A new amino-masking group capable of pH-triggered amino-drug release

Article abstract of DOI:10.1016/j.ejps.2004.11.006

The prodrug approach is potentially useful for mitigating pharmaceutical problems - such as poor membrane permeability or stability - which commonly occur with amino drugs. On the other hand there persists a dearth of useful systems for masking amines tha

Full text of DOI:10.1016/j.ejps.2004.11.006

European Journal of Pharmaceutical Sciences 24 (2005) 315–323
A new amino-masking group capable of pH-triggered amino-drug release
∗
1
´
´
John F. Gilmer , Ana Luısa Simplıcio , John M. Clancy
Department of Pharmaceutical Chemistry, School of Pharmacy, Trinity College, Dublin 2, Ireland
Received 9 September 2004; received in revised form 11 November 2004; accepted 18 November 2004
Abstract
The prodrug approach is potentially useful for mitigating pharmaceutical problems—such as poor membrane permeability or
stability—which commonly occur with amino drugs. On the other hand there persists a dearth of useful systems for masking amines that satisfy
the prodrug requirements of good in vitro stability coupled with predictable and rapid drug release in response to a local tissue condition. This
study describes an evaluation of aminoindanes as bioreversibly masked amines poised to undergo elimination to the parent amine. Several
model amine and amino drug indanone derivatives were synthesised. pKa values were determined by capillary electrophoresis and pH rate
profiles for elimination and amine liberation were measured. The aminoindanone system appears to have particular applicability to secondary
amino substrates whose indanone derivatives are stable at low pH but undergo drug release at rates corresponding to first-order half-lives of
<5 min at pH 7.4.
© 2004 Elsevier B.V. All rights reserved.
Keywords: Amine; Prodrug; pH-sensitive release; Indanone; Indenone
1. Introduction
point of view, is the greater susceptibility of esters to un-
dergo hydrolysis by esterases, and the general chemical and
enzymatic stability of amides and other amine derivatives.
Among the numerous attempts to build chemical or enzy-
matic vulnerability into amino derivatives are the Mannich
bases (Bundgaard and Johansen, 1980), imines (Krause et
al., 1995), carbamates (Gogate and Repta, 1987), enaminones
(Naringrekar and Stella, 1990), tetrahydrothiadizine-2-tiones
(Aboul-Fadl and El-Shorbagi, 1996), as well as coumarin-
based systems that are poised to undergo spontaneous cy-
clisation (Wang et al., 2000) and analogous ‘trimethyl lock’
designs (Amsberry and Borchardt, 1991). Notwithstanding
these endeavours, there remains a dearth of amine-masking
systems with suitable release kinetics to be useful in prodrug
design.
The prodrug strategy involves the temporary masking of
a drug functional group, usually in an effort to improve a
candidate drug’s stability (Gershonov et al., 2000), safety
(Shanbhag et al., 1992) or passive oral drug delivery (Aungst
et al., 1995). In general, the success of the approach hinges
upon a marked contrast between the stability of the prodrug
forminvitro, anditsvulnerabilitytoattackanddrugreleasein
response to an enzymatic or local tissue pH condition follow-
ing administration. The prodrug approach has been particu-
larly effective in addressing pharmaceutical problems with
carboxyl- or hydroxyl-bearing drugs, mainly due to the very
wide possibilities for ester prodrug design. Progress in the
area of amino prodrug design has been somewhat slower.
The major differences between amino derivatives such as
amides or carbamates and carboxylic esters from a prodrug
We have recently reported on the behaviour of certain 3-
aminoindanones (Fig. 1, 1) which are stable at low pH val-
ues but undergo unexpectedly rapid elimination approaching
neutrality, generating in the process, the highly conjugated in-
∗
Corresponding author. Tel.: +353 1 608 2795; fax: +353 1 608 2793.
´
denone 3 (Simplıcio et al., 2004). The objective of the present
E-mail address: gilmerjf@tcd.ie (J.F. Gilmer).
Present address: Laborato´rio Associado, ITQB/IBET/IGC, Av. da
1
study was to determine if this behaviour could be exploited
Repu´blica, 2781-901 Oeiras, Portugal.
in the design of prodrug systems where the amine 2 could be
0928-0987/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.ejps.2004.11.006

316
J.F. Gilmer et al. / European Journal of Pharmaceutical Sciences 24 (2005) 315–323
the preparation of buffers for kinetic studies. Acetonitrile
¨
(Riedel-de Haen) for HPLC was used for the preparation
of stock solutions. Aqueous solutions were prepared with
distilled and deionised water (Milli-Q Water System,
Millipore).
Fig. 1. Scheme showing the elimination/addition equilibrium of aminoin-
danones at 37 ◦C.
2.2. Synthesis
The synthetic approach to the candidate aminoindanones
is outlined in Fig. 2. Indanone (4) was brominated at the ben-
zylic position using the Wohl-Ziegler bromination (Huang
and Williams, 1958) employing N-bromosuccinimide (NBS)
in carbon tetrachloride and a catalytic quantity of diben-
zoylperoxide at reflux for 45 min. The 3-bromoindanone
product (53%) was obtained following evaporation and sepa-
ration by flash chromatography from its 2-bromoindanone
isomer (24%). The aminoindanones were generated by
treating the 3-bromoindanone with the amine in dry
dichloromethane in the presence of a tertiary base. The com-
pounds were obtained in 25–87% yield following evapora-
tion and flash chromatography. A general synthetic method
is given below under Section 2.3. d,l-Tryptophan and d,l-
alanine methyl esters were used because the indanone deriva-
tives of these were more easily synthesised and purified than
the corresponding amino acid derivatives. Similarly, l-dopa
ethyl ester was used in preference to l-dopa. Where the amine
was achiral, the product 3-aminoindanone was obtained as a
mixture of isomers. The other chiral amines (alanine, tryp-
tophan, ephedrine, l-dopa ethyl ester) gave more complex
product mixtures and the approach to dealing with each of
these cases is described in the synthetic section. Generally,
it was found that mixtures of diastereomers could be collec-
tively analysed without any difficulty. The aminoindanones
Table 1
Amines used for the preparation of the indenone derivatives 1a–1k
Compound
RRꢀNH (amine)
1a
1b
1c
1d
1e
1f
1g
1h
1i
Dimethylamine
Ethylamine
Piperidine
Ephedrin
Tryptophan methyl ester
Alanine methyl ester
Phenylethyl amine
Atenolol
Desloratadine
Dopamine
1j
1k
l-Dopa ethyl ester
a candidate drug molecule. Such systems might be employed
in the temporary masking of amines in order to increase drug
stability or lipophilicity, for example, in the protection of
peptides and protein from protease-mediated decomposition
by reversibly protecting basic amino acid residues. This pa-
per describes the synthesis and physicochemical evaluation
of eleven model aminoindanone derivatives as potential pro-
drugs of the parent amine (Table 1).
