Topical iontophoretic delivery of ionizable, biolabile aciclovir prodrugs: A rational approach to improve cutaneous bioavailability

Article abstract of DOI:10.1016/j.ejpb.2015.12.002

The objective was to investigate the topical iontophoretic delivery of a series of amino acid ester prodrugs of aciclovir (ACV-X, where ACV = aciclovir and X = Arg, Gly, Ile, Phe, Trp and Val) as a means to enhance cutaneous delivery of ACV. The newly synthesized prodrugs were characterized by ¹H NMR and high resolution mass spectrometry. Analytical methods using HPLC-UV were developed for their quantification and each method was validated. Investigation of solution stability as a function of pH showed that all ACV-X prodrugs were relatively stable in acid conditions at pH 2.0 and pH 5.5 for up to 8 h but susceptible to extensive hydrolysis at pH 7.4 and under alkaline conditions (pH 10). No ACV-X hydrolysis was observed after contact for 2 h with the external surface of porcine stratum corneum. However, there was significant hydrolysis following contact with the dermal surface of dermatomed porcine skin, in particular, for ACV-Arg. Passive transport of ACV and ACV-X prodrugs from aqueous solution after 2 h was below the limit of detection. Iontophoresis of ACV at 0.5 mA/cm² for 2 h led to modest ACV skin deposition (Q_DEP,ACV) of 4.6 ± 0.3 nmol/cm². In contrast, iontophoresis of ACV-X prodrugs under the same conditions produced order of magnitude increases in cutaneous deposition of ACV species, that is, Q_DEP,TOTAL = Q_DEP,ACV + Q_DEP,ACV-X. Q_DEP,TOTAL for ACV-Gly, ACV-Val, ACV-Ile, ACV-Phe, ACV-Trp and ACV-Arg was 412.8 ± 44.0, 358.8 ± 66.8, 434.1 ± 68.2, 249.8 ± 81.4, 156.1 ± 76.3, 785.9 ± 78.1 nmol/cm², respectively. The extent of bioconversion of ACV-X to ACV in the skin was high and the proportion of ACV present ranged from 81% to 100%. The skin retention ratio, a measure of the selectivity of ACV species for deposition over permeation after iontophoretic delivery of ACV-X prodrugs, was dependent on both the rate of transport and the susceptibility to hydrolysis of the prodrugs. Skin deposition of ACV and its six prodrugs were investigated further as a function of current density (0.125, 0.25 and 0.5 mA/cm²); the effect of duration of current application (5, 10, 30, 60 and 120 min) was evaluated using ACV-Arg and ACV-Ile. Iontophoresis of ACV-Arg and ACV-Ile at 0.25 mA/cm² for only 5 min resulted in the deposition of appreciable amounts of ACV (36.4 ± 5.7 nmol/cm² and 40.3 ± 6.1 nmol/cm², respectively), corresponding to supra-therapeutic average concentrations in skin against HSV-1 or HSV-2. The results demonstrated that cutaneous bioavailability of ACV could be significantly improved after short-duration iontophoresis of ionizable, biolabile ACV-X prodrugs.

Full text of DOI:10.1016/j.ejpb.2015.12.002

European Journal of Pharmaceutics and Biopharmaceutics xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
European Journal of Pharmaceutics and Biopharmaceutics
journal homepage: www.elsevier.com/locate/ejpb
Research Paper
Topical iontophoretic delivery of ionizable, biolabile aciclovir prodrugs:
A rational approach to improve cutaneous bioavailability
⇑
Yong Chen, Ingo Alberti, Yogeshvar N. Kalia
School of Pharmaceutical Sciences, University of Geneva & University of Lausanne, 30 Quai Ernest Ansermet, 1211 Geneva, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history:
The objective was to investigate the topical iontophoretic delivery of a series of amino acid ester prodrugs
of aciclovir (ACV-X, where ACV = aciclovir and X = Arg, Gly, Ile, Phe, Trp and Val) as a means to enhance
cutaneous delivery of ACV. The newly synthesized prodrugs were characterized by 1H NMR and high res-
olution mass spectrometry. Analytical methods using HPLC–UV were developed for their quantification
and each method was validated. Investigation of solution stability as a function of pH showed that all
ACV-X prodrugs were relatively stable in acid conditions at pH 2.0 and pH 5.5 for up to 8 h but susceptible
to extensive hydrolysis at pH 7.4 and under alkaline conditions (pH 10). No ACV-X hydrolysis was
observed after contact for 2 h with the external surface of porcine stratum corneum. However, there
was significant hydrolysis following contact with the dermal surface of dermatomed porcine skin, in par-
ticular, for ACV-Arg. Passive transport of ACV and ACV-X prodrugs from aqueous solution after 2 h was
below the limit of detection. Iontophoresis of ACV at 0.5 mA/cm2 for 2 h led to modest ACV skin
deposition (QDEP,ACV) of 4.6 0.3 nmol/cm2. In contrast, iontophoresis of ACV-X prodrugs under the same
conditions produced order of magnitude increases in cutaneous deposition of ACV species, that is,
Received 29 September 2015
Revised 27 November 2015
Accepted in revised form 1 December 2015
Available online xxxx
Keywords:
Aciclovir
Prodrug
Stability
Iontophoresis
Skin
Skin retention
Bioconversion
QDEP,TOTAL = QDEP,ACV + QDEP,ACV-X. QDEP,TOTAL for ACV-Gly, ACV-Val, ACV-Ile, ACV-Phe, ACV-Trp and
ACV-Arg was 412.8 44.0, 358.8 66.8, 434.1 68.2, 249.8 81.4, 156.1 76.3, 785.9 78.1 nmol/cm2,
respectively. The extent of bioconversion of ACV-X to ACV in the skin was high and the proportion of
ACV present ranged from 81% to 100%. The skin retention ratio, a measure of the selectivity of ACV species
for deposition over permeation after iontophoretic delivery of ACV-X prodrugs, was dependent on both
the rate of transport and the susceptibility to hydrolysis of the prodrugs. Skin deposition of ACV and
its six prodrugs were investigated further as
a function of current density (0.125, 0.25 and
0.5 mA/cm2); the effect of duration of current application (5, 10, 30, 60 and 120 min) was evaluated using
ACV-Arg and ACV-Ile. Iontophoresis of ACV-Arg and ACV-Ile at 0.25 mA/cm2 for only 5 min resulted in the
deposition of appreciable amounts of ACV (36.4 5.7 nmol/cm2 and 40.3 6.1 nmol/cm2, respectively),
corresponding to supra-therapeutic average concentrations in skin against HSV-1 or HSV-2. The results
demonstrated that cutaneous bioavailability of ACV could be significantly improved after short-
duration iontophoresis of ionizable, biolabile ACV-X prodrugs.
Ó 2015 Elsevier B.V. All rights reserved.
1. Introduction
displays higher oral bioavailability (30–50%) [2], gram amounts of
VCV have to be administered orally (1000 mg three times a day) to
Aciclovir (ACV, 9-[(2-hydroxyethoxy)methyl])-9H-guanine,
ZoviraxÒ) is commonly used to treat herpes labialis, but its poor
and variable oral bioavailability (10–20%) decreases efficacy. This
has led to the development of valaciclovir (VCV, ValtrexÒ), the
treat an essentially local disorder. Efficient topical ACV delivery
should enable targeted therapy, reduce circulating drug levels
and attenuate the risk of renal insufficiency [3].
However, despite the therapeutic rationale and high specific
activity against HSV-1 and HSV-2 in virus-infected cells, clinical
studies of several topical ACV formulations against herpes labialis
have given mixed results. Although a reduction in healing time
was reported with 5% ACV cream [4], clinical failures were also
seen [5]. Indeed, ACV ointment (5%) seemed to have less or even
no clinical benefit [6,7]. Successful topical treatment relies on
effective ACV delivery through intact skin since the epidermal
L-valyl ester of aciclovir, which is a substrate for the mammalian
intestinal peptide transporters, PEPT1 and PEPT2 [1]. Although this
⇑
Corresponding author at: School of Pharmaceutical Sciences, University of
Geneva, 30 Quai Ernest Ansermet, 1211 Geneva 4, Switzerland. Tel.: +41 22 379
3355; fax: +41 22 379 3360.
E-mail address: yogi.kalia@unige.ch (Y.N. Kalia).
http://dx.doi.org/10.1016/j.ejpb.2015.12.002
0939-6411/Ó 2015 Elsevier B.V. All rights reserved.
Please cite this article in press as: Y. Chen et al., Topical iontophoretic delivery of ionizable, biolabile aciclovir prodrugs: A rational approach to improve
cutaneous bioavailability, Eur. J. Pharm. Biopharm. (2015), http://dx.doi.org/10.1016/j.ejpb.2015.12.002

2
Y. Chen et al. / European Journal of Pharmaceutics and Biopharmaceutics xxx (2015) xxx–xxx
pathology in the early stages of viral infection does not alter stra-
tum corneum permeability [6]. Drug penetration must be sufficient
to achieve therapeutic concentrations in the basal epidermis,
where the virus is found. Furthermore, drug therapy should be ini-
tiated as early as possible, e.g. in the prodrome or erythematous
lesion stages [7–9]. Initiation of therapy more than 12 h after the
appearance of symptoms, when the classical lesions such as
papule, vesicle, ulcer or crust have been established, was observed
to be ineffective and did not provide any benefit over the natural
resolution of the infection [7,9,10].
