N,N′-Dialkylaminoalkylcarbonyl (DAAC) prodrugs and aminoalkylcarbonyl (AAC) prodrugs of 4-hydroxyacetanilide and naltrexone with improved skin permeation properties

Article abstract of DOI:10.1016/j.bmcl.2011.04.118

N,N′-Dialkylaminoalkylcarbonyl (DAAC) and aminoalkylcarbonyl (AAC) prodrugs of phenolic drugs acetaminophen (APAP) and naltrexone (NTX) are reported. The effects of incorporation of a basic amine group into the promoiety of an acyl prodrug of a phenolic drug on its skin permeation properties are also presented. DAAC-APAP prodrugs were synthesized via a three-step procedure starting with haloalkylcarbonyl esters which were reacted with five different amines: dimethylamine, diethylamine, dipropylamine, morpholine, and piperidine. The spacing between the amino group and the carbonyl group of the acyl group was 1-3 CH₂. After the hydrolysis of the ester, the carboxylic acid product was subsequently coupled with the parent drug via a dicyclohexyl carbodiimide (DCC) mediated coupling to yield the DAAC-APAP-HCl prodrugs in excellent yields. The AAC prodrugs were synthesized using commercially available Boc-protected amino acids using DCC or EDCI as coupling agents. The yields of the prodrugs synthesized using these two different methods have been compared. Half-lives (t_1/2) of a few members of the DAAC and AAC series were measured in buffer (pH 6.0, 20 mM). The members evaluated in hydrolysis experiments exhibit a t_1/2 range of 15-113 min. Among AAC-APAP prodrugs, the isopropyl group in valinate-APAP-HCl exerted a steric effect that increased the t_1/2 value for this prodrug compared to alaninate-APAP-HCl or prolinate-APAP-HCl. The 2-morpholinylacetate-APAP prodrug was able to achieve twice the flux of APAP in in vitro diffusion cell experiments through hairless mouse skin.

Full text of DOI:10.1016/j.bmcl.2011.04.118

Bioorganic & Medicinal Chemistry Letters 21 (2011) 4078–4082
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
N,N0-Dialkylaminoalkylcarbonyl (DAAC) prodrugs and aminoalkylcarbonyl
(AAC) prodrugs of 4-hydroxyacetanilide and naltrexone with improved
skin permeation properties
H. Devarajan-Ketha a, , K. B. Sloan b
⇑
a Department of Biochemistry and Molecular Biology, Mayo Clinic College of Medicine, Rochester, MN 55905, USA
b Department of Medicinal Chemistry, University of Florida, Gainesville, FL 32610, USA
a r t i c l e i n f o
a b s t r a c t
N,N0-Dialkylaminoalkylcarbonyl (DAAC) and aminoalkylcarbonyl (AAC) prodrugs of phenolic drugs acet-
aminophen (APAP) and naltrexone (NTX) are reported. The effects of incorporation of a basic amine group
into the promoiety of an acyl prodrug of a phenolic drug on its skin permeation properties are also pre-
sented. DAAC-APAP prodrugs were synthesized via a three-step procedure starting with haloalkylcarbon-
yl esters which were reacted with five different amines: dimethylamine, diethylamine, dipropylamine,
morpholine, and piperidine. The spacing between the amino group and the carbonyl group of the acyl
group was 1–3 CH2. After the hydrolysis of the ester, the carboxylic acid product was subsequently
coupled with the parent drug via a dicyclohexyl carbodiimide (DCC) mediated coupling to yield the
DAAC-APAP-HCl prodrugs in excellent yields. The AAC prodrugs were synthesized using commercially
available Boc-protected amino acids using DCC or EDCI as coupling agents. The yields of the prodrugs
synthesized using these two different methods have been compared. Half-lives (t1/2) of a few members
of the DAAC and AAC series were measured in buffer (pH 6.0, 20 mM). The members evaluated in hydro-
lysis experiments exhibit a t1/2 range of 15–113 min. Among AAC-APAP prodrugs, the isopropyl group in
valinate-APAP-HCl exerted a steric effect that increased the t1/2 value for this prodrug compared to alan-
inate-APAP-HCl or prolinate-APAP-HCl. The 2-morpholinylacetate-APAP prodrug was able to achieve
twice the flux of APAP in in vitro diffusion cell experiments through hairless mouse skin.
