Design and pharmacological activity of phosphinic acid based NAALADase inhibitors

Article abstract of DOI:10.1021/jm0001774

A novel series of phosphinic acid based inhibitors of the neuropeptidase NAALADase are described in this work. This series of compounds is the most potent series of inhibitors of the enzyme described to date. In addition, we have shown that these compounds are protective in animal models of neurodegeneration. Compound 34 significantly prevented neurodegeneration in a middle cerebral artery occlusion model of cerebral ischemia. In addition, in the chronic constrictive model of neuropathic pain, compound 34 significantly attenuated the hypersensitivity observed with saline-treated animals. These data suggest that NAALADase inhibition may provide a new approach for the treatment of both neurodegenerative disorders and peripheral neuropathies.

Full text of DOI:10.1021/jm0001774

4170
J . Med. Chem. 2001, 44, 4170-4175
Design a n d P h a r m a cologica l Activity of P h osp h in ic Acid Ba sed NAALADa se
In h ibitor s
Paul F. J ackson,*,† Kevin L. Tays,‡ Keith M. Maclin, Yao-Sen Ko, Weixing Li, Dil Vitharana,
Takashi Tsukamoto, Doris Stoermer, Xi-Chun M. Lu, Krystyna Wozniak, and Barbara S. Slusher
Guilford Pharmaceuticals Inc., 6611 Tributary Street, Baltimore, Maryland 21224
Received April 19, 2000
A novel series of phosphinic acid based inhibitors of the neuropeptidase NAALADase are
described in this work. This series of compounds is the most potent series of inhibitors of the
enzyme described to date. In addition, we have shown that these compounds are protective in
animal models of neurodegeneration. Compound 34 significantly prevented neurodegeneration
in a middle cerebral artery occlusion model of cerebral ischemia. In addition, in the chronic
constrictive model of neuropathic pain, compound 34 significantly attenuated the hypersensitiv-
ity observed with saline-treated animals. These data suggest that NAALADase inhibition may
provide a new approach for the treatment of both neurodegenerative disorders and peripheral
neuropathies.
In tr od u ction
damage in diseases in which excess glutamatergic
transmission has been implicated.9 In addition, NAAG
has been shown recently to be an agonist at mGluR3
receptors and a mixed agonist/antagonist at the NMDA
receptor.10 Because mGluR3 agonists are neuroprotec-
tive, elevation of NAAG levels through NAALADase
inhibition could also be beneficial.
The dipeptide N-acetyl-L-aspartyl-L-glutamate (NAAG)
was first identified in the brain in 19651 as one of the
major peptidic components of the central nervous sys-
tem. NAAG was found in high levels in the brain stem
and spinal cord, however, its physiological function was
not determined at the time of its discovery. It was not
until the early 1980s when NAAG was reisolated that
its role as a potential neurotransmitter was explored.2
To prove the dipeptide was a neurotransmitter, a
method of inactivation of NAAG had to be determined.
Toward this end, an enzyme responsible for the catabo-
lism of NAAG was identified in 1987.3 The enzyme was
named NAALADase (N-acetylated-R-linked acidic dipep-
tidase, EC 3.4.17.21) for its specificity as a dipeptidase.
Subsequent in vitro and in vivo studies demonstrated
that NAALADase cleaves NAAG into N-acetylaspartate
and glutamate.4 The membrane bound enzyme has an
apparent molecular weight of 94 kDa and has the
characteristics of a metallopeptidase. Human NAALA-
Dase has been cloned and was shown to be similar to
both prostate specific membrane antigen (PSMA) and
folylpoly-γ-glutamate carboxypeptidase (FGCP).5,6 In
addition, work by Barrett has indicated that this
enzyme is a type II protein of the cocatalytic class of
metallopeptidases and contains two zinc atoms in the
active site.7 Site-directed mutagenesis studies subse-
quently confirmed the active site residues surrounding
the two zinc atoms.8 In addition, the mutagenesis
studies revealed residues potentially responsible for the
substrate specificity of the enzyme.
The first potent and selective inhibitor of NAALA-
Dase, 2-(phosphonomethyl)pentanedioic acid (2-PMPA,
1), has been previously described.11 We have reported
that this prototype inhibitor provides neuroprotection
in an animal model of cerebral ischemia.12 In this paper,
we describe the SAR and synthesis around a series of
phosphinic acid derivatives of 2-PMPA. Importantly, we
also provide data showing that inhibition of NAALA-
Dase with these compounds prevents neurodegeneration
in several animal models of CNS and PNS disorders.
The results described herein demonstrate that NAALA-
Dase inhibition is a new and novel mechanism for
therapeutic intervention in neurodegenerative disor-
ders.13
Although compound 1 was reported to be a potent and
specific inhibitor of the enzyme, the highly polar nature
of the molecule (clogD ) -13, pH ) 7.4) limits penetra-
tion into the brain.12 We began our work by modifying
the acidic portions of compound 1 in order to obtain
more lipophilic compounds and in turn a more favorable
pharmacological profile. In addition, we hoped to gain
an understanding of the steric and electronic require-
ments for effective inhibition of NAALADase.
Syn th esis of P h osp h on a tes a n d P h osp h in ic
Acid s
We became interested in the inhibition of NAALA-
Dase as a potential way of treating neurological disor-
ders. Because NAAG may serve as a storage form of
glutamate, inhibition of NAALADase should decrease
synaptic glutamate levels and be effective in preventing
The phosphonates reported in this work were pre-
pared by the general method shown in Scheme 1.14 By
treating diethyl malonate (2) with sodium ethoxide and
an alkyl halide in ethanol, we obtained the substituted
malonate in good yield. Saponification with potasium
hydroxide resulted in compound 3, which was converted
to an R,â-unsaturated acid with formaldehyde and
† Current address: The R. W. J ohnson Pharmaceutical Research
Institute, Route 202, P.O. Box 300, Raritan, NJ 08869.
