Prodrugs of perzinfotel with improved oral bioavailability

Article abstract of DOI:10.1021/jm8011799

Previous studies with perzinfotel (1), a potent, selective, competitive NMDA receptor antagonist, showed it to be efficacious in inflammatory and neuropathic pain models. To increase the low oral bioavailability of 1 (3-5%), prodrug derivatives (3a-h) were synthesized and evaluated. The oxymethylene-spaced diphenyl analogue 3a demonstrated good stability at acidic and neutral pH, as well as in simulated gastric fluid. In rat plasma, 3a was rapidly converted to 1 via 2a. Pharmacokinetic studies indicated that the amount of systemic exposure of 1 produced by a 10 mg/kg oral dose of 3a was 2.5-fold greater than that produced by a 30 mg/kg oral dose of 1. Consistent with these results, 3a was significantly more potent and had a longer duration of activity than 1 following oral administration in a rodent model of inflammatory pain. Taken together, these results demonstrate that an oxymethylene-spaced prodrug approach increased the bioavailability of 1.

Full text of DOI:10.1021/jm8011799

J. Med. Chem. 2009, 52, 771–778
771
Prodrugs of Perzinfotel with Improved Oral Bioavailability
Reinhardt B. Baudy,† John A. Butera,† Magid A. Abou-Gharbia,†,‡ Hong Chen,† Boyd Harrison,† Uday Jain,† Ronald Magolda,†
Jean Y. Sze,† Michael R. Brandt,§,| Terri A. Cummons,§ Diane Kowal,§ Menelas N. Pangalos,§ Bojana Zupan,§,
Matthew Hoffmann,# Michael May,# Cheryl Mugford,# Jeffrey Kennedy,§ and Wayne E. Childers, Jr.*,†
Chemical & Screening Sciences, Wyeth Research, CN-8000, Princeton, New Jersey 08543, Neuroscience DiscoVery Research, Wyeth Research,
Princeton, New Jersey, and Wyeth Drug Safety & Metabolism, CollegeVille, PennsylVania
ReceiVed September 26, 2008
Previous studies with perzinfotel (1), a potent, selective, competitive NMDA receptor antagonist, showed
it to be efficacious in inflammatory and neuropathic pain models. To increase the low oral bioavailability
of 1 (3-5%), prodrug derivatives (3a-h) were synthesized and evaluated. The oxymethylene-spaced diphenyl
analogue 3a demonstrated good stability at acidic and neutral pH, as well as in simulated gastric fluid. In
rat plasma, 3a was rapidly converted to 1 via 2a. Pharmacokinetic studies indicated that the amount of
systemic exposure of 1 produced by a 10 mg/kg oral dose of 3a was 2.5-fold greater than that produced by
a 30 mg/kg oral dose of 1. Consistent with these results, 3a was significantly more potent and had a longer
duration of activity than 1 following oral administration in a rodent model of inflammatory pain. Taken
together, these results demonstrate that an oxymethylene-spaced prodrug approach increased the bioavailability
of 1.
Introduction
the current study, we prepared and evaluated oxymethylene-
spaced prodrugs of 1 to determine whether they would show
improved oral bioavailability. We required a prodrug that would
be stable in a wide pH range and physiological fluids and yet
decompose rapidly once it reached the plasma. We focused on
steric hindrance around the initial site of hydrolysis as a means
of controlling the rate of decomposition of our prodrug
candidates.
Excitatory amino acids (EAAsa) acting at the N-methyl-D-
aspartate-selective (NMDA) subtype of the glutamate receptor
play a role in both acute and chronic pain. Increases in afferent
input and glutamate release within the spinal cord have been
observed after peripheral injection of irritants (e.g., carrageenan)
or tissue injury.1-3 In preclinical studies, NMDA antagonists
reverse neuronal hyperexcitability as well as hyperalgesia in
several inflammatory and neuropathic pain models associated
with various pathophysiologic mechanisms.4-6 However, among
clinically assessed NMDA antagonists, the narrow separation
between effectiveness and liabilities, such as sedation and
psychotomimetic effects, has severely hampered their utility for
the treatment of neuropathic pain.7
We previously described perzinfotel (1, EAA-090, [2-(8,9-
dioxo-2,6-diazabicyclo[5.2.0]non-1(7)-en-2-yl)ethyl]phospho-
nic acid) as a potent, competitive NMDA receptor antagonist.
It is selective for the NMDA receptor, showing no significant
affinity at over 60 other receptor, enzyme, and ion channel
binding sites.8-10 Previous studies demonstrated 1 to be
efficacious in a variety of pain models.11 However, 1 demon-
strates low bioavailability following oral administration.11 In
Chemistry
A series of bis-phosphonate derivatives of 1 (compounds
3a-h, Scheme 1) was prepared in a straightforward manner
starting from 1 as depicted in Scheme 1. Treatment of 1 with
the appropriate chloromethyl ester in the presence of N,N-
diisopropylethylamine in dimethylformamide at 70 °C for 20 h
provided the desired prodrug candidates 3a-g in moderate to
excellent yield. In order to investigate an oxymethylene spaced
carbonate analogue as a potential prodrug moiety, we prepared
compound 3h in the same manner as 3a-g using chloromethyl
isopropyl carbonate (9).
Noncommercially available chloromethyl ester reagents 4-8
were prepared as shown in Scheme 2 by treatment of the
appropriate acid chloride with a suitable aldehyde in the presence
of zinc chloride. The isopropyl carbonate starting material 9
was synthesized by treating chloromethyl chloroformate with
isopropanol. The branched oxymethylene prodrug candidates 3
d-f (where R1 ) CH3) were obtained as diastereomeric
mixtures. In the case of phenyl derivatives 3d and 3e, the two
diastereomers were separated successfully by HPLC and evalu-
ated separately. In the case of cyclohexyl analogue 3f, the
diastereomeric mixture was evaluated without prior separation.
All compounds were fully characterized by 1H NMR, MS, and
combustion analysis.
* To whom correspondence should be addressed. Phone: 732-274-4253.
Fax: 732-274-4505. E-mail: childew@wyeth.com.
†
Chemical & Screening Sciences, Wyeth Research.
Current address: Temple University School of Pharmacy, Philadelphia,
‡
PA.
§
Neuroscience Discovery Research, Wyeth Research.
Current address: Johnson & Johnson Research and Development,
|
L.L.C., Spring House, PA.
Current address: Weill Cornell Medical School, New York, NY.
Wyeth Drug Safety & Metabolism.
#
a Abbreviations: AMPA, R-amino-3-hydroxy-5-methylisoxazolepropi-
onate; AUC, area under the curve; Cmax, maximum concentration; EAA,
excitatory amino acid; ESI, electrospray ionization; 1H NMR, proton nuclear
magnetic resonance spectroscopy; HPLC, high pressure liquid chromatog-
raphy; LC/MS, liquid chromatography/mass spectrometry; MS, mass
spectrometry; NMDA, N-methyl-D-aspartate; PGE2, prostaglandin E2; po,
oral administration; SGF, simulated gastric fluid; SIBLM, simulated
intestinal bile fluid; SIF, simulated intestinal fluid; t1/2, half-life; TCP, N-[1-
(2-thienyl)cyclohexyl]-3,4-piperidine; Tmax, time after dosing at which
maximum concentration occurs.
