Rationally designed sulfamides as glutamate carboxypeptidase II inhibitors

Article abstract of DOI:10.1111/cbdd.12174

Glutamate carboxypeptidase II (GCPII) is a membrane-bound cell surface peptidase. There is significant interest in the inhibition of GCPII as a means of neuroprotection, while GCPII inhibition as a method to treat prostate cancer remains a topic of further investigation. The key zinc-binding functional group of the well-characterized classes of GCPII inhibitors (phosphonates and phosphoramidates) is tetrahedral and negatively charged at neutral pH, while glutamyl urea class of inhibitors possesses a planar and neutral zinc-binding group. This study explores a new class of GCPII inhibitors, glutamyl sulfamides, which possess a putative net neutral tetrahedral zinc-binding motif. A small library containing six sulfamides was prepared and evaluated for inhibitory potency against purified GCPII in an enzymatic assay. While most inhibitors have potencies in the micromolar range, one showed promising sub-micromolar potency, with the optimal inhibitor in this series being aspartyl-glutamyl sulfamide (2d). Lastly, computational docking was used to develop a tentative binding model on how the most potent inhibitors interact with the ligand-binding site of GCPII. This current study explores a new class of GCPII inhibitors, glutamyl sulfamides, which possess a neutral tetrahedral zinc-binding motif. A small library containing six sulfamides was prepared and evaluated for inhibitory potency against purified GCPII. While most inhibitors have potencies in the micromolar range, one showed promising sub-micromolar potency, with the optimal inhibitor in this series being aspartyl-glutamyl sulfamide.

Full text of DOI:10.1111/cbdd.12174

Received Date : 03-Jan-2013
Revised Date : 07-May-2013
Accepted Date : 28-May-2013
Article type : Research Article
Rationally Designed Sulfamides As Glutamate Carboxypeptidase II (GCPII)
Inhibitors
Cindy J. Choy,a Melody D. Fulton,a Austen L. Davis,a Mark Hopkins,a Joseph K. Choi,a
Marc O. Anderson,b Clifford E. Berkmana,*
aDepartment of Chemistry, Washington State University, Pullman, WA 99164-4630
bDepartment of Chemistry and Biochemistry, San Francisco State University,
San Francisco, CA 94132
*Tel 509-335-7613, Fax 509-335-8867, cberkman@wsu.edu
Keywords: glutamate carboxypeptidase II, GCP2, prostate specific membrane antigen,
PSMA, sulfamide.
Abstract
Glutamate carboxypeptidase II (GCPII) is a membrane-bound cell-surface
peptidase. There is significant interest in the inhibition of GCPII as a means of
neuroprotection, while GCPII inhibition as a method to treat prostate cancer remains a
topic of further investigation. The key zinc-binding functional group of the well-
characterized classes of GCPII inhibitors (phosphonates and phosphoramidates) is
tetrahedral and negatively charged at neutral pH, while glutamyl urea class of inhibitors
possess a planar and neutral zinc-binding group. This current study explores a new
class of GCPII inhibitors, glutamyl sulfamides, which possess a putative net neutral
This article has been accepted for publication and undergone full peer review but has not been
through the copyediting, typesetting, pagination and proofreading process which may lead to
differences between this version and the Version of Record. Please cite this article as an
'Accepted Article', doi: 10.1111/cbdd.12174
This article is protected by copyright. All rights reserved.

tetrahedral zinc-binding motif. A small library containing 6 sulfamides was prepared and
evaluated for inhibitory potency against purified GCPII in an enzymatic assay. While
most inhibitors have potencies in the micromolar range, one showed promising sub-
micromolar potency, with the optimal inhibitor in this series being aspartyl-glutamyl
sulfamide (2d). Lastly, computational docking was used to develop a tentative binding
model on how the most potent inhibitors interact with the ligand-binding site of GCPII.
Introduction
The expression of glutamate carboxypeptidase II (GCPII) in human prostate
epithelium is known as prostate-specific membrane antigen (PSMA). GCPII is a
glycosylated cell surface zinc metallopeptidase overexpressed on prostate cancer cells
and neovasculature of non-prostatic malignancies (1, 2). As a consequence, GCPII has
attracted significant attention as a target for the delivery of imaging and therapeutic
agents and continues to serve as an important biomarker. Several studies have
suggested that targeting GCPII with small molecules (3, 4) and monoclonal antibodies
(4-6) has potential in the development of therapeutic strategies against prostate cancer.
However, the most established use of GCPII inhibitors has been in the diagnostic
imaging of prostate cancer (4, 7). In the nervous system, GCPII has been implicated in
the critical role of regulating synaptic glutamate levels. The presence of excess
glutamate has been implicated in several neurological disorders including ischemia,
traumatic brain injury (TBI), stroke, and amyotrophic lateral sclerosis (ALS) (4, 8, 9).
One source of glutamate in the nervous system is proteolysis of the short neuropeptide
N-acetylaspartylglutamate (NAAG), a reaction catalyzed by GCPII. As such, inhibition
of CNS GCPII causes decreased levels of extracellular glutamate (10), as well as also
increased levels of NAAG which itself has a neuroprotective role (11). Therefore there
is significant interest in the inhibition of GCPII as a means of neuroprotection.
Of the known classes of GCPII inhibitor scaffolds, they are typically characterized
by a zinc-binding group connected to either a glutaryl or the glutamyl group, which
generally serves as the C-terminal or P1’ residue of the chemical scaffold (4, 8). For
most of these inhibitors, the glutarate/glutamate residue appears to occupy the S1’
This article is protected by copyright. All rights reserved.

