Synthesis and biological evaluation of hydroxamate-based inhibitors of glutamate carboxypeptidase II

Article abstract of DOI:10.1016/S0960-894X(03)00407-4

A series of hydroxamic acids has been prepared as potential inhibitors of glutamate carboxypeptidase II (GCP II). Compounds based on a P1′ residue (primed-side inhibitors) were more potent than those based on a P1 group (unprimed-side inhibitors). Inhibitory potency of the primed-side GCP II inhibitors was found to be dependent on the number of methylene units between the hydroxamate group and pentanedioic acid. Succinyl hydroxamic acid derivative, 2-(hydroxycarbamoylmethyl)pentanedioic acid, is the most potent GCP II inhibitor with an IC₅₀ value of 220 nM. The comparison of the results to those of other classes of GCP II inhibitors as well as hydroxamate-based MMP inhibitors provides further insight into the structure-activity relationships of GCP II inhibition.

Full text of DOI:10.1016/S0960-894X(03)00407-4

Bioorganic & Medicinal Chemistry Letters 13 (2003) 2097–2100
Synthesis and Biological Evaluation of Hydroxamate-Based
Inhibitors of Glutamate Carboxypeptidase II
Doris Stoermer, Qun Liu, Monicia R. Hall, Juliet M. Flanary, Ajit G. Thomas,
Camilo Rojas, Barbara S. Slusher and Takashi Tsukamoto*
Guilford Pharmaceuticals Inc., 6611 Tributary Street, Baltimore, MD 21224, USA
Received 25 February 2003; accepted 15 April 2003
Abstract—A series of hydroxamic acids has been prepared as potential inhibitors of glutamate carboxypeptidase II (GCP II).
Compounds based on a P10 residue (primed-side inhibitors) were more potent than those based on a P1 group (unprimed-side
inhibitors). Inhibitory potency of the primed-side GCP II inhibitors was found to be dependent on the number of methylene units
between the hydroxamate group and pentanedioic acid. Succinyl hydroxamic acid derivative, 2-(hydroxy-
carbamoylmethyl)pentanedioic acid, is the most potent GCP II inhibitor with an IC50 value of 220 nM. The comparison of the
results to those of other classes of GCP II inhibitors as well as hydroxamate-based MMP inhibitors provides further insight into the
structure–activity relationships of GCP II inhibition.
# 2003 Elsevier Science Ltd. All rights reserved.
Glutamate carboxypeptidase II (GCP II, EC 3.4.17.21),
also known as N-acetylated a-linked acidic dipeptidase
(NAALADase) and prostate-specific membrane antigen
(PSMA), is a metallopeptidase which cleaves N-acetyl-
aspartylglutamate (NAAG) into N-acetylaspartate and
glutamate in the nervous system.1 This enzyme belongs
to the M28 peptidase family comprised of co-catalytic
metallopeptidases.2 Peptidases in this family contain
two metal ions forming a co-catalytic active site with
ligands consisting of five amino acid residues. Inhibition
of GCP II has gained considerable attention as a strat-
egy to suppress glutamate excitotoxicity leading to
neurological disorders including stroke, spinal cord
injury, amyotrophic lateral sclerosis (ALS), peripheral
neuropathy, chronic pain, schizophrenia, and epilepsy.3
mechanism and physiological role of GCP II as well as
the potential therapeutic effects of GCP II inhibition.
For example, we have demonstrated that 2-PMPA con-
siderably reduces the ischemia-induced rise in extra-
cellular glutamate and vigorously protects against injury
in a neuronal culture model of ischemia and in rats after
transient middle cerebral artery occlusion (MCAO).6
We have extended our SAR studies to other zinc-bind-
ing groups and identified phosphinate and thiol-based
GCP II inhibitors 1b7 and 2,8 which exhibited in vivo
efficacy in animal models of various neurological dis-
orders.
2-(Phosphonomethyl)pentanedioic acid (2-PMPA) 1a4
is one of the first potent inhibitors of GCP II with a Ki
value of 0.2 nM.5 The high potency of compound 1a can
be attributed to the strong chelation of the phosphonate
group to an active site zinc atom as well as the interac-
tion of the glutarate (pentanedioic acid) portion of the
inhibitor with the glutamate recognition site of GCP II.
