Synthesis, chemical and enzymatic hydrolysis, and bioavailability evaluation in rabbits of metronidazole amino acid ester prodrugs with enhanced water solubility

Article abstract of DOI:10.1211/0022357011776199

A series of amino acid esters (3a-e) have been synthesized and evaluated as potential prodrugs of metronidazole with the aim of improving aqueous solubility and therapeutic efficacy. The aqueous solubility and the lipophilicity (expressed as the log P value) of metronidazole and its esters were investigated. In general the prodrugs revealed enhanced water solubility compared with metronidazole. N,N-diethylglycinate hydrochloride (3a) and 4-ethylpiperazinoacetate (3e) derivatives displayed higher aqueous solubility, which exceeded that of the parent drug by factors of approximately 140 and 100, respectively. All the esters revealed lower log P values than metronidazole except for the 4-phenylpiperazinoacetate derivative (3f), which was 6.5-times more lipophilic than metronidazole. The hydrolysis kinetics of the esters were studied in aqueous phosphate buffer (pH 7.4) and 80% human plasma at 37°C. In all cases the hydrolysis followed pseudo-first-order kinetics and resulted in a quantitative reversion to metronidazole as evidenced by HPLC analysis. The prodrugs exhibited adequate chemical stability (half-life, t1/2, 4-16 h) in aqueous phosphate solution of pH 7.4. In 80% human plasma they were hydrolysed within a few minutes to metronidazole. The esters 3d (methylpiperazinoacetate derivative) and 3f were exempted since their t1/2 values were approximately 2.5 and 8.5 h, respectively. A comparative pH-rate profile study of N,N-diethylglycinate hydrochloride (3a) and 4-ethylpiperazinoacetate (3e) derivatives in aqueous buffer solution over the pH range 2.2-10 was investigated. The results indicated that 3a showed marked stability at pH 2-6 followed by accelerated hydrolysis at pH 7.4. The basic ester 3e was found to be less stable at lower pH values but exhibited comparative stability at physiological pH. Moreover, in-vivo experiments in rabbits revealed a higher metronidazole plasma level with sustained release characteristics within the prodrug-treated animals (10- and 2.5-fold) as compared with the parent drug-treated group. In conclusion, the designed amino acid esters 3a and 3c-e might be considered as good candidates for water-soluble prodrug forms of metronidazole.

Full text of DOI:10.1211/0022357011776199

Journal of
Pharmacy and Pharmacology
JPP 2001, 53: 841–848
# 2001 The Authors
Received October 5, 2000
Accepted January 29, 2001
ISSN 0022-3573
Synthesis, chemical and enzymatic hydrolysis, and
bioavailability evaluation in rabbits of metronidazole
amino acid ester prodrugs with enhanced water
solubility
Nadia M. Mahfouz and Maha A. Hassan
Abstract
A series of amino acid esters (3a–e) have been synthesized and evaluated as potential prodrugs
of metronidazole with the aim of improving aqueous solubility and therapeutic efficacy. The
aqueous solubility and the lipophilicity (expressed as the log P value) of metronidazole and its
esters were investigated. In general the prodrugs revealed enhanced water solubility compared
with metronidazole. N,N-diethylglycinate hydrochloride (3a) and 4-ethylpiperazinoacetate (3e)
derivatives displayed higher aqueous solubility, which exceeded that of the parent drug by
factors of approximately 140 and 100, respectively. All the esters revealed lower log P values
than metronidazole except for the 4-phenylpiperazinoacetate derivative (3f), which was
6.5-times more lipophilic than metronidazole. The hydrolysis kinetics of the esters were studied
in aqueous phosphate buffer (pH 7.4) and 80% human plasma at 37mC. In all cases the hydrolysis
followed pseudo-first-order kinetics and resulted in a quantitative reversion to metronidazole
as evidenced by HPLC analysis. The prodrugs exhibited adequate chemical stability (half-life, t12,
4–16 h) in aqueous phosphate solution of pH 7.4. In 80% human plasma they were hydrolysed
within a few minutes to metronidazole. The esters 3d (methylpiperazinoacetate derivative) and
3f were exempted since their t12 values were approximately 2.5 and 8.5 h, respectively. A
comparative pH–rate profile study of N,N-diethylglycinate hydrochloride (3a) and 4-ethyl-
piperazinoacetate (3e) derivatives in aqueous buffer solution over the pH range 2.2–10 was
investigated. The results indicated that 3a showed marked stability at pH 2–6 followed by
accelerated hydrolysis at pH 7.4. The basic ester 3e was found to be less stable at lower pH
values but exhibited comparative stability at physiological pH. Moreover, in-vivo experiments
in rabbits revealed a higher metronidazole plasma level with sustained release characteristics
within the prodrug-treated animals (10- and 2.5-fold) as compared with the parent drug-
treated group. In conclusion, the designed amino acid esters 3a and 3c–e might be considered
as good candidates for water-soluble prodrug forms of metronidazole.