pocket of GCPII (12, 13), while additional motifs (if present) occupy the S1 pocket. Key
classes of inhibitors (Figure 1) include: phosphonates (1a), phosphates (1b),
phosphoramidates (1c-1e), and ureas (1f) (4, 8). In the case of the phosphorous-based
inhibitors, the phosphoryl functionality appears to serve as zinc binding group (ZBG) to
the co-catalytic zinc atoms in the active site (12, 13). A problem with the highly potent
(yet highly charged) phosphonate inhibitor 2-PMPA (1a) and other phosphorous-based
inhibitors is their poor oral bioavailability (14) and high renal clearance (15) thereby
potentially limiting their practical value as clinical therapeutic agents.
One of our goals is to identify alternative functional groups that could serve as a
ZBG group to interact with the co-catalytic zinc atoms in GCPII. In addition, it is
expected that these functional groups could also passively serve as a linker between
P1-glutamate and a P1’ residue to achieve favorable contacts in the active site of
GCPII. One attractive scaffold is the glutamyl urea, pioneered by Kozikowski et al (4),
and has been co-crystallized with GCPII (16). Recent structural investigation shows the
urea oxygen of these inhibitors is approximately 1 Å further away from the catalytic zinc
atoms (compared to the tetrahedral phosphonate moiety), but additional interactions in
the S1 pocket compensate for the weak binding of the urea group to the active site zinc
atoms (16). Another attractive putative zinc-binding motif is the tetrahedral sulfonamide
and sulfamide, which allows for trivial installation into small molecule inhibitors. Both
the sulfonamide and sulfamide moieties are desirable for pharmaceutical development,
compared to similar phosphorous-based groups (1a-1e) due to their aqueous stability
and their net neutral charge. Short alkyl sulfonamides have showed promise as GCPII
inhibitors (17). Although not always functioning as a ZBG, the sulfamide motif has been
employed in the design of biologically active agents against various classes of enzymes
such as aspartic proteases (18, 19) (HIV-1 protease, gamma-secretase), serine
proteases (20) (elastase, chymase, tryptase and thrombin), and metalloproteinase (21)
(carboxypeptidase A [CPA] and matrix metalloproteinases [MMPs]). One biochemical
target for which sulfamides have been successfully used as a zinc-binding groups is
CPA, where single-crystal X-ray diffraction reveals that the sulfamoyl moiety interacts
with the active site zinc ion (21). An important feature of sulfamides (as well as ureas
This article is protected by copyright. All rights reserved.

and sulfonamides) is the ease in which they can be conjugated to chiral and
inexpensive protected glutamate building blocks (8 or 9) to generate an optically active
inhibitor scaffold. In contrast, the highly potent phosphonate inhibitors (e.g. 2-PMPA)
are more challenging to produce in optically active form (22).
The focus of this study was aimed at (1) the development of a facile synthetic
methodology; (2) utilization of methodology to generate a small library of simple
glutamyl sulfamides; (3) modeling of a potential binding mode using computational
docking.
RESULTS AND DISCUSSION
A simple method, amenable to parallel synthesis, was employed to synthesize
the sulfamide inhibitor library (Scheme 1). The common oxazolidinone intermediate
and sulfamides were synthesized as outlined by Borghese and coworkers with minor
modifications (23). Briefly, commercially available chlorosulfonyl isocyanate was
treated with 2-bromoethanol or benzyl alcohol, followed by treatment with di-protected
glutamic acid (8 or 9) to generate the oxazolidinone intermediates 4 or 5 and Cbz-
glutamyl sulfamide 7 respectively. Nucleophilic displacement of the oxazolidinone in 4
or 5 was carried out with amines 6a-e to generate methyl and benzyl protected
sulfamides 3a-e. Amines 6a, 6a’, and 6d-e could be purchased, while 6b-c were
prepared in two steps from commercially available amino acid precursors by coupling to
form the benzamide followed by tert-butyl ester deprotection (see supporting
information). Last, 3b-e were deprotected by either ester hydrolysis or hydrogenolysis
to form inhibitor candidates 2b-2e, while 3a and 3a’ required additional elaboration.
The synthesis of the urea analog 2a and its diasteromeric partner 2a’ required
the installation of the p-iodo-phenyl urea moiety at the ε-amino group prior to global
deprotection procedures (Scheme 2). The Cbz group from 3a was selectively removed
under catalytic transfer hydrogenolysis conditions to generate the free amine. After
overnight drying in vacuo, the amine was treated with p-iodo-phenyl isocyanate and
TEA in DCM to yield 10a, which was subsequently subjected to global deprotection to
yield the target compound 2a. The diasteromeric partner 2a’ was prepared analogously
from 3a’.
This article is protected by copyright. All rights reserved.