Article history:
Received 13 December 2010
Revised 21 April 2011
Accepted 25 April 2011
Available online 11 May 2011
Keywords:
Ionizable amine
Acyl prodrugs
Acetaminophen prodrugs
Naltrexone prodrugs
Ó 2011 Elsevier Ltd. All rights reserved.
Optimization of the topical delivery of phenols is an attractive
goal for research in the areas of pain management (narcotic anal-
getics) and hormone replacement therapy (estradiol). The delivery
of any topically applied drug is defined as the amount of drug dif-
fusing into and through the skin per unit area and time (J, i.e.,
changing the ability of the drug to interact with its target receptor
or enzyme. Prodrugs, which transiently modify the SLIPID and SAQ of
the drug to improve those properties yet revert to the parent drugs
in a controlled and predictable manner, is one strategy to accom-
plish this goal.
In the case of the phenolic drugs, masking the polar OH group
may lead to an increase in SLIPID but not SAQ which will not lead
to optimization of JM. In fact, in a homologous series of prodrugs,
the member of the series exhibiting the best balance of SLIPID and
l
mol cmꢀ2 hꢀ1). Therefore, to enhance the delivery of a topically
applied drug, improvement in its flux (J) needs to be achieved.1
The physicochemical properties of a drug that predict its maximum
flux (JM) are its aqueous and lipid solubilities (SAQ, SLIPID) and its
molecular weight (MW): the Roberts–Sloan equation:2
SAQ, but not necessarily the most lipid soluble member, shows
the best JM. Alkyloxycarbonyl (AOC)3 acyl type and alkylcarbonyl-
oxymethyl (ACOM),4 alkyloxycarbonyloxymethyl (AOCOM)5 and
N-alkyl-N-alkyloxycarbonylaminomethyl (NANAOCAM)6 soft alkyl
type prodrugs of a model phenolic drug, acetaminophen (APAP),
have been synthesized, characterized and evaluated in diffusion
cell experiments in our lab. Similarly, simple acyl type prodrugs
of naltrexone (NTX), a narcotic analgetic, have been synthesized
by others.7,8
log JM ¼ x þ y log SLIPID þ ð1 ꢀ yÞ log SAQ ꢀ zMW
Since y = 0.5–0.6 (above), delivery into (dermal) and through
(transdermal) the skin of phenolic drugs is a challenging goal
due to a lack of balance in SLIPID and SAQ of these drugs. Obvi-
ously, one approach to achieving that balance required by the
Roberts–Sloan equation to optimize JM of the drug is to improve
the SLIPID and SAQ of the drug without greatly increasing MW, or
However, optimization of the prodrugs was not sufficient to
give clinically effective delivery of NTX. The enhancement in J
was primarily due to the increase SLIPID exhibited by the prodrugs
⇑
Corresponding author.
E-mail address: ketha.hemamalini@mayo.edu (H. Devarajan-Ketha).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.04.118

H. Devarajan-Ketha, K. B. Sloan / Bioorg. Med. Chem. Lett. 21 (2011) 4078–4082
4079
while only one exhibited increased SAQ and it gave one of the high-
er JM values.
More recently, studies investigating the effect of incorporation
of ethyleneoxy groups into the promoieties of prodrugs of APAP9
and NTX10 on JM have been reported. Significant increases in SLIPID
and SAQ obtained for promoieties containing three ethyleneoxy
groups was apparently offset by increased MW caused by ethyl-
eneoxy-associated water molecules which compromised their
effectiveness. Thus, in terms of prodrug design, an ideal promoiety
is one that increases SLIPID and SAQ without dramatically increasing
MW.