Department of Pharmaceutical
Medicinal Chemistry, Faculty of
Pharmacy, Assiut University,
Assiut 71526, Egypt
Nadia M. Mahfouz
Department of Pharmaceutics,
Faculty of Pharmacy, Assiut
University, Assiut 71526, Egypt
Maha A. Hassan
Introduction
Correspondence: N. M.
Mahfouz, Department of
Pharmaceutical Medicinal
Chemistry, Faculty of Pharmacy,
Assiut University, Assiut 71526,
Egypt. E-mail:
In the course of our continued interest to improve the therapeutic properties of
metronidazole, we have recently developed two series of identical and non-identical
twin ester prodrugs for oral delivery (Mahfouz et al 1998; Aboul-Fadl & Mahfouz
1998). Those studies afforded a chemical delivery system that was characterized by
nmahfouz!acc.aun.eun.eg
841

842
Nadia M. Mahfouz and Maha A. Hassan
enhanced lipophilicity, adequate chemical stability and
improved bioavailability of metronidazole. A chal-
Materials and Methods
lenging limitation for formulations of metronidazole Chemistry
in a simple injection dosage form is its relatively
Metronidazole was kindly provided by the Alexandria
low water solubility (approximately 1.0%, w\v, at
25mC). The intravenous administration of the drug
is restricted to the infusion of 100–200 mL every
8 h using 0.5% w\v aqueous solution to meet the
required dose (Bundgaard et al 1984a; Jensen et al
1990).
Co. for drugand chemicalindustries, Alexandria, Egypt.
All the other chemicals and reagents were of reagent
grade and those for kinetic studies were of analytical
grade. Freshly double-distilled water was used in the
preparation of the solutions.
Melting points were determined on a Stuart scientific
capillary melting point apparatus and are uncorrected.
Thin layer chromatography was performed using DC-
Alufolein (Kieselgel 60G, F 254, 0.25 mm). Chloro-
form\methanol(9:1)wasusedforthedevelopingsystem
and the spots were visualized by UV at 254 nm. El-
emental analyses were performed at the Chemistry De-
partment, Faculty of Science, Assiut University (Assiut,
Egypt). IR spectra (KBr disc) were recorded on an IR-
Several reports have been published on overcoming
metronidazole’s solubility problem. Firstly, the use of
dicarboxylic acid hemiesters e.g. hemisuccinate, hemi-
glutarate and hemimaleate was studied. However, those
prodrugs showed limited stability in aqueous solution
and were slowly and incompletely converted in-vivo to
metronidazole (Larsen et al 1988; Vermeersch et al
1990). Alternatively, a water-soluble metronidazole
phosphate ester prodrug with adequate chemical sta-
bility was reported by Cho et al (1982), but it was found
that it was not rapidly or quantitatively converted to
the parent drug in-vivo.
1
470 Shimadzu spectrometer (Japan). H NMR spectra
werescanned on a Varian EM-360 L NMR spectrometer
(60 MHz; USA). Chemical shifts were expressed in δ
(ppm) relative to TMS (tetra methylsilane) as an internal
standard.
A third type of water-soluble ester prodrug had an
ionizable amino function in the acid moiety, which
revealed facile hydrolysis in human plasma as illustrated
by propacetamol, 4-(acetamido) phenyl diethylamino-
acetate HCl salt, which is an injectable water soluble
form of paracetamol currently approved by the
European Pharmacopoeia 2000. Application of this
prodrug type on metronidazole has been reported
(Bundgaard et al 1984a, b, 1989; Cho & Haynes 1985;
Jensen et al 1990). This type of water-soluble ester
prodrug was investigated mostly as the hydrochloride
salt but unfortunately it showed high instability in
aqueous solutions (Bundgaard et al 1989; Jensen et al
1990). It appeared from the values of pseudo-first-order
hydrolysis rate constants that the protonated amino
acid ester possessed very high susceptibility to undergo
hydrolysis as compared with the free base form
(Bundgaard et al 1984b).
Solubility and partition coefficients were determined
using a shaker with a thermostatically controlled water
bath (Unitronic 320 OR-Selecta, Spain). All the UV
measurements were performed on a Shimadzu UV
150-02 double beam spectrometer (Tokyo, Japan).