Applying the methodology described in Schemes 1-2, a library of potential GCPII
inhibitors containing a sulfamide moiety was prepared. The objective of the library
design was to test whether the sulfamide core can serve as a ZBG or passively serve
as a bridging unit to connect moieties spanning the S1 and S1’ regions of the active-
site. We generated 2-(sulfamoylamino)pentanedioic acid (2f) to test whether the
terminal sulfamide can serve as a PSMA inhibitor via interaction with the co-catalytic
zincs. In addition, we generated structural analogs of known GCPII inhibitors to test
whether the sulfamide core can serve as a linking unit.
Once prepared, the sulfamide inhibitors 2a-f were assayed for inhibition against
purified GCPII with results indicated as 50% inhibitory concentration (IC50) values
(Table 1) (24). Based on previous reports, the sulfamide moiety is capable of
interacting with the active site Zn in other zinc metalloproteases such as CPA (21) and
carbonic anhydrase (25). Therefore, we hypothesized that sulfamide motif in compound
2f would act as a ZBG and exhibit greater inhibitory potency than glutamate alone (428
μM) (12). Upon comparing the potency of 2f to that of 1c (IC50 = 0.86 nM) (26) against
GCPII, it was concluded that either the sulfamide motif in 2f is not interacting with the
active site Zn or that the interaction with the Zn cation is weak.
The non-classical isosteric replacement of the urea group by the sulfamide motif
in a known GCPII inhibitor 1f (IC50=20 nM) (27) also resulted in sulfamide analog 2a
and its diasteromeric partner 2a’ albeit with decreased inhibitory potency. We
hypothesized that the observed poor potency in the diasteromeric pairs could be due to;
1) the weak zinc interaction of the sulfamide moiety; or 2) lack of functionality capable of
interacting with the arginine patch. The P1 α-carboxylate in the parent urea inhibitor
(1f) makes favorable interactions with Asn519, as well as residues Arg534 and Arg536
of the arginine patch (16). Reorientation of this carboxylate due to the presence of the
tetrahedral sulfamide group may prevent the favorable interactions of the P1 α-
carboxylate present in 2a-2a’ with these Asn and Arg residues, resulting in decreased
affinity. This led us to conclude the non-classical isosteric replacement of the known
urea GCPII inhibitor was detrimental and resulted in loss of potency.
This article is protected by copyright. All rights reserved.

To obtain better understanding of the sulfamide core as a putative ZBG and its
interactions with the active site, we have included 2b in our library design
considerations. Because it is a direct structural analog of a known pseudoirreversible
GCPII inhibitor 1e (IC50=35 nM) (26), it allowed for an interrogation of the sulfamide as a
isostere of the phosphoramidate group. Despite preserving the tetrahedral geometry of
a phosphoramidate as well as features of the P1 and P1’ functionality, this isosteric
replacement resulted in a considerable loss of inhibitory potency towards GCPII,
presumably due to incompatibility of the sulfamide center as a ZBG.
Based on the above observations, we introduced greater flexibility in our inhibitor
design in the P1 residue to promote additional interactions with the well-known
hydrophobic pocket and arginine patch proximal to the active site. Compounds 2d and
2e were designed to allow a carboxylate to probe the arginine patch and/or potentially
chelate to the active site Zn, while compound 2c was introduced to probe for
hydrophobic binding interactions. In general, by introducing greater flexibility, the
inhibitory potencies of 2c-e improved.
To shed light on the mode of binding of the more potent sulfamides in this study,
computational docking was performed on compounds 2d, 2e, and 2f using a high-
resolution X-ray crystal structure of GCPII (pdb = 3D7H), co-crystallized with urea
inhibitor DCIBzL (16), using the software FRED (OpenEye Scientific, Santa Fe, NM)(28,
29). Docking was performed in the absence of a pharmacophore restraint. A summary
of the docking results of 2d and 2e are presented in Figure 2, and rendered images are
provided in Figures S1-S2 (Supporting Information). Un-restrained docking of these
two inhibitors places them in poses inconsistent with numerous other glutamate
containing inhibitors with known modes of binding (1c, DCIBzl, etc) which clearly project
the P1’ Glu γ-carboxylate group into the S1’ pocket to form a salt bridge with Lys699,
and with the P1’ α-carboxylate interacting with Arg210. In our model of 2d and 2e, the
P1’ Glu γ-carboxylate interacts with the co-catalytic zinc atoms, while the P1’ α-
carboxylate interacts with Arg534 and Arg536. In declining affinity for zinc, the order is:
histidine > carboxylate > sulfhydryl > phenol > lysine > main chain carbonyl (30),
This article is protected by copyright. All rights reserved.