1
were characterised by H and 13C NMR, IR and high res-
2. Materials and methods
olution mass spectroscopy (HRMS) and shown to be pure
(>99% by area normalisation) by capillary electrophoresis,
TLC and HPLC. IR spectra were obtained using a Perkin-
Elmer Paragon 1000 FT infrared spectrometer. Liquid or oil
samples were analysed as neat films from dichloromethane
on NaCl plates; solid samples were analysed using KBr
tablets. Wave numbers (IR νmax) are presented for charac-
teristic functional groups and are expressed in cm−1. 1H and
13C NMR were recorded at 20 ◦C on a Brucker DPX 400
spectrophotometer (400.13 MHz 1H, 100.61 MHz 13C) at the
2.1. Materials
´
Atenolol was obtained from Societa Italiana Medic-
inali Scandicci. Desloratadine (micronised USP) was
donated by Schering-Plough (Avondale). Ephedrine, d,l-
tryptophan ethyl ester hydrochloride and alanine ethyl
ester hydrochloride were obtained from Sigma. l-3,4-
Dihydroxyphenylalanine 99%, dimethylamine hydrochlo-
ride, piperidine, 2-phenylethylamine 99%, 1-indanone
99%, 2-methyl-1-indanone 99%, N-bromosuccinimide,
benzoyl peroxide (70% remainder water) and triethylamine
were purchased from Aldrich. Merck silica gel 60 (par-
ticle size 0.040–0.063 mm) was used for flash column
chromatography. Carbon tetrachloride was from Riedel-de
¨
Haen. Phosphoric acid (+99%, Fluka), sodium dihydrogen
orthophosphate (BDH) and tetrabutylammonium dihydro-
genphosphate (97%, Aldrich) were used for the preparation
of running buffers for capillary electrophoresis. Citric
acid monohydrate (99% ACS, Aldrich), boric acid (M&B)
and tripotassium orthophosphate (BDH) were used for
Fig. 2. Scheme showing the synthesis of aminoindanones: (i) N-
bromosuccinimide, CCl4; (ii) R2NH, Et3N, DCM.

J.F. Gilmer et al. / European Journal of Pharmaceutical Sciences 24 (2005) 315–323
317
Department of Chemistry, Trinity College Dublin. Samples
were dissolved in deuterated chloroform (CDCl3) or deuter-
ated dimethyl sulfoxide ((CD3)2SO). Chemical shifts are in
ppm. Coupling constants are in Hz. 1H shifts were assigned
relative to the tetramethylsilane (TMS) peak at 0.00 ppm
and 13C shifts were assigned relative to the central peak
of the CDCl3 triplet at 77.0 ppm or relative to the middle
peak of the (CD3)2SO septet at 39.7 ppm. 1H NMR as-
signments are reported in the following form: shift values
(number of protons, multiplicity and shape, coupling con-
stant (when applicable), proton assignment). 13C NMR as-
signments are reported in the following form: shift values
(number of carbons (if more than one), carbon assignment).
HRMS were acquired on a Micromass spectrometer (TOF,
EI mode) at the Department of Chemistry, Trinity College,
Dublin.
Jgem = 18.6, Jvic = 3.0, CHCH2C O), 2.76(2H, m, CH2CH3),
2.96 (1H, dd, Jgem = 18.6, Jvic = 6.5, CHCH2C O), 4.47
(1H, dd, J = 3.0, 6.5, NHCHCH2C O), 7.42 (1H, t, J = 7.5,
CHCHCC O), 7.61–7.68 (2H, m, CHCHCCHNH) 7.73
(1H, d, J = 7.5, CHCC O). δC (CDCl3) 15.3 (CH3), 41.8
(CH2CH3), 44.7 (CHCH2C O), 56.0 (NHCHCH2), 123.2
(CHCHCC O), 125.8 (CHCCHNH), 128.5 (CHCC O),
134.7 (CHCHCCHNH), 136.7 (CCHNH), 155.9 (CC O),
204.5 (C O). m/z 176.1070 (MH+, expected: 176.1075).
3-Piperidin-1-yl-indan-1-one 1c was prepared from
3-bromoindan-1-one (0.21 g, 1 mmol) and piperidine
(0.08 g, 1 mmol). The product was isolated by flash
chromatography with (hexane/ethyl acetate, 8:2) as a
brown oil (0.18 g, 84%). IR (max (NaCl plate) 1715
(C O), 2933 (aliphatic C C) cm−1. δH (CDCl3) 1.45
(2H, m, NCH2CH2CH2), 1.56 (4H, m, NCH2CH2), 2.30
(2H, m, NHCH2), 2.47 (2H, m, NHCH2), 2.61 (1H,
dd, Jgem = 18.8, Jvic = 7.0, CHCH2C O), 2.77 (1H, dd,
Jgem = 18.8, Jvic = 3.5, CHCH2C O), 4.53 (1H, dd, J = 3.5,
7.0, NCH), 7.42 (1H, t, J = 7.52, CHCHCHCC O),
7.63 (1H, t, J = 7.52, CHCHCC O), 7.73 (2H, m,
CHCHCHCHCC O). δC (CDCl3) 24.5 (NCH2CH2CH2),
26.2 (2C, NCH2CH2), 36.4 (2C, NCH2CH2), 49.7
(CHCH2C O), 63.2 (NCH), 123.0 (CHCC O), 126.7
(CHCHCHCHCC O), 128.4 (CHCHCHCC O), 134.5
(CHCHCC O), 137.5 (CCHCHCHCHCC O), 154.7
(CC O), 204.7 (C O). m/z 216.1393 (MH+, expected:
216.1388).
3-[(2-Hydroxy-1-methyl-2-phenyl-ethyl)-methyl-amino]-
indan-1-one or 3-ephedrin-indan-1-one 1d was prepared
from ephedrine (0.21 g, 1 mmol) and 3-bromoindan-1-one
(0.21 g, 1 mmol). Ephedrine has two chiral centres but con-
sists in only one of the enantiomeric pairs (R,S:S,R). There-
fore, the product of the reaction was obtained as four isomers
or two pairs of enantiomers (R,S,S:R,S,R:S,R,S:S,R,R). The
product was purified by flash chromatography (hexane/ethyl
acetate, 8:2) and isolated as a mixture of isomers (0.12 g,
41%). After a few days, the mixture solidified affording
a yellow solid. IR νmax (KBr) 3429 (OH), 1692 (C O)
cm−1. δH (CDCl3) 1.06, 1.10 (3H, 2d, J = 7.0, CH3CH),
1.91, 2.16 (3H, 2s, NCH3) 2.52–2.75 (2H, m, CH2),
2.92, 3.14 (1H, 2m, NCHCH3), 4.74 (1H, m, NCHCH2),
4.87, 4.95 (1H, 2d, J = 5.5, J = 4.5). ␦C (CDCl3) 11.8,
12.0 (CH3CH), 30.1 33.8 (NCH3) 37.2, 37.5 (CH2), 60.7
(CH2CHN), 63.6, 64.0 (CH3CHN) 73.8, 74.1 (CHOH),
123.0, 123.1 7(CHCHCHCCHOH), 126.0 126.2 (2C,
CHCCHOH) 126.1, 126.2 (CHCHCC O), 127.1, 127.2
(CHCC O), 128.1, 128.2 (2C, CHCHCCHOH), 128.5, 128.6
(CHCHCHCHCC O), 134.8, 134.9 (CHCHCHCC O),
137.1, 137.2 (CCHCHCHCHCC O), 142.0, 142.8
(CCHOH), 155.5, 155.6 (CC O), 204 (C O) m/z 296.1646
(MH+, expected: 296.1651).