[14C]-Labeled 1 and [14C]-labeled 3a were provided by
Amersham Life Sciences12 starting from [14C]-diethyl squarate.
[14C]-1 was prepared using synthetic methodology described
10.1021/jm8011799 CCC: $40.75
2009 American Chemical Society
Published on Web 01/15/2009

772 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 3
Baudy et al.
Scheme 1. Synthesis of Bis-phosphonate Ester Prodrugs 3a-ha
a Reagents and conditions: (i) N,N-diisopropylethylamine, dimethylformamide, *RCOOCHR′Cl, 70 °C, 20 h.
Scheme 2. Synthesis of Chloromethyl Ester Intermediatesa
the course of plasma stability studies using (+ESI) and (-ESI)
LC/MS. The actual hydrolysis step is believed to occur on the
carbon ester, with subsequent elimination of the aldehyde-
derived linker to give the phosphonic acid. Therefore, in order
to test the effects of electronics and steric hindrance on the
hydrolysis, we prepared prodrug candidates with both aromatic
and aliphatic ester groups as well as oxymethylene linkers
derived from both formaldehyde and acetaldehyde.
Compounds 3a-h were initially evaluated for their stability
in rat plasma. From previous studies, it was known that
compound 1 possesses an adequate half-life in vivo (t1/2
)
1.3-7.8 h, depending on dose).11 Therefore, we felt that a
successful prodrug candidate would undergo rapid, complete
conversion to the parent compound 1 in the plasma (within 1-3
h) rather than a slow, continuous release. The results of this
initial test are shown in Figure 2. Hydrolysis of the tert-butyl
(3b) and 4-heptyl (3g) analogues derived from formaldehyde
required 7 h. Analogues derived from acetaldehyde (3d-f) took
even longer, achieving only 30-40% conversion within the 7 h
time frame. Compounds 3a, 3c, and the carbonate derivative
3h displayed adequate kinetics and met the initial criteria for
advancement.
a Reagents and conditions: (i) paraformaldehyde, ZnCl2, 0 °C, followed
by 55 °C, 20 h; (ii) acetaldehyde, ZnCl2, -20°C, 1 h, followed by room
temperature, 20 h; (iii) 2-propanol, pyridine, 0°C, followed by room
temperature, 20 h.
previously;8 [14C]-3a was prepared from [14C]-1 using the
methodology described above.
Pharmacology and Pharmacokinetics
The three advanced prodrug candidates 3a, 3c, and 3h were
evaluated for physiological stability in buffers ranging from pH
1 to pH 9 as well as in SGF, SIF, and SIBLM. Compounds
were dissolved in a 9:1 mixture of the appropriate test reagent
(buffers, SGF, etc.) and acetonitrile according to the procedure
described in the Supporting Information and examined by LC/
MS at regular intervals. The structures of the prodrugs, the
hemiesters, and 1 were confirmed by their MS spectra. At 4 h,
all three compounds demonstrated good stability in solution at
pH values ranging from 1 to 7.4 but were unstable at pH 9. All
three candidates were stable in simulated gastric fluid; however,
compounds 3a and 3h were more stable than 3c when tested in
simulated intestinal fluid (Table 1). The low melting point of
3h (60-79 °C) made compound 3a (mp ) 127-128 °C) the
more attractive drug candidate. Compound 3a was selected for
further evaluation.
Affinities for the competitive NMDA (glutamate), strychnine-
sensitive glycine, kainate, and AMPA binding sites were
assessed as the compounds’ ability to displace the tritiated
standard ligands CGP-39653, glycine, kainic acid, and R-amino-
3-hydroxy-5-methylisoxazole-4-propionic acid, respectively, as
previously described.10 Functional NMDA antagonism was
assessed using a glutamate-stimulated TCP binding assay.10
Efficacy versus inflammatory pain was measured using a PGE2-
induced thermal hypersensitivity model.11 In vitro plasma and
physiological stability assays were carried out at 37 °C as
described below. Pharmacokinetic and radioactive biodistribu-
tion studies were performed in male Sprague-Dawley rats.
Compounds were administered by oral gavage.
Results and Discussion
With the aim of improving oral bioavailability, a series of
oxymethylene-spaced prodrugs of 1 was synthesized and
evaluated (3a-h). We believed that the prodrugs would
decompose to 1 via a stepwise mechanism that involved
intermediate formation of the hemiesters 2a-h (Figure 1). The
structures of the intermediate hemiesters were established during
The results of detailed physiological stability assays at 37
°C for 3a are summarized in Tables 2 and 3. Compound 3a
was relatively stable in buffers ranging from pH 1 to pH 7.4,
as well as SGF and SIBLM. Considerable conversion to 1 was
observed in pH 8.5 buffer and in SIF after 4 h. The conversion

Prodrugs of Perzinfotel
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 3 773
Figure 1. Conversion of bis-phosphonate prodrugs to 1.
Table 2. In Vitro Physiological Stability of 3a in Various Buffers,
Simulated Gastric Fluid (Expressed as Percent Remaining at Indicated
Time)
hour
pH 1.0
pH 4.5
pH 6.6
pH 7.4
SGFa
SIBLMb
0
4
8
12
16
20
24
100.0
99.7
99.3
98.9
98.5
98.0
97.9
100.0
100.2
100.0
99.8
99.7
99.5
100.0
98.1
95.7
93.7
91.5
89.4
87.4
100.0
91.6
83.2
75.6
68.7
62.2
56.7
100.0
99.6
99.3
99.0
98.7
98.3
98.0
100.0
99.5
99.3
98.6
98.6
98.5
98.1
99.5
a Simulated gastric fluid. b Simulated intestinal bile fluid.
Table 3. In Vitro Physiological Stability of 3a in pH 9.0 Buffer and
Simulated Intestinal Fluid (Expressed as Percent Remaining at Indicated
Time)
pH 9.0
SIFa
hours
3a
2a
1
3a
2a
1
0
4
8
12
16
20
24
99.5
21.6
4.7
1.1
0.3
0.5
74.6
87.0
86.4
82.9
79.1
75.3
0.0
3.8
8.3
12.5
16.9
20.9
24.8
99.5
43.4
22.1
12.0
6.7
0.5
55.4
71.5
77.6
78.8
77.7
75.4
0.0
1.2
6.4
10.4
14.6
18.6
22.5
0.1
0.0
3.7
2.2
a Simulated intestinal fluid.
Figure 2. Stability of compounds 3a-h in rat plasma at 37 °C.
Table 1. In Vitro Physiological Stability of 3a, 3c, and 3h (Expressed
as Percent Remaining after 4 h)
compd pH 1.0 pH 4.5 pH 6.6 pH 7.4 pH 9.0 SGFa SIFb
3a
3c
3h
99.7
89.0
100.1
100.2
99.4
100.0
98.1
95.0
97.7
91.6
94.4
94.1
21.2
17.9
7.0
99.6 43.4
91.6 0.0
100.0 33.8
a Simulated gastric fluid. b Simulated intestinal fluid.