therefore it is not surprising to see that carboxylate groups replace the sulfamide groups
to act as a stronger zinc binder. In these inhibitors, the sulfamide functionality does not
appear to be important for binding. In inhibitor 2d, the P1 β-carboxylate interacts with
Arg534 and Arg536, while the P1 α-carboxylate interacts with Arg463 and Arg534. The
slightly less potent inhibitor 2e contains an extra methylene group on its side chain; the
P1 γ-carboxylate interacts with Arg463 and Arg536, while the P1 α-carboxylate is only
within distance of interacting with Arg534. In most cases, the interactions are
considered to be salt bridges between charged functionality (<4 Å between the charged
atoms) (31), while in some cases, we were able to identify potential hydrogen bonds as
well (< 3.2 Å between the donor and acceptor; angle between HBD-H…HBA > 1500)
(32). The specific non-covalent interactions, including distances determined in the
course of our docking simulations of 2d and 2e are presented in Table S1.
A
superimposition of compounds 2d and 2e docked into the catalytic site is presented in
Figure S3, showing a generally consistent mode of binding, aside from the position of
the terminal P1 carboxylate groups, due to the longer side chain in 2e, while the
remainder of the two structures are positioned fairly consistently.
The possibility that 2d and 2e adopt a novel pose is interesting, and would need
to be validated by a more rigorous structural method such as X-ray crystallography. To
provide evidence against the “classical” mode of binding of these inhibitors (P1’ into
S1’), we also performed docking of 2d and 2e in the presence of a pharmacophore
constraint that required the P1’ γ-carboxylate to be within 3.5Å of Lys699 in the S1’
pocket, allowing the compounds to dock in a manner more consistent with other GCPII
inhibitors (data not presented). While such poses would be appealing in that they
share consistency with X-ray crystal structures of phosphorous and urea-based
inhibitors, the seven independent scoring functions in FRED score such poses with
significantly higher energy than the novel poses presented in Figure 2.
Our model does not precisely explain the slightly improved potency of 2d (IC50 =
0.9 μM) versus 2e (IC50 = 1.2 μM), as both compounds make a similar number of
contacts (with similar distances) to the residues in the arginine patch. Nevertheless, it
This article is protected by copyright. All rights reserved.

can be noted: (a) while the P1 α-carboxylate in both compounds can obtain salt bridge
interactions with Arg463, this carboxylate in 2d is within distance to make a second salt
bridge interaction with Arg534, while this group in 2e is bent slightly away from Arg534,
and less likely to interact; (b) the relative potency of the compounds is within
experimental error, thus one must be cautious to conclude 2d is actually more potent
than 2e.
We also performed un-restrained docking of the less potent inhibitor 2f, which is
a sulfamide analog of the extremely potent and structurally analogous phosphoramidate
inhibitor (1c) (Figure S4). Our model of this compound also changes its orientation
from the classical binding mode observed for phosphorous and urea-based inhibitors.
In our model of 2f, one sulfamide S=O group hydrogen bonds with Arg210, while the
Glu α-carboxylate interacts with zinc. The Glu γ-carboxylate interacts with Arg534 and
Arg536 through a salt bridge and through hydrogen bonding interactions. While the
interactions seem reasonable, caution must be exercised due to the relatively poor
potency of this inhibitor (IC50 = 6 μM).
Conclusion and Future Work
A small library of sulfamide analogs of known GCPII inhibitor scaffolds was
prepared to interrogate the value of the sulfamide motif as a zinc-binding group for this
and other metallopeptidases. While the majority of the inhibitors were modestly active,
one inhibitor (2d), which contains an ASP β-carboxylate for the S1 pocket, possessed
sub-micromolar activity. Binding modes for three of the more potent compounds (2d-
2f) was rationalized by a binding model generated by computational docking. Our
model suggests that the two most potent compounds (2d and 2e) obtain a non-
conventional binding mode, in which the sulfamide group does not interact with the co-
catalytic zinc atoms, but with the acidic side chain functionality greatly interacting with
the residues in the S1 arginine patch region. The less potent inhibitor 2f docked in a
conformation that suggests the sulfamide oxygens contribute to zinc binding.
This article is protected by copyright. All rights reserved.

Sulfamides that lack both α- or β-carboxylate in the P1 residue resulted in poor
inhibition. Based on the results of our initial sulfamide library, we conclude the
sulfamide moiety is not a good zinc binding group for GCPII (in contrast with other
proteins like CPA).
Materials and Methods
IC50 Determinations for GCPII Inhibition. Inhibitions studies were performed as
previously described (17, 33, 34).
Synthesis. All solvents used in reactions were both anhydrous and obtained from
commercial sources. 1H and 13C NMR spectra were recorded on a Bruker 400 MHz
1
spectrometer or a Varian 300 MHz. H NMR chemical shifts are relative to TMS (δ=0.00
ppm), CDCl3 (δ=7.26 ppm), CD3OD (δ=3.31 ppm) or D2O (δ=4.79 ppm). 13C NMR
chemical shifts are relative to CDCl3 (δ=77.23 ppm), CD3OD (δ=49.00 ppm). Mass
spectrometry spectra were obtained on an Applied Biosytesms 4800 MALDI- TOF/TOF
mass spectrometer.
General procedure for oxazolidinone-glutamyl sulfamide preparation (4 and 5).
To a solution of chlorosulfonyl isocyanate (1.538 mL, 17.66 mmol) at 0oC in DCM (7
mL), was added a solution of 2-bromoethanol (1.26 mL, 17.78 mmol) in DCM (10 mL)
slowly via addition funnel and maintain reaction temperature range of 0-5oC. Upon
complete addition, the reaction was stirred for an hour at 0-5oC, followed by the addition
of 8 or 9 (19.43 mmol) in DCM (18 mL) and TEA (7.87 mL, 56.52 mmol). The solution
was added slowly to maintain reaction temp 0-5oC. Once addition was complete, the
resulting reaction mixture was allowed to warm to RT and stirred overnight. After
overnight stirring, the reaction mixture was washed with 1N HCl (2X, 50 mL) followed by
brine (50 mL). Water was added to the organic layer and DCM was evaporated to yield
a precipitate (70-86%) and further used without purification.
(S)-dimethyl 2-(2-oxooxazolidine-3-sulfonamido)pentanedioate (4): (85.6%) 1H
NMR (400 MHz, DMSO-d6) δ 9.08 (d, J = 9.0 Hz, 1H), 4.43-4.25 (m, 2H), 4.12 (td, J =
9.5, 4.7 Hz, 1H), 3.97-3.83 (m, 2H), 3.67 (s, 3H), 3.57 (s, 3H), 2.43-2.34 (m, 2H), 2.04-
This article is protected by copyright. All rights reserved.