2.3. General synthesis of amino-indanone analogues
A
solution of 3-bromoindanone (0.21 g) in dry
dichloromethane (20 ml) was cooled to 0 ◦C when tri-
ethylamine (1.9 eq) was added, followed by the amine (1 eq).
The reaction was stirred for 3–12 h at 0 ◦C and monitored
by TLC. The desired aminoindanones were obtained fol-
lowing evaporation of the solvent and flash chromatography
(25–87%) 1a–1k (Table 1). When the hydrochloride form
of the amine was used, the quantity of triethylamine was
doubled. Each reaction was stirred at 0 ◦C for 3–12 h and
the reactions monitored by TLC. The solvent was removed
by evaporation and the residue purified by flash column
chromatography.
3-Dimethylaminoindan-1-one 1a was prepared from
dimethylamine hydrochloride (0.08 g, 1 mmol) and 3-
bromoindanone (0.21 g, 1 mmol). The reaction mixture
was separated by flash chromatography (hexane/ethyl ac-
etate, 8:2) and the title compound isolated as an am-
ber oil (75 mg, 43%). IR νmax (NaCl plate) 1715
(C O) cm−1. δH (CDCl3) 2.06 (6H, s, CH3NCH3),
2.61 (1H, dd, Jgem = 19.0, Jvic = 7.0, CHCH2C O), 2.74
(1H, dd, Jgem = 19.0, Jvic = 3.0, CHCH2C O), 4.61 (1H,
dd, J = 3.0, 7.0, CHCH2C O), 7.47 (1H, t, J = 7.5
CHCHCHCC O), 7.67 (1H, t, J = 7.5, CHCHCC O),
7.73 (1H, d, J = 7.5, CHCHCHCHCC O), 7.78 (1H,
d, J = 7.5, CHCC O). δC (CDCl3) 35.1 (CHCH2C O),
40.5 (2C, CH3), 62.7 (CHCH2C O), 123.2 (CHCC O),
126.7 (CHCHCHCHCC O), 128.8 (CHCHCHCC O),
134.8 (CHCHCC O), n.d. (CCHCHCHCHCC O), 206.7
(CC O), m/z 176.1066 (MH+, expected: 176.1075).
3-Ethylamino-indan-1-one 1b was prepared as described,
from ethylamine hydrochloride (0.16 g, 2 mmol) and 3-
bromoindanone (0.42 g, 2 mmol). The solvent was re-
moved and the residue subjected to flash chromatogra-
phy (dichloromethane/methanol, 95:5) and the title com-
pound obtained as a brown oil (0.14 g, 41%). IR νmax
(NaCl plate) 1713 (C O) cm−1. δH (CDCl3) 1.16 (3H,
d, J = 7.0, CH3), 1.68 (1H, br, S, NH), 2.52 (1H, dd,
3-(1H-Indol-3-yl)-2-(3-oxo-indan-1-ylamino)-propionic
acid methyl ester, or 3-(Tryptophan methyl ester)-indan-
1-one 1e was prepared from d,l-tryptophan methyl ester
hydrochloride (0.50 g, 2 mmol) and 3-bromoindan-1-one

318
J.F. Gilmer et al. / European Journal of Pharmaceutical Sciences 24 (2005) 315–323
(0.42 g, 2 mmol). Two racemic pairs were obtained and they
were not separated. The product purified by flash chro-
matography (hexane/ethyl acetate 6:4) yielding an amber
oil (0.61 g, 87%). IR νmax (NaCl plate) 3040 (indole), 1714
(C O) cm−1. δH (CDCl3) 2.22 (1H, br, s, NHCHCOOCH3),
2.35, (1H, dd, Jgem = 18.6, Jvic = 3.0, NHCHCH2C O), 2.48
(1H, dd, Jgem = 18.6, Jvic = 3.0, NHCHCH2C O), 2.87*
(1H, dd, Jgem = 18.6, Jvic = 6.5, NHCHCH2C O), 2.91*
(1H, dd, Jgem = 18.6, Jvic = 6.5, NHCHCH2C O), 3.14–3.33
(2H, m, CCH2CHNH), 3.68, 3.77 (3H, 2s, CH3), 3.80, 3.89
(1H, 2m, NHCHCOOCH3), 4.39 (1H, m, NHCHCH2C O),
7.00, 7.06 (1H, dd, J = 1.5, J = 25.0 CNHCHC), 7.13–7.26,
(2H, ArH), 7.28–7.43 (2H, ArH), 7.48, 7.46–7.70 (3H, m,
ArH), 8.55, 8.62 (1H, d, J = 25.0, CNHCH). δC (CDCl3)
29.5, 29.6 (CCH2CHNH), 44.7, 45.2 (O CCH2CHNH),
51.8, 51.8 (CH3), 54.6, 55.4 (O CCH2CHNH), 60.2, 60.8
(CCH2CHNH), 110.5, 110.6 (CHCCCHNH), 111.2 (ArCH),
118.3, 118.5 (ArCH), 119.2, 119.2 (ArCH), 121.9 (ArCH),
122.9, 123.0 (ArCH), 123.0, 123.1 (ArCH), 125.5 126.0
(ArCH), 127.1, 127.3 (C), 128.5, 128.6 (ArCH), 134.5,
134.8 (ArCH), 136.1 (C), 136.4, 136.5 (C), 155.2, 155.3
(CHCCHCH2C O), 175.3, 175.4 (O COCH3) 204.2, 204.3
(C O). m/z 349.1563 (MH+, expected: 349.1552).
7.43 (1H, t, J = 7.3, CHCHCHCC O), 7.57–7.65 (2H,
m, CHCHCHCHCC O), 7.74 (1H, d, J = 7.5, CHCC O).