‡ Current address: The R. W. J ohnson Pharmaceutical Research
Institute, 3210 Merryfield Row, San Diego, CA 92121.
10.1021/jm0001774 CCC: $20.00 © 2001 American Chemical Society
Published on Web 10/18/2001

Phosphinic Acid Based NAALADase Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24 4171
Sch em e 1a
Sch em e 3a
a
Reagents: (a) NaOEt, RBr, EtOH, reflux, 8 h; (b) KOH, EtOH,
0 °C to room temperature, 4 h; (c) HCHO, Et2NH, H2O, reflux; (d)
BnBr, DMF, H2O, K2CO3, rt, 3 h; (e) (BnO)2P(O)H, NaH, THF, 1
h, rt; (f) H2, Pd/C, H2O.
Sch em e 2a
a
Reagents: (a) HMPT, 105 °C; (b) ammonium phosphinate,
HMDS; (c) tBuOH, EDC, CH2Cl2; (d) NaH, THF, substituted
acrylate; (e) H2, Pd/C; (f) TFA, DCM; (g) NaH, THF, substituted
benzaldehyde; (h) substituted hexahydro-1,3,5-triazine, toluene,
110 °C.
lysine or arginine in the specificity pocket of NAALA-
Dase.6 Shortening the side chain had a greater negative
effect than lengthening it (i.e., compounds 17 and 18),
potentially indicating that the propionic acid side chain
of compound 1 and the glutamate side chain of NAAG
may bind to NAALADase in an extended rather than a
folded conformation. The phenyl derivatives 19 and 20
were prepared to see if a favorable cation-π interaction
could be achieved with an arginine in the binding
pocket. The two aromatic compounds had suprisingly
high affinity considering they lacked an acidic function-
ality. We next prepared compound 21 but did not see
an increase in affinity compared to compound 19.
Presumably either the pKa of the phenol is too dissimilar
to that of an acid or there exists an unfavorable
orientation of the phenol relative to the basic residues
in the binding pocket. Decreasing the pKa of the phenol
via substitution of the ring or changing the position of
the hydroxyl group may result in increased affinity for
the enzyme. Two additional compounds, nitrile 22 and
tetrazole 23, were prepared but also showed modest
activity. Overall, our efforts to alter the propionic acid
side chain at C-2 of compound 1 resulted in less potent
NAALADase inhibitors.
We then focused on changing the zinc-binding phos-
phonic acid of compound 1 to a phosphinic acid, leaving
the propionic acid side chain unchanged. Table 3 sum-
marizes the compounds prepared via the methods
described above and their corresponding IC50 values.
Within the series of derivatives 24-27, the benzyl
derivative 26 demonstrated the best activity. Large
groups directly attached to the phosphorus decreased
activity (compound 25). The benzyl series was elabo-
rated to include compounds 28-34 but resulted in only
a 2-fold decrease in the IC50 value. By placing an alkyl
a
Reagents: (a) HMDS, 100 °C, 2 h; CH2Cl2, -78 °C to room
temperature, 15 h; (b) HCl, CH2Cl2; BnOH, EDC, CH2Cl2; (c) NaH,
THF, 0 °C to room temperature, 25 min; (d) H2, Pd/C, H2O.
diethylamine. The R,â-unsaturated acid was then con-
verted to benzyl ester 4. Conjugate addition of dibenzyl
phosphite to 4 with NaH (0.4 equiv) as an initiator, fol-
lowed by hydrogenation of the benzyl protecting groups,
provided the monocarboxylic acid phosphonate 5.
The dicarboxylic phosphinic acids utilized in this work
were prepared via several methods, one of which is
shown in Scheme 2. A benzyl protected phosphinic acid
8 was prepared by reacting ammonium phosphonate
with an alkyl bromide.15 Phosphinic acid 8 was treated
with NaH and the benzyl acrylate dimer 9a in THF to
provide compound 10. Finally, hydrogenation of the
benzyl esters afforded phosphinic acid 11. Alternatively,
an appropriately substituted acrylate (12) could be
dimerized to give the protected diester 9 (Scheme 3).
This was then reacted with ammonium phosphinate and
hexamethyldisilazane followed by treatment with tBuOH
and EDC to give the protected phosphinate 13. This
could then be utilized as an intermediate in the syn-
thesis of tri and tetra acid derivatives (14), R-hydroxy
derivatives (15), and amino derivatives (16) (Table 1).
In Vitr o Resu lts
Analysis of the data presented in Table 2 indicates
that modification of the propionic acid side chain of
compound 1 greatly reduces enzyme inhibition efficacy.
The side chain acid is thought to interact with either a

4172 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24
J ackson et al.
Ta ble 1. Representative Analytical Data
compound
analytical data
dibenzyl 2-methylenepentanedioate (9a )
di-tert-butyl 2-methylenepentanedioate (9b)
1H NMR (CDCl3) 2.6 (m, 4H), 5.1 (s, 2H), 5.2 (s, 2H), 5.6 (s, 1H), 6.2 (s, 1H), 7.3 (m, 10H)
1H NMR (CDCl3) 1.4 (m, 18H), 2.4 (t, 2H), 2.6 (t, 2H), 5.5 (s, 1H), 6.0 (s, 1H)
di-tert-butyl 2-[(tert-butoxyphosphinyl)methy]- 1H NMR (CDCl3) 1.4 (m, 27 H), 1.8 (m, 1H), 1.9 (m, 2H), 2.1 (m, 1H), 2.3
pentanedioate (13b)
(m, 2H), 2.7-2.8 (m, 1H), 6.7 & 8.0 (d, 1H, P-H)
2-[[(2-flurobenzylhydoxyphosphinoyl]methyl]- 1H NMR (D2O) 1.8-1.9 (m, 3H), 2.0-2.2 (m, 1H), 2.3-2.4 (m, 2H), 2.6-2.7
pentanedioic acid (28)
(m, 1H), 3.28 (d, J ) 16.6 Hz, 2H), 7.1-7.5 (m, 4H). Anal. Calcd for
C
13H16FO6P‚0.1H2O: C, 48.79; H, 5.10. Found: C, 48.84; H, 5.14.
