe cHPMPC (2b), D-Val-L-Ser-OMe
cHPMPC (2c), D-Val-D-Ser-OMe cHPMPC (2d), L-Ser-OiPr
cHPMPC (3a), D-Ser-OiPr cHPMPC (3b), L-Ser-L-Ala-OiPr
cHPMPC (4), L-Ser-OiPr cHPMPA (5), L-Ser-OMe cHP-
MPA (6a), D-Ser-OMe cHPMPA (6b), L-Val ACV (7) and
L-Val-L-Val ACV (8). At the concentrations used, none of
the compounds induced currents in control oocytes (data not
shown). Unless otherwise noted, experiments were repeated
on at least three oocytes from two different donor frogs.
Gly-Sar Uptake Assays. Oocytes were incubated for 30
min in the presence of 5 µM Gly-Sar (0.1 µM [3H]-Gly-
Sar) in buffer containing (in mM) 100 NaCl, 2 KCl, 1 MgCl2,
1 CaCl2 and 10 N-(2-hydroxyethyl)piperazine-N′-(2-ethane-
sulfonic acid) (Hepes)/Tris (pH 7.5) or in Na+ buffer at pH
Results and Discussion
Chemistry. The synthesis of 2-4 required gram amounts
of HPMPC, which was prepared as previously described.4,5
The Boc-protected dipeptide promoieties 11a-d and 13 were
prepared using EDC coupling, whereas the amino acid
promoieties 12a and 12b were synthesized by acid-catalyzed
esterification and standard Boc-protection of the amino acid.
Conjugates 2-4 were synthesized as described in Scheme
1. The Boc-protected intermediates were prepared using
PyBOP to convert HPMPC to 1 and attach the promoiety in
one pot.4,5 The reaction was easily monitored by 31P NMR,
and when neither HPMPC nor 1 remained, the reaction was
stopped. The PyBOP byproduct and excess promoiety were
removed with a diethyl ether wash before purification by
preparative TLC. Following Boc-deprotection with TFA, the
target compounds 2-4 were purified by preparative TLC.
The bis(trifluoroacetic acid) salts of 2a-d were obtained by
precipitation with diethyl ether from a saturated methanolic
solution (overall yields 21-58%). The structures of all the
1
cHPMPC conjugates were confirmed by H and 31P NMR
and by HRMS. Purity was verified by LC-MS. A pair of
diastereomers is formed (owing to the chiral phosphorus
atom) when the promoiety is conjugated to 1, detectable by
31P NMR. It should be noted that each cHPMPC conjugate
was isolated and evaluated as the diastereomeric mixture.
Thus, pairs of 31P NMR signals are observed, but a single
parent ion is seen in the mass spectrum. The stereoisomer
with the more upfield chemical shift is formed predominantly,
in a ratio of from 7:3 to 4:1 relative to the more downfield
stereoisomer. The latter is relatively less polar, based on its
longer retention on a C-18 LC column.4
The cHPMPA prodrugs were prepared using a solid-phase
method (Scheme 2), wherein the Boc protecting group is
essentially replaced with a resin.21 This new method has the
advantage of easier product purification, avoiding exposure
to silica gel. Briefly, the side chain hydroxyl group of the
(19) Nashed, Y. E.; Mitra, A. K. Synthesis and characterization of novel
dipeptide ester prodrugs of acyclovir. Spectrochim. Acta, Part A
2003, 59A (9), 2033–2039.
(20) Mackenzie, B.; Loo, D. D.; Fei, Y.; Liu, W. J.; Ganapathy, V.;
Leibach, F. H.; Wright, E. M. Mechanisms of the human intestinal
H+-coupled oligopeptide transporter hPEPT1. J. Biol. Chem.
1996, 271 (10), 5430–5437.
(21) Krylov, I. S.; Peterson, L. W.; Kashemirov, B. A.; Breitenbach,
J.; Borysco, K.; Drach, J. C.; Kim, J. S.; Hilfinger, J. M.;
McKenna, C. E. In vitro transport, activation and antiviral
evaluation of new HPMPA prodrugs synthesized on a solid
support. AntiViral Res. 2009, 82 (2), A75.