δC (CDCl3) 36.3 (NCHCH2), 44.4 (NCH2CH2), 48.3
(NCH2CH2), 55.8 (NCHCH2), 123.2 (CHCC O), 125.7
(CHCHCHCCH2CH2), 126.2 (CHCHCHCHCC O), 128.4
(3C, CHCHCCH2CH2 and CHCHCHCC O), 128.6 (2C,
CHCCH2CH2), 134.8 (CHCHCC O), 136.7 (CCH2CH2),
139.4 (CCHCHCHCHCC O), 155.5 (CC O), 204.4 (C O).
m/z 252.1386 (MH+, expected: 252.1388).
2-(4-{2-Hydroxy-3-[isopropyl-(3-oxo-indan-1-yl)-
amino]-propoxy}-phenyl)-acetamide, or, 3-atenolol-indan-
1-one 1h was prepared from atenolol (RS) (0.27 g, 1 mmol)
and 3-bromoindanone (0.21 g, 1 mmol). The evaporated
residue was purified by flash chromatography using
(DCM/methanol, 92.5:7.5) which yielded a single pair
of enantiomers uncontaminated (the second pair were
discarded) (0.094 g, 0.24 mmol, 11.8%). IR νmax (NaCl
plate) 1670, 1707, (C O) 3354 (NH2) cm−1. δH (CDCl3)
1.13 (3H, d, J = 6.4, CH3), 1.25 (3H, d, J = 6.4, CH3), 2.44
(1H, dd, Jgem = 13.6, Jvic = 6.1, NCH2CH), 2.63 (1H, dd,
Jgem = 13.6, Jvic = 6.1, NCH2CH), 2.70 (1H, dd, Jgem = 19.1,
Jvic = 3.4, CHCH2C O), 2.77 (1H, dd, Jgem = 19.1,
Jvic = 6.8, CHCH2C O), 3.14 (1H, m, CH3CHCH3), 3.52
(2H, s, CH2C ONH2), 3.66 (2H, m, CHCH2O), 3.90 (1H, m,
CHOH), 4.70(1H, dd, J = 3.4, 6.8, NCHCH2C O), 5.48, (1H,
s, br) 5.72 (1H, s, br), 6.75 (2H, d, J = 8.9, OCCHCHCCH2),
7.16 (2H, d, J = 8.9, OCCHCHCCH2), 7.40 (1H, t,
J = 7.5, CHCHCC O), 7.57 (1H, m, CHCHCHCC O),
7.65 (1H, d, J = 4.1, CHCHCHCHCC O), 7.73 (1H,
d, J = 7.5, CHCC O). m/z 397.2121 (MH+, expected:
397.2127).
2-(3-Oxo-indan-1-ylamino)-propionic acid methyl ester
or 3-(alanine methyl ester)-indan-1-one 1f was prepared
from d,l-alanine methyl ester (0.279 g, 2 mmol) and
3-bromoindanone (0.42 g, 2 mmol). The product was
purified by flash chromatography (hexane/ethyl acetate,
2:1). An amber oil was obtained (0.12 g, 0.50 mmol) as
a mixture of two pairs of enantiomers, which were not
separated. 25%. IR νmax (NaCl plate) 1715 (C O) cm−1
.
δH (CDCl3) 1.18, (3H, d, Jvic = 4, CHCH3) 1.22* (3H,
d, Jvic = 4, CHCH3), 2.11 (1H, br, s, NH), 2.42 (1H, dd,
Jgem = 18.7, Jvic = 3.5, CH2), 2.44 (1H, dd, Jgem = 18.7,
Jvic = 2.9, CH2), 2.85* (1H, dd, Jgem = 18.7, Jvic = 6.4, CH2),
2.92* (1H, dd, Jgem = 18.7, Jvic = 7.0, CH2), 3.45 (1H, q,
J = 7.0, CH3CHNH), 3.55* (1H, q, J = 7.0, CH3CHNH),
4.31 (1H, dd, J = 2.9, 6.4, CHCH2C O), 4.41* (1H, dd,
J = 3.5, 7.0, CHCH2C O), 7.41 (1H, m, CHCHCHCC O),
7.63 (3H, m, CHCHCHCHCC O), δC (CDCl3) 19.2,
19.7 (CH3CH), 44.9, 45.4 (CH2), 52.0 (OCH3),
54.3 (NCHCH2), 54.7, 55.1 (NCHCH3), 123.2, 123.4
(CHCC O), 125.4, 126.2 (CHCHCHCHCC O), 128.7,
128.8 (CHCHCHCC O), 134.7, 135.0 (CHCHCC O),
136.55, 136.6 (CCHCHCHCHCC O), 155.2, 155.4
(CC O), 176.0, 176.2 (COOCH3), 204.0, 204.2 (C O). m/z
234.1135 (MH+, expected: 234.1130).
3-Phenethylamino-indan-1-one 1g was prepared from
phenylethylamine (0.12 g, 1 mmol) and 3-bromoindan-1-one
(0.21 g, 1 mmol). The product isolated by flash chromatog-
raphy (hexane/ethyl acetate, 8:2) as a brown oil (0.09 g,
36%). IR νmax (NaCl plate) 1712 cm−1 (C O). δH (CDCl3)
2.50 (1H, dd, Jgem = 18.8, Jvic = 3.2, NHCHCH2C O),
2.84–3.05 (5H, NHCHCH2C O, NHCH2CH2), 4.48 (1H,
dd, J = 3.2, 6.8, NHCHCH2C O), 7.22–7.28 (3H, m,
CCHCHCHCHCH), 7.30 (2H, m, CCHCHCHCHCH),
3-[4-(8-Chloro-5,6-dihydro-benzo[5,6]cyclohepta[1,2-
b]pyridin-11-ylidene)-piperidin-1-yl]-indan-1-one,
or
3-desloratadine-indan-1-one 1i was prepared from deslo-
ratadine (0.31 g, 1 mmol) and 3-bromoindan-1-one (0.21 g,
1 mmol). After evaporation, the reaction mixture residue was
purified by flash column chromatography (DCM/methanol,
96:4). A reddish crystalline solid was obtained (0.38 g,
86%). IR νmax (KBr) 1710 cm−1 (C O). δH (CDCl3),
2.27–2.50 (6H, m, CH2), 2.5–2.77 (2H, m, CH2) 2.79–2.96
(4H, m, CH2), 3.31–3.42 (2H, m, CH2), 4.62 (1H, m,
CH2NCH), 7.05–7.15 (4H, m, ArCH), 7.41–7.45 (2H,
m, ArCH), 7.65 (1H, t, J = 7.9, ArCH), 7.72–7.77 (2H,
m, ArCH), 8.38 (1H, m, ArCH). δC (CDCl3) 31.3, 31.7
(2C, CCH2CH2C), 36.1, 36.3 (2C, CH2CCH2), 45.8
(NCHCH2), 48.5, 48.6 (2C, NCH2), 62.5 (NCHCH2), 122.1
(ArCH), 123.1 (ArCH), 125.9 (ArCH), 126.7 (ArCH), 128.7
(ArCH), 128.9 (ArCH), 130.7 (ArCH), 132.6 (C), 133.0
(C), 133.4 (C), 134.8 (ArCH), 137.3 (ArCH), 137.5 (C),
138.3 (C), 139.4 (C), 146.5 (Ar-NCH), 157.3 (C), 175.0
(CC O), 207.7 (C O). m/z 441.1750 (MH+, expected:
441.1734).