Synthesis of metronidazole chloroacetate (2)
To a stirred suspension of metronidazole (1), 5.13 g
(30 mmol) in 50 mL methylene chloride, dried pyridine
2.74 g (35 mmol) was added followed by dropwise ad-
dition of chloroacetyl chloride 3.96 g (35 mmol) within
30 min. The mixture was refluxed for 2 h. The resulting
solution was cooled, washed with water (3i100 mL)
and dried over anhydrous sodium sulfate. The filtrate
was evaporated to dryness under reduced pressure and
crystallized from ether to yield 85% metronidazole
chloroacetate (2), mp 71–73mC, reported 70–72mC
However, we reported a series of bucetin ester pro-
drugs bearing a nonprotonated α-amino acid moiety.
Some of those derivatives exhibited enhanced water
solubility, good stability in aqueous buffer solutions
and rapid bioconversion (El-Faham & Mahfouz 1994).
This study reports on a series of metronidazole amino
acid esters as water-soluble prodrugs. The physico-
chemical properties, kinetics of hydrolysis and bio-
1
(Bundgaard et al 1984a). H NMR (CDCl3) δ (ppm):
2.4 (s, 3H, CH3); 3.9 (s, 2H,CH2Cl), 4.5 (m, 4H,
NCH2CH2O) and 7.7 (s, 1H, imidazole).
General method for synthesis of amino acid
ester prodrugs (3a–f)
availability in rabbits of the synthesized metronidazole A mixture of the respective amine (20 mmol) and metro-
amino acid ester prodrugs were investigated. A com- nidazole chloroacetate (2), 1.2 g (4 mmol) in 50 mL dry
parative study of the pH–rate profiles for a salt and dioxane, was stirred at room temperature for 24 h. The
basic form of the prodrug was undertaken also.
solvent was evaporated under reduced pressure. The

843
Metronidazole amino acid ester prodrugs
residue was dissolved in 100 mL methylene chloride, CH2NCH2), 3.2 (m, 6H, COCH2NjCH2NCH2), 4.5
washed with a saturated solution of sodium chloride (m, 4H, NCH2CH2O) 6.9 (d, J l 8 Hz, 3H, phenyl
(3i50 mL), dried over anhydrous sodium sulfate and H3,4,5), 7.25 (t, J l 8 Hz, 2H, phenyl H2,6), and 7.9 (s,
filtered. The filtrate was evaporated under reduced pres- 1H, imidazole). Anal. calcd. for C18H24N5O4 : C 57.74, H
sure and the residue recrystallized from the proper 6.46, N 18.71%. Found: C 58.14, H 6.30, N 18.74.
solvent to give the title ester prodrugs. Compound 3a
was obtained as the hydrochloride salt.
Aqueous solubility
Metronidazole N,N-diethylglycinate hydrochloride (3a)
1
The aqueous solubility of metronidazole and its amino
acid ester prodrugs (3a–f) was determined at 25mC. An
excess amount of the respective compound was added to
2 mL water in a screw-capped test tube. The suspension
was rotated on a thermostatically controlled mechanical
Yield 85%, mp 173–181mC (ether\ethanol). H NMR
(CDCl3\DMSO-d6) δ (ppm): 1.3 (t, J l 7 Hz, 6H,
2CH2CH3), 2.6 (s, 3H, CH3); 3.2 (q, J l 7 Hz, 4H,
2CH2CH3), 4.0 (m, 4H, NCH2CH2O), 11.5 (broad ex-
changeable s, 1H, protonated diethylamino) and 8.0
(1H, s, imidazole). Anal. calcd. for C12H21N4O4Cl: C
44.93, H 6.60, N 17.47%. Found: C 44.54, H 6.44, N
17.32.
1
−
shaker switched on 30 strokes min for 24 h to attain
equilibrium. The mixture was filtered, and a portion was
diluted with an appropriate amount of water and ana-
lysed for metronidazole or its amino acid ester prodrug
content spectrophotometrically at 320 nm.
Metronidazole 1-piperidinoacetate (3b)
1
Yield 85%, mp 111–114mC (ether). H NMR (CDCl3)
Lipophilicity
δ (ppm): 1.5 (m, 6H, piperidino CH2CH2CH2), 2.5
(m, 7H, CH3jpiperidino CH2NCH2), 3.15 (s, 2H,
COCH2N), 4.5 (m, 4H, NCH2CH2O) and 8.0 (s, 1H,
imidazole). Anal. calcd. for C13H20N4O4 : C 52.69, H
6.80, N 18.91%. Found: C 52.49, H 6.80, N 18.78.