2-PMPA has been extensively utilized to study the
In order to further diversify the pharmacophore of GCP
II inhibitors, we have examined hydroxamate group as
an alternative zinc-binding group. The hydroxamate
group of metalloprotease inhibitors is known to bind in
a bidentate fashion to the active site zinc(II) ion and has
proven to be one of the most effective zinc-binding
*Corresponding author. Fax: +1-410-631-6797; e-mail: tsukamotot@
guilfordpharm.com
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00407-4

2098
D. Stoermer et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2097–2100
groups particularly in the area of matrix metallopro-
tease (MMP) inhibitors.9 As shown in Figure 1, our
design of hydroxamate-based GCP II inhibitors is
based on an endogenous substrate NAAG that can
be divided into the three segments, 1-acetylamino-2-
carboxyethyl group, scissile amide bond, and glutaric
acid. Incorporation of a hydroxamic acid group into
the first segment leads to a P1 group-containing
inhibitor 3, also conventionally referred to as an
unprimed-side inhibitor. Attachment of a hydroxamic
acid group to the glutaric acid segment through a
methylene linker results in derivatives 4a–d, which con-
tain a P10 group and thus are referred to as primed-side
inhibitors. In this paper, we describe synthesis and bio-
logical evaluation of these hydroxamate-based inhibi-
tors of GCP II.
group of compound 10b was selectively cleaved by
the treatment with lithium hydroxide in THF/H2O.
Subsequent decarboxylation and coupling to benzyl
alcohol afforded 2-allylglutaric acid dibenzyl ester
11b. Oxidation of the terminal olefin into a car-
boxylic acid, followed by coupling to benzyloxyamine
gave the fully protected precursor 12b. The benzyl
groups were removed by catalytic hydrogenolysis to
yield the desired compound 4b. The same method was
successfully applied to the synthesis of two other ana-
logues 4c (n=2) and 4d (n=3) by replacing allyl bro-
mide with 4-bromo-1-butene and 5-bromo-1-pentene,
respectively.
Inhibitory potencies of hydroxamic acids 3, 4a–d and
other structurally relevant compounds (S)- and (R)-
13,14 (S)-14,14 15,15 and 1614 were evaluated using N-
acetyl-l-aspartyl-[3H]-l-glutamate as a substrate5 and a
purified recombinant GCP II.16 For those compounds
with an IC50 value less than 1 mM, a median effective
concentration (EC50) value in a cell culture model of
cerebral ischemia17 was also determined. The results are
summarized in Table 1 along with previously reported
representative GCP II inhibitors 1a–b and 2 for com-
parison purposes.
Synthesis of (S)-3 is illustrated in Scheme 1. N-Acetyl-l-
aspartic acid b-benzyl ester (S)-5,10 was coupled to
O-benzylhydroxylamine to yield (S)-6. Subsequent cat-
alytic hydrogenolysis afforded N-acetyl-l-aspartic acid
a-hydroxamate (S)-3. Ready access to (R)-511 also
allowed us to prepare the corresponding d-isomer (R)-3
in a similar manner.
Synthesis of 4a–d is outlined in Scheme 2. Conjugate
addition of Meldrum’s acid 7 to benzyl acrylate gave
5-substituted Meldrum’s acid 8.12 Alcoholysis of 8 with
benzyl alcohol13 followed by coupling to O-benzyl-
hydroxylamine afforded compound 9. Removal of all
benzyl groups from 9 by catalytic hydrogenolysis gave
the malonyl hydroxamic acid derivative 4a. The succinyl
hydroxamic acid derivative 4b was also synthesized
from Meldrum’s acid. One-pot tandem alkylation of
Meldrum’s acid with benzyl acrylate and allyl bromide
gave disubstituted Meldrum’s acid 10b. The acetonide
The rationale for testing both enantiomers of 3 is based
on the previously reported studies18 on amino acid
hydroxamate inhibitors of another metalloprotease
from family M28, Aeromonas proteolytica aminopepti-
dase. Wilkes and Prescott found that this aminopepti-
dase was inhibited to a greater extent by d-leucine
hydroxamic acid (Ki=2.0 nM) than by l-leucine
hydroxamic acid (Ki=350 nM) despite its substrate
specificity for l-amino acid residue in the N-terminal
position. In the GCP II assay, neither (S)-3 nor (R)-3
exhibited inhibitory activity in concentration up to
100 mM. Similarly, no significant activity was observed
with other commercially available amino acid hydrox-
amates (S)-13, (R)-13, and (S)-14, which potentially
interact with S1 site of the enzyme.