Effective treatment of herpes labialis needs early intervention
and rapid delivery of sufficient drug to not only shorten the dura-
tion of the episode [9] but also to decrease virus entry into sensory
neurons and the copy number of latent genome cloned. A smaller
reservoir of latent virus would decrease its ability to reactivate
periodically from latency [11]. However, ACV is not a good candi-
date for passive topical delivery. Its polarity (log DpH7 = ꢀ1.76)
and sparing solubility in both aqueous and lipid media render its
formulation difficult and limit its partitioning into the stratum cor-
neum. A slow rate of delivery and poor bioavailability result in
delayed antiviral intervention in the basal epidermis and sub-
inhibitory concentrations which translate into poor clinical efficacy
[12].
Iontophoresis is a technique that involves the application of a
mild electric current and can be used to substantially increase
the transport of molecules into and across the skin [13,14]. It is
best-suited to water soluble, charged molecules and ACV (pKa1
2.27; pKa2 9.25) is uncharged under physiologically acceptable
conditions and has poor aqueous solubility, ꢁ0.2% at 25 °C [15] –
thus, it is far from being an ideal candidate for iontophoresis.
Although cutaneous delivery of therapeutic amounts of ACV by
iontophoresis has been reported, it could only be achieved with
formulations at pH 3 or 11 [16,17], which are unsuitable for human
use. In contrast, we have previously shown that anodal iontophore-
sis of VCV, with a formulation pH of 5.24 or 5.65, was able to pro-
duce order of magnitude increases in ACV delivery [18].
Iontophoretic transport was facilitated by the positively charged
valyl group, which was then cleaved by esterases to release ACV
and valine during cutaneous transport.
The objective of the project was to investigate the electrotrans-
port of a series of amino acid ester prodrugs of ACV (ACV-X, where
X = Arg, Gly, Ile, Phe, Trp and Val) and to determine the influence of
the different amino acid components on ACV delivery. The prodrug
strategy has been widely used to develop topical and transdermal
formulations – most routinely in the case of corticosteroid esters.
However, in the case of passive administration, the objective is to
increase molecular lipophilicity, for example, by masking hydroxyl
Fig. 1. Schematic representations illustrating the different rationale for using
prodrugs to improve (a) passive and (b) iontophoretic transport into and across the
skin. (c) Water soluble amino acid ester prodrugs of aciclovir (ACV-X) containing
ionized moieties benefit from electromigration as the principal transport mecha-
nism. They are hydrolyzed in the skin to release ACV and the amino acid. The rate of
transport and the rate of hydrolyze influence the biodistribution of the prodrug and
the relative amounts permeated and retained within the membrane.
groups at the 16
(e.g. triamcinolone acetonide) or through esterification of the
hydroxyl groups at the 17 and 21 positions with short chain car-
a and 17a positions through acetonide formation
a
aqueous solution as a function of pH and in the presence of skin,
(iv) to compare passive and iontophoretic transport by measuring
skin deposition and cumulative permeation of ACV species (that is,
ACV and each ACV-X prodrug) and to evaluate skin retention and
total bioconversion ratios, and (v) to optimize iontophoretic
conditions for milder and shorter current application.
boxylic acids (e.g. betamethasone dipropionate) (Fig. 1a). Con-
versely, in the case of these ‘‘iontophoretic” prodrugs, the aim
was to introduce polar ionizable moieties that increase
hydrophilicity, aqueous solubility and most importantly, impart a
charge to the molecule enabling it to electromigrate into the skin
and so benefit more efficiently from iontophoresis (Fig. 1b). In this
case, the hypothesis was that the presence of the positively
charged moieties in the ACV-X prodrugs would improve their
aqueous solubility and facilitate their electromigration into the
skin, where ACV would be released following enzymatic cleavage
by cutaneous esterases; this would enable greater amounts of
ACV to be delivered more rapidly to the skin (Fig. 1c).
2. Materials and methods
2.1. Materials
ACV was purchased from Sequoia Research Products Ltd.
(Pangbourne, UK). N,N0-Diisopropylcarbodiimide (DIC) was
purchased from TCI (Tokyo, Japan). Boc-Gly-OH, Boc-Ile-OH,
Boc-Trp-OH, Boc-Val-OH and 1-hydroxybenzotriazole hydrate
(HOBt) were supplied by Sigma–Aldrich (Buchs, Switzerland).
The specific aims of this study were (i) to synthesize, purify and
characterize the amino acid ester prodrugs, (ii) to develop and val-
idate robust HPLC–UV methods to quantify ACV and each prodrug
simultaneously, (iii) to determine the stability of the prodrugs in
Please cite this article in press as: Y. Chen et al., Topical iontophoretic delivery of ionizable, biolabile aciclovir prodrugs: A rational approach to improve
cutaneous bioavailability, Eur. J. Pharm. Biopharm. (2015), http://dx.doi.org/10.1016/j.ejpb.2015.12.002

Y. Chen et al. / European Journal of Pharmaceutics and Biopharmaceutics xxx (2015) xxx–xxx
3
Boc-Phe-OH was obtained from Acros (Hungary). Boc-Arg(Boc)2-OH
was purchased from Bachem AG (Bubendorf, Switzerland). 4-
solid. De-protection of the intermediates was carried out using a
1:1 mixture of dichloromethane and trifluoroacetic acid (30 ml)
in an ice bath for 2 h. After removal of the solvent by rotary evap-
oration, the resulting oil was then added dropwise to vigorously
stirred cold diethyl ether (20 ml) and ACV-X prodrug was immedi-
ately precipitated. The precipitate was filtered with a Büchner fun-
nel and then washed with cold diethyl ether (20 ml). The filter cake
was dissolved in 10 ml of H2O and lyophilized for 30 h using an
AlphaÒ 2–4 LD Plus freeze dryer (Martin Christ GmbH; Osterode
am Harz, Germany). A hygroscopic, fluffy white solid with a purity
of >95% (HPLC–UV, wavelength 254 nm) was obtained. The 1H
NMR spectrum of each prodrug was recorded on a Varian Gemini
NMR spectrometer (300 MHz) at ambient temperature. The high
resolution mass spectrum of each prodrug was acquired using an
API QSTARÒ Pulsar, ESI-Qtof mass spectrometer (Applied Biosys-
tems; Rotkreuz, Switzerland).
(dimethylamino)
pyridine
(DMAP)
and
2-morpholino-
ethanesulfonic acid monohydrate (MES) were obtained from Fluka
(Buchs, Switzerland). ULC/MS grade acetonitrile (ACN) and PVC
tubing (ID 3.17 mm; OD 4.97 mm) for preparing salt bridges were
bought from VWR International AG (Dietikon, Switzerland). ULC/MS
grade formic acid (FA) was purchased from Biosolve (Valkenswaard,
Netherlands). Silver wire and silver chloride used for making
electrodes were bought from Sigma–Aldrich (St. Louis, US). All aque-
ous solutions were prepared using Milli-QÒ water (resistiv-
ity > 18 MX cm). All other chemicals were at least of analytical grade.
2.2. Prodrug synthesis
I. (ACV): 1H NMR (DMSO-d6, 300 MHz): d 3.46(m, 4 H, 12-H, 13-
H), 4.67(s, 1 H, 13-OH), 5.34 (s, 2 H, 10-H), 6.46(s, 2 H, 2-NH2), 7.81
(s, 1 H, 8-H), 10.72(s, 1 H, 1-NH). HRESI-MS (C8H11N5O3+H): Calc.
226.0935; Exp. 226.0931.
Amino acid ester prodrugs of ACV – ACV-Arg, ACV-Gly, ACV-Ile,
ACV-Phe, ACV-Trp and ACV-Val, were prepared by a two-step syn-
thesis as outlined in Scheme 1 starting from ACV using a method
adapted from the literature [15,19,20]. Briefly, N-Boc protected L-
arginine, L-glycine, L-isoleucine, L-phenylalanine, L-tryptophan or
IVa. (ACV-Gly): Yield: 54%. 1H NMR (DMSO-d6, 300 MHz): d 3.71
(m, 2 H, 12-H), 3.80 (s, 2 H, 20-H), 4.25 (m, 2 H, 13-H), 5.37 (s, 2 H,
10-H), 6.61 (s, 2 H, 2-NH2), 7.91 (s, 1 H, 8-H), 8.29 (br s, 3 H, 20-
NH+3), 10.92 (s, 1 H, 1-NH). HRESI-MS (C10H15N6O4+H): Calc.
283.1149; Exp. 283.1141.
L
-valine (3.55 ꢂ 10ꢀ3 mol), DIC (0.48 g, 3.80 ꢂ 10ꢀ3 mol) and HOBt
(0.48 g, 3.55 ꢂ 10ꢀ3 mol) were dissolved in DMF (10 ml) under N2.
The mixture was stirred for 2 h at 0 °C, and was then added drop-
wise to a solution of ACV (0.8 g, 3.55 ꢂ 10ꢀ3 mol) and DMAP
(0.005 g, 0.4 ꢂ 10ꢀ3 mol) in DMF (110 ml) at room temperature.
The solution was stirred continuously and warmed up to 50 °C.