Figure 3. Conversion rate of 3a in rat plasma at 37 °C: X-axis, time
in hour after the addition of 3a; Y-axis, moles remaining expressed as
percent.
products identified included benzoic acid, the hemiester inter-
mediate 2a, and the parent molecule 1. After 24 h of incubation,
the prodrug 3a was essentially gone. However, the hemiester
2a appeared to be significantly more stable, with 75% still
remaining after 24 h.
The stability of 3a in rat plasma was examined in detail at
various time points (0, 0.25, 0.5, 1, 3, 5, and 7 h). The
conversion of 3a to the hemiester 2a was rapid (Figure 3). In
less than 15 min the concentration of 3a was below the level
of detection. The conversion of 2a to 1 was relatively slower,
requiring 3 h to generate around 90% of the total expected molar
percent of 1. The identities of 2a and 1 in this assay were
confirmed by LC/MS. Evaluation of the kinetics of conversion
of 3a to 1 in pH 8.5 buffer, SIF, and rat plasma indicates that
they fit the model of a two-step, first-order reaction. The
calculated rate constants for the two steps (K1 and K2), t1/2 values
for loss of 3a and appearance of 1, and the Tmax for generation
of intermediate 2a are shown in Table 4. These data confirm
the observation that the first step in rat plasma is rapid (K1 >
10 L/h), while the second conversion proceeds at a more modest
rate (K2 ) 0.85 L/h). Conversion rates in pH 8.5 buffer and
SIF were significantly slower than those seen in rat plasma but
still followed the same pattern. On the basis of these favorable
data, compound 3a was examined in vitro and in vivo for its
pharmacological activity and compared to direct administration
of 1.

774 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 3
Baudy et al.
Table 4. PK Parameters for the Conversion of 3a to 1 in Rat Plasma,
pH 8.5 Buffer, and SIF
Table 5. Average Pharmacokinetic Parameters for the Appearance of 1
Following Equieffective Oral Doses of 1 (30 mg/kg) or 3a (10 mg/kg)
in Male Rats
t1/2 of
3a (h)
tmax of
2a (h)
t1/2 of
1 (h)
medium
K1 (1/h)
K2 (1/h)
parameter
1
3a
10
2.4
1
rat plasma
pH 8.5 buffer
SIFa
>10b
0.377
0.194
0.854
0.0136
0.0144
<0.07b
0.19
9.14
14.48
0.87
>24
>24
dose (mg/kg)
t1/2 (h)
30
1.84
1.3
0.5
218
342
342
3.57
Tmax (h)
Cmax (ng/mL)
AUC0-4 (ng·h/mL)
AUC0-last (ng·h/mL)
254
682
857
a Simulated intestinal fluid. b Could not be accurately determined due to
insufficient data points.
On the basis of these data, roughly equieffective doses of 3a
(10 mg/kg, po) and 1 (30 mg/kg, po) were evaluated in
pharmacokinetic studies to examine the pharmacokinetic/
pharmcodynamic relationship of 3a and its conversion to 1. The
results are shown in Table 5. Although the 10 mg/kg, po, dose
of 3a yielded the parent compound 1 with a Cmax value similar
to that seen from direct administration of a 30 mg/kg, po, dose
of 1, the time to achieve Cmax (i.e., Tmax) and the t1/2 were
observed later with administration of 3a, resulting in a larger
AUC for 1 when administered as the prodrug. These data
suggest that higher exposure levels of 1 are being achieved when
given as the prodrug form and are consistent with the 7-fold
greater potency seen with 3a in the behavioral model.
To further characterize the pharmacokinetics of 3a, biodis-
tribution studies were conducted using 10 mg/kg, po, of [14C]-
3a and 30 mg/kg, po, of [14C]-1. Six time points were examined
(0.167, 0.5, 1, 4, 8, and 24 h). Results are presented in Figure
5. Analysis of whole blood, plasma, and brain samples revealed
that 88% of the total radioactivity was in the form of compound
1 at the 0.167 time point and that essentially all of the circulating
radioactivity (98%) was in the form of compound 1 by 0.5 h.
Compound 3a was not detected at any time point and the
hemiester 2a (10% of total radioactivity) was seen only at the
0.167 h time point. In vitro stability tests confirmed that [14C]-
3a was not stable in control rat whole blood or plasma, with no
[14C]-3a detected after 10 min in plasma and after 30 min in
whole blood (data not shown). Therefore, measurements of total
radioactivity at or after the 0.5 h time point can be attributed to
the concentration of compound 1. Total radioactivity in both
plasma and brain were higher after administration of [14C]-3a
Figure 4. Reversal of PGE2-induced thermal hypersensitivity after
administration of compounds 3a and 1: X-axis, time in hour after oral
compound administration; Y-axis, percent reversal of thermal hyper-
sensitivity.
To confirm that the in vivo mechanism of action of 3a
involves conversion to the parent compound 1, both compounds
were evaluated in vitro for their NMDA affinity and functional
NMDA antagonist activity. In a rat cortical binding assay, both
3a and the intermediate hemiester 2a had low affinity for the
competitive NMDA binding site (IC50 ) 4.2 and 17 µM,
respectively) compared to 1 (IC50 ) 23 nM). Likewise,
compound 3a showed no appreciable inhibition of stimulated
TCP binding at high concentrations (8% at 300 µM) compared
to 1, which inhibited this functional measure of NMDA binding
with an IC50 value of 17 µM. The selectivity of 1 for the
competitive glutamate binding site on the NMDA receptor
agreed with earlier data8,11 demonstrating that the compound
had little affinity for strychnine-sensitive glycine, kainate, and
AMPA receptors (percent inhibition of 33%, 6%, and 17%,
respectively, at a concentration of 100 µM).
Compound 3a reversed PGE2-induced thermal hypersensitiv-
ity in a dose-dependent and time-dependent manner (Figure 4)
following oral administration. Importantly, the duration of action
of 3a was substantially longer than that seen with 1. A low
dose of 3a (3 mg/kg) did not substantially modify thermal
hypersensitivity. Higher doses (5.6-30 mg/kg) dose-dependently
blocked thermal hypersensitivity with peak effects occurring
5 h after administration and significant effects seen out to at
least 24 h. In comparison, an oral dose of 100 mg/kg of 1
produced peak effects at 0.5 h after administration, which
declined by 5 h. Potency comparisons evaluated at the time of
peak effect (0.5 h for 1; 5 h for 3a) indicated that 3a (ED50
)
10 mg/kg) was 3.5-fold more potent than 1 (ED50 ) 35 mg/kg)
in this assay. When converted into molar equivalence (molecular
weights: 260 for 1; 529 for 3a), this result translated into a 7-fold
improvement in potency when 1 was administered as the
prodrug 3a. Moreover, the magnitude of activity shown by 3a
in this model is consistent with the profile of 1 previously
described11 and is not shared by all NMDA antagonists.1
Figure 5. Total radioactivity in plasma (a) and brain (b) after the oral
administration of 30 mg/kg [14C]-1 or 10 mg/kg [14C]-3a.

Prodrugs of Perzinfotel
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 3 775
squarate using previously described synthetic methodology8 and
the method described below for the synthesis of unlabeled 3a.