The partition coefficients of metronidazole and its pro-
drugs (3a–f) were determined in an n-octanol\water
system at 25mC. The two phases had been previously
saturated with each other and then separated. To ensure
solubility an accurately weighed amount of each com-
pound was added to the aqueous phase. A 3-mL sample
of this solution was transferred to a 25-mL stoppered
bottle and 3 mL of the saturated n-octanol was added.
Metronidazole 1-morpholinoacetate (3c)
Yield 87%, mp 75–77mC (ether\ethylacetate), reported
1
77–78mC (Bundgaard et al 1984a). H NMR (CDCl3) δ
1
(ppm): 2.5 (m, 7H, CH3jCH2NCH2), 3.2 (s, 2H,
COCH2N), 3.6 (t, J l 5 Hz, 4H, CH2OCH2) 4.5 (m, 4H,
NCH2CH2O) and 7.9 (s, 1H, imidazole).
−
The bottle was then shaken for 4 h at 35 strokes min .
The aqueous layer was carefully separated and 1 mL of
this layer was suitably diluted with water. The con-
centration of each compound was determined spectro-
photometrically at λmax l 320 nm by virtue of cali-
Metronidazole 4-methyl-1-piperazinoacetate (3d)
Yield 86%, mp 97–99mC (ether), reported 97–99mC bration curves correlating substance absorbance and
(Bundgaard et al 1984a). H NMR (CDCl3) δ (ppm): known concentrations. All experiments were conducted
1
2.15 (s, 3H, NCH3), 2.4 (broad s, 11H, CH3j piperazine in triplicate and the mean values were taken. The
protons), 3.0 (s, 2H, COCH2N), 4.5 (m, 4H, NCH2 partition coefficients were calculated.
CH2O) and 7.7 (s, 1H, imidazole).
Hydrolysis kinetics
Metronidazole 4-ethyl-1-piperazinoacetate (3e)
1
Yield 90%, mp 64–65mC (ether\petroleum ether). H
NMR (CDCl3) δ (ppm): 1.07 (t, J l 7 Hz, 3H, CH2CH3)
2.5 (m, 13H, CH3jCH2CH3jpiperazine protons), 3.15
(s, 2H, COCH2N), 4.6 (m, 4H, NCH2CH2O) and 8.0 (s,
1H, imidazole). Anal. calcd. for C14H23N5O4 : C 51.68,
H 7.13, N 21.53%. Found: C 52.12, H 7.00, N 21.00.
The kinetic studies were carried out using an HPLC
system. This consisted of a pump (Knauer pump 64,
Berlin, Germany), a Knauer variable-wavelength de-
tector, a reversible-phase C18 column (Eurospher 80
RP-18, 25i0.5 cm i.d.) equipped with a cartridge guard
column, a Shimadzu C–R 6A chromatopac recording
integrator and a 20-µL injection loop. Chromatographic
separations were achieved using a mobile phase of
Metronidazole 4-phenyl-1-piperazinoacetate (3f )
1
Yield 85%, mp 88–90mC (ether\ethanol). H NMR methanol\aqueousphosphatebuffermixturecontaining
(CDCl3) δ (ppm): 2.6 (m, 7H, CH3jpiperazine (% w\v) potassium dihydrogen phosphate (0.43 g) and

844
Nadia M. Mahfouz and Maha A. Hassan
Bioavailability study in rabbits
disodium hydrogen phosphate (0.37 g). The relative
ratio of methanol and phosphate aqueous solution was
adjusted in each case to attain complete separation of
the peaks of the prodrug, parent drug and the expected
degradation products. All measurements were carried
Rabbits (approximately 2 kg), from the authorized
animal house (Faculty of Medicine, Assiut University),
were divided into three groups of three animals each.
The rabbits were used to compare the bioavailability of
the prodrugs 3a and 3e with that of metronidazole
following intravenous administration in the marginal
ear vein. In group one, each rabbit received a dose
1
−
out at λmax l 320 nm, and the flow rate of 1.0 mL min
was adjusted for all prodrugs except the ester 3c, where
1
the flow rate was set at 1.5 mL min . The kinetic data
−
were the average of at least three experiments.