1.96 (m, 1H), 1.78-1.69 (m, 1H). 13C NMR (100 MHz, DMSO-d6) δ 172.86, 171.86,
153.18, 62.88, 55.64, 52.89, 51.88, 45.77, 29.65, 27.09.
1
(S)-dibenzyl 2-(2-oxooxazolidine-3-sulfonamido)pentanedioate (5): (69.8%) H
NMR (400 MHz, DMSO-d6) δ 9.14 (d, J = 8.8 Hz, 1H), 7.49-7.14 (m, 10H), 5.25-4.92
(m, 4H), 4.37-4.12 (m, 3H), 4.01-3.67 (m, 2H), 2.46-2.35 (m, 2H), 2.08-2.00 (m, 1H),
1.83-1.74 (m, 1H). 13C NMR (100 MHz, DMSO-d6) δ 172.27, 171.37, 153.16, 136.48,
135.98, 128.87, 128.85, 128.62, 128.45, 128.40, 128.38, 67.08, 66.05, 62.85, 55.79,
45.75, 29.85, 27.18.
General procedure for methyl ester deprotection.(2a, 2a’, 2b, and 2c). To a
solution of 3a-c in 1,4-Dioxane (1 mL), was added 1 M LiOH (1.5 eq/methyl ester). The
resulting solution was stirred at room temperature until completion. Upon completion,
the reaction mixture was acidified (pH=3) with 4 M HCl. The organic solvent was
removed and the crude mixture was taken up to 0.5 mL with HPLC grade methanol,
micro filtered then peak purified with a preparative Thermo Scientific* Hypersil* GOLD
preparative C18 column (20mm x 50mm, particle size 5 um) running a 5% to 97%
gradient of water/methanol with 0.1% formic acid with a 15 mL/min flow rate on a
SpectraSystem HPLC. Chromatographs were monitored with single wavelength UV
detector at 254 nm. Fractions were concentrated down to yield a white solid (57-65%).
(2S)-2-((N-(1-carboxy-5-(3-(4-iodophenyl)ureido)pentyl)sulfamoyl)amino)
pentanedioic acid (2a): (65.4%) 1H NMR (400 MHz, CD3OD) δ 7.52 (d, J = 8.8 Hz, 1H),
7.17 (d, J = 8.7 Hz, 1H), 4.00-3.96 (m, 1H), 3.94-3.91 (m, 1H), 3.18 (t, J = 6.5 Hz, 1H),
2.55-2.34 (m, 1H), 2.17-2.10 (m, 1H), 1.98-1.77 (m, 1H), 1.76-1.67 (m, 1H), 1.61-1.38
(m, 2H). 13C NMR (100 MHz, CD3OD) δ 175.06, 174.35, 173.90, 173.55, 156.54,
139.67, 137.28, 120.50, 83.63, 55.53, 54.88, 54.82, 50.75, 39.11, 32.13, 29.37, 29.30,
29.27, 27.69, 27.61, 22.31, 22.28.
NMR spectra for 2a’ is identical to 2a. M/Z
calculated for C18H24IN4O9S [M-H] 599.0314; found [M-H] 599.0011.
(S)-2-((N-((S)-2-benzamido-2-carboxyethyl)sulfamoyl)amino)pentanedioic
acid
(2b): (58.6%) 1H NMR (400 MHz, CD3OD) δ 7.93-7.85 (m, 2H), 7.61-7.51 (m, 1H),
7.51-7.44 (m, 2H), 4.72-4.69 (m, 1H), 3.96-3.92 (m, 1H), 3.66-3.46 (m, 2H), 2.53-2.41
This article is protected by copyright. All rights reserved.