VOL. 7, NO. 6 MOLECULAR PHARMACEUTICS 2355

brief articles
Peterson et al.
Scheme 1. Synthesis of the Peptidomimetic cHPMPC Conjugates 2-4a
a Reagents and conditions: (a) PyBOP, DIPEA, DMF, 35 °C, 4 h; (b) TFA, CH2Cl2, 35 °C, 5 h.
amino acid was first protected by a tert-butyldimethylsilyl
(TBDMS) group, and the resulting TBDMS-protected serine
alkyl ester (13, 14a and 14b) was then immobilized by
reaction with a tritylchloride polystyrene (TCP) resin. After
removal of the TBDMS protecting group with TBAF, the
coupling reaction to the drug was accomplished using PyBOP
to cyclize HPMPA and conjugate the amino acid, which can
be visualized by 31P gel NMR (Figure 2). The product
conjugates were cleaved from the resin using either TFA in
dichloromethane (17a and 17b) or HCl in dioxane (5).
Purification of 17a and 17b by preparative TLC, followed
by treatment with 3N HCl in MeOH afforded final products
6a and 6b as hydrochloride salts (overall yields 26-30%).
was verified by 1H and 13C NMR, which were in agreement
with the reported values,19 and the purity was confirmed by
LC-MS.
Interactions between hPEPT1 and the Peptidomimetic
cHPMPC and cHPMPA Prodrugs 2-6. Cidofovir prodrugs
were tested for recognition by hPEPT1 using the two-
electrode voltage clamp (TEVC) technique and radiotracer
uptake assays. First, the compounds were examined for
hPEPT1 translocation, based on their ability to generate
inward currents in Xenopus oocytes. The interactions of the
prodrugs with hPEPT1 were further investigated by measur-
ing their effect on the inward currents induced by 0.5 mM
of the known hPEPT1 substrate Gly-Sar (Figure 3B), and
on the uptake of 5 µM Gly-Sar (Figure 4). All prodrugs were
tested at 10 mM; test solutions were prepared immediately
before each assay, and the pH was checked before analysis.
Solutions of 10 mM L-valine (L-Val), L-serine (L-Ser), and
cHPMPC (1) were used as controls.
Figure 3A shows a record from a representative TEVC
experiment, in which an hPEPT1-expressing oocyte voltage-
clamped at -50 mV was superfused with 0.5 mM Gly-Sar
or 10 mM 3a at pH 6. Addition of Gly-Sar resulted in the
generation of an inward H+ current, 100 nA. 3a did not
induce a detectable current, indicating that it is not trans-
ported by hPEPT1 (data not shown). When the oocyte was
1
The structures of all conjugates were confirmed by H and
31P NMR and by HRMS. Purity was verified by LC-MS.
As demonstrated by their 31P NMR spectra, like the cHPMPC
conjugates they are obtained as diastereomeric mixtures,
however the major stereoisomer (which is present only in
small excess) is that one having the more downfield 31P
resonance. The cHPMPA conjugates were prepared as
hydrochloride salts because the reduced basicity of the
adenine ring appears to make the TFA salts less stable.
L-Val-L-Val ACV (8) was synthesized according to the
literature procedure,19 and obtained as a trifluoroacetic acid
salt, which was purified by preparative TLC. The structure
2356 MOLECULAR PHARMACEUTICS VOL. 7, NO. 6

Peptidomimetic Conjugates of Cyclic HPMPC and HPMPA
brief articles
Scheme 2. Solid-Phase Synthesis of Peptidomimetic cHPMPA Conjugatesa
a Reagents and conditions: (a) TBDMSCl, imidazole, CH2Cl2, 0 °C to rt, 24 h; (b) TCP-resin, DIPEA, CH2Cl2, 0 °C to rt, 18 h; (c) TBAF, THF, rt, 5 h; (d)
PyBOP, DIPEA, DMF, 38 °C, 18 h; (e) TFA, CH2Cl2 and 3N HCl in MeOH or HCl, dioxane, rt, 18 h. Isolated products are HCl salts.
evoke an inward current, indicating they are not transported
by hPEPT1 (data not shown). Compounds 2a and 2b, both
N-terminal L-valine dipeptide conjugates, abolished 100%
and 93% of the Gly-Sar current, respectively, while the
D-valine analogues 2c and 2d decreased the Gly-Sar current
by only about 40% (Figure 3B).