1
−
of 2 mL kg 0.5% w\v aqueous solution of metro-
1
nidazole. An equivalent amount (0.5 mL kg ) of an
−
In aqueous buffers
aqueous solution (3.8%, w\v) of the prodrug 3a or 3e
The hydrolysis kinetics of metronidazole amino acid was administered to the rabbits of groups two and three,
ester prodrugs (3a–f) was studied at pH 7.4. In addition respectively. At appropriate time intervals up to 3 h,
to this, the pH–rate profile of the prodrugs 3a and 3e 2-mL blood samples were withdrawn from each rabbit
was determined in aqueous buffer solutions over the pH into heparinized tubes and centrifuged for 15 min to
range 2.2–10, using McIlvaine buffer (pH 2.2–8) and separate the plasma. Plasma samples (100 µL) were
Clark and Lubs buffer (pH 7.8–10). The total buffer diluted with an equivalent volume of methanol, and
concentration was 0.02  and a constant ionic strength treated as detailed above for the hydrolysis in human
(µ) of 0.5 for each buffer was maintained by adding a plasma and assayed for metronidazole content. The
calculated amount of potassium chloride. A sample mean concentration of metronidazole at the respective
(250 µL) of the methanolic solution of the respective time interval in each group was taken.
3
−
prodrug (2i10 ) was added to 4.75 mL preheated
buffer solution in screw-capped test tubes. The solutions
were kept at 37mC in a water bath and at appropriate
time intervals 20-µL samples were withdrawn and ana-
Results and Discussion
lysed by HPLC for the remaining ester prodrugs. Least-
square equations, derived by correlating areas in HPLC
chromatograms to known concentrations of each com-
pound were used for calculation of the residual ester
concentrations in the studied samples. The correlation
coefficients of the standard curves were 0.999. Pseudo-
first-order-rate constants for the hydrolysis were then
obtained from linear plots of the log concentration of
remaining ester vs time.
Chemistry
The reaction sequence used for the preparation of
compounds 3a–f was achieved in two steps (Figure 1).
The precursor metronidazole chloroacetate (2) was ob-
tained by refluxing metronidazole with chloroacetyl
chloride in methylene chloride using pyridine as acid
N
N
CICOCH2CI
In human plasma
O2N
N
CH3
O2N
N
CH3
The rate of enzymatic hydrolysis of the ester prodrugs
3a–f in human plasma diluted to 80% with 0.02 
phosphate buffer pH 7.4 was determined. Samples of
CH2CH2OH
CH2CH2OCOCH2Cl
1
2
R
HN
stock solution (2i10 4 ; 200 µL) of the respective
−
R′
prodrug were incubated with 1.80 mL 80% human
plasma at 37mC. At appropriate time intervals 100-µL
samples of plasma\buffer mixture were withdrawn, de-
proteinated by addition of 200 µL methanol, immedi-
N
O2N
N
CH3
R
R′
CH2CH2OCOCH2N
4
1
−
3a-e
ately mixed and centrifuged at 1i10 rev min for
10 min. Samples (20 µL) of the resulting clear super-
natants were withdrawn and analysed for the remaining
prodrugs by an HPLC method. The pseudo-first-order-
rate constants for the hydrolysis and the half-lives (t12) of
prodrugs were calculated from linear plots of the log
concentrations of the remaining prodrugs vs time.
C2H5
b:
NRR’ = a: N
c:
N
N
O
C2H5
N C6H5
d:
N
N CH3 e:
N
N CH3
f:
N
Figure 1 The synthesis of the metronidazole ester prodrugs 3a–f.

845
Metronidazole amino acid ester prodrugs
scavenger. Subsequent condensation of the intermediate was observed as a general pattern that the synthesized
(2) with the respective amine at room temperature gave prodrugs possessed higher aqueous solubility compared
the title compounds 3a–f in good yield. The purity was with metronidazole. The hydrochloride salt of metro-
assessed by TLC and HPLC, and the assignments of the nidazole N,N-diethylglycinate ester (3a) displayed a
structures were based on elemental and spectroscopic higher aqueous solubility that exceeded the solubility of
methods of analyses. The IR spectral data revealed the the parent drug in water by a factor of approximately
characteristic carbonyl stretching vibration bands at 140. In addition, the solubility of the piperazine deriva-
1
−
1722–1755 cm of the ester groups in addition to the tives (3d–f) paralleled with the availability of the lone
1
ether bands at approximately 1170 and 1255 cm . In pair of electrons on the N4 atom of the piperazine
−
the 1H NMR spectra the signals of the respective moiety. Accordingly, the ethyl piperazine derivative (3e)
protons of the prepared esters were verified on the basis showed higher aqueous solubility in contrast to the
of their chemical shifts, multiplicities and coupling con- phenyl piperazine derivative (3f), which displayed poor
stants. The spectra showed a singlet signal at δ " water solubility. The solubility of 3f was even poorer
2.5 ppm correlated to the methyl group located at C-2 of than that of the parent drug. Generally, the observed
the imidazole nucleus; four methylene protons of metro- increase in aqueous solubility of the investigated pro-
nidazole appeared as multiplet at δ " 4.5 ppm in ad- drugs was accompanied by a decrease in lipophilicity
dition to a singlet signal at δ " 3 ppm corresponding to relative to the parent drug as illustrated by log P values
the COCH2N group. This signal was shifted downfield (Table 1). The prodrug 3f, derived from 4-phenyl-
to δ " 4 ppm in compound 4a due to the withdrawing piperazinoacetate, showed an expected extremely high
effect of the protonated amino group. A downfield partition coefficient, higher than metronidazole (4.73
shifted signal at δ " 7.9 ppm was assigned to the imi- and 0.74, respectively). This enhanced lipophilicity of 3f
dazole ring proton. The protons of the amine moiety represented a basis for its consideration as a prodrug
were resonated at different δ-values of approximately form for oral, dermal and rectal delivery. The pH values
1.2–4.3 ppm and a downfield signal at δ " 7 ppm was of an aqueous solution (2.5–20%, w\v) of N,N-diethyl-
due to the aromatic protons of the phenyl group of the glycinate hydrochloride salt (3a) and the base 4-ethyl-1-
ester 3f.