The incorporation of basic amine groups into the promoiety
should also lead to increased SLIPID (as the free base), and especially
SAQ, of prodrugs of phenols. The increased SAQ is the consequence
of the ionization of the amine group in water which in turn de-
pends on the pKa of the amine. Only a few reports describing the
synthesis of acyl type aminoalkylcarbonyl (AAC) prodrugs of phe-
nolic drugs containing a basic amine group in the promoiety have
been published,11–13 and in none of those reports was transdermal
delivery observed. On the other hand, Milosovich et al. have shown
a N,N-dialkylaminoalkylcarbonyl ester of the aliphatic hydroxy
group in testosterone did increase the transdermal delivery of tes-
tosterone.14 However, no synthesis of N,N-dialkylaminoalkylcar-
bonyl (DAAC) prodrugs of phenols with different alkyl chain
lengths between the amine and the carbonyl groups have been re-
ported. Increasing the alkyl chain lengths should result in increas-
ing pKa values of the amine groups as they become insulated from
the electron withdrawing effect of the carbonyl group and should
result in increased SAQ values. Thus, it was of interest to synthesize
DAAC and more AAC prodrugs of phenolic drugs and to determine
what effect changes in pKa may have on SAQ and hence on JM.
Here, we report on efficient synthesis of DAAC and AAC pro-
drugs of APAP and an AAC prodrug of NTX (Fig. 1). The half lives
(t1/2) of some of the DAAC-APAP, AAC-APAP and AAC-NTX prodrugs
are reported. The pKa values of the amine group in each prodrug
are estimated using the pKa prediction software pKa DB 12.0
(ACD Labs, Inc.) (Table 3). The results from an in vitro diffusion cell
experiment on one member of the series and its hydrochloride salt
using hairless mouse skin and an isopropyl myristate (IPM) vehicle
are also reported.
The DAAC prodrugs of APAP (5a–5k) were synthesized in three
steps (Scheme 1). The choice of the amine incorporated into the
promoiety was based on the pKa and the steric bulk of the amine.
Five secondary amines were chosen for the synthesis of 3b, 3c,
3e–3k which exhibited pKa values between 10.2 and 8.3: di-
methyl-, diethyl-, and dipropylamine, piperidine and morpholine.
3a and 3d were commercially available. An alpha (
a), beta (b) or
gamma ( ) halo ester was then reacted with the secondary amines
c
to give the corresponding N,N-dialkylaminoalkyl esters 2c, 2e–2k:
2b was commercially available. In the cases of the synthesis of 2c
and 2e (more sterically hindered amines), the ethyl chloroacetate
was converted to ethyl iodoacetate by treatment with 1 equivalent
of NaI in 10 mL acetone prior to reaction with the amine. The ethyl
iodoacetate was not isolated but immediately used. For the synthe-
sis of 2g, 2h, 2g, 2j and 2k the corresponding bromo esters were
commercially available. The esters were hydrolyzed with HCl to
give their corresponding amino acid salts (3b, 3c, 3e–3k) which,
along with 3a and 3d, were coupled to APAP using dicyclohexylcar-
bodiimide (DCC) in pyridine to give the hydrochloride salts of the
prodrugs (4a–4k). The free bases of the prodrugs were isolated
by rapid (<30 s) partitioning of the hydrochlorides between satu-
rated aqueous NaHCO3 and dichloromethane at 0–4 °C as the
starting temperature. If this procedure was not used significant
(20%–60%) hydrolysis was observed.
Figure 1. Chemical structures of (a) APAP; (b) NTX and the synthesized DAAC and
AAC prodrugs.
NTX) in dichloromethane to give the Boc protected AAC prodrug
(6a–6c). Deprotection with HCl/ether gave the prodrugs as HCl
salts (7a–7c) in good yield (Scheme 2).
The AAC-APAP prodrugs were synthesized by coupling the cor-
responding N-Boc protected amino acid with the phenol (APAP or
When DCC/DMAP was used as the coupling reagent (Method A)
to give the Boc protected compounds, purification and further

4080
H. Devarajan-Ketha, K. B. Sloan / Bioorg. Med. Chem. Lett. 21 (2011) 4078–4082
Scheme 1. Synthesis of DAAC prodrugs of APAP. Reagents and conditions: (a) ACN, 60 °C, 8–12 h; (b) concd HCl, excess, 100 °C, 12 h; (c) DCC/Py, 5, rt, 24–48 h; (d) aq satd
NaHCO3, 0–4 °C, 30 s.