The reaction ran overnight. DMF was removed by rotary evapora-
tion using a RotavaporÒ (Büchi Labortechnik AG; Flawil, Switzer-
land) and the residue was placed on a silica gel column and
eluted with 8:1:0.1–15:1:0.1 mixture of dichloromethane, metha-
nol and acetic acid to give the desired intermediate as a white
IVb. (ACV-Val): Yield: 61%. 1H NMR (DMSO-d6, 300 MHz): d 0.88
(m, 6 H, 40-H, 50-H), 2.05(m, 1 H, 30-H), 3.71(m, 2 H, 12-H), 3.88(m,
1 H, 20-H), 4.31(m, 2 H, 13-H), 5.36(s, 2 H, 10-H), 6.52(s, 2 H,
2-NH2), 7.82(s, 1 H, 8-H), 8.18(br s, 3 H, 20-NH3+), 10.72(s, 1 H, 1-
NH). HRESI-MS (C13H21N6O4+H): Calc. 325.1618; Exp. 325.1628.
IVc. (ACV-Ile): Yield: 46%. 1H NMR (DMSO-d6, 300 MHz):
d 0.80(m, 6 H, 50-H, 60-H), 1.24(m, 2 H, 40-H), 1.69(m, 1 H, 30-H),
Scheme 1. Synthesis of amino acid ester prodrugs of ACV.
Please cite this article in press as: Y. Chen et al., Topical iontophoretic delivery of ionizable, biolabile aciclovir prodrugs: A rational approach to improve
cutaneous bioavailability, Eur. J. Pharm. Biopharm. (2015), http://dx.doi.org/10.1016/j.ejpb.2015.12.002

4
Y. Chen et al. / European Journal of Pharmaceutics and Biopharmaceutics xxx (2015) xxx–xxx
3.72(m, 2 H, 12-H), 3.87(m, 1 H, 20-H), 4.32(m, 2 H, 13-H),
5.36(s, 2 H, 10-H), 6.64(s, 2 H, 2-NH2), 7.87(s, 1 H, 8-H), 8.39(br s,
3 H, 20-NH3+), 10.83(s, 1 H, 1-NH). HRESI-MS (C14H23N6O4+H): Calc.
339.1775; Exp. 339.1773.
Table 1
HPLC conditions for quantification of ACV and ACV-X prodrugs.
ACV/ACV-X
Mobile phase
Retention time (min)
LOQ (lM)
ACV
11.2
11.2
Prodrug
IVd. (ACV-Phe): Yield: 31%. 1H NMR (DMSO-d6, 300 MHz): d
3.04(d, 2 H, 30-H), 3.61(m, 2 H, 12-H), 4.18(m, 2 H, 13-H), 4.30(m,
1 H, 20-H), 5.33(s, 2 H, 10-H), 6.55(s, 2 H, 2-NH2), 7.18–7.31(m, 5
H, 50-H, 60-H, 70-H, 80-H, 90-H), 7.83(s, 1 H, 8-H), 8.44(br s, 3 H,
20-NH+3), 10.72(s, 1 H, 1-NH). HRESI-MS (C17H21N6O4+H): Calc.
373.1618; Exp. 373.1621.
ACV & ACV-Gly
0–15.0 min
–
0.2
0.5
99% A + 1% B
8.4
0–15.5 min
99% A, 1% B
11.5
11.5
22.0
25.0
0.5
0.5
ACV-Val
ACV-Ile
15.6–16.5 min
99% A + 1% B to
92% A + 8% B
16.6–27.6 min
92% A + 8% B
27.7–28 min
92% A + 8% B to
99% A + 1% B
28–40 min
IVe. (ACV-Trp): Yield: 35%. 1H NMR (DMSO-d6, 300 MHz): d 3.23
(d, 2 H, 30-H), 3.65(m, 2 H, 12-H), 4.18(m, 2 H, 13-H), 4.24(m, 1 H,
20-H), 5.32(s, 2 H, 10-H), 6.54(s, 2 H, 2-NH2), 7.00–7.46(m, 4 H,
90-H, 100-H, 110-H, 120-H), 7.22(m, 1 H, 50-H), 7.83(s, 1 H, 8-H),
8.35(br s, 3 H, 20-NH3+), 10.71(s, 1 H, 1-NH), 11.06(s, 1 H, 60-NH).
HRESI-MS (C19H22N7O4+H): Calc. 412.1727; Exp. 412.1729.
IVf. (ACV-Arg): Yield: 52%. 1H NMR (DMSO-d6, 300 MHz): d 1.53
(m, 2 H, 40-H), 1.68(m, 2 H, 30-H), 3.13(m, 50-H), 3.71(m, 2 H, 12-H),
4.01(m, 1 H, 20-H), 4.26(m, 2 H, 13-H), 5.37(s, 2 H, 10-H), 6.64(s, 2
H, 2-NH2), 7.32(br s, 4 H, 70-(NH+3)NH), 7.86(s, 1 H, 8-H), 7.93(m, 1
H, 60-H), 8.34(br s, 3 H, 20-NH+3), 10.86(s, 1 H, 1-NH). HRESI-MS
(C10H15N6O4+H): Calc. 382.1945; Exp. 382.1947.
99% A + 1% B
0–15.5 min
99% A, 1% B
ACV-Phe
12.0
25.3
1.0
15.6–16.5 min
99% A + 1% B to
90% A + 10% B
16.6–27.6 min
90% A + 10% B
27.7–28 min
90% A + 10% B to
99% A + 1% B
28–40 min
ACV-Trp
ACV-Arg
12.0
11.0
26.6
9.6
1.0
2.0
2.3. Analytical procedure
An ASI-100 auto-sampler equipped with a P680A LPG-4 pump
and UVD 170U detector (formerly Dionex AG, now Thermo Fisher
Scientific AG; Reinach, Switzerland) was used to quantify
simultaneously the amounts of ACV and each ACV-X prodrug
permeated across and retained within the skin. The system was
controlled by ChromeleonÒ Chromatography Management
Software. The isocratic or gradient separations were achieved
on a LiChrospherÒ (BGB Analytik AG; Boeckten, Switzerland)
99% A + 1% B
0–14.0 min
98.7% A + 1.3% B
Table 2
Precision and accuracy of the analytical methods used to quantify ACV and ACV-X
prodrugs.
C18 reverse-phase analytical column (250 ꢂ 4.0 mm, 5
l
m). A
LiChrospherÒ guard column (10 ꢂ 4.0 mm, 5
lm) with the same
ACV/ACV-X
Theoretical
concentration
Experimental
concentration
Precision
(%)a
Accuracy
(%)b
packing material was mounted upstream from the analytical
column. The column temperature was kept at 30 °C, the flow rate
was maintained at 1.0 ml/min, and the UV absorbance wave-
(
l
M)
10
50
(
lM)
ACV
9.83 0.04
51.34 0.29
102.49 0.68
0.41
0.56
0.66
98.30
102.68
102.49
length was set at 254 nm. The injection volume was 20 ll. The
100
mobile phase comprised phase A, Na2HPO4 buffer (20 mM,
pH 3.0), and phase B, acetonitrile. The composition of the mobile
phase, the retention time for each compound, and the limit of
quantification (LOQ) are shown in Table 1. The methods were
ACV-Gly
ACV-Val
ACV-Ile
ACV-Phe
ACV-Trp
ACV-Arg
10
50
100
9.61 0.06
48.23 0.51
96.39 0.79
0.62
1.05
0.82
96.10
96.46
96.39
10
50
100
9.81 0.05
50.74 0.43
103.43 0.80
0.51
0.85
0.77
98.10
101.48
103.43
validated with 3 replicates at 10, 50 and 100 lM and showed
good intra-day precision and accuracy (Table 2). The specificity
of the analytical methods was confirmed with respect to the
peak resolution between ACV and each ACV-X prodrug, also
between ACV species and endogenous substances extracted from
porcine skin.
10
50
100
10.77 0.07
49.02 0.39
102.56 1.01
0.65
0.79
0.98
107.70
98.04
102.56
10
50
100
9.59 0.10
47.49 0.33
94.23 0.69
1.04
0.69
0.73
95.90
94.98
94.23
2.4. Physicochemical properties
10
50
100
10.65 0.10
52.83 0.51
99.17 0.97
0.94
0.96
0.98
106.50
105.66
99.17
2.4.1. Log D, pKa and aqueous solubility
10
50
100
11.51 0.10
53.77 0.47
108.15 1.03
0.87
0.87
0.95
115.10
107.54
108.15
The log D values at pH 5.5 (log DpH5.5) and the pKa values of ACV
and the ACV-X prodrugs were predicted using ACD Labs ChemS-
ketch software. The aqueous solubility of ACV and the ACV-X pro-
drugs was measured in MES buffer (10 mM, pH 5.5). Excess ACV or
a
Precision = (SD/mean) ꢂ 100.
b
Accuracy = (obtained concentration/theoretical concentration) ꢂ 100.
ACV-X was added to 200
temperature for 2 h. Then, the samples were centrifuged at
12,000 rpm for 15 min. An aliquot (10 l) was taken from the
ll MES buffer and vortexed at ambient
l
2.4.2. Stability
supernatant and diluted 1:1,000,000 by volume with fresh MES
buffer, which was then subjected to HPLC analysis. The extremely
high solubility of ACV-Arg meant that it was not possible to obtain
a saturated solution.