These results demonstrate the importance of an L-config-
uration at the N-terminal amino acid for hPEPT1 affinity,
whereas the configuration of the C-terminal, drug-linked
serine has little influence on binding to the transporter.
Despite the occurrence of interaction between hPEPT1 and
2a and 2b, it appears that the steric profile of these prodrugs,
Figure 2. Gel 31P NMR of immobilized 5 prior to
cleavage from solid-phase support.
which approach tetrapeptides in size, and/or their polarity,
prevents transport of the bound molecule.23
In order to better define structural effects on the interac-
tions of serine side chain-linked peptidomimetic prodrugs
with hPEPT1, we prepared the analogues 3-6. First, the
effect of removing the nonlinking, second amino acid from
the dipeptide moiety, as in the cHPMPC conjugates 3a and
3b, was ascertained. These compounds did not evoke an
inward current, indicating they are not substrates for hPEPT1,
and inhibited the Gly-Sar current only by about 50% (Figure
3B). In addition, the difference between the L- and D-isomers
was small, most likely due to fewer points of bonding
interaction with the transporter.
exposed to 0.5 mM Gly-Sar, the addition of 10 mM 3a
produced a 55% reduction in the current.
Neither L-Val nor L-Ser evoked detectable currents, and
both had little effect on Gly-Sar transport (Figure 3B), as
expected since amino acids are not hPEPT1 substrates.22
Cyclic cidofovir (1) was not transported and inhibited Gly-
Sar transport weakly (Figures 3B and 4), suggesting poor
recognition by hPEPT1. cHPMPC conjugates 2a-d did not
(22) Liang, R.; Fei, Y. J.; Prasad, P. D.; Ramamoorthy, S.; Han, H.;
Yangfeng, T. L.; Hediger, M. A.; Ganapathy, V.; Leibach, F. H.
Human intestinal H+/peptide cotransporter - cloning, functional
expression, and chromosomal localization. J. Biol. Chem. 1995,
270 (12), 6456–6463.
(23) Rubio-Aliaga, I.; Daniel, H. Peptide transporters and their roles
in physiological processes and drug disposition. Xenobiotica 2008,
38 (7-8), 1022–1042.
VOL. 7, NO. 6 MOLECULAR PHARMACEUTICS 2357

brief articles
Peterson et al.
Figure 4. Effect of prodrugs on Gly-Sar uptake.
Uptake of 5 µM Gly-Sar (0.1 µM [3H]-Gly-Sar) was
measured in Na+ buffer at pH 7.5 (gray bar) or pH 6
(black bars), in the absence or in the presence of 10
mM Gly-Sar, cHPMPC (1), L-Val-L-Ser-OMe cHPMPC
(2a), D-Val-D-Ser-OMe cHPMPC (2d), L-Ser-OiPr
cHPMPC (3a), D-Ser-OiPr cHPMPC (3b), L-Ser-OMe
cHPMPA (6a), and D-Ser-OMe cHPMPA (6b), L-Val
ACV (7), or L-Val-L-Val ACV (8). Results are shown
as means
( SEM of 5-8 oocytes and were
normalized to the Gly-Sar uptake in the absence of
external inhibitors (control), 15 ( 1 pmol h-1 oocyte-1
(mean ( SEM of 4 experiments with different donor
frogs). Uptake in noninjected oocytes was 0.6% (
0.1% of control, regardless of pH and was not
affected by any of the test compounds. Compounds 1
to 8: data are from individual experiments, but were
confirmed in at least one additional trial with oocytes
from a different frog.
Figure 3. Interactions between prodrugs and hPEPT1.
Four to seven days postinjection with hPEPT1
cRNA, X. laevis oocytes were mounted in
a
two-microelectrode voltage clamp setup, superfused
with Na+ buffer at pH 6, and held at -50 mV.