2-[[(4-methylbenzyl)hydroxyphosphinoyl]-
methyl]pentanedioic acid (30)
1H NMR (D2O) 1.7-1.9 (m, 3H), 2.13 (ddd, J ) 9.3 Hz, 13.3 Hz, 15.5 Hz, 1H),
2.33 (d, J ) 2.1 Hz, 3H), 2.40 (dt, J ) 2.1 Hz, 7.4 Hz, 2H), 2.5-2.7 (m, 1H),
3.18 (d, J ) 16.9 Hz, 2H), 7.23 (s, 4H). Anal. Calcd for C14H19O6P‚0.30H2O:
C, 52.60; H, 6.18. Found: C, 52.60; H, 6.28.
2-[[(4-methoxybenzyl)hydroxyphosphinoyl]-
methyl]pentanedioic acid (31)
1H NMR (D2O) 1.8-1.9 (m, 3H), 2.0-2.2 (m, 1H), 2.3-2.4 (m, 2H), 2.55-2.70 (m, 1H),
3.16 (d, J ) 16.7 Hz, 2H): 3.81 (s, 3H), 6.98 (d, J ) 8.7 Hz, 2H), 7.25 (d, J ) 8.7 Hz,
2H). Anal. Calcd for C14H19O7P‚0.30H2O: C, 50.09; H, 5.89. Found: C, 49.98; H, 5.80.
1H NMR (D2O) 1.8-2.0 (m, 3H), 2.1-2.3 (m, 1H), 2.3-2.5 (m, 2H), 2.7-2.9
(m, 1H), 3.29 (d, J ) 15.4 Hz, 2H). Anal. Calcd for C13H12F5O6P‚0.45H2O; C,
39.20; H, 3.26. Found C, 39.17; H, 3.28.
2-[[(pentaflurobenzyl)hydroxyphosphinoyl]-
methyl]pentanedioic acid (34)
2-[[[2-(benzylcarboxy)propyl]hydroxyphos-
phinoyl]methyl]pentanedioic acid (37)
1H NMR (D2O) 1.2 (m, 3 H), 1.6-1.8 (m, 4H), 2.1 (m, 2H), 2.2 (m, 2H), 2.6
(m, 1H), 2.8 (m, 1H), 5.0 (m,2H), 7.3 (m, 5H). Anal. Calcd for
C
17H23O8P‚1.0H2O: C, 50.50; H, 6.23. Found: C, 50.52; H, 5.92.
2-[[(carboxy)propyl]hydroxyphosphinoyl]-
methyl]pentanedioic acid (38)
1H NMR (D2O) 1.2 (d, 3H), 1.9 (m, 4H), 2.2 (m, 2H), 2.4 (m, 2H), 2.8 (m, 2H).
Anal. Calcd for C10H17O8P‚0.2CH3CN: C, 41.03; H, 5.83. Found: C, 41.05;
H, 5.92.
2-[(methylhydroxyphosphinoyl)methyl]-
pentanedioic acid (39)
1H NMR (CDCl3) 1.9 (m, 3H), 2.2 (m, 1H), 2.4 (m, 2H), 2.7 (m, 1H) 5.0 (d, 1H),
7.4 (m, 5H). Anal. Calcd for C13H17O7P‚0.6H2O: C, 47.74; H, 5.61. Found:
C, 47.73; H, 5.68.
2-[[[benzylamino]methyl](hydoxyphos-
phinoyl)methyl]pentanedioic acid (42)
2-[[[phenylamino]methyl](hydoxyphos-
phinoyl)methyl]pentanedioic acid (43)
2-[[[4-methoxyphenylamino]methyl]-
(hydoxyphosphinoyl)methyl]-
1H NMR (D2O) 1.6-2.0 (m, 4H), 2.2-2.4 (m, 2H), 2.5-2.7 (m, 1H), 3.00 (d,
J ) 9.0 Hz, 2H), 4.21 (s, 2H), 7.3-7.5 (m, 5H)
1H NMR (D2O) 1.6-2.0 (m, 4H), 2.2-2.4 (m, 2H), 2.5-2.7 (m, 1H), 3.53 (d,
J ) 8.8 Hz, 2H), 7.3-7.5 (m, 5H)
1H NMR (D2O) 1.2-1.3 (m, 1H), 1.6-1.7 (m, 3H), 2.2 (m, 2H), 2.3-2.5 (m, 1H),
3.4 (d, J ) 8.9 Hz, 2H), 3.7 (s, 3H), 7.0 (d, J ) 12 Hz, 2H), 7.4 (d, J )
12 Hz, 2H)
pentanedioic acid (44)
Ta ble 2a
Ta ble 3a
compd
R
IC50 (nM)
0.3
2200
SD
1
17
18
19
20
21
22
23
CH2CH2COOH
CH2CO2H
CH2CH2CH2COOH
CH2Ph
(0.05
(1960
(14
compd
R
IC50 (nM)
SD
185
548
199
508
335
175
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
n-C3H7
Ph
CH2Ph
CH2CH2Ph
CH2(2-F-Ph)
360
2930
53
149
156
143
68
90
55
49
82
0.5
2
95
1.5
55
10
16
59
4
3
(124
(1250
(33
(78
(35
(53
(0
(160
(106
(412
(64
CH2CH2Ph
CH2(3-OH)Ph
CH2CH2CN
CH2CH2(CN4)
(35
CH2(3-NH2-Ph)
CH2(4-CH3-Ph)
a
IC50 values were determined as previously described in ref 11,
with each value representing two to four independent determina-
tions. All of the data reported are for the racemic compounds.