2.4.2.1. Effect of pH. The stability of the ACV-X prodrugs as a func-
tion of pH was investigated by dissolving each prodrug (100 lM)
into 4 buffers at different pH: (i) Glycine buffer at pH 2.0 (10 mM
glycine); (ii) MES buffer at pH 5.5 (10 mM MES monohydrate);
Please cite this article in press as: Y. Chen et al., Topical iontophoretic delivery of ionizable, biolabile aciclovir prodrugs: A rational approach to improve
cutaneous bioavailability, Eur. J. Pharm. Biopharm. (2015), http://dx.doi.org/10.1016/j.ejpb.2015.12.002

Y. Chen et al. / European Journal of Pharmaceutics and Biopharmaceutics xxx (2015) xxx–xxx
5
(iii) Phosphate buffered saline at pH 7.4 (136.9 mM NaCl, 2.7 mM
KCl, 1.7 mM KH2PO4 and 8.1 mM Na2HPO4) and (iv) Na2CO3 buffer
at pH 10.0 (10 mM Na2CO3). Aliquots were withdrawn at regular
time intervals over 2 h. The natural logarithm of the remaining
percentage of each ACV-X prodrug (ln ACV-X%) was plotted as a
function of time (t). The first order rate constant (kobs) of hydrolysis
was estimated from the slope of the regression curve and the cor-
responding half-life (t1/2) calculated (t1/2 = ln2/kobs). All experi-
ments were performed in triplicate at 8 °C.
2.5.3. Sample preparation for analysis by HPLC–UV
A sample (1 ml) was withdrawn from each receiver compart-
ment upon termination of current application. The diffusion cells
were dismantled and the skin was washed with running water to
remove any residual formulation on the surface. The formulation-
exposed area was cut into small pieces and extracted for 3 h with
0.1% formic acid (this was used to deactivate skin esterases). Given
that the ACV-X prodrugs were too polar to penetrate the skin pas-
sively and were relatively unstable in contact with dermis, it was
not possible to validate their extraction efficiency. However, it
was assumed that the extraction method was efficient since ACV
and ACV-X prodrugs are highly soluble in acidic solution, and the
previously reported incubation time to extract ACV from human,
porcine and rabbit skin was less than 1 h [12,24,25]. The skin
extraction medium and the samples from the receiver compart-
ment were centrifuged at 12,000 rpm for 15 min. The supernatant
was filtered through an ExapureÒ nylon filter membrane with a
2.4.2.2. Effect of skin tissue. The effect of skin metabolism on the
stability of ACV-X prodrugs was examined by (i) placing 1 ml of
a 5 mM solution of each ACV-X prodrug in MES buffer (pH 5.5) in
contact with the external surface of porcine stratum corneum
and (ii) exposing 1 ml of a 100 lM solution of each ACV-X prodrug
in PBS buffer (pH 7.4) to the dermal surface of dermatomed porcine
skin for 2 h. Aliquots were withdrawn at regular intervals. Experi-
ments were performed in triplicate at 32 °C. The remaining fraction
of ACV-X was plotted as a function of time (t). All experiments
were performed in triplicate.
pore size of 0.22 lm (Alys Technologies; Bussigny-près-Lausanne,
Switzerland). All samples were analyzed by HPLC–UV.
2.6. Data analysis
2.5. Transport studies
Data were expressed as the mean SD. Outliers determined
using the Dixon test were discarded. Results were evaluated statis-
tically using either analysis of variance (ANOVA) or Student’s t-test.
Student Newman Keuls test was used when necessary as a post hoc
procedure. The level of significance was fixed at p = 0.05.
2.5.1. Skin source
Fresh porcine ears were obtained from a local abattoir (CARRE;
Rolle, Switzerland). The skin was processed (ꢁ750
lm) with an air
dermatome (Zimmer Holdings Inc; Münsingen, Switzerland). Skin
disks (diameter of 26 mm) were punched out from skin mem-
branes and stored at ꢀ20 °C. Porcine ear skin has been extensively
investigated and is considered to be a good model for human skin
[21–23].
3. Result and discussion
3.1. Synthesis of ACV-X prodrugs
A series of amino acid ester prodrugs of ACV was successfully
synthesized via Steglich esterification (Scheme 1). DIC was pre-
ferred to N,N0-dicyclohexylcarbodiimide (DCC) as the coupling
reagent since the reaction by-product, N,N0-diisopropylurea, was
soluble in most organic solvents and easier to remove than the
DCC by-product, N,N0-dicyclohexylurea [26,27]. Moreover, being a
liquid rather than a waxy solid, DIC was also easier to handle. HOBt
was used as a racemization suppressant since DMAP was sus-
2.5.2. Iontophoretic setup
Iontophoretic transport studies were conducted to quantify
skin deposition and cumulative permeation of ACV (QDEP,ACV and
QPERM,ACV, respectively) and each ACV-X prodrug (QDEP,ACV-X and
QPERM,ACV-X, respectively). Passive transport experiments using
the same setup but in the absence of current served as the control.
The dermatomed skin was clamped in vertical Franz diffusion cells
(Glass Technology; Geneva, Switzerland) and allowed to equili-
brate for 30 min before the transport experiments. The area of skin
exposed to the donor formulation was ꢁ0.8 cm2. The anodal and
cathodal (receiver) compartments were filled with 20 and 10 ml
of PBS solution (pH 7.4), respectively. The receiver solution was
continuously stirred, and the temperature was maintained at
32 °C using a dynamic water bath system. Ag and AgCl electrodes
were placed in the anodal and receiver compartments, respec-
tively. Salt bridge assemblies (3% agarose in 0.1 M NaCl) were
employed to electrically connect the physically isolated anodal
and donor compartments. The salt bridges were used to prevent
competition between Na+ ions (from the NaCl used to supply Clꢀ
for anodal Ag/AgCl electrochemistry) and the positively charged
prodrugs. ACV and ACV-X prodrugs were prepared at a concentra-
tion of 5 mM in MES buffer (10 mM, pH 5.5). 1 ml of solution was
pipetted into each donor compartment and current application
initiated (the donor compartment was not occluded). Two series
of iontophoretic transport studies were carried out: (i) to investi-
gate the effect of current on molecular transport using three
constant current densities, 0.125, 0.25 and 0.5 mA/cm2, applied
for 120 min and (ii) to investigate the effect of duration of
iontophoresis on molecular transport, the application time was
decreased to 60, 30, 10 and 5 min using a current density of
0.25 mA/cm2. Current was applied using a APH 1000 M power
generator (Kepco Inc; Flushing, US). All transport studies were
carried out in quintuplicate.
pected to lead to slight racemization of the L-amino acid during
esterification [19,28]. All products gave unambiguous 1H NMR
assignments, accurate m/z [M+H]+ and showed good purity. 1H
NMR spectra, showed that the peak due to the proton of the –OH
group in ACV (d 4.67 ppm) was absent in the spectra of each pro-
drug. Proton deshielding of the 13-H (immediately adjacent to
the –OH group) was also observed: ACV d 3.44, ACV-Arg d 4.26,
ACV-Gly d 4.25, ACV-Ile d 4.32, ACV-Phe d 4.18, ACV-Trp d 4.18
and ACV-Val d 4.31, which was due to the electron-withdrawing
(-I) inductive effect of the newly formed ester group. The proto-
nated primary amine from the amino acid moiety was also seen
at low field (d > 8) because all prodrugs were obtained as the TFA
salt. The m/z [M+H]+ values of the prodrugs showed that the abso-
lute value of the difference between observed and calculated value
was less than 6 ꢂ 10ꢀ4 units. The purity of all prodrugs was deter-
mined by HPLC–UV and was >95%.
3.2. Investigation of ACV-X prodrug physicochemical properties
3.2.1. Log D, pKa and aqueous solubility
The log D, pKa and aqueous solubility of ACV and ACV-X pro-
drugs are shown in Table 3. The predicted log DpH5.5 of ACV was
ꢀ1.48, and the log DpH5.5 values of most of the ACV-X prodrugs
were lower than those of ACV. The very low log DpH5.5 value of
ACV-Arg (ꢀ6.08) was attributed to the introduction of two polar
Please cite this article in press as: Y. Chen et al., Topical iontophoretic delivery of ionizable, biolabile aciclovir prodrugs: A rational approach to improve
cutaneous bioavailability, Eur. J. Pharm. Biopharm. (2015), http://dx.doi.org/10.1016/j.ejpb.2015.12.002

6
Y. Chen et al. / European Journal of Pharmaceutics and Biopharmaceutics xxx (2015) xxx–xxx
Table 3
Table 4
Physicochemical properties of ACV and ACV-X prodrugs at pH 5.5.
pH stability of ACV-X prodrugs: remaining fraction of each prodrug (%ACV-X) after
8 h, kobs (hꢀ1) and t1/2 (h).