The lack of GCP II inhibition by N-acetyl-l-aspartic
acid a-hydroxamate 3 is in sharp contrast with the
potent inhibition of A. proteolytica aminopeptidase by
leucine hydroxamate.19 The marked difference in inhib-
itory profile between the two metalloproteases can be
explained by the substrate specificity of each enzyme. A.
proteolytica aminopeptidase is specific for peptides pos-
sessing leucine as an NH2-terminal residue20 and thus
potently inhibited by a P1 residue-containing inhibitors
such as leucine hydroxamic acid. On the other hand,
GCP II is known to hydrolyze not only NAAG but also
g-glutamyl linkage of folate and antifolate poly-g-glu-
tamates21 in which case its S1 site is occupied by g-glu-
tamyl residue. The lower substrate specificity with
respect to the P1 residue is presumably attributable to
the lack of GCP II inhibition by the unprimed-side
inhibitors.
Figure 1. Design of hydroxamate-based GCP II Inhibitors.
In contrast to the unprimed-side inhibitors, all of the
primed-side inhibitors 4a–d showed IC50 values below
10 mM. The results are in a good agreement with the
.
Scheme 1. (a) BnONH2 HCl, EDC, DMAP, HOBt, Et3N, CH2Cl2, rt,
51%; (b) H2 (30 psi), Pd/C, MeOH, rt, 99%.

D. Stoermer et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2097–2100
2099
À
ꢂ
ꢂ
Scheme 2. (a) Benzyl acrylate, K2CO3, BnEt3N+Cl , CH3CN, 60 C, 98%; (b) (i) BnOH, toluene, 90 C; (ii) BnONH2 HCl, EDC, DMAP, Et3N,
CH2Cl2, rt; 40%; (c) H2 (30 psi), Pd(OH)2/C, MeOH, rt, 90%; (d) (i) benzyl acrylate, K2CO3, BnEt3N+ClÀ, CH3CN, 60 ꢂC; (ii) CH2¼CH(CH2)nBr,
50 ꢂC; 72% for 10b, 41% for 10c, 54% for 10d; (e) (i) LiOH, THF/H2O, rt; (ii) neat, 130 ꢂC; (iii) BnOH, EDC, DMAP, CH2Cl2, rt; 71% for 11b,
.
.
82% for 11c, 39% for 11d; (f) (i) RuO2, NaIO4, acetone/H2O; (ii) BnONH2 HCl, EDC, DMAP, Et3N, CH2Cl2, rt; 78% for 12b, 68% for 12c, 45%
for 12d; (g) H2 (30 psi), Pd(OH)2/C, MeOH, rt, 98% for 4b, 98% for 4c, 73% for 4d.
substrate specificity and inhibitory profile of the enzyme
previously reported. GCP II only hydrolyzes peptides
with glutamate as the COOH-terminal residue.1 It sug-
gests that the predominant binding determinants of
peptide substrates for GCP II reside on the primed side
of the scissile bond, and that the primed-side inhibitors
based on a 2-substituted glutaric acid should exhibit
more potency compared to the corresponding
unprimed-side inhibitors. Indeed, recent SAR studies
on GCP II inhibitors conducted by us and other
groups demonstrated that a 2-substituted glutaric acid is
an essential component for the potent inhibition of
GCP II regardless of the choice of zinc-binding
group.4,7,8,22À25
group of 4b as a zinc-binding group in interacting with
GCP II.
The compound 4b also provided dose-dependent pro-
tection in an established tissue culture model of cerebral
ischemia with an EC50 value of 220 nM while the less
potent GCP II inhibitor 4d exhibited only weak neuro-
protection. Despite their structural diversity, there is a
good correlation for relative potency (IC50 value) of
GCP II inhibitors (1a–b, 2, 4b, and 4d) and their neu-
roprotective potency (EC50 value). This provides further
evidence to support the hypothesis that the neuropro-
tective effect of these compounds is due to their ability
to inhibit GCP II activity.
Inhibitory potency of 4a–d against GCP II was found
to be dependent on the number of methylene units
between the hydroxamate group and pentanedioic acid.