Drugs. Prostaglandin E2 was purchased from Sigma-Aldrich (St.
Louis, MO). For oral dosing, test compounds were suspended in a
vehicle consisting of 2% Tween-80 and 0.5% methylcellulose in
sterile water. Drug concentration doses (calculated as mg/kg) were
calculated using the molecular weight of the base form and were
administered in a volume of 1 mL/kg.
Simulated Physiological Fluids. Simulated Gastric Fluid. The
fluid was generated by preparing a solution of sodium chloride (0.2
g), concentrated HCl (0.7 mL), and pepsin (Sigma, P-7000, 0.32 g)
in deionized water (95 mL). The final volume was adjusted to 100
mL by addition of deionized water.
Simulated Intestinal Fluid. The fluid was generated by prepar-
ing a solution of monobasic potassium phosphate (KH2PO4, 0.68 g),
0.1 N aqueous sodium hydroxide (38 mL), and pancreatin (Sigma,
P-1625, 1 g) in deionized water (57 mL). The pH of the solution
was adjusted to 7.5 by addition of 1 N aqueous sodium hydroxide,
and then the final volume was adjusted to 100 mL by addition of
deionized water.
compared to [14C]-1. Although plasma and brain elimination
t1/2 values were approximately equal after administration of
[14C]-3a and [14C]-1, dose-normalized AUC values for total
radioactivity in plasma and brain were 7.1-fold and 5.4-fold
higher, respectively, following administration of [14C]-3a com-
pared to [14C]-1. This result was apparently due to a higher Cmax
value seen for [14C]-3a. Brain-to-plasma ratios of total radio-
activity were similar after administration of either compound.
The pharmacokinetic data are consistent with the pharmaco-
dynamic results seen in the behavioral model. Equieffective
doses of 3a and 1 yielded plasma levels of 1 with similar Cmax
values, and yet 3a produced a 2-fold higher AUC largely
becasue of longer t1/2. Together with the lack of substantial
affinity for 3a or the intermediate 2a at the NMDA receptor,
these results suggest that the pharmacodynamic activity of 3a
is related to its conversion to 1.
Conclusion
Simulated Intestinal Bile Fluids. Two stock solutions were first
prepared. Stock A consisted of monobasic potassium phosphate
(KH2PO4, 12.2705 g, 0.0902 M) and sodium chloride (1.7435 g,
0.0298 M) dissolved in 1 L of deionized water. Stock B consisted
of dibasic potassium phosphate (K2HPO4, 1.7129 g, 0.0098 M) and
sodium chloride (1.7435 g, 0.0298 M) dissolved in 1 L of deionized
water. SIBLM fed state (pH 5.5) was generated by mixing 72 mL
of stock A, 26 mL of stock B, and sodium taurocholate (0.54 g, 10
mM) and then adjusting the final volume to 100 mL with stock B.
SIBLM fasted state (pH 6.0) was generated by mixing 44 mL of
stock A, 55 mL of stock B, and sodium taurocholate (0.054 g, 1
mM) and then adjusting the final volume to 100 mL with Stock B.
Chemistry. Preparation of 3-{2[8,9-Dioxo-2,6-diazabicyclo[5.2.0-
[non-1(7)-en-2-yl}-3-oxido-7-oxo-7-phenyl-2,4,6-trioxa-3-phospha-
hept-1-yl Benzoate (3a). Chloromethyl Benzoate (4). Zinc chloride
(50 mg) was added to paraformaldehyde (4.5 g) at 0 °C. Under
stirring and protection from moisture, benzoyl chloride (20 g, 16.5
mL, 142.28 mmol) was slowly added to the reaction mixture over
1 h. The reaction mixture was then heated to 55 °C and stirred at
55 °C for 20 h, cooled to ambient temperature, and left overnight.
The reaction mixture was chromatographed on silica gel. Elution
with ethyl acetate/hexane (1:1) gave 17 g (∼70%) of the desired
ester as a clear oil. 1H NMR (DMSO): δ 7.95-8.0 (d, 2H),
7.68-7.75 (t, 1H), 7.52-7.58 (t, 2H), 6.06 (s, 2H).
3-{2[8,9-Dioxo-2,6-diazabicyclo[5.2.0[non-1(7)-en-2-yl}-3-oxido-
7-oxo-7-phenyl-2,4,6-trioxa-3-phosphahept-1-yl Benzoate (3a). Com-
pound 1 (5.25 g, 20.16 mmol) was dissolved in dimethylformamide
(120 mL) at ambient temperature. N,N-Diisopropylethylamine (14
mL, 80.64 mmole) and compound 4 (11.67 g, 68.5 mmol) were
added at once, and the reaction mixture was heated to 70 °C for
20 h. After the mixture was cooled to ambient temperature, ethyl
acetate (100 mL) was added. The organic solution was washed with
saturated aqueous sodium bicarbonate solution followed by brine.
The organic layer was dried over anhydrous magnesium sulfate,
filtered, and evaporated in vacuo. The residue was subjected to silica
gel flash chromatography. Elution with methanol/chloroform/
ammonia (7:92:1) gave the desired product, which was crystallized
from dichloromethane/ethyl acetate/hexane (6:60:35) to yield 10.5 g
(97%) of white microcrystals, mp 124-125 °C. 1H NMR (DMSO):
δ 8.46 (s, 1H), 7.95-7.98 (d, 4H), 7.65-7.72 (t, 2H), 7.47-7.52
(t, 4H), 5.81-5.92 (m, 4H), 3.86-3.95 (m, 2H), 3.26-3.34 (t, 2H
in the water peak), 3.18-3.23 (t, 2H), 2.36-2.43 (m, 2H),
1.78-1.86 (m, 2H). MS (+ESI) m/z ) 528 (M + H)+. Anal.
(C25H25N2O9P) C, H, N.
In conclusion, we designed and prepared a series of oxym-
ethylene-spaced prodrug candidates of the known competitive
NMDA antagonist 1 in an attempt to improve the oral
bioavailability of that compound. Initial plasma stability assays
identified three molecules (3a, 3c, and 3h) that released parent
compound 1 with the appropriate kinetics to be considered as
good potential prodrug candidates. Further physiological stability
measurements and consideration of druglike properties (e.g.,
melting point) led to the selection of compound 3a for detailed
evaluation. Compound 3a yields 1 via the intermediate hemiester
2a in vitro in a two-step reaction sequence, with the initial
conversion of 3a to 2a being considerably more rapid than the
decomposition of 2a to 1. In a PGE2-induced thermal hyper-
sensitivity model, orally administered 3a was more potent at
producing antinociceptive effects than 1 (ED50 ) 10 and 35
mg/kg, respectively). The duration of action of a single,
equieffective dose of 3a was also longer than that seen with 1.
In vitro pharmacological characterization, biodistribution studies,
and pharmacokinetic studies indicate that the greater potency
and duration of action seen with 3a correlate with its conversion
to 1 and the higher exposure of 1 that is achieved when 3a is
administered. When converted to molar equivalence, these
results suggest a 7-fold improvement in potency when 1 is
administered in the prodrug form.