CH2(4-OCH3-Ph)
CH2(3,5-CF3-Ph)
CH2(3,5-F-Ph)
CH2 (2,3,4,5,6-F-Ph)
CH2CH(CH2CH2CO2H)CO2H
CH2CH(CH2Ph)CO2H
CH2CH(CH3)CO2Bn
CH2CH(CH3)CO2H
CH(OH)Ph
CH(OH)-4-pyridyl
CH(OH)-(3-F-Ph)
CH2 NHCH2Ph
CH2NHPh
CH2NH(4-OCH3-Ph)
(22
(6
(16
(14
(0.1
(1.8
(7.1
(0.71
(9
acid group on the phosphorus, we were able to subse-
quently increase activity (i.e., compounds 35-38).16
Because NAALADase contains two zincs in the active
site, we next examined a series of compounds containing
an additional functional group that could interact with
the metal ions. We hoped to recover some of the potency
lost by modifying the phosphonic acid in 1 to a phos-
phinic acid. Toward this end, a series of hydroxy- and
methylaminophosphinyl derivatives (39-44) were pre-
pared. The most potent compounds in the series were
43 and 44, both aniline derivatives. The low IC50 values
(4 nM) of the aniline derivatives could be due to an
additional interaction between a zinc atom and the
amino group.17 Currently, we are using modeling tech-
niques to examine the interactions of these compounds
with both zinc atoms in the active site.
(1.8
(9
(14
(0
(2.6
a
IC50 values were determined as previously described in ref 11,
with each value representing two to four independent determina-
tions. All of the data reported are for the racemic compounds.
NAALADa se In h ibition Is P r otective in a Ra t
Mid d le Cer ebr a l Ar ter y Occlu sion Mod el
Compound 1 was initially used as a prototype com-
pound in the rat transient middle cerebral artery

Phosphinic Acid Based NAALADase Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24 4173
F igu r e 1. Effect of compound 34 in a transient MCAO model
of cerebral ischemia. As described previously, ischemia was
produced using a monofilament nylon suture inserted through
the internal carotid artery for 2 h. When administered 1 h post
ischemia, compound 34 decreased total infarct volume.
F igu r e 2. Chronic constrictive injury model of neuropathic
pain. Sciatic nerve ligation, consisting of four ligatures being
tied loosely around the sciatic nerve at 1 mm intervals
proximal to the nerve trifurcation, was performed on rats. As
shown, compound 34 significantly attenuated chronic constric-
tive injury-associated hyperalgesia.
occlusion model using the intraluminal filament tech-
nique. This compound was shown to be neuroprotective
when administered at 60 min post-occlusion (100 mg/
kg bolus followed by 20 mg/kg/h iv infusion, 30%
protection, p < 0.05). A high dose was required due to
the low brain levels obtained with this compound.12
Compound 34 also showed efficacy in this model. For
example, when this compound was dosed at 30 mg/kg
iv bolus, followed by 6 mg/kg/h iv infusion for 4 h,
compound 34 reduced mean total infarct injury by 50%
(p < 0.0001), as compared to the saline treatment group
(Figure 1). In addition, this compound was efficacious
at this dose when administered up to 2 h post occlusion,
whereas compound 1 was only effective when adminis-
tered up to 90 min post occlusion. This is remarkable
due to the fact that compound 34 is at least 200 times
less effective at inhibiting NAALADase in vitro when
compared to compound 1.
Su m m a r y
We have synthesized a series of phosphinic acid based
NAALADase inhibitors. The compounds require an
acidic moiety in the propionic acid portion of the
molecule to obtain good in vitro potency. Alternatively,
in the opposing section of the phosphinic acids, a polar
group enhances activity. The enhanced activity may be
due to an additional interaction with a zinc atom or an
amino acid residue near the catalytic site. We have
demonstrated for the first time that phosphinic acid
containing NAALADase inhibitors show efficacy in
animal models of neurodegeneration. Inhibition of this
enzyme represents a novel strategy for neuroprotection
and may have utility in a variety of disorders such as
peripheral neuropathy.
Exp er im en ta l Section
Ch em istr y. NMR spectra were recorded on a Bruker 400
instrument. Chemical shifts are reported in parts per million
relative to tetramethylsilane as internal standard. Analytical
thin-layer chromatography (TLC) was conducted on prelayered
silica gel GHLF plates (Analtech, Newark, DE) Visualization
of the plates was accomplished by using UV light, phospho-
molybdic acid-ethanol, and/or iodoplatinate charring. Flash
chromatography was conducted on Kieselgel 60, 230-400
mesh (E. Merck, Darmstadt, West Germany). Solvents are
either reagent or HPLC grade. Reactions were run at ambient
temperature and under a nitrogen atmosphere unless other-
wise noted. Solutions were evaporated under reduced pressure
on a Buchi rotary evaporator. Elemental analysis was per-
formed by Atlantic Microlabs, Norcross, GA.
Diben zyl 2-Meth ylen ep en ta n ed ioa te (9a ). Benzyl acry-
late (500 g, 3.0 mol) was heated to 100 °C under an atmosphere
of nitrogen. The heating was stopped, and HMPT (10 g, 61
mmol) was added dropwise while maintaining an internal
temperature of 135-145 °C. Once addition was complete, the
mixture was cooled to room temperature and a slurry of silica
with 5:1 hexanes/EtOAc was added. The slurry was then
transferred to a column containing a plug of dry silica. The
column was then washed with 1:1 Hex/EtOAc and the solvent
was collected and evaporated. The clear yellow liquid was
distilled under high vacuum (200 µHg) to give an initial
fraction of 8 g distilling at 45 °C and then the desired product
at 180-185 °C. This afforded 212 g (42%) of a clear and
colorless liquid.