piperazinoacetate (3e) were determined and found to be
1.5–2.2 and 9–9.5, respectively.
Physicochemical properties
The aqueous solubility and partition coefficient (log P)
of metronidazole (1) and its amino acid ester prodrugs
(3a–f) are listed in Table 1. It is evident that esterification Aqueous buffers
In-vitro stability
of the hydroxyl group of metronidazole with different Hydrolysis kinetics of metronidazole amino acid ester
amino acids provided an additional handle to modify prodrugs 3a–f were studied in aqueous buffer solution at
the physicochemical properties of the parent drug. It pH 7.4. Under the experimental conditions used the
Table 1 Physicochemical and kinetic data* of metronidazole and its prodrugs (3a–f) in aqueous
phosphate buffer solution and 80% human plasma.
1
−
Compound
Solubility (mg mL
)
log P
pH 7.4
Plasma
3
t12 (h)
k
t12 (min)
−
ki10
Metronidazole (1)
10.5
1448.4
3.5
0.75
0.03
0.51
0.27
0.29
0.51
4.73
–
–
–
–
3a
3b
3c
3d
3e
3f
2.846
4.836
1.488
0.954
0.691
1.536
4.06
2.39
7.76
12.11
16.73
7.52
0.021
0.014
0.013
0.001
0.015
0.005
33.56
49.51
52.59
511.2
46.52
138.0
337.4
98.6
994.2
1.19
*Average of three experiments.

846
Nadia M. Mahfouz and Maha A. Hassan
*
Table 2 pH rate profile data for half-lives (t12) of the prodrugs 3a
N
N
k
+ HOOCCH2NRR’
CH3
and 3e in aqueous buffer solutions (pH 2.2–10) at 37mC.
O2N
N
CH3
N
O2N
R
pH
3a
3e
CH2CH2OCCH2N
CH2CH2OH
R’
O
kobs
t12 (h)
kobs
t12 (h)
3a-f
1
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
Figure 2 The degradation of the metronidazole ester prodrugs
3a–f.
2.2
1.157i10
1.316i10
0.917i10
1.439i10
2.035i10
2.846i10
6.708i10
9.652i10
0.012
9.98
8.78
12.61
8.03
5.68
4.06
1.72
1.19
1.00
1.114i10
0.191i10
0.369i10
0.372i10
1.064i10
0.669i10
1.480i10
8.557i10
0.019
10.37
60.60
31.27
31.07
10.86
16.73
7.81
3.0
4.0
5.0
6.0
7.4
8.0
9.0
10.0
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1.0
1.35
0.60
*Average of three experiments.
related to some electronic and steric properties of the
amine moiety. The observed higher susceptibility of the
diethylamino and piperidino derivatives 3a and 3b was
clearly attributed to the higher basic character of the
amine partial structure of the molecule relative to the
morpholino and piperazino derivatives 3c–f. The effect
of the electronic properties on the rate of hydrolysis at
the studied pH has been emphasized by comparison of
the kobs values of the hydrochloride salt 3a and the base
3b, whereby 3a exhibited relative less susceptibility
(kobsl 2.85) for hydrolysis than 3b (kobsl 4.84). On the
basis of this kinetic data, it was apparent that the basic
form of esters 3c–f were stable enough for several hours
at pH 7.4 to be used as water soluble formulations.