ACV/ACV-X MWa Charge:mass log D pKa
ratio (ꢂ103)
Solubilityb
(mM)
ACV/ACV-X
pH
Remaining fraction
of prodrug (%ACV-X)
kobs (hꢀ1
)
t1/2 (h)
ACV
225.2
0
ꢀ1.48 2.55, 9.35
9.2
ACV-Gly
ACV-Val
ACV-Ile
ACV-Phe
ACV-Trp
ACV-Arg
282.3 3.54
324.3 3.08
338.4 2.96
372.4 2.69
411.4 2.43
381.4 5.24
ꢀ3.30 2.53, 7.20, 9.34
ꢀ2.64 2.54, 7.75, 9.34
ꢀ2.16 2.54, 7.78, 9.34
ꢀ1.22 2.54, 6.93, 9.34
ꢀ0.73 2.54, 6.70, 9.34
512.2
441.3
347.8
304.4
51.3
ACV-Gly
2.0
5.5
7.4
100.8 0.5
99.4 3.4
76.5 0.7
20.6 1.1
–
–
–
–
0.037 0.001
0.211 0.007
18.8 0.5
3.3 0.1
10.0
ACV-Val
ACV-Ile
2.0
5.5
7.4
100.4 0.5
98.9 1.5
92.5 0.6
79.2 0.1
–
–
–
–
ꢀ6.08 2.54, 7.28, 9.34, 13.36 >820
a
0.009 0.001
0.029 0.001
74.5 4.0
24.1 0.2
Free base.
b
10.0
Measured in MES buffer (10 mM, pH 5.5).
2.0
5.5
7.4
99.6 0.1
99.3 0.2
94.9 0.2
86.0 0.3
–
–
–
–
0.007 0.001
0.019 0.001
101.1 4.4
36.4 0.6
charged groups – the primary amine group and the guanidyl group.
ACV-Phe has a comparable log DpH5.5 to ACV because the conjuga-
tion with a phenylalanine introduced both a hydrophilic primary
amine group and a lipophilic phenyl group. According to the Hen-
derson–Hasselbalch equation, all prodrugs are >90% ionized and
positively charged at pH 5.5 (approximate skin surface pH) – this
enables them to benefit from electromigration during iontophore-
sis which is a far more efficient electrotransport mechanism than
electroosmosis [13,18]. Furthermore, although the aqueous solu-
bility of ACV at pH 5.5 is limited, ꢁ9.2 mM, all of the ACV-X pro-
drugs were very soluble at this pH, which was also an advantage
for iontophoresis. ACV-Trp, ACV-Phe, ACV-Ile, ACV-Val, ACV-Gly,
and ACV-Arg had solubility enhancements of 6, 33, 38, 48, ꢁ56,
and >90-fold (by molar concentration), respectively, as compared
to ACV. Indeed, ACV-Arg was so soluble that it was not possible
to determine its solubility limit, which exceeded 820 mM. The sol-
ubility enhancement can be attributed to the effect of hydrogen-
bonded solvation of the ionizable group(s) [29].
10.0
ACV-Phe
ACV-Trp
ACV-Arg
2.0
5.5
7.4
99.8 0.6
98.3 0.1
76.9 0.4
20.6 0.6
–
–
–
–
0.033 0.001
0.200 0.005
21.4 0.4
3.5 0.1
10.0
2.0
5.5
7.4
100.1 0.9
99.5 0.5
78.8 0.9
36.9 0.9
–
–
–
–
0.029 0.001
0.132 0.016
24.2 0.4
5.3 0.6
10.0
2.0
5.5
7.4
100.0 0.5
100.8 2.1
77.5 1.8
24.1 0.1
–
–
–
–
0.032 0.003
0.177 0.001
22.0 2.2
3.9 0.0
10.0
105
104
103
102
101
100
99
ACV-Gly
ACV-Phe
ACV-Val
ACV-Trp
ACV-Ile
ACV-Arg
3.2.2. Stability
The investigation into the effect of pH and porcine skin on the
stability of ACV-X prodrugs showed that the ACV-X prodrugs were
converted to ACV by both chemical and enzymatic hydrolyses. The
remaining unhydrolyzed proportion of each prodrug (ACV-X) after
8 h, kobs (hꢀ1) and t1/2 (h) for the pH stability experiments is shown
in Table 4. The rate of chemical hydrolysis appeared to be pH-
dependent. All ACV-X prodrugs were more stable in acidic solution.
At pH 2.0 and pH 5.5 more than 98.5% of the prodrugs was
unchanged after 8 h at 8 °C, and the t1/2 was not measurable. How-
ever, incubation at pH 7.4 and pH 10.0 for 8 h at 8 °C resulted in
appreciable hydrolysis of certain prodrugs – the unchanged frac-
tion was 75–95% and 20–85%, respectively. The prodrugs were
clearly more susceptible to base-catalyzed chemical hydrolysis as
the concentrations of nucleophilic hydroxyl ions increased at
higher pH. Among the six ACV-X prodrugs, ACV-Arg was quite
labile at pH 7.4 and pH 10.0 (t1/2 21.9 h and 3.9 h, respectively),
and ACV-Ile showed the highest stability at either pH (t1/2
100.4 h and 36.3 h, respectively). A formulation pH of 5.5 was
selected for use in the transport studies, which ensured relative
stability of all ACV prodrugs and protected the skin from irritation.
All ACV-X prodrugs were very stable after incubation with the
external surface of porcine stratum corneum for 2 h (Fig. 2a), but
they were extensively hydrolyzed to ACV and the respective amino
acid when in contact with the dermal surface of dermatomed por-
cine skin for 2 h (Fig. 2b). For example, ACV-Arg underwent com-
plete bioconversion after exposure to the dermis for 45 min,
whereas negligible reactivity was found following contact with
the stratum corneum for 2 h. It has been reported that very limited
enzymatic activity was observed in the keratinized stratum
corneum [30], in contrast, considerable amounts of non-specific
esterase and other phase I metabolic enzymes were found in the
98
97
96
95
94
93
92
91
(a)
90
0
15
30
45
60
75
90
105
120
Time (min)
ACV-Gly
ACV-Ile
ACV-Trp
ACV-Val
ACV-Phe
ACV-Arg
100
90
80
70
60
50
40
30
20
10
0
(b)
0
15
30
45
60
75
90
105
120
Time (min)
Fig. 2. Solution stability of ACV-X prodrugs in the presence of (a) external surface of
porcine epidermis (32 °C, 2 h) and (b) dermis (32 °C, 2 h) (Mean SD; n = 3).
Please cite this article in press as: Y. Chen et al., Topical iontophoretic delivery of ionizable, biolabile aciclovir prodrugs: A rational approach to improve
cutaneous bioavailability, Eur. J. Pharm. Biopharm. (2015), http://dx.doi.org/10.1016/j.ejpb.2015.12.002

Y. Chen et al. / European Journal of Pharmaceutics and Biopharmaceutics xxx (2015) xxx–xxx
7
250
250
200
150
100
50
(a)
(b)
ACV-X
ACV
ACV-X
ACV
200
150
100
50
0
0
600
500
400
300
200
100
0
(c)
(d)
600
500
400
300
200
100
0
ACV-X
ACV
ACV-X
ACV
900
800
700
600
500
400
300
200
100
0
(e)
(f) 900
800
ACV-X
ACV
ACV-X
ACV
700
600
500
400
300
200
100
0
Fig. 3. Delivery of ACV and ACV-X following iontophoresis of ACV-X prodrugs (5 mM, pH 5.5) at (a and b) 0.125, (c and d) 0.25 and (e and f) 0.5 mA/cm2 for 2 h. Data are given
in terms of amounts deposited in the skin (QDEP = QDEP,ACV + QDEP,ACV-X) (a, c, d) and the cumulative permeation (QPERM = QPERM,ACV + QPERM,ACV-X) (b, d, f) of ACV species
(Mean SD; n = 5).
dermis [31]. The enzymatic hydrolysis of ACV-X prodrugs by der-
mis was also prodrug-dependent. In contrast to ACV-Arg, which
was extremely susceptible to biotransformation in contact with
the dermis, only 21.5% of ACV-Ile was hydrolyzed after exposure
to dermis for 2 h. The different hydrolytic rates must be attributed
to the different structures of the substrates; for example, steric
hindrance encountered by the esterase might have been influenced
by the amino acid moiety [32].
after 2 h; neither skin retention nor cumulative permeation was
quantifiable. The amount of ACV present in skin was inferior to
0.25 nmol/cm2 since the LOQ of ACV was 0.2
lM and the volume
of skin extraction medium was 1 ml. As indicated in Section 3.2.1,
ACV has poor oil/water solubility, which limits its partitioning into
the highly lipidic intercellular space in the keratinized stratum
corneum. Formulation strategies have been extensively described
to improve the cutaneous bioavailability of ACV either by incorpo-
ration of chemical enhancers such as DMSO [33,34] and high
quantities of propylene glycol [35] or by fabrication of lipid- or
polymer-based formulations such as microemulsions [36],
liposomes [37], microparticles [24] and nanoparticles [38];
however, the risk of skin irritation or complicated preparative
procedures may affect their clinical development.
3.3. Transport studies
3.3.1. Passive delivery
The passive diffusion of ACV and ACV-X prodrugs into and
across porcine skin from 5 mM aqueous solutions was negligible
Please cite this article in press as: Y. Chen et al., Topical iontophoretic delivery of ionizable, biolabile aciclovir prodrugs: A rational approach to improve
cutaneous bioavailability, Eur. J. Pharm. Biopharm. (2015), http://dx.doi.org/10.1016/j.ejpb.2015.12.002

8
Y. Chen et al. / European Journal of Pharmaceutics and Biopharmaceutics xxx (2015) xxx–xxx
Table 5
Total delivery of ACV species (QDEP,ACV + QPERM,ACV + QDEP,ACV-X + QPERM,ACV-X) (nmol/cm2), transport efficiency (%) and delivery efficiency (%) after iontophoresis of ACV-X prodrugs
for 2 h.