NAALADa se In h ibition Sh ow s Effica cy in
Neu r op a th ic P a in Mod els
We have also evaluated the utility of NAALADase
inhibition for the potential treatment of neuropathic
pain. Glutamate is thought to be involved in nociception,
and several reports have suggested that glutamate
antagonists show efficacy in animal models of pain.18
More recently, NAALADase has been shown to be
present in the peripheral nervous system. Specifically,
NAALADase has been shown to be involved in Schwann
cell-signaling and appears to be downregulated as
myelination begins.19 Using the chronic constrictive
injury model of neuropathic pain, we examined the
efficacy of compound 34.20 Following sciatic nerve liga-
tion, rats were treated with compound 34 and tested
using a Hargreaves apparatus. The difference score
(between the latency of the response for the paw on the
operated side versus the control side) was determined.
Animals received compound 34 (10 mg/kg ip daily) or
vehicle starting 10 days post surgery. Treatment with
this compound dramatically normalized the difference
scores between the two paws compared to the continued
hyperalgesic vehicle-treated controls (Figure 2). This
effect was significant starting at 12 days of drug
treatment and persisted through to the end of the study
(for 23 days of daily dosing).
Di-ter t-bu tyl 2-Meth ylen ep en ta n ed ioa te (9b). tert-Butyl
acrylate (465 g, 3.6 mol) was warmed to 100 °C under nitrogen,
then HMPT (10 g, 61.2 mmol) was added dropwise, and the

4174 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24
J ackson et al.
addition rate was adjusted to maintain the reaction temper-
ature at 100 °C. The reaction mixture was allowed to cool, then
poured over a plug of silica with 4:1 hexane/ethyl acetate. The
solvent was removed under reduced pressure, and the result-
ing oil was distilled to afford 300 g of the product (65%, bp
67-70 °C at 300 µm) as a clear oil.
then filtered through a pad of Celite and concentrated under
reduced pressure. The resulting residue was dissolved in
distilled water (5 mL), passed through a column of AG 50W-
X8 resin (H+ form), and lyophilized to give 0.21 g of 2-[[(4-
methylbenzyl)hydroxyphosphinoyl]methyl]pentanedioic acid as
a white solid (55% yield).
Di-ter t-b u t yl 2-[(ter t-Bu t oxyp h osp h in yl)m et h y]p en -
ta n ed ioa te (13b). A mixture of ammonium phosphinate
(162.6 g, 1.96 mol) and hexamethyldisilazane (316 g, 1.96 mol)
was heated to 105 °C for 2 h. The reaction mixture was cooled
in an ice bath and di-tert-butyl 2-methylenepentane-1,5-dioate
(251 g, 0.979 mol) dissolved in dichloromethane (1000 mL) was
added dropwise. The reaction mixture was allowed to warm
to room temperature overnight. The reaction mixture was then
quenched with distilled water (500 mL), the layers were
separated, and the aqueous layer was washed with dichlo-
romethane. The combined organic layers were dried over
MgSO4, and the solvent was removed under reduced pressure.
This afforded a slightly yellow oil (315 g), which was used
without further purification.
To a solution of di-tert-butyl 2-[(hydroxyphosphinyl)methyl]-
pentane-1,5-dioate (315 g, 0.98 mol) in dichloromethane (1000
mL) cooled in an ice bath and under nitrogen were added tert-
butyl alcohol (123 g, 1.7 mol), 4-(dimethylamino)pyridine (1
g, 8.2 mmol), and 1-ethyl-3-(3-dimethylaminopropyl)carbodi-
imide (281 g, 1.47 mol). The reaction was allowed to stir
overnight. Water was added to the reaction mixture, the
dichloromethane layer was retained and dried, and the solvent
was removed under reduced pressure. The resulting residue
was purified by column chromatography, and the desired
product was eluted with 1:1 to 2:3 hexane/ethyl acetate. This
afforded 260 g of the desired product as a clear oil (70%).
P r ep a r a tion of 2-[[(4-Meth ylben zyl)h yd r oxyp h osp h i-
n oyl]m eth yl]p en ta n ed ioic Acid (30). Hexamethyldisilazane
(21.1 mL, 100 mmol) was added to vigorously stirred am-
monium phosphate (8.30 g, 100 mmol), and the resulting
suspension was stirred at 105 °C for 2 h. A solution of
4-methylbenzyl bromide (5.00 g, 27.0 mmol) was then added
dropwise to the suspension at 0 °C. The mixture was stirred
at room temperature for 19 h. The reaction mixture was then
diluted with dichloromethane (50 mL) and washed with 1 N
HCl (50 mL). The organic layer was separated, dried over Na2-
SO4, and concentrated to give 4.72 g of a white solid. This was
dissolved in dichloromethane (50 mL), and benzyl alcohol (3.24
g, 30 mmol) was added to the solution. 1,3-Dicyclohexylcar-
bodiimide (DCC) (6.19 g, 30 mmol) was then added to the
solution at 0° C, and the suspension was stirred at room
temperature for 14 h. The solvent was removed under reduced
pressure, and the residue was suspended in EtOAc. The
resulting suspension was filtered, and the filtrate was con-
centrated. The residue was purified by silica gel chromatog-
raphy (hexanes:EtOAc, 4:1 to 1:1) to give 2.40 g of 4-methyl-
benzyl-o-benzylphosphinic acid as a white solid (34% yield):
Rf 0.42 (EtOAc); 1H NMR (DMSO-d6) 2.30 (s, 3H), 3.29 (d, J )
16.6 Hz, 2H), 5.2 (m, 2H), 7.0 (d, J ) 543 Hz, 1H), 7.1-7.2 (m,
4H), 7.3-7.4 (m, 5H).