Currents in response to the addition of 0.5 mM of the
hPEPT1 substrate glycylsarcosine (Gly-Sar) and/or
10 mM of the following compounds were recorded:
L-Val, L-Ser, cHPMPC (1), L-Val-L-Ser-OMe cHPMPC
(2a), L-Val-D-Ser-OMe cHPMPC (2b), D-Val-L-Ser-
OMe cHPMPC (2c), D-Val-D-Ser-OMe cHPMPC (2d),
L-Ser-OiPr cHPMPC (3a), D-Ser-OiPr cHPMPC (3b),
L-Ser-L-Ala-OiPr cHPMPC (4), L-Ser-OiPr cHPMPA
(5), L-Ser-OMe cHPMPA (6a), and D-Ser-OMe
cHPMPA (6b). (A) Individual trace for Gly-Sar and 3a.
(B) Effect of the compounds on inward currents
evoked by Gly-Sar. Data were normalized to the
current due to 0.5 mM Gly-Sar (control), 112 ( 4 nA
(mean ( SEM of at least 5 measurements in 3
oocytes from different frogs), and represent the
percentage of Gly-Sar current that remained upon
addition of each compound. Results are from
individual oocytes (L-Val, L-Ser, 1, and prodrugs 2 to
5); all observations were confirmed in at least one
additional oocyte from a different batch. For 6a and
6b, data are shown as mean ( SEM of three oocytes
from two donor frogs.
A possible explanation of this effect could be that HPLC
purified samples contain an excess of TFA compared to those
purified by TLC. This conclusion is based on 19F NMR and
combustion analysis for CHN, which support the presence
of excess TFA in the HPLC-purified preparations. Thus, to
avoid this artifact, we recommend that compounds tested in
the oocyte system be carefully purified to remove excess
TFA, or alternatively, that HCl salts be used.
We also prepared and evaluated in the model three single
serine side chain-linked conjugates of the adenine analogue
of cHPMPC, cHPMPA: 5 (L-Ser-OiPr), 6a (L-Ser-OMe), and
6b (D-Ser-OMe). These compounds also did not evoke an
inward current and thus were not transported. However, they
inhibited Gly-Sar currents (Figure 3B) to a higher extent than
3a and 3b, suggesting that the larger size of the drug
heterocycle does not greatly affect interaction with the
transporter. In the L/D stereoisomer pair 6a and 6b, inhibition
of the Gly-Sar current by the conjugate with an L-configu-
ration at the serine was modestly more pronounced than with
the corresponding cHPMPC conjugates, 3a/3b (Figure 3B).
We next prepared the “reversed” dipeptide cHPMPC
conjugate 4 to investigate the effect of the linking serine
position in the dipeptide sequence. Compound 4 again
evoked no inward current, indicating absence of translocation
by hPEPT1, but it abolished the current induced by Gly-
It is important to mention that weak inward currents were
produced in the model when 3a was purified by HPLC using
0.1% TFA followed by lyophilization (data not shown).
Nevertheless, these signals were not reproducible and the
same compound, purified by preparative TLC, gave no signal.
2358 MOLECULAR PHARMACEUTICS VOL. 7, NO. 6

Peptidomimetic Conjugates of Cyclic HPMPC and HPMPA
brief articles
Sar, suggesting good recognition by the transporter. These
observations show that the position of the linked L-serine
does not affect hPEPT1 binding or improve the ability of
the prodrug to undergo transport.
Next, we evaluated the effect of 1, 2a, 2d, 3a/3b and 6a/
6b on the uptake of radiolabeled Gly-Sar; results are
summarized in Figure 4. In hPEPT1 oocytes, influx of 5 µM
Gly-Sar (0.1 µM [3H]-Gly-Sar) at pH 6 was 15 ( 1 pmol
h-1 oocyte-1 (control), 150-fold higher than in noninjected
oocytes, and dropped 98% in the presence of 10 mM
unlabeled Gly-Sar. At pH 7.5, Gly-Sar uptake was 20% of
control at 3 ( 0.2 pmol h-1 oocyte-1 (Figure 4, gray bar),
indicating that hPEPT1 has a limited, but significant. ability
to transport Gly-Sar in the absence of an inwardly directed
H+ gradient. hPEPT1-mediated Gly-Sar uptake decreased
75% in the presence of 2a but only ∼30% due to 1 or 2d
(Figure 4), in accordance with poor recognition of the parent
drug by hPEPT1, and further shows stereospecificity in the
interactions between the carrier and our dipeptide prodrugs.