ACV-X
Current (mA/cm2)
Total deliverya (nmol/cm2)
Skin retention ratiob
Bioconversion ratioc
Transport efficiencyd (%)
Delivery efficiencye (%)
ACV-Gly
0.125
0.25
0.5
220.1 29.3
515.5 123.6
708.4 71.2
64.9 10.0
62.0 9.0
58.3 3.8
96.0 1.4
97.6 2.1
87.3 1.9
2.36
2.76
1.80
3.52
8.25
11.33
ACV-Val
ACV-Ile
0.125
0.25
0.5
375.9 22.7
791.3 103.9
1034.6 110.8
50.0 7.7
46.3 3.7
42.9 3.3
95.1 1.4
93.1 2.1
90.4 1.6
4.02
4.23
2.62
6.01
12.66
16.55
0.125
0.25
0.5
401.3 62.0
875.9 341.1
1243.4 150.6
50.3 5.6
42.4 6.2
41.6 5.8
87.8 2.1
87.5 2.0
81.6 2.5
4.30
4.69
3.15
6.42
14.01
19.89
ACV-Phe
ACV-Trp
ACV-Arg
0.125
0.25
0.5
231.5 28.1
388.3 51.5
530.5 86.4
46.7 6.1
48.0 7.4
48.3 9.4
97.7 1.4
98.3 0.4
97.4 0.5
2.48
2.08
1.34
3.70
6.21
8.49
0.125
0.25
0.5
120.8 53.0
167.4 16.4
303.5 102.4
57.0 31.1
44.2 4.3
51.9 14.6
95.0 2.2
96.1 0.4
95.6 1.0
1.29
0.90
0.77
1.93
2.68
4.86
0.125
0.25
0.5
353.8 40.2
647.6 33.3
1172.9 104.2
65.5 10.1
65.5 6.1
67.1 5.0
100.0 0.0
100.0 0.0
100.0 0.0
7.57
6.93
5.95
5.66
10.36
18.77
a
Total delivery of ACV species (QDEP,ACV + QPERM,ACV + QDEP,ACV-X + QPERM,ACV-X).
b
Skin retention of total ACV species (sum of ACV and ACV-X) is given by the ratio of the amounts deposited to the total delivery (QDEP,ACV + QDEP,ACV-X/(QDEP,ACV + QDEP,ACV-X
+ QPERM,ACV + QPERM,ACV-X) ꢂ 100%).
c
d
e
Bioconversion of ACV-X to ACV is given by (QDEP,ACV + QPERM,ACV/(QDEP,ACV + QPERM,ACV + QDEP,ACV-X + QPERM,ACV-X) ꢂ 100%).
Transport efficiency (transport number ꢂ 100%).
Delivery efficiency (fraction of the applied dose delivered ꢂ 100%).
3.3.2. Iontophoretic delivery of ACV
given current (Fig. 3). Iontophoresis of the two aromatic amino acid
ester prodrugs, ACV-Trp and ACV-Phe, led to the lowest skin depo-
sition of ACV species. Iontophoresis of ACV-Gly, ACV-Val and ACV-
Ile delivered intermediate, comparable levels of ACV species into
skin. Di-protonated ACV-Arg had the highest charge-mass ratio
among the six prodrugs; therefore, its electromigration was pre-
sumably the highest under a given electric potential gradient.
Mono-protonated, ACV-Trp and ACV-Phe had lower charge-mass
ratio and thus presumably had lower electric mobility. Moreover,
the aromatic rings may also have facilitated hydrophobic interac-
tions with the skin microenvironment. ACV-Val and ACV-Ile had
slightly higher charge-mass ratios than ACV-Trp and ACV-Phe
but displayed much greater deposition in skin. They enabled the
highest level of cumulative permeation of ACV species (Fig. 3).
The mechanisms used to explain iontophoretic transport across
the skin consider the ions to be stable and to remain unchanged
during transit. Obviously, this was not the case here – ACV-X pro-
drugs underwent extensive hydrolysis during cutaneous transport.
More esterase-sensitive prodrugs, such as ACV-Arg, were readily
hydrolyzed after entering the viable epidermis, and lost the amino
acid promoiety and the ACV remained in the skin. Prodrugs, such
as ACV-Val and ACV-Ile, which were less susceptible to enzymatic
hydrolysis, retained their charged moiety and were able to electro-
migrate further into the skin before cleavage of the ester linkage.
ACV-Trp and ACV-Phe were also resistant to esterase activity but
the aromatic amino acid side-chains facilitated interactions with
the skin and hindered transport. These observations provide an
insight into the rational design of prodrugs for iontophoresis and
their optimization for either topical or transdermal delivery based
on the stability of linkage to the charged moiety. Comparatively
stable prodrugs might have a greater probability of reaching the
systemic circulation during iontophoresis, i.e. better suited to
transdermal delivery and systemic action. In contrast, enzymati-
cally labile prodrugs would be more likely to undergo hydrolysis
earlier during cutaneous transit, release uncharged parent drug
that would be retained within the membrane. They may be consid-
ered as better suited for the treatment of dermatological diseases.
However, the intrinsic stability (i.e. ease of formulation) and
Deposition of ACV (QDEP,ACV) after iontophoresis of a 5 mM solu-
tion at pH 5.5 for 2 h at 0.125, 0.25 and 0.5 mA/cm2 was 0.9 0.3,
2.1 0.3 and 4.6 0.3 nmol/cm2, respectively. Cumulative perme-
ation of ACV (QPERM,ACV) was below the LOQ. Given that ACV is
essentially neutral at pH 5.5 (Table 3), it is transported by elec-
troosmosis, a solvent volume flow induced by the movement of
ions across a charged membrane [13]. Electroosmosis is a less
effective transport mechanism than electromigration; thus, there
is only a modest increase in ACV delivery [25]. Given the pKa of
its ionizable groups, iontophoretic enhancement of ACV delivery
by electromigration can only be observed with formulations at a
pH outside the physiological range. For example, at pH 3 where
ACV was present at ꢁ20% as a cation, anodal iontophoresis for
7 h at 0.5 mA/cm2 using human skin in vitro resulted in epidermal
and dermal ACV concentrations in the range of 80–150
l
g/cm3
(corresponding to 8–15
l
g of ACV per cm2, as the thickness of
the skin was ꢁ1.0 mm) [16]. Similarly, at pH 11, where ACV was
ꢁ99% deprotonated and present as an anion, cathodal iontophore-
sis for 10 min at 0.2 mA/cm2 in rats in vivo resulted in skin levels of
27.27 3.53 l
g/cm2 [17].
3.3.3. Iontophoretic delivery of ACV-X prodrugs
Iontophoresis of ACV-X prodrugs at 0.5 mA/cm2 for 2 h
produced order of magnitude increases in cutaneous deposition
of ACV species (QDEP = QDEP,ACV + QDEP,ACV-X
) as compared to
either passive diffusion or iontophoresis of ACV (ACV-Gly:
412.8 44.0, ACV-Val: 358.8 66.8, ACV-Ile: 434.1 68.2,
ACV-Phe: 249.8 81.4, ACV-Trp: 156.1 76.3, and ACV-Arg:
785.9 78.1 nmol/cm2, respectively) (Fig. 3). It should also be
emphasized that all prodrug formulations tested were easily pre-
pared in aqueous solution at a patient-friendly pH 5.5. The total
delivery of ACV species (QTOTAL = QDEP,ACV + QPERM,ACV + QDEP,ACV-X
+
QPERM,ACV-X), transport efficiency (transport number ꢂ 100%) and
delivery efficiency (fraction of the applied dose delivered ꢂ 100%)
after iontophoresis of ACV-X prodrugs for 2 h are shown in Table 5.
Iontophoresis of the di-protonated prodrug, ACV-Arg, resulted
in the greatest amount of skin deposition of ACV species at any
Please cite this article in press as: Y. Chen et al., Topical iontophoretic delivery of ionizable, biolabile aciclovir prodrugs: A rational approach to improve
cutaneous bioavailability, Eur. J. Pharm. Biopharm. (2015), http://dx.doi.org/10.1016/j.ejpb.2015.12.002

Y. Chen et al. / European Journal of Pharmaceutics and Biopharmaceutics xxx (2015) xxx–xxx
9
600
500
400
300
200
100
0
450
(a)
(a)
400
ACV-Ile
ACV
ACV-Arg
ACV
350
300
250
200
150
100
50
0
5
10
30
60
120
5
10
30
60
120
Time (min)
Time (min)
600
500
400
300
200
100
0
300
250
200
150
100
50
(b)
(b)
ACV-Arg
ACV
ACV-Ile
ACV
0
5
10
30
60
120
Time (min)
5
10
30
60
120
800
700
600
500
400
300
200
100
0
Time (min)
(c)
ACV-Arg
ACV
1000
900
800
700
600
500
400
300
200
100
0
(c)
ACV-Ile
ACV
5
10
30
60
120
Time (min)
5
10
30
60
120
Fig. 4. (a) Skin deposition (QDEP), (b) cumulative permeation (QPERM) and (c) total
delivery (QPERM) as a function of the duration of iontophoresis at 0.25 mA/cm2 for
ACV-Arg (Mean SD; n = 5).
Time (min)
Fig. 5. (a) Skin deposition (QDEP), (b) cumulative permeation (QPERM) and (c) total
delivery (QPERM) as a function of the duration of iontophoresis at 0.25 mA/cm2 for
ACV-Ile (Mean SD; n = 5).
molecular properties that influence potential interactions along
the transport pathways must also be considered in conjunction
with these ‘‘guidelines” based on the stability of the linkage
between the parent molecule and prodrug moiety.