To a solution of 4-methylbenzyl-o-benzylphosphinic acid
(2.16 g, 8.3 mmol) in THF (15 mL) was added sodium hydride
(0.10 g, 60% dispersion in oil) followed by dibenzyl 2-methyl-
enepentanedioate at 0° C, and the mixture was stirred at room
temperature for 4 h. The reaction mixture was then diluted
with EtOAc (50 mL) and poured into 1 N HCl (50 mL). The
organic layer was separated, dried over Na2SO4, and concen-
trated. This material was purified by silica gel chromatography
(hexanes:EtOAc, 4:1 to 1:1) to give 3.41 g of 2,4-di-(benzyloxy-
carbonyl)butyl(4-methylbenzyl)-o-benzylphosphinic acid as a
colorless oil (70% yield): Rf 0.56 (EtOAc); 1H NMR (CDCl3)
1.6-1.8 (m, 1H), 1.9-2.0 (m, 2H), 2.1-2.4 (m, 6H), 2.7-2.9
(m, 1H), 3.05 (dd, J ) 9.0, 16.8 Hz, 2H), 4.8-5.1 (m, 6 H),
7.0-7.1 (m, 4H), 7.2-7.4 (m, 15 H).
2-[[[2-(Ben zylca r b oxy)p r op yl]h yd r oxyp h osp h in oyl]-
m eth yl]p en ta n ed ioic Acid (37). To a solution of di-tert-butyl
2-[(tert-butoxyphosphinyl)methyl]pentanedioate (13.6 g, 36
mmol) and benzyl methacrylate (6.35 g, 36 mmol) in THF (100
mL) under nitrogen was added sodium hydride (0.14 g, 60 5
dispersion in oil, 3.60 mmol). After 3 h, the reaction mixture
was poured into water (300 mL), and ether (100 mL) was
added. The organic layer was separated, and the combined
organic extracts were dried over MgSO4 and the solvent was
removed under reduced pressure. The resulting residue was
purified by column chromatography, and the product was
eluted with 2:3 EtOAc/hexane. This afforded 10.5 g (53%) as
a clear oil. 1H NMR (CDCl3) 1.3 (m, 3 H), 1.5 (m, 27H), 1.7 (m,
2H), 1.9 (m, 2H), 7.3 (m, 5H). To a solution of di-tert-butyl
2-[[[2-(benzylcarboxy)propyl]tert-butoxyphosphinyl]methyl]-
pentanedioate (1.6 g, 2.9 mmol) in dichloromethane (10 mL)
under nitrogen was added trifluoroacetic acid (10 mL). The
reaction mixture was stirred for 2 h and then concentrated
under reduced pressure. Additional DCM was added to the
reaction residue, and the solvent was removed under reduced
pressure. The product was dissolved in EtOAc, washed with
water, dried over MgSO4, and the solvent was removed under
reduced pressure leaving 800 mg of the desired product as a
clear oil (72%).
2-[[(Ca r b oxy)p r op yl]h yd r oxyp h osp h in oyl]m e t h yl]-
p en ta n ed ioic Acid (38). A solution of 2-[[[2-(benzylcarboxy)-
propyl]hydroxyphosphinyl]methyl]pentanedioic acid (8.9 g,
16.1 mmol), palladium on carbon catalyst (10%, 1.0 g), and
ethyl acetate (100 mL) was shaken under hydrogen (60 psi)
for 16 h. The reaction mixture was filtered over Celite, and
the filtrate was concentrated under reduced pressure leaving
7.5 g of a clear oil. To a solution of di-tert-butyl 2-[[[2-(carboxy)-
propyl]tert-butoxyphosphinyl]methyl]pentanedioate (2.1 g, 4.53
mmol) in DCM (10 mL) under nitrogen was added TFA (10
mL). The reaction mixture was stirred for 2 h and then
concentrated under reduced pressure. Additional DCM was
added and removed under reduced pressure. The resulting
residue was triturated with acetonitrile, then dried under
reduced pressure leaving the product as a thick clear oil (1.2
g, 89%).
P r ep a r a tion of 2-[(Meth ylh yd r oxyp h op sh in oyl)m eth -
yl]p en ta n ed ioic Acid (39). Dry phosphinic acid (100 g, 1.52
mol) was dissolved in 100 mL of CHCl3 and treated with
triethylamine (155 g, 1.52 mol). The mixture was evaporated
and transferred to a 3 L flask, containing 750 mL of CHCl3.
The solution was stirred by means of a mechanical stirrer and
the flask cooled to 0 °C. The clear solution was treated with
triethylamine (277 g, 2.72 mol) followed by trimethylsilyl
chloride (281 g, 2.58 mol). Once addition was complete,
dibenzyl 2-methylenepentanedioate in 150 mL of CHCl3 was
added, and the mixture warmed to room temperature. After 6
h the thick slurry was filtered and the filtrate cooled to 0 °C.
The filtrate was then quenched with 5% HCl and the organic
layer removed. The aqueous layer was extracted with CHCl3,
the combined organic layers were dried, and the solvent was
removed under reduced pressure to afford 55 g of 2,4-di-
(benzyloxycarbonyl)butylphosphinic acid as a light yellow
liquid. This was purified by flash chromatography and eluted
using 3:1 hexanes/EtOAc containing 5% TFA to give 40 g of
2,4-di(benzyloxycarbonyl)butylphosphinic acid. 1H NMR (CDCl3)
2.0 (d, 3H), 2.2 (m, 1H), 2.4 (t, 2H), 2.9 (m, 1H), 5.1 (s, 2H),
7.2 (d, 1H), 7.3 (m, 10H).
To a solution of 2,4-di(benzyloxycarbonyl)butylphosphinic
acid (19.3 g, 49,4 mmol) in THF was added benzyl alcohol (5.3
g, 49 mmol) and DMAP (0.5 g). Dicyclohexylcarbodiimide (12
g, 58 mmol) was added, and a white precipitate formed. After
30 min, the white suspension was filtered and the filtrate
evaporated under reduced pressure. The clear and colorless
To a solution of 2,4-di(benzyloxycarbonyl)butyl(4-methyl-
benzyl)-o-benzylphosphinic acid (0.70 g, 1.2 mmol) in ethanol
(30 mL) was added Pd/C (5%, 0.10 g), and the suspension was
shaken under hydrogen (50 psi) for 18 h. The suspension was

Phosphinic Acid Based NAALADase Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24 4175
(3) Robinson M. B.; Blakely R. D.; Couto R.; Coyle J . T. Hydrolysis
of the brain dipeptide N-acetyl-L-aspartyl-L-glutamate. Identi-
fication and characterization of a novel N-acetylated alpha-
linked acidic dipeptidase activity from rat brain. J . Biol. Chem.