2a and 2b inhibited nearly 100% of the Gly-Sar induced
H+-currents (Figure 3B), but uptake in the presence of 2a
was similar to that at pH 7.5. This may indicate that, while
in the presence of 2a or 2b H+-coupled Gly-Sar transport is
fully blocked (Figure 3B), uncoupled transport of the
dipeptide is still possible. On the other hand, 10 mM 6a and
6b blocked 97% of the 5 µM [3H]-Gly-Sar uptake (Figure
4), but only 70% of the current generated by 0.5 mM Gly-
Sar (Figure 3B). Thus, 6a and 6b seem to be less effectively
recognized by the transporter than 2a and 2b, but when
present at a sufficiently high prodrug:substrate concentration
ratio, they are capable of inhibiting Gly-Sar transport entirely.
Finally, 6a and 6b inhibited [3H]-Gly-Sar uptake more than
3a and 3b (Figure 4), further suggesting a preference of the
transporter for the 6 isomer pair over the 3 isomers.
Figure 5. Transport of acyclovir prodrugs by hPEPT1. X.
laevis oocytes expressing the transporter were
superfused with Na+ buffer at pH 6, and held at -50
mV. White bars: inward currents induced by 10 mM
L-Val ACV (7), 10 mM L-Val-L-Val ACV (8), or 0.5 mM
Gly-Sar. Black bars: effect of 10 mM 7 or 8 on the
currents due to 0.5 mM Gly-Sar. Data were normalized
to the current induced by 10 mM Gly-Sar (Imax), 175 (
18 nA. Results are shown as mean ( SEM of at least
three oocytes from two different frogs.
approximately 3-fold that of 7,24 and is in agreement with
our observations with the cHPMPA and cHPMPC prodrugs
that suggest preference of the transporter for dipeptide drug
conjugates compared to single amino acid analogues.
The effect of ACV conjugates 7 and 8 on Gly-Sar transport
was also investigated. Both compounds inhibited Gly-Sar
uptake (Figure 4), 94% (7) and 87% (8), and reduced Gly-
Sar currents (Figure 5), 80% (7) and 35% (8). Inhibition of
substrate currents by transported drugs has been observed
previously in hPEPT1 and other transporters, such as the
renal H+-coupled oligopeptide cotransporter hPEPT2 and the
Na+/glucose cotransporter hSGLT1.13,25,26 In these, inhibition
is due to the drug being transported at a significantly lower
rate than the substrate, which has the effect of slowing down
the system and reducing the overall current when cosuper-
fused with the substrate. The lower inhibitory effects of 8
on Gly-Sar currents (Figure 5, black bars) are consistent with
its higher apparent transport rate (Figure 5, white bars), and
may be due to the additional amino acid, making it a
dipeptide promoiety and giving more points for interaction
with the transporter. Therefore, the coadministration of 8 does
not interfere with the transport of Gly-Sar as much as that
hPEPT1-Mediated Transport and Affinity of
Peptidomimetic ACV Conjugates. We wished to verify that
transport of a prodrug known to be transporting by hPEPT1
could be observed in our X. laeVis oocyte model system.
For this purpose we chose L-Val ACV (7)9,12 and also
included a dipeptide analogue of this prodrug, L-Val-L-Val
ACV (8). First, we compared the currents induced by 10
mM 7 or 8 to those due to 10 mM Gly-Sar, a saturating
concentration of the dipeptide (Figure 5) that generates the
maximum inward current, Imax. Compounds 7 and 8 induced
currents equal to 6% and 15% Imax, respectively (Figure 5).