Experimental Section
Materials and Methods. Melting points were determined on a
Thomas-Hoover capillary or an Electrothermal melting point
apparatus and are uncorrected. 1H NMR spectra were recorded on
a Varian Unity Plus 400 spectrometer using the residual DMSO
signal at 2.49 ppm (ppm) as an internal standard. The chemical
shifts are reported in ppm downfield from zero, and coupling
constants are reported in hertz (Hz). Solvate, hydrate, and HCl
protons are not included. Mass spectra were recorded on a Hewlett-
Packard 5995A, a Finnigan Trace MS, or a Micromass LCT
spectrometer. C, H, N combustion analyses either were determined
on a Perkin-Elmer 2400 analyzer or were performed by Robertson
Microlit (Madison, NJ). All analyzed compounds are within (0.4%
of the theoretical value unless otherwise indicated. Flash column
chromatography was performed on EM Science silica gel (230-400
mesh). Commercially available reagents and reaction solvents
(acetonitrile, dichloromethane, ethyl acetate, hexane) were pur-
chased from Aldrich. The reaction solvents were packaged in Sure/
Seal bottles. Solvents for extraction and chromatography (dichlo-
romethane, ethyl acetate) were HPLC-grade OmniSolv reagents
purchased from EM Science. The synthesis of compound 1 has
been described elsewhere.8 [14C]-1 and [14C]-3a were provided by
Amersham Biosciences12 and were prepared from [14C]-diethyl
2,2-Dimethylpropionic Acid (2,2-Dimethylpropioyloxymethoxy)-
[2-(8,9-dioxo-2,6-diazabicyclo[5.2.0]non-1(7)-en-2-yl)ethyl)phosphi-
noyloxymethyl Ester (3b). The title compound was prepared in the
same manner as compound 3a using commercially available
chloromethyl 2,2-dimethylpropionate. The desired product was
obtained as a viscous oil in 64% yield. 1H NMR (DMSO): δ 8.53
(s, 1H), 5.57-5.65 (m, 4H), 3.87-3.93 (m, 2H), 3.35-3.38 (m,

776 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 3
Baudy et al.
2H), 3.25-3.28 (m, 2H), 2.28-2.35 (m, 2H), 1.90-1.94 (m, 2H),
1.98 (s, 18H). MS (+ESI) m/z 498 (M + H)+. Anal. (C21H33N2O9P):
expected, C (51.64), H (6.81), N (5.73); found, C (50.17), H (6.99),
N (5.50).
7-Cyclohexyl-3-{2-[8,9-dioxo-2,6-diazabicyclo[5.2.0]non-1(7)-en-
2-yl]ethyl}-3-oxido-7-oxo-2,46-trioxa-3-phosphahept-1-ylcyclohex-
ane Carboxylate (3c). Chloromethyl Cyclohexanecarboxylate (5).
This reagent was prepared in the same fashion as compound 4
starting with cyclohexanecarboxylic acid chloride. The product was
obtained as a clear oil in 36% yield. 1H NMR (DMSO): δ 5.85 (s,
2H), 1.80-1.87 (m, 2H), 1.64-1.72 (m, 2H), 1.55-1.60 (m, 1H),
1.15-1.42 (m, 6H). MS (+ESI) m/z 176 (M + H)+.
workup, the crude reaction mixture was purified by flash chroma-
tography on silica gel. Elution with 3% methanol/ethyl acetate gave
the desired product as a yellow oil in 16% yield. 1H NMR (DMSO):
δ 8.51 (s, 1H), 6.43-6.46 (m, 1H), 6.38-6.42 (m, 1H), 3.90-3.97
(m, 1H), 3.75-3.82 (m, 1H), 3.34-3.38 (m, 2H), 3.23-3.26 (m,
2H), 2.30-2.37 (m, 2H), 2.20-2.28 (m, 2H), 1.87-1.92 (t, 2H),
1.78-1.83 (m, 4H), 1.62-1.67 (m, 4H), 1.54-1.59 (m, 2H),
1.42-1.45 (2d, 6H), 1.12-1.27 (m, 10H). MS (ESI-) m/z ) 567
(M - H)-. Anal. (C27H41N2O9P) C, H, N.
3-{2-[8,9-Dioxo-2,6-diazabicyclo[5.2.0]non-1(7)-en-2-yl]ethyl}-3-
oxido-7-oxo-8-propyl-2,4,6-trioxa-3-phosphaundec-1-yl 2-Propyl-
pentanoate (3g). Chloromethyl 2-Propylpentanoate (6). The reagent
was prepared in the same manner as intermediate 4 starting with
2-propylbutanoic acid. The product was obtained as an oil in 75%
yield and was used without further purification. MS (+ESI) m/z )
193 (M + H)+.
7-Cyclohexyl-3-{2-[8,9-dioxo-2,6-diazabicyclo[5.2.0]non-1(7)-en-
2-yl]ethyl}-3-oxido-7-oxo-2,46-trioxa-3-phosphahept-1-yl Cyclohex-
anecarboxylate (3c). The title compound was prepared from 1 in
the same fashion as compound 3a starting with intermediate 4. The
desired product was obtained after crystallization from ethyl acetate/
diethyl ether/hexane (6:1:3) as a white solid in 53% yield, mp
3-{2-[8,9-Dioxo-2,6-diazabicyclo[5.2.0]non-1(7)-en-2-yl]ethyl}-3-
oxido-7-oxo-8-propyl-2,4,6-trioxa-3-phosphaundec-1-yl 2-Propyl-
pentanoate (3g). This compound was prepared from 1 in the same
fashion as compound 3a using intermediate 6. After workup, the
crude reaction mixture was purified by flash chromatography on
silica gel. Elution with 8% methanol/ethyl acetate gave the desired
product in 56% yield as a colorless oil which solidified upon
1
64-65 °C. H NMR (DMSO): δ 8.54 (s, 1H), 5.7-6.2 (m, 4H),
3.85-3.93 (m, 2H), 3.35-3.4 (m, 2H), 3.25-3.27, (t, 2H),
2.38-2.41 (m, 2H), 2.28-2.33 (m, 2H), 1.83-1.98 (m, 4H),
1.65-1.72 (d, 4H), 1.58-1.62 (d, 2H), 1.17-1.40 (m, 10H). MS
(+ESI) m/z 541 (M + H)+. Anal. (C25H37N2O9P) C, H, N.
3-{2-[8,9-Dioxo-2,6-diazabicyclo[5.2.0]non-1(7)-en-2-yl]ethyl}-1,5-
dimethyl-3-oxido-7-oxo-7-phenyl-2,4,6-trioxa-3-phosphahept-1-yl
Benzoates 3d (Diastereomer 1) and 3e (Diastereomer 2). 1-Chlo-
roethyl Benzoate (7). Zinc chloride (0.2 g) was added to benzoyl
chloride (20 g, 20.63 mL, 177.85 mmol) at -20 °C. Under stirring
and protection from moisture, acetaldehyde (10 mL, 178.88 mmol)
was added dropwise. After the addition, the reaction mixture was
stirred at -20 °C for 1 h. Thereafter, the reaction mixture was
allowed to reach ambient temperature and stirring was continued
for 20 h. The reaction mixture was concentrated in vacuo and flash-
chromatographed on silica gel. Elution with ethyl acetate/hexane
(1:9) gave 22.4 g (68%) of the desired ester as a clear oil. 1H NMR
(DMSO): δ 8.00-8.05 (d, 2H), 7.72-7.77 (t, 1H), 7.57-7.61 (t,
2H), 6.8-6.85 (q, 1H), 1.9-1.93 (d, 3H). MS (+ESI) m/z ) 184
(M + H)+.