1987, 262, 14498-14506.
(4) Stauch, B. L.; Robinson, M. B.; Forloni, G.; Tsai, G.; Coyle, J . T.
The effects of N-acetylated alpha-linked acidic dipeptidase
(NAALADase) inhibitors on [3H]NAAG catabolism in vivo.
Neurosci. Lett. 1989, 100, 295-300.
(5) (a) Carter, R. E.; Feldman, A. R.; Coyle, J . T. Prostate-specific
membrane antigen is a hydrolase with substrate and pharma-
cological characteristics of a neuropeptidase. Proc. Natl. Acad.
Sci. 1996, 93, 749-753. (b) Luthi-Carter, R.; Barczak, A. K.;
Speno, H.; Coyle, J . T. Molecular characterization of human
brain N-acetylated alpha-linked acidic dipeptidase (NAALA-
Dase). J . Pharmacol. Exp. Ther. 1998, 286, 1020-1025. (c)
Devlin, A. M.; Ling, E. H.; Peerson, J . M.; Fernando, S.; Clarke,
R.; Smith, A. D.; Halsted, C. H. Glutamate carboxypeptidase II:
a polymorphism associated with lower levels of serum folate and
hyperhomocysteinemia. Hum. Mol. Genet. 2000, 22, 2837-2844.
(6) Due to the similarity between NAALADase and PSMA, the
enzyme is also referred to as Glutamate Carboxypeptidase II.
See: Barrett, A. J .; Rawlings, N. D.; Woessner, J . F. Glutamate
Carboxypeptidase II. In Handbook of Proteolytic Enzymes;
Academic Press: New York, 1998; pp 1434-1437.
(7) Rawlings, N. D.; Barrett, A. J . Structure of membrane glutamate
carboxypeptidase. Biochim. Biophys. Acta 1997, 1339, 247-252.
(8) Speno, H. S.; Luthi-Carter, R.; Macias, W. L.; Valentine, S. L.;
J oshi, A. R.; Coyle, J . T. Site-Directed mutagenesis of predicted
Active Site Residues in Glutamate Carboxypeptidase II. Mol.
Pharmacol. 1999, 55, 179-185.
(9) For a review, see: Fagg, G. E.; Foster, A. C. Excitatory amino
acid synaptic mechanisms and neurological function. Trends
Pharmacol. Sci. 1986, 7, 357-363.
(10) Wroblewski, B.; Wroblewski, J . T.; Pshenichkin, S.; Surin, A.;
Sullivan, S. E.; Neale, J . H. N-Acetylaspartylglutamate selec-
tively activates mGluR3 receptors in transfected cells. J . Neu-
rochem. 1997, 69, 1116-1119.
(11) J ackson, P. F.; Cole, D. C.; Slusher, B. S.; Stetz, S. L.; Ross, L.
E.; Donzanti, B. A.; Trainor, D. A. Design, synthesis, and
biological activity of a potent inhibitor of the neuropeptidase
N-acetylated-R-linked acidic dipeptidase. J . Med. Chem. 1996,
39, 619-622.
(12) Slusher, B. S.; Vornov, J . J .; Thomas, A. G.; Hurn, P. D.;
Harukuni, I.; Bhardwaj, A.; Traystman, R. J .; Robinson, M. B.;
Britton, P.; May Lu, X. C.; Tortella, F. C.; Wozniak, K. M.;
Yudkoff, M.; Potter, B. M.; J ackson, P. F. Selective Inhibition of
NAALADase, Which Converts NAAG to Glutamate, Reduces
Ischemic Brain Injury. Nat. Med. 1999, 5, 1396-1402.
(13) For a general description, see: J ackson, P. F.; Slusher, B. S.
Design of NAALADase Inhibitors: A Novel Neuroprotective
Strategy. Curr. Med. Chem. 2001, 8, 949-957.
(14) Thottathil, J . K.; Ryono, D. E.; Przybyla, C. A.; Moniot, J . L.;
Neubeck, R. Preparation of Phosphinic Acids: Michael Addition
of Phosphonous Acids/Esters to Conjugated Systems. Tetrahe-
dron Lett. 1984, 25, 4741-4744.
(15) Boyd, E. A.; Regan, A. C. Synthesis of γ-keto substituted
phosphinic acids from bis(trimethylsilyl)phopshonite and R,â-
unsaturated ketones Tetrahedron Lett. 1992, 31, 2933-2936.
(16) Recently, compound 35 has been reported to have weak mGluR3
activity. See: Nan, F.; Bzdega, T.; Pshenichkin, S.; Wroblewski,
J . T.; Wroblewska, B.; Neale, J . H.; Kozikowski, A. P. J . Med.
Chem. 2000, 43, 772-774.
(17) There is evidence for this type of interaction in the literature.
In the crystal structure of the dizinc aminopeptidase from
Aeromonas proteolytica (AAP) complexed with the inhibitor
L-leucinephosphonic acid (LPA), LPA binds in the active site as
an η-1,2-µ-phosphonate with the zinc atoms. In addition, the
N-terminal amine of LPA binds one of the two zincs. Stamper,
C.; Bennett, B.; Edwards, T.; Holz, R. C.; Ringe, D.; Petsko, G.
Inhibition of the Aminopeptidase from Aeromonas proteolytica
by L-Leucinephosphonic acid. Spectroscopic and Crystallographic
Characterization of the Transition State of Peptide Hydrolysis.