Scheme 2. Synthesis of AAC prodrugs of APAP and NTX. Reagents and conditions: Method A: DCC/DMAP, rt, 8–12 h; Method B: EDCI/CH2Cl2, rt, 6 h; (a) 2 N HCl in
diethylether, rt, 8 h.
characterization of these compounds was a problem because of the
presence of the DCC coupling by-product: Dicyclohexyl urea (DCU).
So the Boc-AAC-APAP prodrugs (6a–6c) were deprotected without
purification to give 7a–7c in good yields. In order to obtain pure
Boc protected AAC prodrugs, N-(3-dimethylaminopropyl)-N0-ethyl-
carbodiimide (EDCI) was used as the coupling reagent. However,
the yields of 7a–7c were lower. For the coupling of APAP with
Boc-amino acid, the presence of DMAP as a catalyst was essential.
In the absence of the catalyst the reaction did not go to completion
even after 48 h of reaction time. An attempt to synthesize the Boc-
AAC-APAP compounds using DCC coupling in pyridine also led to
an incomplete (40%–50%) reaction.
some water solubility, the lower yield using the EDCI mediated
coupling may be due to the loss of the Boc-AAC prodrugs during
the aqueous work-up of the reaction mixture.
The relative yields of the compounds synthesized using Method
A or Method B have been shown in Table 2. Generally, the yields of
DAAC-APAP prodrugs were much higher (Table 1) than the AAC
series of prodrugs (Table 2). The results from the elemental analy-
sis of the synthesized prodrugs have been shown in Table S1 (Sup-
plementary data).
The trend in the pKa values of DAAC-APAP prodrugs (Table 3)
can be explained in terms of inductive effect of the carbonyl group.
The presence of an electron withdrawing carbonyl group in the
beta position causes the electron pair on the N atom to be less
‘available’. The increase in the distance between the carbonyl car-
bon and the amine group makes the amine increasingly more ba-
The overall yields of AAC prodrugs synthesized by DCC medi-
ated coupling were relatively higher than that obtained with EDCI
mediated coupling. Since the Boc-AAC-APAP prodrugs exhibit

H. Devarajan-Ketha, K. B. Sloan / Bioorg. Med. Chem. Lett. 21 (2011) 4078–4082
4081
Table 1
To investigate the transient nature of DAAC and the AAC pro-
drugs of APAP and NTX, kinetics of hydrolysis of some members
of the series was studied. The t1/2 values of the prodrugs have been
shown in Table 3. The presence of an amine group activates the
acyl group towards hydroxide ion attack because of the (ꢀI) effect
of the amine group and promotes its intramolecular general base
catalyzed hydrolysis.15,16 The t1/2 of these activated esters is much
smaller than unactivated ester groups which may make these com-
pounds attractive prodrug candidates. Kinetic behavior of esters
containing such an activated acyl group (metranidazole N,N-di-
methyl glycinate ester,17 hydrocortisone lysine ester18 and acet-
Percentage yields of the DAAC prodrugs
(DAAC-APAP-HCl) and the corresponding free base
forms (DAAC-APAP) from the hydrochlorides
% Yield
DAAC-APAP HCl
4a
4b
4c
4d
4e
4f
4g
4h
4i
75
85
70
80
50
71
65
78
70
50
61
aminophen glycine,
a )
-aspartic acid and b-aspartic acid esters12
has been studied in detail. Bungaard et al. and other investigators
have showed that the protonated and unprotonated form of the
prodrug undergoes hydrolysis at different rates.17–19 The half-lives
of the amino acid esters show a dependence on the pKa of the
amine in the molecule which governs the ratio of the protonated
to the unprotonated form at a particular pH.