1
standing. H NMR (DMSO): δ 8.57 (s, 1H), 5.58-5.66 (m, 4H),
3.94-3.93 (m, 2H), 3.34-3.37 (m, 2H), 3.26-3.29 (m, 2H),
2.39-2.43 (m, 2H), 2.25-2.35 (m, 2H), 1.89-1.95 (m, 2H),
1.52-1.58 (m 4H), 1.4-1.45 (4H), 1.21-1.30 (m, 8H) 0.82-0.87
(t, 12H). MS (+ESI) m/z ) 573 (M + H)+. Anal. (C27H45N2O9P)
C, H, N.
7-Bis{[(isopropoxycarbonyl)oxy]methyl}-{2-[8,9dioxo-2,6-
diazabicyclo[5.2.0]non-1(7)-en-2-yl]ethyl}phosphonate (3h). Chlo-
romethyl Isopropylcarbonate (9). A solution of chloromethyl
chloroformate (6.5 mL, 73.85 mmol) and 2-propanol (5.6 mL, 73.14
mmole) in diethyl ether (100 mL) was cooled to 0 °C. Pyridine (6
mL, 74.18 mmol) was added dropwise. Thereafter, the ice bath
was removed and the reaction mixture was allowed to reach ambient
temperature and was stirred for 20 h. The formed solids were filtered
and washed with ethyl ether. The the combined supernatant was
washed with aqueous 1% citric acid followed by aqueous 1%
NaHCO3 and brine. After evaporation of the solvent in vacuo, the
desired product was obtained as a clear oil (10.26 g, 92%). 1H NMR
(DMSO): δ 5.82 (s, 2H); 4.77-4.86 (m, 1H), 1.2-1.23 (d, 6H).
MS (ESI+) m/z ) 152 (M + H)+.
7-Bis{[(isopropoxycarbonyl)oxy]methyl}-{2-[8,9dioxo-2,6-
diazabicyclo[5.2.0]non-1(7)-en-2-yl]ethyl}phosphonate (3h). The
title compound was prepared from 1 in the same fashion as
compound 3a using intermediate 9. After workup the crude reaction
mixture was crystallized from ethyl acetate/diethyl ether/hexane (6:
1:3) to yield the desired product as an off-white solid in 44% yield.
1H NMR (DMSO): δ 8.46 (s, 1H), 4.75-4.83 (m, 2H), 3.80-3.87
(m, 2H), 3.32-3.35 (m, 2H), 3.20-3.24 (m, 2H), 2.23-2.35 (m,
2H), 1.82-1.88 (m, 2H), 1.18-1.22 (d, 12H). MS (ESI+) m/z )
493 (M + H)+. Anal. (C19H29N2O11P) C, H, N.
Pharmacology Methods. Animals. Animal maintenance and
research were conducted in accordance with the National Research
Council’s policies and guidelines for the handling and use of
laboratory animals outlined in the Guide for the Care and Use of
Laboratory Animals (National Research Council, 1996). The
laboratory facility was licensed by the United States Department
of Agriculture and accredited by the American Association for
Accreditation of Laboratory Animal Care. Research protocols were
approved by the Wyeth Institutional Animal Care and Use Com-
mittee in accordance with the guidelines of the Committee for
Research and Ethical Issues of the International Association for
the Study of Pain.13
Binding Assays. Membranes derived from rat whole brain were
prepared as previously described for [3H]-CGP-39653 (NMDA
recognition site), [3H]-TCP (PCP site), and [3H]-glycine binding
assays,14 whereas rat frontal cortex and forebrain membranes were
used as previously reported for [3H]-kainate and [3H]-AMPA
binding assays, respectively.15 Competition binding experiments
3-{2-[8,9-Dioxo-2,6-diazabicyclo[5.2.0]non-1(7)-en-2-yl]ethyl}-1,5-
dimethyl-3-oxido-7-oxo-7-phenyl-2,4,6-trioxa-3-phosphahept-1-yl
Benzoates 3d (Diastereomer 1) and 3e (Diastereomer 2). The
diastereomeric mixture of compounds 3d and 3e was prepared from
compound 1 in the same fashion as compound 3a using intermediate
7. After the reaction workup the diastereomers were separated by
flash column chromatography on silica gel.
Elution with 8% methanol/chloroform gave 7% of the first
diastereomer compound 3d as a yellow glasslike material. 1H NMR
(DMSO): δ 8.41 (s, 1H), 7.92-7.96 (d, 4H), 7.63-67 (t, 2H),
7.44-7.53 (t, 4H), 6.21-6.26 (2H), 3.8-3.87 (m, 2H), 3.22.3.30
(m, 2H), 3.15-3.19 (m, 2H), 2.24-2.34 (m, 2H), 1.77-1.81 (m,
2H), 1.53-1.58 (d, 6H). MS (+ESI) m/z ) 557 (M + H)+. Anal.
(C27H29N2O9P) C, H, N.
Further elution with 8% methanol/chloroform afforded compound
3e as a yellow glasslike material in 16% yield. 1H NMR (DMSO):
δ 8.42 (s, 1H), 7.9-7.96 (m, 4H), 7.62-7.68 (q, 2H), 7.45-7.54
(m, 4H), 6.7-6.75 (m, 1H), 6.61-6.67 (m, 1H), 3.93-4.02 (m,
1H), 3.73-3.84 (m, 1H), 3.24-3.30 (m, 2H), 3.15-3.20 (m, 2H),
2.23-2.36 (m, 2H), 1.77-1.82 (t, 2H), 1.56-1.59 (d, 3H),
1.43-1.46 (s, 3H). MS (+ESI) m/z ) 557 (M + H)+. Anal.
(C27H29N2O9P) C, H, N.
7-Cyclohexyl-3-{2-[8,9-dioxo-2,6-diaza[5.2.0]non-1(7)-en-2-yl]ethyl}-
1,5-dimethyl-3-oxido-7-oxo-2,4,6-trioxa-3-phosphahept-1-yl Cyclo-
hexanecarboxylate (3f). 1-Chloroethyl Cyclohexanecarboxylate
(8). This reagent was prepared in the same manner as intermediate
7 starting with cyclohexanecarboxylic acid chloride. The product
was obtained as a clear oil in 53% yield and was used without
further purification. MS (+ESI) m/z ) 191 (M + H)+.