This demonstrates that 7 and 8 are transported by hPEPT1,
and suggests that 8 is transported more efficiently. In the
absence of detailed kinetic data, however, it is not possible
to determine whether this is due to a higher affinity, a higher
turnover rate, or both. The larger signal observed with the
dipeptide 8 is consistent with the report of Talluri et al. that
the permeability of 8 across a Caco-2 monolayer was
(24) Talluri, R. S.; Samanta, S. K.; Gaudana, R.; Mitra, A. K. Synthesis,
metabolism and cellular permeability of enzymatically stable
dipeptide prodrugs of acyclovir. Int. J. Pharm. 2008, 361 (1-2),
118–124.
(25) Loo, D. D. F.; Hirayama, B. A.; Sala-Rabanal, M.; Wright, E. M.
How Drugs Interact with Transporters: SGLT1 as a Model. J.
Membr. Biol. 2008, 223 (2), 87–106.
(26) Sala-Rabanal, M.; Loo, D. D. F.; Hirayama, B. A.; Wright, E. M.
Molecular mechanism of dipeptide and drug transport by the
human renal H+/oligopeptide cotransporter hPEPT2. Am. J.
Physiol. 2008, 294 (6, Pt.2), F1422–F1432.
VOL. 7, NO. 6 MOLECULAR PHARMACEUTICS 2359

brief articles
Peterson et al.
Table 1. Calculated log D values at pHs 6.5 and 7.4a
of 7. In turn, 7 is recognized by hPEPT1, but its properties
may not be optimal as it only contains one amino acid, which
results in a lower transport rate.
log D
compound
pH 6.5
pH 7.4
1
2
3
4
5
6
7
8
-3.52
-1.31
-1.32
-1.43
-1.21
-1.97
-1.59
-2.20
-3.52
-1.31
-0.99
-1.43
-0.88
-1.64
-0.92
-1.37
Conclusion
In summary, several peptidomimetic prodrugs incorporat-
ing different promoieties were synthesized and evaluated for
their ability to be transported actively by hPEPT1, using
tracer uptake and electrophysiology in Xenopus oocytes
overexpressing the transporter. Under the conditions studied,
none of the cHPMPC conjugates 2-4 evoked a current,
indicating that there is no detectable hPEPT1-mediated
transport of these compounds. However, they all inhibited
the transport of hPEPT1 substrate Gly-Sar, suggesting that
they are recognized by the transporter. At 10 mM, com-
pounds 2a, 2b, and 4 blocked 93-100% of the current due
to 0.5 mM Gly-Sar, whereas 2c and 2d were weaker
inhibitors. This suggests that, although these compounds are
not hPEPT1 substrates, they may bind to the transporter with
low to high affinity, depending on the stereochemistry at the
terminal amino acid, but without much sensitivity to the
position of the drug-linked serine within the dipeptide
conjugate. The single amino acid cHPMPC conjugates 3a
and 3b showed low inhibitory potency and little stereospeci-
ficity, a pattern reproduced with the analogous cHPMPA
conjugates 5 and 6a,b. Thus affinity for the transporter also
appears to be promoted by the presence of a dipeptide in
the conjugate, which is consistent with the known specificity
of hPEPT1. The ability of the transporter model to detect
prodrug hPEPT1 transport activity was verified using the
known hPEPT1 substrate valacyclovir (7) and its dipeptide
analogue, L-Val-L-Val acyclovir (8). Both compounds were
transported in our model system, but 8 exhibited a higher
transport rate, which may be attributed to its dipeptide
promoiety. The results suggest the L-amino-terminal prodrugs
2a, 2b and 4 are stereospecifically recognized by hPEPT1,
but are not transported due to the steric and/or polar structural
properties of the linked cHPMPC drug cargo. Similar results
were obtained with three single amino acid conjugates of
cHPMPA (5, 6a and 6b).
a Values calculated using Marvin Sketch v. 5.2.2.
does not play a decisive role. Additional studies using acyclic
analogues are needed to determine the effect the ring has on
transport.