4j
4k
DAAC-APAP
5a
5b
5c
5d
5e
5f
5g
5h
5i
95
95
78
61
—
71
75
85
70
80
61
The shortest t1/2 amongst all the prodrugs evaluated was
14.9 min for the PRO-APAP-HCl prodrug, 7c. This is an unexpected
result based on the steric factors alone. However, it has been previ-
ously observed that piperazine-2, 5-diones are formed readily from
dipeptide esters particularly those containing N-methylamino
acids or proline. In the case of peptides containing proline, geomet-
rical factors take control.20 For the cyclization of dipeptide esters to
piperazines to occur, the peptide bond must be in the cis conforma-
tion so that the terminal amino group and the ester carbonyl carbon
can interact to form the six-membered ring. For example, glycylpr-
oline ethyl ester cyclizes more readily than the prolylglycine ester
because of the ease in assumption of the cis conformation. The case
of the 7c can be explained by a similar explanation. Presence of the
proline ring system ‘fixes’ the –NH group in an orientation that fa-
vors more effective general base catalysis by the –NH group. In this
regard it is worth while to mention that Wu et al.13 recently re-
ported a decrease in the t1/2 of the proline ester prodrug of APAP
from 65 min at pH 5.0 to 3.5 min at pH 7.4, which correlates well
with our t1/2 value of 14.9 min at pH 6.0.
5j
5k
Table 2
Percentage yields of the Boc-AAC-APAP compounds and the
corresponding AAC prodrugs synthesized from Method A or
Method B
% Yield Method A
% Yield Method B
Boc-AAC-APAP
6a
6b
6c
—
—
—
60
46a
31a
AAC-APAP-HCl
The procedure for the diffusion cell experiments has been
described earlier.3 The measurement of the steady state flux of
the prodrug or the parent drug was followed by the measurement
of the steady state flux of a standard drug/vehicle combination
(theophylline/propylene glycol) as a control experiment to validate
the absence of any damage caused to the skin sample during the
7a
7b
7c
74
69
84
44
57
27
AAC-NTX
8
10
—
—
52
90
a
course of the experiment: control value of 1.02 l .
mol cmꢀ2 hꢀ1 21
Average of two experiments.
The results from diffusion cell experiments through hairless
mouse skin with MORn1-APAP (5i) and its hydrochloride
(MORn1-APAP.HCL, 4i) are listed in Table 4. The experimental flux
value (EXP log JM) for 5i was two times higher compared to APAP
whereas the EXP log JM value for the corresponding hydrochloride,
Table 3
Half lives (t1/2, min) and predicted pKa values of DAAC-APAP-
HCl, AAC-APAP-HCl and VAL-NTX-TFA prodrugs in buffer (pH
6.0, Phosphate, 50 mM, I = 0.5) at 37 2 °C
Table 4
a
pKa
t1/2 (min)
Maximum flux of the prodrugs through hairless mouse skin from
isopropyl myritate (IPM) donor phase (JMIPM), maximum flux of theophylline through
hairless mouse skin from saturated propylene glycol donor phase (JS), log
experimental (EXP) JMIPM calculated using Roberts–Sloan equation (CALC log JMIPM),
absolute difference between EXP and CALC log JMIPM values ( log JMIPM
a saturated
4a
4b
4c
4d
4f
4g
7a
7b
7c
10
6.94
8.66
8.08
9.59
6.98
8.65
8.62
7.46
8.65
—
76.92 0.24
113.03 2.85
105.71 4.45
79.32 11.92
—
79.67 1.25
32.33 0.28
91.37 3.45
14.90 0.81
31.78 1.52
a
D
)
a
a
JMIPM
JS
EXP
log JMIPM
CALC
log JMIPM
Dlog JMIPM
c
5i
4i
1.05 0.17 1.35 0.65
0.021
ꢀ0.203
—
0.262
0.224
—
0.472
0.54 0.02 0.96 0.25 ꢀ0.26b
APAP 0.51d
0.74d
ꢀ0.19
a
a
b
c
l .
mol cmꢀ2 hꢀ1
All experiments were run in triplicate.
Units of
Experiment run in duplicate.
Calculated from log CALC JMIPM = ꢀ0.599 + 0.502 ꢁ log SIPM + 0.498 ꢁ log
sic. The pKa of ethyl N,N-dimethylaminobutyrate is very close to di-
methyl amine indicating a weak inductive effect over three meth-
ylene groups.15
S4.0 ꢀ 0.00235 ꢁ MW, where SIPM = 2.52 mM and S4.0 = 50.51 mM (values from Ref.
18).
d
Values from Wasdo et al. (2000).