(m, 2H), 2.19-2.09 (m, 1H), 1.95-1.82 (m, 1H). 13C NMR (100 MHz, CD3OD) δ 176.48,
176.46, 175.53, 175.43, 173.30, 170.12, 170.06, 135.06, 132.93, 129.57, 128.52,
128.51, 56.43, 54.91, 44.69, 31.01, 29.16. M/Z calculated for C15H18N3O9S [M-H]
416.0769; found [M-H] 416.0920.
(S)-2-((N-((S)-3-benzamido-3-carboxypropyl)sulfamoyl)amino)pentanedioic acid
(2c): (57.2%) 1H NMR (400 MHz, CD3OD) δ 7.91-7.82 (m, 2H), 7.57-7.50 (m, 1H), 7.50
-7.41 (m, 2H), 4.67-4.62 (m, 1H), 3.92-3.88 (m, 1H), 3.23-3.02 (m, 2H), 2.50-2.39 (m,
13
2H), 2.28 -2.16 (m, 1H), 2.15-1.94 (m, 1H), 1.92-1.79 (m, 1H). C NMR (100 MHz,
CD3OD) δ 175.00, 174.31, 173.82, 169.05, 133.77, 131.41, 128.09, 127.14, 127.13,
54.94, 50.83, 39.49, 30.98, 29.56, 27.71, 27.69. M/Z calculated for C16H20N3O9S [M-H]
430.0926; found [M-H] 430.0882.
General procedure for hydrogenolysis (2d, 2e, and 2f). Ammonium formate (5
eq/benzyl ester) was added to a rapidly stirred suspension of 3d-f, Pd/C (0.03 eq) and
potassium bicarbonate (1.0 eq/carboxylate) in 1,4-Dioxane: water solution (2:1 v/v, 2
mL). The reaction was stirred until completion. Upon completion, the reaction mixture
was filtered and concentrated down to produce the desired compound in quantitative
yield as the potassium salt.
(S)-2-((N-((S)-1,2-dicarboxyethyl)sulfamoyl)amino)pentanedioic
acid
(2d):
(95.3%) 1H NMR (400 MHz, D2O) δ 3.82 -3.75 (m, 1H), 3.55-3.48 (m, 1H), 2.50-2.39 (m,
1H), 2.35-2.20 (m, 1H), 2.10-1.96 (m, 2H), 1.82-1.60 (m, 2H). 13C NMR (100 MHz, D2O)
δ 182.20, 179.28, 178.86, 178.48, 58.00, 57.97, 56.11, 56.01, 40.68, 33.69, 33.59,
29.24, 29.20. M/Z calculated for C9H13N2O10S [M-H] 341.0296; found [M-H] 340.8914.
1
(2S,2'S)-2,2'-(sulfonylbis(azanediyl))dipentanedioic acid (2e): (97.3%) H NMR
(400 MHz, D2O) δ 3.72-3.55 (m, 2H), 2.29-2.13 (m, 4H), 2.01-1.73 (m, 4H). 13C NMR
(100 MHz, D2O) δ 182.08, 179.20, 58.07, 33.73, 29.28.
M/Z calculated for
C10H15N2O10S [M-H] 355.0453; found [M-H] 355.0450.
(S)-2-(sulfamoylamino)pentanedioate (2f): (100%) 1H NMR (400 MHz, Deuterium
Oxide) δ 3.74 (td, J = 8.5, 5.0 Hz, 1H), 2.42-2.28 (m, 2H), 2.10-1.95 (m, 1H), 1.88-1.79
This article is protected by copyright. All rights reserved.

(m, 1H). 13C NMR (100 MHz, D2O) δ 180.01, 178.68, 170.50, 62.38, 57.75, 31.97,
28.46. M/Z calculated for C5Hl9N2O6S [M-H] 225.0187; found [M-H] 225.0301.
References and Notes
1.
2.
Haffner MC, Kronberger IE, Ross JS, Sheehan CE, Zitt M, Muhlmann G, et al.
Prostate-specific membrane antigen expression in the neovasculature of gastric
and colorectal cancers. Hum Pathol. 2009;40(12):1754-61. Epub 2009/09/01.
Chang SS, O'Keefe DS, Bacich DJ, Reuter VE, Heston WD, Gaudin PB.
Prostate-specific membrane antigen is produced in tumor-associated
neovasculature. Clinical cancer research : an official journal of the American
Association for Cancer Research. 1999;5(10):2674-81.
3.
4.
Tang H, Brown M, Ye Y, Huang G, Zhang Y, Wang Y, et al. Prostate targeting
ligands based on N-acetylated alpha-linked acidic dipeptidase. Biochem Biophys
Res Commun. 2003;307(1):8-14.
Zhou J, Neale JH, Pomper MG, Kozikowski AP. NAAG peptidase inhibitors and
their potential for diagnosis and therapy. Nat Rev Drug Discov. 2005;4(12):1015-
26.
5.
6.
7.
8.
Chang SS. Monoclonal antibodies and prostate-specific membrane antigen. Curr
Opin Investig Drugs. 2004;5(6):611-5.
Chang SS. Overview of prostate-specific membrane antigen. Rev Urol. 2004;6
Suppl 10:S13-8. Epub 2006/09/21.
Bander NH. Technology insight: monoclonal antibody imaging of prostate cancer.
Nature clinical practice. 2006;3(4):216-25.
Tsukamoto T, Wozniak KM, Slusher BS. Progress in the discovery and
development of glutamate carboxypeptidase II inhibitors. Drug Discov Today.
2007;12(17-18):767-76.
9.
Whelan J. NAALADase inhibitors: A novel approach to glutamate regulation.
Drug Discov Today. 2000;5(5):171-2. Epub 2000/05/02.
10.
Vornov JJ, Wozniak K, Lu M, Jackson P, Tsukamoto T, Wang E, et al. Blockade
of NAALADase: a novel neuroprotective strategy based on limiting glutamate and
elevating NAAG. Ann N Y Acad Sci. 1999;890:400-5. Epub 2000/02/11.
Kamiya H, Shinozaki H, Yamamoto C. Activation of metabotropic glutamate
receptor type 2/3 suppresses transmission at rat hippocampal mossy fibre
synapses. J Physiol. 1996;493 ( Pt 2):447-55. Epub 1996/06/01.
Barinka C, Rovenska M, Mlcochova P, Hlouchova K, Plechanovova A, Majer P,
et al. Structural insight into the pharmacophore pocket of human glutamate
carboxypeptidase II. J Med Chem. 2007;50(14):3267-73.
Mesters JR, Henning K, Hilgenfeld R. Human glutamate carboxypeptidase II
inhibition: structures of GCPII in complex with two potent inhibitors, quisqualate
and 2-PMPA. Acta Crystallogr D Biol Crystallogr. 2007;63(Pt 4):508-13.
Peng XQ, Li J, Gardner EL, Ashby CR, Jr., Thomas A, Wozniak K, et al. Oral
administration of the NAALADase inhibitor GPI-5693 attenuates cocaine-induced
11.
12.
13.
14.
This article is protected by copyright. All rights reserved.