Succinyl hydroxamic acid derivative 4b was found to
be the most potent inhibitor of GCP II with an IC50
value of 220 nM. Increasing or decreasing the number
of methylene units between the hydroxamate group
and pentanedioic acid led to a decrease in inhibitory
potency as observed in compounds 4a and 4c–d. The
results are consistent with early SAR analysis by John-
son’s group who demonstrated that the succinyl
hydroxamate is the optimal backbone for the inhibition
of MMP-1.26
Comparison of different zinc-binding groups within 2-
substituted glutarate derivatives arrived at the following
preference in terms of GCP II inhibition: phosphonate
(1a) ꢀ phosphinate (1b)ꢁsulfhydryl (2)>hydroxamate
(4b) ꢀ carboxylate (16). Interestingly, this preference
differs from that of MMPs, which are most effectively
inhibited by hydroxamates but only weakly by the cor-
responding phosphonates.9 While MMPs require one
zinc ion for catalysis, GCP II utilizes two zinc ions that
act cocatalytically. Therefore, the catalytic structure of
MMPs is notably different from that of GCP II, which
may explain the difference in the preference for zinc-
binding group.
A truncated analogue of 4b, N-hydroxy-succinamic
acid 15, was over 100-fold less potent than 4b in
inhibiting GCP II. The notable decline in potency upon
removal of a propionate portion from 4b suggests cri-
tical contribution of the P10 side chain of 4b to its
strong affinity to GCP II. A similar trend was observed
during our SAR studies on thiol and phosphonate-
based GCP II inhibitors.8,27 The poor inhibition by
tricarboxylic acid derivative 16 can be attributed to its
inability to effectively interact with an active site zinc
ion and suggests a critical role for the hydroxamate
Crystallographic studies on a hydroxamate-based
inhibitor bound to MMP revealed that the hydroxamate
group acts as a bidentate ligand to the active site zinc
ion with its two oxygen atoms.28 The hydroxamate
group provides additional interactions with the enzyme
via the hydroxy group and nitrogen atom which form a
hydrogen bond with the proximate Glu and Ala resi-
dues of the enzyme backbone, respectively. The rela-
tively weak GCP II inhibition by the hydroxamate-
based inhibitor 4b may be attributed to its inability to
guide the hydroxamate group in such way to achieve all

2100
D. Stoermer et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2097–2100
Table 1. Inhibition of GCP II by 1–4 and 13–16
References and Notes
Compd
Structure
IC50 (nM)a
EC50 (nM)a
1. Carter, R. E.; Coyle, J. T. In Handbook of Proteolytic
Enzymes; Barrett, A. J., Rawlings, N. D., Woessner, J. F.,
Eds.; Academic: New York, 1998, p 1434.
0.30ꢃ0.05b
0.62b
2. Rawlings, N. D.; Barrett, A. J. Biochim. Biophys. Acta
1997, 1339, 247.
1a
3. Whelan, J. Drug Discov. Today 2000, 5, 171.
4. Jackson, P. F.; Cole, D. C.; Slusher, B. S.; Stetz, S. L.;
Ross, L. E.; Donzanti, B. A.; Trainor, D. A. J. Med. Chem.
1996, 39, 619.
5. Rojas, C.; Frazier, S. T.; Flanary, J.; Slusher, B. S. Anal.
Biochem. 2002, 310, 50.
6. Slusher, B. S.; Vornov, J. J.; Thomas, A. G.; Hurn, P. D.;
Harukuni, I.; Bhardwaj, A.; Traystman, R. J.; Robinson, M. B.;
Britton, P.; Lu, X. C.; Tortella, F. C.; Wozniak, K. M.; Yudkoff,
M.; Potter, B. M.; Jackson, P. F. Nat. Med. 1999, 5, 1396.
7. Jackson, P. F.; Tays, K. L.; Maclin, K. M.; Ko, Y. S.; Li,
W.; Vitharana, D.; Tsukamoto, T.; Stoermer, D.; Lu, X. C.;
Wozniak, K.; Slusher, B. S. J. Med. Chem. 2001, 44, 4170.
8. Majer, P.; Jackson, P. F.; Delahanty, G.; Grella, B. S.; Ko,
Y.-S.; Li, W.; Liu, Q.; Maclin, K. M.; Polakova, J.; Shaffer, K.
A.; Stoermer, D.; Vitharana, D.; Wang, E. Y.; Zakrzewski, A.;
Slusher, B. S.; Wozniak, K. M.; Burak, E.; Limsakun, T.;
Tsukamoto, T. J. Med. Chem. 2003, 46, 1989.