2-[2-(3,4-Dihydroxy-phenyl)-ethylamino]-indan-1-one, or
3-dopamine-indan-1-one 1j was prepared from dopamine
hydrochloride (0.19 g, 1 mmol) and 3-bromoindan-1-one
(0.21 g, 1 mmol) according to the described procedure

J.F. Gilmer et al. / European Journal of Pharmaceutical Sciences 24 (2005) 315–323
319
with the exception that the mixture was stirred for 3
2.4. Kinetics experiments
days. The product was purified by flash chromatography
(DCM/methanol 95:5) affording a yellowish solid (50 mg,
18%) with a melting point of 198–200 ◦C (with degrada-
tion). IR νmax (KBr) 3311 (OH) 1693 (C O) cm−1. H H
and C H COSY were used for assignment of the NMR
signals. δH (CDCl3/(CD3)2SO) 2.84 (2H, m, NHCH2CH2),
2.85 (1H, dd, Jgem = 19.0, Jvic = 3.0, CHCH2C O), 3.04
(1H, dd, Jgem = 19.0, Jvic = 7.5, CHCH2C O), 3.15 (2H,
m, NHCH2CH2), 5.03 (1H, d, J = 6.0, NCH), 6.48 (1H,
dd, J = 2.0, 8.0, CH2CCHCHCOH), 6.65 (1H, d, J = 2.0,
CH2CCHCHCOH), 6.68 (1H, d, J = 8.0, CH2CCHCOH),
7.59 (1H, t, J = 7.52, CHCHCC O), 7.72–7.77 (2H, m,
CHCHCHCHCC O), 8.05 (1H, d, J = 8.0, CHCC O), 8.46
(1H, s, COH), 8.48(1H, s, COH). ␦C (CDCl3/(CD3)2SO)30.0
(CH2CH2NH), 38.0 (CHCH2C O), 45.9 (CH2CH2NH),
53.5 (CHCH2C O), 114.3 (HOCCHCCH2CH2NH), 114.6
(CHCHCCH2CH2NH), 118.0 (CHCHCCH2CH2NH),
121.8 (CHCC O), 125.9 (CCH2CH2NH), 126.2
(CHCHCHCHCC O), 129.0 (CHCHCHCC O), 133.6
(CHCHCC O), 136.1 (CCHCHCHCHCC O), 142.7
(HOCCHCHC), 143.8 (HOCCHC), 146.2 (CC O), 198.7
(C O). m/z 284.1295 (MH+, expected: 284.1287).
2.4.1. Aqueous kinetics samples
The disappearance of the prodrugs was studied in the pH
range of 0.5–11.7. For the preparation of the working buffers
in the range 2–11.7, two stock solutions were used. Solution
A was 0.05 M of citric acid monohydrate and 0.02 M of boric
acid in distilled and deionised water. Solution B was 0.1 M in
tripotassium-orthophosphate in distilled and deionised water.
The two solutions were mixed and diluted in the necessary
proportions to achieve the desired pHs and ionic strength of
0.154. For pH values under 2, HCl solutions were used: ionic
strength was set by addition of NaCl where appropriate. Typ-
ically, stock solutions of the compounds under investigation
of approximately 1–5 mg/ml in acetonitrile were prepared as
appropriate. Stock solution (20–100 ␮l) were diluted in 1 ml
of the working buffer and pre-warmed at 37 ◦C to obtain solu-
tions with a final concentration of approximately 100 ␮g/ml.
Each solution was introduced into the autosampler of the CE
warmed to 37 ◦C and injections started immediately using the
method described in 2.3.3.
2.4.2. Plasma kinetics samples
3-(3,4-Dihydroxy-phenyl)-2-(3-oxo-indan-1-ylamino)-
propionic acid ethyl ester, or 3-etilevodopa-indan-1-one
1k was prepared from l-dopa ethyl ester (etilevodopa)
(0.26 g, 1 mmol) and 3-bromoindan-1-one (0.21 g, 1 mmol).
The product was obtained following flash chromatography
(DCM/methanol, 97.5:2.5) as a brown paste (0.16 g, 45%).
According to the NMR integration, two diastereomers were
obtained in a proportion of 34/66. IR νmax (NaCl plate)
1707 cm−1 (C O). Assignment of the NMR shifts to protons
and carbons was confirmed by C H and H H COSY. δH
(CDCl3) 1.19* (3H, t, J = 7.2, CH3), 1.25 (3H, t, J = 7.2,
CH3), 2.33–2.43* (1H, m, NHCHCH2C O), 2.76–2.92*
(3H, m, CH2CHCOOC2H5, NHCHCH2C O), 3.55–3.61*
(1H, m, CH2CHCOOC2H5), 4.07–4.12*, 4.19–4.25 (2H, m,
CH2CH3), 4.33–4.39* (1H, m, NHCHCH2C O), 6.52–6.54
(1H, dd, J = 1.8, 7.6, CH2CCHCHCOH), 6.59–6.61*
(1H, dd, J = 1.8, 8.2, CH2CCHCHCOH), 6.65 (1H, m,
CH2CCHCHCOH), 6.71* (1H, m, CH2CCHCHCOH)
6.73–6.78* (1H, d, J = 8.2, CH2CCHCOH), 7.39–7.42*
(1H, m, CHCHCHCC O), 7.53–7.57* (1H, m,
CHCHCHCHCC O), 7.58–7.62* (1H, m, CHCHCC O),
7.69–7.72* (1H, m, CHCC O). δC (CDCl3) 14.1*, 14.2
(CH3), 39.1, 39.2* (CH2CHCOOC2H5), 44.5, 45.1*
(NHCHCH2C O), 54.6, 55.4* (NHCHCH2C O), 61.2*,
61.3 (CH2CH3), 61.4, 61.4* (CHCOOC2H5), 115.2*
(CH2CCHCOH), 116.1,116.2* (CH2CCHCHCOH), 121.2,
121.3* (CH2CCHCHCOH), 123.4*, 123.5 (CHCC O),
125.7, 126.3* (CHCHCHCHCC O), 128.8*, 128.9
(CHCHCHCC O), 129.1 (CH2CCHCOH), 135.0*, 135.3
(CHCHCC O), 136.6 (CCHCHCHCHCC O), 143.0
(CH2CCHCHCOH), 144.0 (CH2CCHCOH), 155.1, 155.2*
(CC O), 174.7,175.0* (COOC2H5), 205.0 (CH2C O) m/z
356.1497 (MH+, expected: 356.1498).