Biochemistry 2001, 40, 7035-7046.
oil was purified by flash chromatography and eluted with 1:1
hexanes/EtOAc to give 2,4-di(benzyloxycarbonyl)butylben-
zylphosphinic acid (11.5 g) as a clear and colorless oil. 1H NMR
(CDCl3) 1.9 (m, 3H), 2.2 (m, 3H), 2.9 (t, 1H), 5.0 (m, 6H), 7.2
(d, 1H), 7.3 (m, 15H).
2,4-Di-(benzyloxycarbonyl)butylbenzylphosphinic acid in 5
mL of THF was added dropwise to a stirring, cooled mixture
of sodium hydride (0.09 g, 2.3 mmol) in 15 mL of THF. After
15 min benzaldehyde (0.23 g, 2.2 mmol) was added via syringe
while maintaining a temperature of 0 °C. After 30 min, the
reaction was quenched with water and extracted with DCM.
The solvent was removed under reduced pressure to give a
clear colorless oil which was purified using flash chromatog-
raphy using 1:1 hexane/EtOAc. This afforded 0.4 g (33%) of
2,4-di(ben zyloxyca r bon yl)bu t yl[h ydr oxy(ph en yl)m et h yl]-
benzylphosphinic acid as a clear and colorless oil. 1H NMR
(CDCl3) 1.9 (m, 3H), 2.2 (m, 3H), 2.8 (dm, 1H), 4.9 (m, 6H),
5.2 (m, 1H), 7.3 (m, 20H).
2,4-Di(benzyloxycarbonyl)butyl[hydroxy(phenyl)methyl]-
benzylphosphinic acid (0.37 g, 0.6 mmol) in 25 mL of water
containing 0.10 g of 10% Pd/C was hydrogenated at 40 psi for
6 h. The mixture was filtered through a pad of Celite and
lyophilized to give 0.14 g (70%) of 2-[(methylhydroxyphosphi-
nyl)methyl]pentanedioic acid as a white solid.
Syn th esis of 2-[[[Ben zyla m in o]m eth yl](h yd oxyp h os-
p h in oyl)m eth yl]p en ta n ed ioic Acid (42). A solution of 1,3,5-
tribenzylhexahydro-1,3,5-triazine (14.30 g, 40.0 mmol) and di-
tert-butyl 2-[[[tert-butoxy]phosphinyl]methyl]pentane-1,5-dioate
(37.85 g, 100 mmol) in toluene (200 mL) was stirred at 110 °C
for 14 h. The solvent was removed under reduced pressure,
and the residual yellow oil was purified by silica gel chroma-
tography (hexanes/ethyl acetate 2:1) to give 23.40 g of di-tert-
butyl 2-[[[tert-butoxy](benzyl amino)methyl]phosphinoyl]methyl]-
pentane-1,5-dioate as a light yellow oil (43%). 1H NMR (CDCl3)
1.40-1.48 (m, 27H), 1.7-2.1 (m, 4H), 2.2-2.4 (m, 3H), 2.6-
3.0 (m, 3H) 3.8-4.0 (m, 2H), 7.2-7.4 (m, 5H).
To a solution of 2-[[[tert-butoxy](benzyl amino)methyl]-
phosphoryl]methyl]pentane-1,5-dioate (0.498 g, 1.0 mmol) in
DCM (10 mL) was added TFA (5 mL) at O °C, and the mixture
was stirred at room temperature for 18 h. The solvent was
removed under reduced pressure, the resulting taken up in
DCM (10 mL), and the solvent was removed under reduced
pressure. This process was repeated three times to remove the
excess TFA. The resulting oil was crystallized from methanol
to give 0.174 g of 2-[[[benzylamino]methyl](hydroxyphos-
phinoyl)methyl]pentanedioic acid as a white solid.
In Vivo Effica cy Stu d ies. The rat middle cerebral artery
occlusion model was performed as previously described.12 The
chronic constrictive injury model of neuropathic pain was
performed following the procedure of Bennett.20 In brief, sciatic
nerve ligation, consisting of four ligatures being tied loosely
around the sciatic nerve at 1 mm intervals proximal to the
nerve trifurcation, was performed on rats. Following this
treatment, rats exhibit a thermal hyperalgesia and allodynia.
Animals were habituated to the Hargreaves apparatus, and
the infrared heat source was directed onto the dorsal surface
of the hindpaw and the time taken for the animal to withdraw
its paw noted. The difference score (between the latency of the
response for the paw on the operated side versus the control
side) was determined. Animals received the test compound (10
mg/kg i.p. daily) or vehicle, starting 10 days post surgery.
Treatment with compound 34 normalized the difference scores
between the two paws compared to the continued hyperalgesic
vehicle-treated controls.
(18) Sang, C. N. NMDA-receptor antagonists in neuropathic pain:
experimental methods to clinical trials. J . Pain Symptom
Manage. 2000, 19, S21-5.
(19) Berger, U. V.; Schwab, M. E. N-acetylated alpha-linked acidic
dipeptidase may be involved in axon-Schwann cell signalling.
J . Neurocytol. 1996, 499-512.
(20) Bennett, G. J .; Xie, Y. K. A peripheral mononeuropathy in rat
that produces disorders of pain sensation like those seen in man.
Pain 1998, 33, 87-107.
Refer en ces
(1) Curatolo, A.; D’Archangelo, P.; Lino, A.; Brancati, A. Distribution
of N-acetyl-aspartic acid and N-acetyl-aspartyl-glutamic acids
in nervous tissue. J . Neurochem. 1965, 12, 339-342.
(2) a) Coyle, J . T. The Nagging Question of the Function of
N-Acetylaspartylglutamate. Neurobiol. Dis. 1997, 4, 231-238.
(b) Neale, J . H.; Bzdega, T.; Wroblewska, B. N-acetylyapartyl-
glutamate: the Most Abundant Peptide Neurotransmitter in the
Mammalian Central Nervous System. J . Neurochem. 2000, 75,
443-452.
J M0001774