1b
2
82ꢃ14b
90ꢃ26b
9.5ꢃ2.1
13ꢃ6
(S)-3
(R)-3
>100,000
>100,000
—
—
9. Whittaker, M.; Floyd, C. D.; Brown, P.; Gearing, A. J. H.
Chem. Rev. 1999, 99, 2735.
10. Cox, R. J.; Jenkins, H.; Schouten, J. A.; Stentiford, R. A.;
Wareing, K. J. J. Chem. Soc., Perkin Trans. 1 2000, 2023.
11. Liotta, L. J.; Gibbs, R. A.; Taylor, S. D.; Benkovic, P. A.;
Benkovic, S. J. J. Am. Chem. Soc. 1995, 117, 4729.
12. Hin, B.; Majer, P.; Tsukamoto, T. J. Org. Chem. 2002, 67,
7365.
4a (n=0)
4b (n=1)
4c (n=2)
4d (n=3)
4000ꢃ1400
220ꢃ40
—
220c
—
1100c
1200ꢃ400
940ꢃ80
13. Chorev, M.; Rubini, E.; Gilon, C.; Wormser, U.; Selinger,
Z. J. Med. Chem. 1983, 26, 129.
14. Compounds (S)- and (R)-13, (S)-14, and 16 were pur-
chased from Sigma-Aldrich.
15. Ames, D. E.; Grey, T. F. J. Chem. Soc. 1955, 631.
16. Barinka, C.; Rinnova, M.; Sacha, P.; Rojas, C.; Majer, P.;
Slusher, B. S.; Konvalinka, J. J. Neurochem. 2002, 80, 477.
17. Vornov, J. J. J. Neurochem. 1995, 65, 1681.
18. Wilkes, S. H.; Prescott, J. M. J. Biol. Chem. 1983, 258, 13517.
19. Baker, J. O.; Wilkes, S. H.; Bayliss, M. E.; Prescott, J. M.
Biochemistry 1983, 22, 2098.
20. Wagner, F. W.; Wilkes, S. H.; Prescott, J. M. J. Biol.
Chem. 1972, 247, 1208.
21. Pinto, J. T.; Suffoletto, B. P.; Berzin, T. M.; Qiao, C. H.;
Lin, S.; Tong, W. P.; May, F.; Mukherjee, B.; Heston, W. D.
Clin. Cancer Res. 1996, 2, 1445.
22. Nan, F.; Bzdega, T.; Pshenichkin, S.; Wroblewski, J. T.;
Wroblewska, B.; Neale, J. H.; Kozikowski, A. P. J. Med.
Chem. 2000, 43, 772.
(S)-13
(R)-13
(S)-14
15
>100,000
>100,000
>100,000
43,000c
—
—
—
—
—
16
20,000ꢃ2000
23. Kozikowski, A. P.; Nan, F.; Conti, P.; Zhang, J.; Rama-
dan, E.; Bzdega, T.; Wroblewska, B.; Neale, J. H.; Pshe-
nichkin, S.; Wroblewski, J. T. J. Med. Chem. 2001, 44, 298.
24. Rodriguez, C. E.; Lu, H.; Martinez, A. R.; Hu, Y.; Bru-
nelle, A.; Berkman, C. E. J. Enzyme Inhib. 2001, 16, 359.
25. Tsukamoto, T.; Flanary, J. M.; Rojas, C.; Slusher, B. S.;
Valiaeva, N.; Coward, J. K. Bioorg. Med. Chem. Lett. 2002,
12, 2189.
26. Johnson, W. H.; Roberts, N. A.; Borkakoti, N. J. Enz.
Inhib. 1987, 2, 1.
27. Jackson, P. F.; Slusher, B. S. Curr. Med. Chem. 2001, 8,
949.
28. Babine, R. E.; Bender, S. L. Chem. Rev. 1997, 97, 1359.
aValues are the meansꢃSD of at least two separate experiments unless
otherwise noted.
bValues have been previously reported6À8 and are included herein for
reference purposes.
cValues are from a single experiment.
of these interactions with the enzyme. Further structural
optimization with respect to other parts of the molecule
may enhance the interaction of a hydroxamate group
with the active site and lead to more potent hydrox-
amate-based inhibitors of GCP II.