Citrated human blood was obtained from volunteers in
the School of Pharmacy, Trinity College Dublin. Plasma
was separated by centrifugation (3000 rpm for 10 min).
Aminoindanone stock solutions (10–100 ␮l) were diluted in
plasma (1–3 ml) to achieve the target concentration (0.4 mM).
Aliquots from plasma samples were taken at time intervals
and quenched with a 5% (v/v) perchloric acid solution or
a 4% (v/v) perchloric acid and 1 g/l EDTA solution (1.5/1,
v/v). These quenched samples were centrifuged for 3 min at
10,000 rpm and the supernatant analysed by CE or HPLC.
2.4.3. Capillary electrophoresis
CE was performed in a Beckman P/ACE system 5510
equipped with a UV filter detector set at 200 or 214 nm. The
system also had an autosampler with cooling facility which
was modified by connection to a water circulator. For ki-
netic tests the water temperature was set so that, inside the
sample vials, a temperature of 37 1 ◦C was achieved. The
fused silica capillary was 12 cm effective length (20 cm to-
tal length) and an internal diameter of 50 ␮m. Samples were
loaded by pressure injection for 5 s. Runs were carried out
at 25 ◦C and at constant current of 100 or 150 ␮A in the
direction of the cathode. New capillaries were conditioned
beforehand with 0.1 M NaOH followed by deionised water
for 5 min and the running buffer for 5 min. Between runs, the
capillary was rinsed with the running buffer for 1 min. The
running buffer was a phosphate buffer (pH 3, 100 mM) to
which 100 mM tetrabutylammonium phosphate (TBA) was
added. The buffer was filtered through a 0.45 ␮m membrane
filter before use. Data acquisition was performed by the sys-
tem Gold and peak areas were recorded at 200 or 214 for the
original compound and the parent amine (when applicable).

320
J.F. Gilmer et al. / European Journal of Pharmaceutical Sciences 24 (2005) 315–323
For pKa determination, CE and the method described in
´
Simplıcio et al. (2004) was used.
2.4.4. HPLC
HPLC was performed using a system consisting of a
Spectra System SCM1000 ultrasound, Spectra System P4000
pump and controller, Spectra System AS3000 autosam-
pler and a Spectra System UV1000 detector controlled
by Chromquest Chromatography Manager. The stationary
phase was a C18 (4.6 mm × 250 mm) Waters Spherisorb with
a 10 ␮m particle size column. A mixed gradient-isocratic
mobile phase was employed which consisted of aqueous
42.5 mM orthophosphoric acid/6.1 mM triethylamine and
acetonitrile in a ratio of 80:20 for 5 min then graded to 20:80
(aqueous: MeCN) over 10 min, maintained at that for 5 min
then stepped back to initial conditions over 1 min and re-
equilibrated for 9 min. The flow rate was 1.2 ml/min.
2.4.5. Non-linear regression
Graph Pad PRISM® 3.02 was used for fitting experimental
data of the pH/rate profiles and the pH/mobility profiles.
3. Results and discussion
3.1. Elimination kinetics
Fig. 3. Degradation profile of 1i at pH 7.9 ( , desloratadine; ꢀ 1i).
The disappearance of the test compounds in aqueous solu-
tion over the pH range 0.5–11.7 at 37 ◦C was measured using
CE or HPLC as appropriate; for some compounds both HPLC
and CE were used. The identity of the amine was confirmed
by running external standards in CE and HPLC. The presence
of indenone was established by its PDA spectrum and reten-
tion time (HPLC), which coincided with that of an authentic
sample of indenone prepared by treating 3-bromoindanone in
diethyl ether with triethylamine (Marvel and Hinman, 1954).
The disappearance of aminoindanones was generally well
described by apparent first-order kinetics up to around neu-
trality. First-order plots were constructed from the logarithm
of % remaining compound versus time and the half-lives es-
timated using Eq. (1):
the formation of two later eluting unidentified components,
resulting in complex kinetics. However, the degradation of
1c followed apparent first-order kinetics up to pH 7.
The rate of the forward reaction (kf) under apparent equi-
librium conditions was measured using the following expres-
sion (Eq. (2)):
ꢀ
ꢁ
a0beq + baeq
a0(beq − b)
2a0 − beq
ln
=
kft
(2)
beq
Since the concentration of 2 at equilibrium (beq) and at
time t (b) can be related to the original concentration of 1
(a0) by beq = a0–aeq and b=a0 − a, Eq. (2) can be rewritten
strictly as a function of the concentration of 1 (Eq. (3)):
0.693
t1/2
=
(1)
ꢂ
ꢃ
a20 − aaeq
kobs
a0 − aeq
a0 − aeq
ln
=
kft
(3)
a0(a − aeq)
This straightforward picture was not always adhered to.
At higher pH, in some cases, there was evidence of the re-
verse reaction between the liberated amine and the indenone
leading to an equilibrium (Fig. 1). The extent to which this
occurred and the pH at which it began depended on the com-
pound involved, but the behaviour was particularly marked
for compounds 1i (Fig. 3) and 1c, compounds which share
a nucleophilic piperidine moiety. Indeed, in the case of 1c
there was evidence of a further reaction between the liber-
ated piperidine and the indenone.
The graphical representation of this equation should pro-
vide a straight line for times preceding equilibrium. The first-
order rate constant (kf) for the forward reaction can be derived
from the slope, m, of that line from Eq. (4):
a0 − aeq
kf = m
(4)
a0 + aeq
The decay plots of each compound at different pHs were
evaluated and assigned to one of these assumptions; i.e. to
completion or to an equilibrium situation. The extent of the
When the degradation of 1c was monitored at pH > 7 by
HPLC, the indenone reaction product was consumed with

J.F. Gilmer et al. / European Journal of Pharmaceutical Sciences 24 (2005) 315–323
321
a particular case of a Boltzmann sigmoid (Eq. (5)):
k2 − k1
kobs = k2 −
(5)
10(pH–pK ) + 1
a
where k1 and k2 represent, respectively, the elimination rates
from the protonated and the unprotonated aminoindanone,
pKa is the ionization constant of the compound and kobs is
the observed degradation rate constant at a particular pH.
Non-linear regression applied to the pH-rate profiles ob-
tained for compounds 1a and 1c–1k, to Eq. (5), yielded values
for k1, k2 and pKa presented in Table 2. pKa values were also
determined using electrophoretic mobilities according to the
Fig. 4. Plot showing the pH-rate profile for the disappearance of compound
1h ( ) at 37 ◦C and µ = 0.154. The solid line was constructed using Eq. (5).