reinstatement of drug-seeking behavior in rats. European journal of
pharmacology.627(1-3):156-61.
Lapi SE, Wahnishe H, Pham D, Wu LY, Nedrow-Byers JR, Liu T, et al.
Assessment of an 18F-labeled phosphoramidate peptidomimetic as a new
prostate-specific membrane antigen-targeted imaging agent for prostate cancer.
J Nucl Med. 2009;50(12):2042-8. Epub 2009/11/17.
Barinka C, Byun Y, Dusich CL, Banerjee SR, Chen Y, Castanares M, et al.
Interactions between human glutamate carboxypeptidase II and urea-based
inhibitors: structural characterization. J Med Chem. 2008;51(24):7737-43. Epub
2008/12/05.
15.
16.
17.
18.
19.
20.
21.
22.
Blank BR, Alayoglu P, Engen W, Choi JK, Berkman CE, Anderson MO. N-
substituted glutamyl sulfonamides as inhibitors of glutamate carboxypeptidase II
(GCP2). Chem Biol Drug Des. 2011;77(4):241-7. Epub 2011/01/12.
Backbro K, Lowgren S, Osterlund K, Atepo J, Unge T, Hulten J, et al.
Unexpected binding mode of a cyclic sulfamide HIV-1 protease inhibitor. J Med
Chem. 1997;40(6):898-902. Epub 1997/03/14.
Sparey T, Beher D, Best J, Biba M, Castro JL, Clarke E, et al. Cyclic sulfamide
gamma-secretase inhibitors. Bioorg Med Chem Lett. 2005;15(19):4212-6. Epub
2005/08/02.
Zhong J, Gan X, Alliston KR, Lai Z, Yu H, Groutas CS, et al. Potential protease
inhibitors based on a functionalized cyclic sulfamide scaffold. J Comb Chem.
2004;6(4):556-63. Epub 2004/07/13.
Park JD, Kim DH, Kim SJ, Woo JR, Ryu SE. Sulfamide-based inhibitors for
carboxypeptidase A. Novel type transition state analogue inhibitors for zinc
proteases. J Med Chem. 2002;45(24):5295-302. Epub 2002/11/15.
Vitharana D, France JE, Scarpetti D, Bonneville GW, Majer P, Tsukamoto T.
Synthesis
(phosphonomethyl)pentanedioic
carboxypeptidase II. Tetrahedron: Asymmetry. 2002;13(15):1609-14.
and
biological
evaluation
acids as
of
(R)-
and
of
(S)-2-
glutamate
inhibitors
23.
Borghese A, Antoine L, Van Hoeck JP, Mockel A, Merschaert A. Mild and safer
preparative method for nonsymmetrical sulfamides via N-sulfamoyloxazolidinone
derivatives: Electronic effects affect the transsulfamoylation reactivity. Organic
Process Research & Development. 2006;10(4):770-5.
24.
25.
Wu LY, Anderson MO, Toriyabe Y, Maung J, Campbell TY, Tajon C, et al. The
molecular pruning of a phosphoramidate peptidomimetic inhibitor of prostate-
specific membrane antigen. Bioorg Med Chem. 2007;15(23):7434-43.
Abbate F, Supuran CT, Scozzafava A, Orioli P, Stubbs MT, Klebe G.
Nonaromatic sulfonamide group as an ideal anchor for potent human carbonic
anhydrase inhibitors: role of hydrogen-bonding networks in ligand binding and
drug design. J Med Chem. 2002;45(17):3583-7. Epub 2002/08/09.
26.
27.
Liu T, Toriyabe Y, Kazak M, Berkman CE. Pseudoirreversible Inhibition of
Prostate-Specific Membrane Antigen by Phosphoramidate Peptidomimetics.
Biochemistry. 2008;47(48):12658-60.
Maresca KP, Hillier SM, Femia FJ, Keith D, Barone C, Joyal JL, et al. A series of
halogenated heterodimeric inhibitors of prostate specific membrane antigen
This article is protected by copyright. All rights reserved.