4082
H. Devarajan-Ketha, K. B. Sloan / Bioorg. Med. Chem. Lett. 21 (2011) 4078–4082
4i was only as high as APAP. The second application flux values JS
were not greater than the control value indicating that the flux
enhancement was not a result of damage to the skin sample by
the vehicle. The Roberts–Sloan equation21 was used to calculate
the flux of 5i (CALC log JM). The absolute difference between EXP
log JM and CALC log JM for 5i was 0.22 log units. A detailed
evaluation of the solubility and permeation properties of the pro-
drug candidates will be published elsewhere.
In conclusion, a highly efficient synthesis of DAAC and AAC pro-
drugs of phenol containing drugs, APAP and NTX has been re-
ported. The DAAC-APAP compounds can be scaled up without the
requirement of tedious column purification process by a simple
recrystallization step. In terms of prodrug design, an important
advantage of incorporating a basic amine functional group into
an acyl prodrug is that the physicochemical properties, for exam-
ple, basicity, aqueous or lipid solubilities (unpublished results)
and chemical stability of the corresponding prodrugs, can be tai-
lored by varying the amine group, adjusting the alkyl chain length
in the promoiety of the DAAC or an appropriate choice or R group
in the AAC prodrugs.
References and notes
1. Sloan, K. B.; Wasdo, S. C.; Rautio, J. Pharm. Res. 2006, 23, 2729.
2. Roberts, W. J.; Sloan, K. B. J. Pharm. Sci. 1999, 88, 515.
3. Wasdo, S. C.; Sloan, K. B. Pharm. Res. 2004, 21, 940.
4. Thomas, J. D.; Sloan, K. B. Int. J. Pharm. 2008, 346, 80.
5. Thomas, J. D.; Sloan, K. B. Int. J. Pharm. 2009, 371, 25.
6. Majumdar, S.; Sloan, K. B. Int. J. Pharm. 2007, 337, 48.
7. Vaddi, H. K.; Banks, S. L.; Chen, J.; Hammell, D. C.; Crooks, P. A.; Stinchcomb, A.
L. J. Pharm. Sci. 2009, 98, 2611.
8. Vaddi, H. K.; Hamad, M. O.; Chen, J.; Banks, S. L.; Crooks, P. A.; Stinchcomb, A. L.
Pharm. Res. 2005, 22, 758.
9. Thomas, J. D.; Majumdar, S.; Sloan, K. B. Molecules 2009, 14, 4231.
10. Milewski, M.; Yerramreddy, T. R.; Ghosh, P.; Crooks, P. A.; Stinchcomb, A. L. J.
Controlled Release 2010, 146, 37.
11. Ostacolo, C.; Marro, F.; Laneri, S.; Sacchi, A.; Nicoli, S.; Padula, C.; Santi, P. J.
Controlled Release 2004, 99, 403.
12. Kovach, I. M.; Pitman, I. H.; Higuchi, T. J. Pharm. Sci. 1981, 70, 881.
13. Wu, Z.; Patel, A.; Dave, R.; Yuan, X. Bioorg. Med. Chem. Lett. 2010, 20,
3851.
14. Milosovich, S.; Hussain, A.; Dittert, L.; Aungst, B.; Hussain, M. J. Pharm. Sci.
1993, 82, 227.
15. Hay, R. W.; Porter, L. J.; Morris, P. J. Aust. J. Chem. 1966, 19, 1197.
16. Kirby, A. J.; Lloyd, G. J. J. Chem. Soc., Perkin Trans. 2 1976, 1748.
17. Bundgaard, H. In Design of Prodrugs; Bundgaard, H., Ed.; Elsevier Science
Publishers B. V. (Biomedical Division), 1985; pp 1–30.
18. Johnson, K.; Amidon, G. L.; Pogany, S. J. Pharm. Sci. 1985, 74, 87.
19. Dhareshwar, S. S.; Stella, V. J. In Prodrugs: Challenges and Rewards; Springer,
2008; Vol. 3,
Supplementary data
20. Benoiton, L.; Purdie, J. E. J. Chem. Soc., Perkin Trans. 2 1973, 1845.
21. Devarajan-Ketha, H. Ph. D. Theses, University of Florida, August 2010.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2011.04.118.