0
50
100 150
200 250
300
350
Time (min)
Figure 3 First-order plots for the hydrolysis of metronidazole ester
prodrugs (3a $; 3b #; 3c X; 3d W; 3e *; 3f ꢀ) in phosphate buffer
(pH 7.4). Each value is the average of three experiments.
target compounds hydrolysed to release the parent drug
(Figure 2) as evidenced by HPLC analysis. At constant pH–rate profile
pH and temperature, the reaction displayed strict first- The esters 3a and 3e, as representatives for salt and basic
order kinetics.
forms of the prodrugs, were selected for the pH–rate
The rate constants (kobs) and the corresponding half- profile study over the pH range 2.2–10 at 37mC. The
lives (t12) for the respective prodrugs were calculated investigation aimed to determine the optimum pH for
from the linear regression equations correlating the log maximumstability. Thevalues ofthe pseudo-first-order-
concentration of the residual prodrug vs time. The rate constants (kobs) for hydrolysis of both compounds
results are summarized in Table 1 and Figure 3. The at the respective pH value and their corresponding t12 are
kinetic data revealed that the susceptibility of various given in Table 2 and the pH–rate profiles are shown in
amino acid esters to undergo hydrolysis in isotonic Figure 4.
phosphate buffer solution varied appreciably. Un-
The results indicated that the hydrochloride salt 3a
expectedly the ester derived from 1-piperidinoacetate revealed a marked stability at the pH range 2–6, which
(3b) showed a reasonable facile cleavage at pH 7.4 (t12 corresponded to the pH of its aqueous solution at
approximately2.5 h). This isanalogous tothe previously concentrations ranging from 2.5 to 20%, w\v. However,
reported data for bucetin 1-piperidinoacetate (El- this prodrug showed an accelerated degradation rate
Faham & Mahfouz 1994). Generally, it could be postu- near the physiological pH (U shape segment 6–8). These
lated that the rate of hydrolysis of the amino acid esters observations coincided with previously reported results
3a–f at pH 7.4 (as illustrated in Table 1) was mainly of the pH-rate profile of metronidazole N,N-dimethyl-

847
Metronidazole amino acid ester prodrugs
1.5
1.4
1.3
1.2
1.1
1.0
0.9
0.8
0.7
0.6
0.5
–1.5
–1.8
–2.1
–2.4
–2.7
–3.0
–3.3
–3.6
–3.9
0
50
100
150
200
250
1
2
3
4
5
6
7
8
9
10
Time (min)
pH
Figure 5 First-order plots for the hydrolysis of metronidazole ester
prodrugs (3a $; 3b #; 3c X; 3d W; 3e *; 3f ꢀ) in 80% human
plasma. Each value is the average of three experiments.
Figure 4 The pH–rate profiles for the hydrolysis of metronidazole
prodrugs (3a $; 3e #) at 37mC in aqueous buffers. Each value is the
average of three experiments.
(Figure 5). The ester derived from 4-phenyl piperazino-
acetate (3f) showed only a small acceleration in the
cleavage rate in the presence of plasma. Likewise, the
hydrolysis of the 4-methyl piperazinoacetate (3d) was
catalysed to only a minor extent. In contrast, a reason-
able rapid cleavage in plasma was obtained for the esters
3a–c and 3e. Consequently, these compounds were good
substrates for plasma enzymes and appeared to be
suitable candidates as prodrugs for metronidazole. It
was interesting to observe from the kinetic study that
metronidazole N,N-diethylglycinate hydrochloride (3a)
was found to be much more readily hydrolysed in 80%
human plasma and aqueous buffer solution (pH 7.4)
than the investigated basic amino acid esters (3b–f).
glycinate hydrochloride (Bundgaard et al 1984b). Con-
sequently, the hydrochloride salt of metronidazole N,N-
diethylglycinate could be formulated to be reconstituted
as a solution for parenteral administration before use.
The basic ester prodrug 3e was found to be less stable
at lower pH (2–6), but exhibited marked stability relative
to the hydrochloride salt 3a at physiological pH. Formu-
lations of this prodrug in aqueous solution of pH 7.4
attained a non-irritating stable parenteral preparation.
The observed enhanced hydrolysis susceptibility
above the physiological pH might be explained on the
basis of two factors; the elevated pOH value of the
medium and the intramolecular nucleophilic catalysis
by the lone pair of electrons of the amine moiety (Jensen
et al 1990).