´
method described in Simplıcio et al. (2004). Results of these
determinationsarepresentedinTable3andwhilethereisgen-
erally good agreement between the two methods, the latter
values are likely to be more accurate. Reductions of 3–4 pKa
units were observed for the aminoindanones in comparison
with the corresponding free amines. This is likely to be ad-
vantageous for prodrugs intended for peroral administration
because of the corresponding suppression of ionisation at
intestinal pH.
The elimination rate for the ionised form (k1) was gen-
erally very low with no significant reduction in concentra-
tion or emergence of parent amine at pH 0.5–1 over 24 h at
37 ◦C. The elimination rate was found to increase rapidly
as the pH of the aqueous solution approached the pKa of the
amine. These rates varied significantly across the small amine
set presented here. The compounds fell into two groups: (i)
secondary derivatives of primary amines for which k2 was
low, leading to half-lives of 100–800 min, and; (ii) tertiary
amines (1d, 1h) for which half-lives at around neutrality
were <5 min. Two further tertiary derivatives 1c and 1i, had
longer half-lives but measurement in these cases was com-
plicated by a prominent reverse reaction. These observations
are consistent with the behaviour of a group of six promis-
ing anti-allergic/anti-inflammatory aminoindanones (Walsh
reverse reaction exhibited a marked dependence on pH. Un-
der acid conditions the reverse reaction is suppressed because
of amine protonation. Correspondingly, equilibria were most
frequently observed at pH values in excess of amino pKa
when the amino group is mostly unionised. Amino nucle-
ophilicity also seems likely to be important in determining
the rate of the reverse reaction as evidenced here by the be-
haviour of 1i and 1c.
Further complications in measuring degradation rates
were encountered with the amino esters 1e and 1f. At basic
pH values these compounds degraded via a mixture of ester
hydrolysis and amine elimination; only the behaviour up to
the maximum pH at which elimination was apparently the
exclusive mode of degradation is considered here. Finally,
although the degradation of compound 1j followed typical
kinetics, the yield of liberated dopamine was low (∼40%).
pH-rate profiles were experimentally determined for all
compounds in the study except for 1b, which only slowly de-
graded across the pH range making measurement difficult. As
´
previously stated (Simplıcio et al., 2004), pH rate profiles of
´
the type depicted in Fig. 4 can be accounted for by assuming
unimolecular decomposition of the protonated and unionised
forms and may be mathematically represented in the form of
et al., 2001) on which we have recently reported (Simplıcio
et al., 2004). In that group, three tertiary amines had half-
lives of <10 min at pH 7.4 whereas the corresponding
Table 2
Elimination rates k1 and k2 and pKa values obtained by non-linear regression of experimental data to Eq. (5) along with projected half-lives based on the
theoretical pseudo-first-order k2 rate constants
k1 (min−1
)
k2 (min−1
)
t1/2 (min) k2
t1/2 (min) [experimental]
37 ◦C, low pH
t1/2 (min) [experimental]
37 ◦C, pH 7.4
pKa
1a
1b
1c
1d
1e
1f
1g
1h
1i
0
0.0004
0.0029 0.0003
nd
238.97
nd
210
Stable at pH < 4
Stable at pH 0.5
Stable at pH < 4
693 at pH 2.8
866 min at pH 3
Stable at pH < 3
Stable at pH < 3
192 at pH 3 2310 at pH 0.5
Stable at pH < 2
Stable at pH < 4
Stable at pH 0.5
266 (pH 7.8)
836
76
<5
277 (pH 5)
213 (pH 6.14)
215
1.1
nd
120
240
5.1 0.4
nd
nd
1e−7 3e−5
0.0033 3e−5
5.9 0.03
4.5 0. 6
3.3 0.9
4.3 0.4
6.6 0.6
5.3 0.2
5.0 0.4
6.6 0.9
3.7 0.4
0
0
0.03
0.001
0.13 0.01
5.33
0.0026 0.0011
0.0032 0.0004
0.0030 0.0007
0.18 0.01
0.0064 0.001
0.0067 0.0003
0.0027 0.0004
266.54
216.56
231
1e−7 0.0003
0
0.0006
0.005 0.007
0.0003 0.0005
0.0 0.002
3.85
108.28
103.43
256.67
1j
1k
0
0.0003
The experimentally determined k2 disappearance rate is also presented along with stability characteristics at low pH. The statement that the material was stable
means that there was no statistically significant change in concentration when a sample was incubated at that pH for 24 h.

322
J.F. Gilmer et al. / European Journal of Pharmaceutical Sciences 24 (2005) 315–323
Table 3
Experimental pKa values of the amines and their indanone derivatives determined by CE, along with pKa values available in the literature
Amine
pKa amine (literature)
pKa amine (experimental)
Compound
pKa indanone (experimental)
Dimethylamine
Ethylamine
Piperidine
10.73a
10.65a
11.12a
9.6a
7.29b
7.78b
9.83a
9.6a
n.d.
n.d.
n.d.
9.4 0.8
6. 9 0.2
n.d.
10.0 0.2
9.6 0.7
n.d.
1a
1b
1c
1d
1e
1f
1g
1h
1i
6.81 0.06
7.67 0.07
7.15 0.08
6.16 0.07
3.9 0.2
5.04 0.06
6.82 0.05
5.75 0.08
5.07 0.06
6.8 0.1, 8.5 ± 0.3
4.2 0.2, 8.1 ± 0. 2
Ephedrine
Tryptophan methyl ester
Alanine methyl ester
Phenethylamine
Atenolol
Desloratadine
Dopamine
8.8, 10a
7.12c
7.6 0.2, n.d.
6.74 0.09, 8.9 0.20
1j
1k
Etilevodopa
a
Albert and Serjeant (1984).
Hay and Porter (1967).
Beilstein.
b
c
secondary analogues degraded at pH 7.4 with half-lives of
200–500 min.
pH > 7 the two diastereomers underwent degradation at dif-
ferent rates, but analysis was complicated at this pH by a com-
peting ester hydrolysis. Thus, it was difficult to tell whether
the isomers underwent elimination or hydrolysis at different
rates.