7000
6000
5000
4000
3000
2000
1000
0
700
600
500
400
300
200
100
0
QDEP,ACV; ACV-Arg
DEP,ACV; ACV-Ile
CDEP,ACV; ACV-Arg
DEP,ACV; ACV-Ile
Q
The skin retention ratio (QDEP,ACV + QDEP,ACV-X/QTOTAL ꢂ 100%),
was used to represent the selectivity for deposition over perme-
ation (Table 5). The skin retention ratio of ACV-Arg after ion-
tophoresis for 2 h at any given current was statistically
significantly higher than those of the other prodrugs. For example,
at 0.5 mA/cm2 iontophoresis of ACV-Arg, ACV-Gly, ACV-Trp, ACV-
Phe, ACV-Val and ACV-Ile resulted in skin retention ratios of
67.1 5.0%, 58.3 3.8%, 51.9 4.6%, 48.3 9.4%, 42.9 3.3% and
41.6 5.8%, respectively. The skin retention ratio of ACV species
did not seem to be affected by the applied current density (Table 5).
The bioconversion ratio (QDEP,ACV + QPERM,ACV/QTOTAL ꢂ 100%) after
iontophoresis of the ACV-X prodrugs is shown in Table 5. ACV-
Arg was the most biolabile prodrug, and complete bioconversion
(100%) was achieved during ACV-Arg iontophoresis, which was
statistically significantly higher than that of any of the other
C
5
10
30
60
120
Time (min)
Fig. 6. Skin deposition of ACV (QDEP,ACV) (columns) and corresponding concentra-
tion CDEP,ACV (filled circles) observed as a function of the duration of iontophoresis at
0.25 mA/cm2 for ACV-Arg and ACV-Ile (nmol/cm2) (Mean SD; n = 5).
Please cite this article in press as: Y. Chen et al., Topical iontophoretic delivery of ionizable, biolabile aciclovir prodrugs: A rational approach to improve
cutaneous bioavailability, Eur. J. Pharm. Biopharm. (2015), http://dx.doi.org/10.1016/j.ejpb.2015.12.002

10
Y. Chen et al. / European Journal of Pharmaceutics and Biopharmaceutics xxx (2015) xxx–xxx
ACV-X prodrugs. In contrast, ACV-Ile, the most enzymatically
stable prodrug, had the lowest total bioconversion ratio.
References
[1] M.E. Ganapathy, W. Huang, H. Wang, V. Ganapathy, F.H. Leibach, Valacyclovir:
a substrate for the intestinal and renal peptide transporters PEPT1 and PEPT2,
Biochem. Biophys. Res. Commun. 246 (1998) 470–475.
[2] S. Weller, M.R. Blum, M. Doucette, T. Burnette, D.M. Cederberg, P. de Miranda,
M.L. Smiley, Pharmacokinetics of the acyclovir pro-drug valaciclovir after
escalating single-and multiple-dose administration to normal volunteers, Clin.
Pharmacol. Ther. 54 (1993) 595–605.
[3] M.H. Sawyer, D.E. Webb, J.E. Balow, S.E. Straus, Acyclovir-induced renal failure:
clinical course and histology, Am. J. Med. 84 (1988) 1067–1071.
[4] A.P. Fiddian, J.M. Yeo, R. Stubbings, D. Dean, Successful treatment of herpes
labialis with topical acyclovir, Br. Med. J. (Clin. Res. Ed.) 286 (1983) 1699–1701.
[5] M. Shaw, M. King, J.M. Best, J.E. Banatvala, J.R. Gibson, M.R. Klaber, Failure of
acyclovir cream in treatment of recurrent herpes labialis, Br. Med. J. (Clin. Res.
Ed.) 291 (1985) 7–9.
[6] D.J. Freeman, N.V. Sheth, S.L. Spruance, Failure of topical acyclovir in ointment
to penetrate human skin, Antimcriob. Agents Chemother. 29 (1986) 730–732.
[7] S.L. Spruance, L.E. Schnipper, J.C. Overall, E.R. Kern, B. Wester, J. Modlin, G.
Wenerstrom, C. Burton, K.A. Arndt, G.L. Chiu, C.S. Crumpacker, Treatment of
herpes simplex labialis with topical acyclovir in polyethylene glycol, J. Infect.
Dis. 146 (1982) 85–90.
[8] S.L. Spruance, C.S. Crumpacker, L.E. Schnipper, E.R. Kern, S. Marlowe, K.A.
Arndt, J.C. Overall, Early patient-initiated treatment of herpes labialis with
topical 10% acyclovir, Antimicrob. Agents Chemother. 25 (1984) 553–555.
[9] S.L. Spruance, R. Nett, T. Marbury, R. Wolff, J. Johnson, T. Spaulding, Acyclovir
cream for treatment of herpes simplex labialis: results of two randomized,
double-blind, vehicle-controlled, multicenter clinical trials, Antimicrob. Agents
Chemother. 46 (2002) 2238–2243.
[10] S.L. Spruance, C.S. Crumpacker, H. Haines, C. Bader, K. Mehr, J. MacCalman, L.E.
Schnipper, M.R. Klauber, J.C. Overall, Ineffectiveness of topical adenine
arabinoside 50-monophosphate in the treatment of recurrent herpes simplex
labialis, N. Engl. J. Med. 300 (1979) 1180–1184.
[11] J. Piret, J. Lamontagne, A. Désormeaux, M.G. Bergeron, Efficacies of gel
formulations containing foscarnet, alone or combined with sodium lauryl
sulfate, against establishment and reactivation of latent herpes simplex virus
type 1, Antimicrob. Agents Chemother. 45 (2001) 1030–1036.
[12] G.E. Parry, P. Dunn, V.P. Shah, L.K. Pershing, Acyclovir bioavailability in human
skin, J. Invest. Dermatol. 98 (1992) 856–863.
Given their contrasting properties and behavior, it was decided
to investigate the iontophoretic transport of ACV-Arg and ACV-Ile
more closely. The variation of the relative contributions of
deposition (QDEP) and permeation (QPERM) to the total iontophoretic
delivery of ACV-Arg and ACV-Ile as a function of the duration of
current application is shown in Figs. 4 and 5. Given the rapid bio-
conversion of ACV-Arg to ACV, it is not surprising to observe that
total delivery was dominated by deposition of ACV (QDEP,ACV). A
greater absolute amount and percentage of the applied ACV-Ile
permeated across the skin as a function of time; this was consis-
tent with the greater resistance to hydrolysis of the ester linkage
in ACV-Ile. After iontophoresis of ACV-Arg for 60 min, QDEP,ACV
was 454.9 55.9 nmol/cm2. In the case of ACV-Ile, QDEP,ACV + QDEP,
was 285.5 22.1 nmol/cm2. The profiles started to plateau
ACV-ILE
which may have indicated that deposition was a saturable process
[39].
The cutaneous deposition of ACV (QDEP,ACV) and corresponding
concentration of ACV in the skin (CACV) following iontophoresis of
ACV-Arg and ACV-Ile as a function of time are shown in Fig. 6. There
is considerable variability in the reported IC50 of ACV against HSV-1
and HSV-2, 0.01–0.7 lg/ml and 0.01–3.2 lg/ml, respectively [19],
which is equivalent to 0.04–3.1 nmol/ml and 0.04–14.2 nmol/ml,
respectively. The average cutaneous concentrations of ACV after
iontophoresis of ACV-Ile and ACV-Arg at each time were orders of
magnitude greater than IC50 of ACV against HSV-1 and HSV-2. For
example, iontophoresis of ACV-Arg and ACV-Ile at 0.25 mA/cm2
for only 5 min resulted in the deposition of appreciable amounts
ACV (36.4 5.7 nmol/cm2 and 40.3 6.1 nmol/cm2, respectively),
corresponding to supra-therapeutic levels of ACV being achieved
in the skin (485.2 75.9 nmol/ml and 537.4 80.7 nmol/ml,
respectively).
[13] Y.N. Kalia, A. Naik, J. Garrison, R.H. Guy, Iontophoretic drug delivery, Adv. Drug
Deliv. Rev. 56 (2004) 619–658.
[14] T. Gratieri, Y.N. Kalia, Mathematical models to describe iontophoretic
transport in vitro and in vivo and the effect of current application on the
skin barrier, Adv. Drug Deliv. Rev. 65 (2012) 315–329.
[15] L. Colla, E. De Clercq, R. Busson, H. Vanderhaeghe, Synthesis and antiviral
activity of water-soluble esters of acyclovir [9-[(2-hydroxyethoxy)methyl]
guanine], J. Med. Chem. 26 (1983) 602–604.
4. Conclusion
[16] N.M. Volpato, S. Nicoli, C. Laureri, P. Colombo, P. Santi, In vitro acyclovir
distribution in human skin layers after transdermal iontophoresis, J. Control.
Release 50 (1998) 291–296.
[17] S. Siddoju, V. Sachdeva, P.M. Friden, Y.Y. Yu, A.K. Banga, Acyclovir skin depot
characterization following in vivo iontophoretic delivery, Skin Res. Technol. 17
(2011) 234–244.
[18] N. Abla, A. Naik, R.H. Guy, Y.N. Kalia, Topical iontophoresis of valaciclovir
hydrochloride improves cutaneous aciclovir delivery, Pharm. Res. 23 (2006)
1842–1849.