(PSMA) as radiolabeled probes for targeting prostate cancer. J Med Chem.
2009;52(2):347-57. Epub 2008/12/30.
McGann M. FRED Pose Prediction and Virtual Screening Accuracy. Journal of
Chemical Information and Modeling. 2011;51(3):578-96.
McGaughey GB, Sheridan RP, Bayly CI, Culberson JC, Kreatsoulas C, Lindsley
S, et al. Comparison of topological, shape, and docking methods in virtual
screening. J Chem Inf Model. 2007;47(4):1504-19. Epub 2007/06/27.
Jaffe EK. Predicting the Zn(II) Ligands in Metalloproteins: Case Study,
Porphobilinogen Synthase. Comments on Inorganic Chemistry. 1993;15(2):67-
92.
28.
29.
30.
31.
32.
33.
Kumar S, Nussinov R. Close-range electrostatic interactions in proteins.
Chembiochem. 2002;3(7):604-17. Epub 2002/09/27.
Jeffrey GA. An introduction to hydrogen bonding. New York: Oxford University
Press; 1997. vii, 303 p. p.
Maung J, Mallari JP, Girtsman TA, Wu LY, Rowley JA, Santiago NM, et al.
Probing for a hydrophobic a binding register in prostate-specific membrane
antigen
with
phenylalkylphosphonamidates.
Bioorg
Med
Chem.
2004;12(18):4969-79.
34.
Anderson MO, Wu LY, Santiago NM, Moser JM, Rowley JA, Bolstad ES, et al.
Substrate specificity of prostate-specific membrane antigen. Bioorg Med Chem.
2007;15(21):6678-86.
Supporting Information
Scheme S1, Figures S1-S3, experimental procedure, and characterization data
for intermediates 3a-3e, and 7.
Figure 1. Known GCPII inhibitors.
Scheme 1. Synthesis of glutamyl sulfamides. Reagents and conditions: (a) 2-
bromoethanol; (b) benzyl alcohol; (c) HCl.Glu(OMe).OMe (8) or Tosyl.Glu(OBn).OBn
(9); (d) amino acids 6a-e, TEA, ACN, 50oC; (e) LiOH or Pd/C, NH4COOH, KHCO3, 1:2
H2O:Dioxane (v/v)
Scheme 2: Synthesis of sulfamide 2a and 2a’. Reagents: (a) Pd/C, NH4COOH,
KHCO3, 1:2 H2O:Dioxane (v/v); (b) 4-iodo-phenyl isocyante, TEA, DCM; (c) LiOH,
Dioxane.
Table 1: Inhibitory potency of sulfamides against purified GCPII
This article is protected by copyright. All rights reserved.

Figure 2: Summary of computational docking results for the most potent compounds
2d (panel A) and 2e (panel B) into the active site of an X-ray crystal structure of PSMA
(PDB=3D7H). Rendered images are provided in Figures S1-S2 (Supporting
Information).
Scheme S1. Synthesis of 6b and 6c. Reagents and conditions: (a) HBTU, TEA,
Benzoic acid, DMF; (b) 4M HCl in Dioxane.
Figure S1: Un-restrained computational docking of compound 2d into the active site of
an X-ray crystal structure of GCP2 (PDB=3D7H). The PyMOL Molecular Graphics
System (http://www.pymol.org/) was used to prepare the molecular drawings.
Figure S2: Un-restrained computational docking of compound 2e into the active site of
an X-ray crystal structure of GCP2 (PDB=3D7H).
Figure S3: Representation of docked compound 2d (red) superimposed with 2e (blue).
Figure S4: Un-restrained computational docking of compound 2f into the active site of
an X-ray crystal structure of GCP2 (PDB=3D7H).
Table S1: Interactions between ligands 2d-2f, and the GCP2 active site residues, as
determined by computational docking. Interaction type: A = salt bridge; B = metal
coordination; C = moderate-to-strong hydrogen bond as defined by Jeffrey et al (5). For
interactions A-B, the distance is between the associated atoms. For interaction C, the
distance is calculated between the hydrogen bond acceptor (typically on the ligand) and
a proton on the hydrogen bond donor (in most cases arginine). In most cases, the
position of the proton was not shown in Figures S1-S2 and S4, for the purpose of
clarity.
This article is protected by copyright. All rights reserved.

Compound
Structure
IC50 (µM)
2a
5(1)
2a’
2b
2c
2d
10(2)
>100
15(2)
0.9(1)
2e
2f
1.2(1)
5.9(4)
Table 1: Inhibitory potency of sulfamides against purified GCPII
This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.