Steric factors are likely to play a significant role in de-
termining the rate of elimination of amino drugs from their
indanone derivatives. We were unable to establish any quan-
titative relationships in the present amino set because of
the unavailability of appropriate parameters (Taft or Char-
ton) for many of its members. Sets of stereoisomers were
analysed in five cases; 1d, 1e, 1f, 1h and 1k. The k2 rates
were fairly similar for both enantiomeric pairs of 1d (0.0011,
0.0008 min−1) and 1e (0.0009, 0.0007 min−1). The isomeric
pairs of 1f (the alanine compound) were not sufficiently well
separated to determine if one pair degraded faster than the
other. In the case of the l-dopa ethyl ester analogue, 1k at
In terms of prodrug design, the amine release kinetics
might be regarded as especially favourable for 1d and 1h,
the indanone forms of ephedrine and the ␤-adrenergic block-
ing agent atenolol, respectively. Compound 1h is stable at low
pH but undergoes rapid elimination of the parent atenolol at
higher pH (t1/2 = 1.1 min at pH 7.4). The decomposition of 1h
was also studied in diluted human plasma at pH 7.4, where
elimination of atenolol is quantitative (Fig. 5). For these tests,
the samples had to be quenched in order to eliminate proteins
before CE or HPLC tests. Perchloric acid was used for this
according to the procedure described in 2.3.2. Buffered so-
lutions were treated in the same way for comparison. The
degradation rate in plasma was a little slower than in pH 7.4
buffer alone, probably due to protection of the compound
while bound to plasma proteins.
4. Conclusions
The technology presented here adds to the armoury of
chemistries available for reversibly masking amines. 3-
AminoindanonesarestableatlowpHbutundergoelimination
atbloodpH, liberatingtheparentamine. Indanoneattachment
leads to significant suppression of ionisation relative to the
parent amine and the compounds are also likely to be more
lipophilic given the nature of the appended group. Caveats to
the design include the following: little data is available con-
cerning the toxicity characteristics of indenone; the half-lives
of the tertiary compounds might be too short for peroral ad-
ministration while the secondary compounds are potentially
too stable.
The potential for ␤-amino ketones to undergo elimination
is fairly well established (and predictable), though there have
been few reports of this behaviour in an aqueous environ-
ment and the phenomenon has not been exploited previously
Fig. 5. Progress curves for the disappearance of 1h (᭹) in plasma buffered
at pH 7.4 and the appearance of atenolol (ꢁ) (mean, n = 3).

J.F. Gilmer et al. / European Journal of Pharmaceutical Sciences 24 (2005) 315–323
323
in the design of prodrug systems for amines (Tramontini and
Angiolini, 1994). The unexpectedly rapid rates of elimina-
tion of the neutral tertiary aminoindanones appears to be
connected with the extended conjugation available to the in-
denone side-product. In that sense there is probably a strong
thermodynamic component to the elimination of aminoin-
danones. It is easy to imagine other arrangements in which
amine elimination is similarly driven, and potentially at more
readily modulated rates. We are currently evaluating struc-
turally different carrier systems which fit this description.
Gershonov, E., Goldwaser, I., Fridkin, M., Shechter, Y., 2000. A novel ap-
proach for a water-soluble long-acting insulin prodrug: design, prepa-
ration, and analysis of [(2-sulfo)-9-fluorenylmethoxycarbonyl](3)-
insulin. J. Med. Chem. 43, 2530–2537.
Gogate, U., Repta, A., 1987. N-(acyloxyalkoxycarbonyl) derivatives as
potential prodrugs for amines. II. Esterase-catalysed release of parent
amines from model prodrugs. Int. J. Pharm. 40, 249–255.
Hay, R., Porter, L., 1967. Proton ionization constants and kinetics of base
hydrolysis of some (-amino-acid esters in aqueous solution. J. Chem.
Soc. (B), 1261–1264.
Huang, R.L., Williams, P., 1958. The relative stabilising influences of
substituents on free alkyl radicals. Part V. Selective bromination by
N-bromosuccinimide. J. Chem. Soc., 2637–2640.
Krause, M., Rouleau, A., Stark, H., Luger, P., Lipp, R., Garbarg, M.,
Schwart, J.C., Schunack, W., 1995. Synthesis, X-ray crystallography,
and pharmacokinetics of novel azomethine prodrugs of (R)-alpha-
methylhistamine: highly potent and selective histamine H3 receptor
agonists. J. Med. Chem. 38, 4070–4079.
Acknowledgements
This project was accomplished with the financial support
˜
ˆ
of the Fundac¸ao para a Ciencia e a Tecnologia, Portugal and
Marvel, C., Hinman, C., 1954. The synthesis of indenone and some related
compounds. J. Am. Chem. Soc. 76, 5435–5437.
the European Social Fund in the form of a scholarship.
Naringrekar, V.H., Stella, V.J., 1990. Mechanism of hydrolysis and
structure-stability relationship of enaminones as potential prodrugs
of model primary amines. J. Pharm. Sci. 79, 138–146.
Shanbhag, V.R., Crider, A.M., Gokhale, R., Harpalani, A., Dick, R.M.,
1992. Ester and amide prodrugs of ibuprofen and naproxen: synthesis,
anti-inflammatory activity, and gastrointestinal toxicity. J. Pharm. Sci.
81, 149–154.
Simpl´ıcio, A.L., Gilmer, J.F., Frankish, N., Sheridan, H., Walsh, J.J.,
Clancy, J.M., 2004. Ionisation characteristics and elimination rates
of some aminoindanones determined by capillary electrophoresis. J.
Chromatogr. A 1045/1–2, 233–238.
References
Aboul-Fadl, T., El-Shorbagi, A., 1996. New prodrug approach for
amino acids and amino-acid-like drugs. Eur. J. Med. Chem. 31,
165–169.
Albert, A., Serjeant, E., 1984. The Determination of Ionisation
Constants—A Laboratory Manual. Chapman and Hall.
Amsberry, K.L., Borchardt, R.T., 1991. Amine prodrugs which utilize hy-
droxy amide lactonization. I. A potential redox-sensitive amide pro-
drug. Pharm. Res. 8, 323–330.
Tramontini, M., Angiolini, L., 1994. Mannich Bases. Chemistry and Uses.
CRC Press.
Aungst, B.J., Blake, J.A., Rogers, N.J., Saitoh, H., Hussain, M.A., En-
singer, C.L., Pruitt, J.R., 1995. Prodrugs to improve the oral bioavail-
ability of a diacidic nonpeptide angiotensin II antagonist. Pharm. Res.
12, 763–767.
Wang, W., Camenisch, G., Sane, D.C., Zhang, H., Hugger, E., Wheeler,
G.L., Borchardt, R.T., Wang, B., 2000. A coumarin-based prodrug
strategy to improve the oral absorption of RGD peptidomimetics. J.
Con. Rel. 65, 245–251.
Bundgaard, H., Johansen, M., 1980. Prodrugs as drug delivery systems
IV: N-Mannich bases as potential novel prodrugs for amides, ureides,
amines, and other NH-acidic compounds. J. Pharm. Sci. 69, 44–46.
Walsh J., Frankish N., Sheridan H., Byrne W., 2001. Indane compounds
with smooth muscle relaxing and/or anti-inflammatory activity, US
Patent 6297399 B1, 2001.