Bioavailability
Human plasma
The rates of hydrolysis of the target esters were studied The amino acid ester prodrugs 3a and 3e were chosen
in 80% human plasma (pH 7.4) at 37mC as predictive for in-vivo evaluation as representative examples for the
for their in-vivo behaviour. The bioconversion pro- studied water soluble prodrugs of metronidazole for
ceeded in a similar sequence as in the case of the aqueous parenteral administration. Metronidazole and the pro-
buffer solution. The rate constants (kobs) for the in- drugs 3a and 3e were given intravenously to rabbits and
dividual reactions were calculated from the linear re- at different time intervals; the concentration of metro-
gression equations correlating log residual concen- nidazole was estimated in plasma (Figure 6). A com-
trations of the prodrugs vs time. The corresponding parison of the plasma concentration–time curves of
half-life (t12) for the respective prodrug was then calcu- metronidazole indicated that the esters were rapidly
lated (Table 2). The rate data clearly revealed that the hydrolysed to the parent drug. Maximum plasma con-
susceptibility for enzymatic hydrolysis of the studied centrations of the liberated metronidazole from the
metronidazole amino acid ester prodrugs varied widely esters 3a and 3e were attained after approximately 9 and

848
Nadia M. Mahfouz and Maha A. Hassan
11
10
9
References
Aboul-Fadl, T., Mahfouz, N. M. (1998) Metronidazole twin ester
prodrugs II: non identical twin esters of metronidazole and some
antiprotozoal halogenated hydroxyquinoline derivatives. Sci.
Pharm. 66: 309–324
8
7
Bundgaard, H., Larsen, C., Thorbek, P. (1984a) Prodrugs as drug
delivery system XXVI. Preparation and enzymatic hydrolysis of
various water-soluble amino acid esters of metronidazole. Int. J.
Pharm. 18: 67–77
6
5
Bundgaard, H., Larsen, C., Arnold, E. (1984b) Prodrugs as drug
delivery system XXVII. Chemical stability and bioavailability of a
water-soluble prodrug of metronidazole for parenteral admin-
istration. Int. J. Pharm. 18: 79–87
4
3
2
Bundgaard, H., Falch, E., Jensen, E. (1989) A novel solution-stable
prodrug type for drugs containing a hydroxyl or an NH-acidic
group. J. Med. Chem. 32: 2503–2507
1
0
20 40 60 80 100 120 140 160 180
Time (min)
Cho, M. J., Haynes, L. C. (1985) Serum-catalyzed hydrolysis of
metronidazole amino acid esters. J. Pharm. Sci. 74: 883–885
Cho, M. J., Kurz, R. R., Lewis, C., Machkovech, S. M., Houser, D. J.
(1982) Metronidazole phosphate – a water soluble prodrug for
parenteral solutions of metronidazole. J. Pharm. Sci. 71: 410–414
El-Faham, T. H., Mahfouz, N. M. (1994) Synthesis, hydrolysis kin-
etics and physicochemical properties of ester prodrugs of bucetin
Alex. J. Pharm. Sci. 8: 33–37
Figure 6 Plasma levels of metronidazole in rabbits after intravenous
administration of the prodrugs (3a X; 3e *; metronidazole $). Each
value represents the meanps.e. (n l 3).
10 min, respectively, compared with 8 min for the
parent-drug-treated animals.
Jensen, J., Bundgaard, H., Falch, E. (1990) Design of a water soluble,
solution stable and biolabile prodrug of metronidazole for par-
enteral administration: N-substituted aminomethylbenzoate. Int.
J. Pharm. 58: 143–153
At all time intervals, the concentrations of metronida-
zole in the prodrug-treated animals were relatively
higher compared with the animals receiving the parent
drug. Furthermore, 3 h after administration of 3a the
metronidazole level was 10-times higher compared with
the parent drug, indicating a sustained release. The
prodrug 3e resulted in a 2.5-fold increase in the metro-
nidazole level after 3 h. In conclusion, the in-vivo evalu-
ation study indicated that the amino acid esters 3a and
3e might be considered as potential biolabile water-
soluble prodrug forms of metronidazole.
Larsen, C., Kurtzhals, P., Johnansen, M. (1988) Kinetics of regen-
eration of metronidazole from hemiesters of maleic acid, succinic
acid and glutaric acid in aqueous buffer, human plasma and pig
liver homogenate. Int. J. Pharm. 41: 121–129
Mahfouz, N. M., Aboul-Fadl, T., Diab, A. K. (1998) Metronidazole
twin ester prodrugs: synthesis, physicochemical properties, hy-
drolysis kinetics and antigiardial activity. Eur. J. Med. Chem. 33:
675–683
Vermeersch, H., Remon, J. P., Permentier, D., Schacht, E. (1990) In
vitro antitrichomonal activity of water-soluble prodrug esters of
metronidazole. Int. J. Pharm. 60: 253–260