7-Cyclohexyl-3-{2-[8,9-dioxo-2,6-diaza[5.2.0]non-1(7)-en-2-yl]ethyl}-
1,5-dimethyl-3-oxido-7-oxo-2,4,6-trioxa-3-phophsphahept-1-yl Cy-
clohexanecarboxylate (3f). The title compound was prepared from
1 in the same fashion as compound 3a using intermediate 8. After

Prodrugs of Perzinfotel
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 3 777
utilizing [3H]-CGP-39653, [3H]-glycine, [3H]-kainic acid, and [3H]-
AMPA were carried out as previously described.16-19 Functional
[3H]-TCP binding assays were performed as previously described19
with modifications.10 Competition binding curves were fitted with
a four-parameter logistic model, which defined IC50 values for
compounds 1 and 3a (GraphPad Prism, San Diego, CA).
parent compounds and conversion products were confirmed by LC/
MS (-ESI, scan range 200-1000 m/z, 60 psi, 12.0 L/min, 350
°C). The kinetics of the conversion reactions were analyzed by
MicroMath Scientist software.
Plasma Stability. The in vitro physiological stability of com-
pounds 3a-h in rat plasma was determined by dissolving the
compounds in rat plasma to a concentration of 0.05 mg/mL.
Aliquots of 0.5 mL of the compound/plasma mixture were placed
separately into extraction tubes and incubated at 37 °C for 0, 0.25,
0.5, 1, 3, 5, 7, and 24 h, respectively. At the appropriate time, the
plasma sample was treated with 0.5 mL of acetonitrile and
centrifuged. The supernatant was diluted with deionized water and
injected onto the HPLC system (Inertsil C8-3, 3 µm, 150 mm ×
4.6 mm; 1 mL/min; mobile phase of 0.1% formic acid in water/
0.1% formic acid in acetonitrile). The percentage of the parent
compound and conversion products in plasma were calculated on
the basis of area comparison of their peaks at 304 nm, corrected
by an internal standard. The identities of the parent compounds
and conversion products were confirmed by LC/MS (-ESI, scan
range 200-1000 m/z, 60 psi, 12.0 L/min, 350 °C). The kinetics of
the conversion reactions were analyzed by MicroMath Scientist
software.
Pharmacokinetic Studies. Male Sprague-Dawley rats were
administered either 1 (30 mg/kg po) or 3a (10 mg/kg po) via oral
gavage. These doses were chosen because they are pharmacologi-
cally equivalent. Plasma samples were taken at 0, 0.25, 0.5, 1, 2,
4, and 6 h following administration. The number of animals per
time point were n ) 2 for 0, 2, 4, and 6 h and n ) 3 for 0.25, 0.5,
and 1 h. Plasma samples were stabilized by adding 5 µL of glacial
acetic acid and 10 µL of 300 mM aqueous potassium fluoride
solution. For MS, the plasma samples were acidified using 20 mL
of 1 N HCl and extracted into 4 volumes of n-butanol. The butanol
layer was concentrated to dryness in vacuo and reconstituted into
100 µL of water for injection onto the HPLC system (Aquasil C18,
3 µm, 2 mm × 20 mm; temperature of 50 °C; autosampler
temperature of 10 °C; 1 mL/min; 4:1 split into mass spectrometer,
200 µL/min into the mass spectrometer; mobile phase of 0.1%
formic acid in water/0.1% formic acid in acetonitrile). Pharmaco-
kinetic parameters were calculated using IBIS (version 1.5.0),
DMAS (version 1.2), or WinNonlin (version 2.1) software.
Radiolabeled Oral Bioavailability Studies. Male Sprague-
Dawley rats were administered either [14C]-1 (30 mg/kg) or [14C]-
3a (10 mg/kg) via oral gavage. Blood, plasma, and brain (n ) 3
rats/time point) were collected at 0.167, 0.5, 1, 4, 8, and 24 h
postdose. Radioactivity concentrations were determined in each
matrix. Metabolite profiles in plasma and brain were determined
using HPLC (Aquasil C18, 3 µm, 2 mm × 20 mm; temperature of
50 °C; autosampler temperature of 10 °C; 1 mL/min; mobile phase
of 0.1% formic acid in water/0.1% formic acid in acetonitrile) with
in-line radioactivity detection. The stability of [14C]-3a in control
rat whole blood and plasma incubated at 37 °C was also determined.
PGE2-Induced Thermal Hypersensitivity. Male Charles River
Sprague-Dawley rats (Kingston/Stoneridge, NY) weighing 200-250
g at time of arrival were individually housed in wire cages in a
climate-controlled room. A 12 h light/dark cycle (lights on at 06:
30) was in effect for all animals, and water was available ad libitum.
For all oral dosing studies, rats were fasted for approximately 16 h
before drug administration. To assess baseline thermal sensitivity,
the terminal 10 cm of the tail was placed into water warmed to 34,
38, 42, 46, 50, 54, or 58 °C. The latency in seconds for the animal
to remove the tail from the water was used as a measure of
nociception. If the animal did not remove the tail within 20 s, the
experimenter removed the tail and a maximum latency of 20 s was
recorded. Following the assessment of baseline thermal sensitivity,
thermal hypersensitivity was produced by an intradermal 50 µL
injectionof0.1mgofPGE2(Sigma,St.Louis,MO).Temperature-effect
curves were generated before (baseline) and after (30 min) PGE2.
On the basis of previous results,20 subjects were tested a maximum
of three times with a minimum of 5 days between tests (one baseline
0.1 mg PGE2 assessment and one or two compound tests in the
presence of 0.1 mg of PGE2). The ability of compounds to block
0.1 mg PGE2-induced thermal hypersensitivity was assessed using
a single dose procedure. Under this procedure, a single dose of
compound was administered ip or po at 0.5, 1.7, 3, 5, 24, or 48 h
before the injection of PGE2 and thermal sensitivities were assessed
30 min after PGE2 injection.
Temperature-effect curves were generated for each experimental
condition for individual rats. The temperature that produced a half-
maximal increase in the tail-withdrawal latency (i.e., T10) was
calculated from each temperature-effect curve. The T10 was
determined by interpolation from a line drawn between the point
above and the point below 10 s on the temperature-effect curve.
For these studies, thermal hypersensitivity was defined as a leftward
shift in the temperature-effect curve and a decrease in the T10 value.
Reversal of thermal hypersensitivity was defined as a return to
baseline of the temperature-effect curve and the T10 value.
Blockade of irritant-induced thermal hypersensitivity was quantified
as the percentage return to baseline values (% reversal) according
to the following equation:
(Td10rug+irritant) - (Tirritant
)
10
% reversal )
× 100
(Tb1a0seline) - (Tirritant
)
10
drug+irritant
in which T
is the T10 after a drug in combination with PGE2
is the T10 after PGE2 or capsaicin alone, and
10
irritant
10
or capsaicin, T
baseline
T
is the T10 under control conditions. Statistical analysis was
10
Acknowledgment. The authors thank Wyeth Discovery
Analytical Chemistry for their assistance in the structural
confirmation of the compounds described in this manuscript and
for their help with HPLC separations. The authors also thank
Amersham Biosciences for supplying [14C]-1 and [14C]-3a.
done using a within-subjects-repeated-measures analysis of variance
on T10 values. The criterion for significant reversal of the T10 value
from the chemical irritant alone was p < 0.05.
In Vitro Stability Studies. Physiological Stability. The physi-
ological stability of derivative 3a was determined by examining
the stability of the compound in a series of buffers with pH values
ranging from 1 to 9 as well as in SGF, SIF, and SIBLM at 37 °C.