Thus, the enhanced absorption of prodrug 2a relative to
the parent drug cannot be explained by active transport
mediated by hPEPT1, although the prodrug is bound and
inhibits Gly-Sar transport, and although this interaction with
the transporter exhibits specificitiessdipeptide over single
amino acid, N-terminal residue in L-configurationssimilar
to those displayed by known substrates. This raises the
question whether some other active transporter, perhaps
unknown, could be responsible. To assess the alternative of
passive transport, a qualitative27 estimate of permeability can
be made by comparing the log D values of the drugs and
prodrugs studied here (Table 1). Keeping in mind the
limitations of this and other single parameter permeability
estimators,27 it can be seen that the prodrugs have log D
values that are about 2 logs more positive than HPMPC, and
similar to those of valacyclovir and its dipeptide derivative,
which are actively transported by hPEPT1. Further work will
be necessary to establish whether an as yet unknown active
process augments passive transport of prodrugs such as 2a,
and whether replacement of HPMPC or HPMPA by a
different drug cargo would allow the peptidomimetic to not
merely bind, but also be transported by, hPEPT1.
Abbreviations Used
HPMPC, cidofovir; cHPMPC, cyclic cidofovir; HPMPA,
9-[(2S)-3-hydroxy-2-phosphonomethoxypropyl]adenine; ACV,
acyclovir; DCC, dicyclohexylcarbodiimide; DCM, dichlo-
romethane; DMAP, N,N-dimethylaminopyridine; PyBOP,
(benzotriazol-1-yloxy)tripyrrolidinophosphonium hexafluo-
rophosphate; DIPEA, diisopropylethylamine; DMF, N,N-
dimethylformamide; EDC, N-(3-dimethylaminopropyl)-N′-
ethylcarbodiimide; HOBt, 1-hydroxybenzotriazole; SAR,
structure-activity relationship; STR, structure-transport
relationship; TBAF, tetrabutylammonium fluoride; TCP,
tritylchloride polystyrene resin; TEA, triethylamine; TFA,
trifluoroacetic acid.
It is interesting that the point of attachment of the large
drug side chain within the dipeptide did not observably affect
the potent affinity of the L-Val-L-Ser prodrug conjugates. It
should be noted that the nucleoside base (G) in valacyclovir
is larger than that in our HPMPC and HPMPA prodrugs;
the sugar-mimicking moiety is smaller and is linked differ-
ently to the amino acid, via esterification of the terminal
hydroxy group of the drug to the valine carboxylic acid,
instead of by formation of a phosphonic ester bond with a
serine side chain, as in our compounds. The molecular
rigidity introduced by the presence of the cyclic phosphonate
ring may also affect transport by hPEPT1, but because the
compounds were tested in diastereomeric pairs (racemic at
the chiral phosphorus), the absence of a signal indicating
proton-coupled transport implies that this stereochemistry
(27) Dressman, J. B.; Thelen, K.; Jantratid, E. Towards quantitative
prediction of oral drug absorption. Clin. Pharmacokinet. 2008,
47 (10), 655–667.
2360 MOLECULAR PHARMACEUTICS VOL. 7, NO. 6

Peptidomimetic Conjugates of Cyclic HPMPC and HPMPA
brief articles
1
Acknowledgment. This work was supported by Na-
tional Institutes of Health Grants AI061457 and AI056864
to C.E.M. L.W.P. was a 2007-2008 Stauffer Fellow and a
2008-2009 WiSE Merit Fellow. The authors wish to thank
Professor Ernest M. Wright (UCLA Department of Physiol-
ogy) for kindly making it possible to conduct the voltage-
clamped X. laeVis oocyte assays in his laboratory, and for
helpful advice.
Supporting Information Available: H NMR spectra of
the intermediates: compounds 9a, 9b, 10, 11a-d, 12a, 12b,
13, 14, 15a, and 15b. 1H and 31P NMR spectra of final products:
2a-d, 3a and 3b, 4, 5, 6a, 6b, 17a, and 17b. 1H and 13C NMR
spectra of compound 8. LC-MS traces for the target com-
pounds: 2a, 2d, 3a, 4, 5, 8, 17a, and 17b. This material is
available free of charge via the Internet at http://pubs.acs.org.
MP100186B
VOL. 7, NO. 6 MOLECULAR PHARMACEUTICS 2361