[19] L.M. Beauchamp, G.F. Orr, P. De Miranda, T. Burnette, T.A. Krenitsky, Amino
acid ester prodrugs of acyclovir, Antivir. Chem. Chemother. 3 (1992) 157–164.
[20] C. Yang, H. Gao, A.K. Mitra, Chemical stability, enzymatic hydrolysis, and nasal
uptake of amino acid ester prodrugs of acyclovir, J. Pharm. Sci. 90 (2001) 617–
624.
[21] I.P. Dick, R.C. Scott, Pig ear skin as an in-vitro model for human skin
permeability, J. Pharm. Pharmacol. 44 (1992) 640–645.
[22] F.P. Schmook, J.G. Meingassner, A. Billich, Comparison of human skin or
epidermis models with human and animal skin in in-vitro percutaneous
absorption, Int. J. Pharm. 215 (2001) 51–56.
[23] U. Jacobi, M. Kaiser, R. Toll, S. Mangelsdorf, H. Audring, N. Otberg, W. Sterry, J.
Lademann, Porcine ear skin: an in vitro model for human skin, Skin Res.
Technol. 13 (2007) 19–24.
[24] E.G. de Jalón, M.J. Blanco-Príeto, P. Ygartua, S. Santoyo, Topical application of
acyclovir-loaded microparticles: quantification of the drug in porcine skin
layers, J. Control. Release 75 (2001) 191–197.
Prodrugs have been widely used to increase the topical bioavail-
ability of molecules in the skin [40]. For example, mono- and di-
ester derivatives of corticosteroid esters are routinely used to
increase lipophilicity by masking hydroxyl groups and so facilitate
partition into the intercellular lipid matrix in the stratum corneum.
In contrast, prodrugs for topical iontophoresis used in this study
were designed to do the exact opposite – namely, (i) to increase
hydrosolubility and (ii) to introduce ionizable group(s). Improved
percutaneous transport of dehydroepiandrosterone [41] and
dexamethasone [42,43] was observed when the drugs were conju-
gated to ionizable, biolabile promoieties, to form charged, water-
soluble prodrugs, and delivered across skin by iontophoresis.
However, this is not always the case [44]. The results described
in this first part of the study confirmed that iontophoresis of
ionized amino acid ester prodrugs of ACV was an extremely
effective means to increase cutaneous delivery of ACV. The next
step was to optimize the iontophoretic conditions and formulation
and to demonstrate that topical iontophoresis not only increased
absolute amounts of drug in the skin but enabled targeted
intraepidermal delivery to the deep epidermis where the virus is
found. This is described in the next paper.
[25] C. Padula, F. Sartori, F. Marra, P. Santi, The influence of iontophoresis on
acyclovir transport and accumulation in rabbit ear skin, Pharm. Res. 22 (2005).
1519-1124.
[26] M. Ginisty, M.N. Roy, J.C. Poupon, A. Picard, A.B. Charette,
Tetraarylphosphonium-supported
carbodiimide
reagents:
synthesis,
Acknowledgments
structure optimization and applications, J. Org. Chem. 4 (2008) 2542–2547.
[27] F.F. Du, H. Wang, Y.Y. Bao, B. Liu, H.T. Zheng, R.K. Bai, Conjugated coordination
polymers based on 8-hydroxyquinoline ligands: impact of polyhedral
oligomericsilsesquioxaneson solubility and luminescence, J. Mater. Chem. 21
(2011) 10859–10864.
We thank the University of Geneva and the Ernst and Lucie Sch-
midheiny Foundation for financial support. YC also thanks the
China Scholarship Council (CSC) for a Ph.D. fellowship. We thank
Dr. Sachin Dubey and Dr. Dhaval R. Kalaria for helpful discussions.
[28] G.J. Ho, K.M. Emerson, D.J. Mathre, R.F. Shuman, E.J. Grabowski, Carbodiimide-
mediated amide formation in a two-phase system: a high-yield and low-
Please cite this article in press as: Y. Chen et al., Topical iontophoretic delivery of ionizable, biolabile aciclovir prodrugs: A rational approach to improve
cutaneous bioavailability, Eur. J. Pharm. Biopharm. (2015), http://dx.doi.org/10.1016/j.ejpb.2015.12.002

Y. Chen et al. / European Journal of Pharmaceutics and Biopharmaceutics xxx (2015) xxx–xxx
11
racemization procedure for peptide synthesis, J. Org. Chem. 60 (1995) 3569–
3570.
[29] E. Grunwald, E.K. Ralph III, Kinetic studies of hydrogen-bonded solvation
complexes of amines in water and hydroxylic solvents, Acc. Chem. Res. 4
(1971) 107–113.
[37] E. Horwitz, S. Pisanty, R. Czerninski, M. Helser, E. Eliav, E. Touitou, A clinical
evaluation of a novel liposomal carrier for acyclovir in the topical treatment of
recurrent herpes labialis, Oral Surg. Oral Med. Oral Pathol. Oral Radiol. Endod.
87 (1999) 700–705.
[38] A. Hasanovic, M. Zehl, G. Reznicek, C. Valenta, Chitosan-tripolyphosphate
[30] F. Beisson, M. Aoubalaa, S. Marullb, A.M. Moustacas-Gardiesa, R. Voultouryc, R.
Vergera, V. Arondela, Use of the tape stripping technique for directly
quantifying esterase activities in human stratum corneum, Anal. Biochem.
290 (2001) 179–185.
[31] Y. Dancik, C. Thompson, G. Krishnan, M.S. Roberts, Cutaneous metabolism and
active transport in transdermal drug delivery, in: Nancy A. Monteiro-Riviere
(Ed.), Toxicology of the Skin, Informa Healthcare, New York, 2010, pp. 69–82.
[32] N.L. Pochopina, W.N. Charmanb, V.J. Stella, Amino acid derivatives of dapsone
as water-soluble prodrugs, Int. J. Pharm. 121 (1995) 157–167.
[33] S.L. Spruance, M.B. McKeough, J.R. Cardinal, Penetration of guinea pig skin by
acyclovir in different vehicles and correlation with the efficacy of topical
therapy of experimental cutaneous herpes simplex virus infection, Antimicrob.
Agents Chemother. 25 (1984) 10–15.
nanoparticles as a possible skin drug delivery system for aciclovir with
enhanced stability, J. Pharm. Pharmacol. 61 (2009) 1609–1616.
[39] H. Wagner, K.H. Kostka, C.M. Lehr, U.F. Schaefer, Drug distribution in human
skin using two different in vitro test systems: comparison with in vivo data,
Pharm. Res. 17 (2000) 1475–1481.
[40] K.B. Sloan, S. Wasdo, Designing for topical delivery: prodrugs can make the
difference, Med. Res. Rev. 23 (2003) 763–793.
[41] S. Laneri, A. Sacchi, E.A. di Frassello, E. Luraschi, P. Colombo, P. Santi, Ionized
prodrugs of dehydroepiandrosterone for transdermal iontophoretic delivery,
Pharm. Res. 16 (1999) 1818–1824.
[42] J.P. Sylvestre, R.H. Guy, M.B. Delgado-Charro, In vitro optimization of
dexamethasone phosphate delivery by iontophoresis, Phys. Ther. 88 (2008)
1177–1185.
[34] J. Piret, A. Désormeaux, P. Gourde, J. Juhász, M.G. Bergeron, Efficacies of topical
formulations of foscarnet and acyclovir and of 5-percent acyclovir ointment
[43] J. Cázares-Delgadillo, C. Balaguer-Fernández, A. Calatayud-Pascual, A. Ganem-
Rondero, D. Quintanar-Guerrero, A.C. López-Castellano, V. Merino, Y.N. Kalia,
Transdermal iontophoresis of dexamethasone sodium phosphate in vitro and
in vivo: effect of experimental parameters and skin type on drug stability and
transport kinetics, Eur. J. Pharm. Biopharm. 75 (2010) 173–178.
[44] O.W. Ackaert, J. De Graan, R. Capancioni, O.E. Della Pasqua, D. Dijkstra, B.H.
Westerink, M. Danhof, J.A. Bouwstra, The in vitro and in vivo evaluation of new
synthesized prodrugs of 5-OH-DPAT for iontophoretic delivery, J. Control.
Release 144 (2010) 296–305.
(Zovirax) in
a murine model of cutaneous herpes simplex virus type 1
infection, Antimicrob. Agents Chemother. 44 (2000) 30–38.
[35] O. Díez-Sales, T.M. Garrigues, J.V. Herráez, R. Belda, A. Martín-Villodre, M.
Herráez, In vitro percutaneous penetration of acyclovir from solvent systems
and Carbopol 971-P hydrogels: influence of propylene glycol, J. Pharm. Sci. 94
(2005) 1039–1047.
[36] S. Rajan Shishu, Kamalpreet, Development of novel microemulsion-based
topical formulations of acyclovir for the treatment of cutaneous herpetic
infections, AAPS PharmSciTech 10 (2009) 559–565.
Please cite this article in press as: Y. Chen et al., Topical iontophoretic delivery of ionizable, biolabile aciclovir prodrugs: A rational approach to improve
cutaneous bioavailability, Eur. J. Pharm. Biopharm. (2015), http://dx.doi.org/10.1016/j.ejpb.2015.12.002