The compounds were prepared in a 9:1 mixture of the appropriate
test component (buffer, SGF, SIF, or SIBLM) and acetonitrile to a
final concentration of 0.01 mg/mL according to the procedure
described in the Supporting Information. The samples were
thoroughly mixed and placed in an autosampler maintained at 37
°C. Each sample was injected consecutively onto the system (Inertsil
C8-3, 3 µm, 150 mm × 4.6 mm; 1 mL/min; mobile phase of 0.1%
formic acid in water/0.1% formic acid in acetonitrile; MS scan range
of 200-1000 m/z, 60 psi, 12.0 L/min, 350 °C) every 4 h over a
24 h period. The percent remaining at each injection time was
calculated on the basis of the area comparison of compound peak
at 304 nm, corrected by an internal standard. The identities of the
Supporting Information Available: Results from elemental
analysis for compounds 3a-h and experimental protocol for the
preparation of samples for physiological stability measurements.
This material is available free of charge via the Internet at http://
pubs.acs.org.
References
(1) Sluka, K. A.; Westlund, K. N. Spinal cord amino acid release and
content in an arthritis model: the effects of pretreatment with non-
NMDA, NMDA, and NK1 receptor antagonists. Brain Res. 1993, 627,
89–103.
(2) Kawamata, M.; Omote, K. Involvement of increased excitatory
amino acids and intracellular Ca2+ concentration in the spinal dorsal

778 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 3
Baudy et al.
horn in an animal model of neuropathic pain. Pain 1996, 68, 85–
(11) Brandt, M. R.; Cummons, T. A.; Potestio, L.; Sukoff, S. J.; Rosenz-
weig-Lipson, S. Effects of the N-methyl-D-aspartate receptor antagonist
perzinfotel [EAA-090; [2-(8,9-dioxo-2,6-diazabicyclo[5.2.0 non-1(7)-
en-2-yl)-ethyl]phosphonic acid] on chemically-induced thermal hy-
persensitivity. J. Pharmacol. Exp. Ther. 2005, 313, 1379–1386.
(12) Amersham Biosciences, part of GE Healthcare Life Sciences, 800
Centennial Avenue, P.O. Box 1327, Piscataway, NJ 08855.
(13) Zimmermann, M. Ethical guidelines for investigations of experimental
pain in conscious animals. Pain 1983, 16, 109–110.
(14) Murphy, D. E.; Schneider, J.; Boehm, C.; Lehmann, J.; Williams, J.
Binding of [3H]3-(2-carboxypiperazin-4-yl)propyl-1-phosphonic acid
to rat membranes: a selective, high-affinity ligand for N-methyl-D-
aspartate receptors. J. Pharmacol. Exp. Ther. 1987, 240, 778–783.
(15) London, E. D.; Coyle, J. T. Specific binding of [3H]kainic acid receptor
sites in rat brain. Mol. Pharmacol. 1979, 15, 492–505.
(16) Sills, M. A.; Fagg, G.; Pozza, M.; Angst, C.; Brundish, D. E.; Hurt,
S. D.; Wilusz, E. J.; Williams, M. [3H]CGP 39653; a new N-methyl-
D-aspartate antagonist radioligand with low nanomolar affinity in rat
brain. Eur. J. Pharmacol. 1991, 192, 19–24.
(17) Monahan, J. B.; Corpus, V. M.; Hood, W. F.; Thomas, J. W.; Compton,
R. P. Characterization of a [3H]glycine recognition site as a modulatory
site of the N-methyl-D-aspartate receptor complex. J. Neurochem. 1989,
53, 370–375.
96.
(3) Dickenson, A. H.; Chapman, V.; Green, G. M. The pharmacology of
excitatory and inhibitory amino acid-mediated events in the transmis-
sion and modulation of pain in the spinal cord. Gen. Pharmacol. 1997,
28, 633–638.
(4) Mao, J.; Price, D. D.; Hayes, R. L.; Lu, J.; Mayer, D. J.; et al.
Intrathecal treatment with dextrorphan or ketamine potently reduces
pain-related behaviors in a rat model of peripheral mononeuropathy.
Brain Res. 1993, 605, 164–168.
(5) Chaplan, S. R.; Malmberg, A. B.; Yaksh, T. L. Efficacy of spinal nmda
receptor antagonism in formalin hyperalgesia and nerve injury evoked
allodynia in the rat. J. Pharmacol. Exp. Ther. 1997, 280, 829–838.
(6) Suzuki, R.; Matthews, E. A.; Dickenson, A. H. Comparison of the
effects of MK-801, ketamine and memantine on responses of spinal
dorsal horn neurones in a rat model of mononeuropathy. Pain 2001,
91, 101–109.
(7) Kristensen, J. D.; Svensson, B.; Gordh, T., Jr. The NMDA-receptor
antagonist CPP abolishes neurogenic “wind-up pain” after intrathecal
administration in humans. Pain 1992, 51, 249–253.
(8) Kinney, W. A.; Abou-Gharbia, M.; Garrison, D. T.; Schmid, J.; Kowal,
D. M.; et al. Design and synthesis of [2-(8,9-dioxo-2,6-diazabicyclo-
[5.2.0]non-1(7)-en-2-yl)-ethyl]phosphonic acid (EAA-090), a potent
N-methyl-D-aspartate antagonist, via the use of 3-cyclobutene-1,2-dione
as an achiral alpha-amino acid bioisostere. J. Med. Chem. 1998, 41,
236–246.
(9) Childers, W. E.; Abou-Gharbia, M. A.; Moyer, J. A.; Zaleska, M. M.
EAA-090, neuroprotectant, competitive NMDA antagonist. Drugs
Future 2002, 27, 633–638.
(10) Sun, L.; Chiu, D.; Kowal, D.; Simon, R.; Smeyne, M.; et al.
Characterization of two novel N-methyl-D-aspartate antagonists:
EAA-090 (2-[8,9-dioxo-2,6-diazabicyclo [5.2.0]non-1(7)-en2-yl-
]ethylphosphonic acid) and EAB-318 (R-alpha-amino-5-chloro-1-
(phosphonomethyl)-1H-benzimidazole-2-propanoic acid hydrochlo-
ride). J. Pharmacol. Exp. Ther. 2004, 310, 563–570.
(18) Murphy, D. E.; Showhill, E. W.; Williams, M. Characterization of
quisqualate recognition sites in rat train tissue using DL-[3H]-R-amino-
3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA) and a filtration
assay. Neurochem. Res. 1987, 12, 775–782.
(19) Ransom, R. W.; Stec, N. L. Cooperative modulation of [3H]MK-801
binding to the N-methyl-D-aspartate receptor-ion channel complex by
L-glutamate, glycine, and polyamines. J. Neurochem. 1988, 51, 830–
836.
(20) Brandt, M. R.; Furness, M. S.; Mello, N. K.; Rice, K. C.; Negus, S. S.
Antinociceptive effects of delta-opioid agonists in rhesus monkeys:
effects on chemically-induced thermal hypersensitivity. J. Pharamcol.
Exp. Ther. 2001, 296, 939–946.
JM8011799