S-glycosyl primary sulfonamides - A new structural class for selective inhibition of cancer-associated carbonic anhydrases

Article abstract of DOI:10.1021/jm900914e

In this paper, we present a new class of carbonic anhydrase (CA) inhibitor that was designed to selectively target the extracellular domains of the cancer-relevant CA isozymes. The aromatic moiety of the classical zinc binding sulfonamide CA inhibitors is

Full text of DOI:10.1021/jm900914e

J. Med. Chem. 2009, 52, 6421–6432 6421
DOI: 10.1021/jm900914e
S-Glycosyl Primary Sulfonamides-A New Structural Class for Selective Inhibition of
Cancer-Associated Carbonic Anhydrases
Marie Lopez,† Blessy Paul,† Andreas Hofmann,† Julia Morizzi,‡ Quoc K. Wu,‡ Susan A. Charman,‡ Alessio Innocenti,§
Daniela Vullo,§ Claudiu T. Supuran,§ and Sally-Ann Poulsen*,†
†Eskitis Institute for Cell and Molecular Therapies, Griffith University, Nathan, Queensland 4111, Australia, ‡Centre for Drug Candidate
Optimisation, Faculty of Pharmacy and Pharmaceutical Sciences, Monash University, 381 Royal Parade, Parkville, Victoria 3052, Australia, and
§Polo Scientifico, Laboratorio di Chimica Bioinorganica, Rm. 188, Universitaꢀ degli Studi di Firenze, Via della Lastruccia 3, 50019 Sesto
Fiorentino, Florence, Italy
Received June 21, 2009
In this paper, we present a new class of carbonic anhydrase (CA) inhibitor that was designed to
selectively target the extracellular domains of the cancer-relevant CA isozymes. The aromatic moiety of
the classical zinc binding sulfonamide CA inhibitors is absent from these compounds and instead they
incorporate a hydrophilic mono- or disaccharide fragment directly attached to the sulfonamide group to
give S-glycosyl primary sulfonamides (1-10). The inhibition properties of these compounds at the
physiologically abundant human CA isozymes I and II and cancer-associated IX and XII were
determined, and all compounds had moderate potency with Kis in the micromolar range. We present
the crystal structures of anomeric sulfonamides 4, 7, and 10 and the sugar sulfamate drug topiramate in
complex with human recombinant CA II. From these structures, we have obtained valuable insights into
ligand-protein interactions of these novel carbohydrate-based sulfonamides with CA.
Introduction
known since the 1930s (with isozymes discriminated in the
1960s). Other hCA isozymes are much more recent discov-
eries, and interestingly they were characterized well after CA
inhibitors were established as a mainstay of antihypertensive,
antiglaucoma, antithyroid, and hypoglycemic drug treatment
regimes. For example, the aromatic sulfonamide CA inhibi-
tors acetazolamide, methazolamide, ethoxzolamide, and the
bis-sulfonamide, dichlorophenamide, have been used clini-
cally for over 40 years as CA inhibitors, Figure 2a.1 Acetazo-
lamide is the par excellence CA inhibitor and was the first
nonmercurial diuretic to be used clinically commencing in
1956. The side effects of this “dirty” broad specificity CA
inhibitor and other systemically administered CA inhibitors
are now attributed to the inhibition of multiple CA isozymes
throughout the body.1 The primary sulfonamide (-SO2NH2)
of these drugs binds to the active site Zn2þ of CAs (Figure 2b)
and blocks access to the zinc coordination sphere for
the substrate water molecule, thus impeding the hydration
of CO2.1
The optimization of molecular recognition interactions
between ligand and enzyme is an important approach for
increasing the potency of a ligand, however in isolation this
approach is insufficient when the primary goal of our research
efforts is to develop useful drug-like compounds.2 Drug
design must take into account the relevant biological land-
scape for both normal and diseased cells/tissues associated
with the drug target. While different hCA isozymes have a
high degree of active site conservation, the isozymes have
variable tissue distribution, subcellular locations, and expres-
sion profiles,1 and it is these differentiating properties that
underpin added opportunities for medicinal chemistry pro-
grams to target CA isozymes selectively. This strategy should
Carbonic anhydrases (CAs,a EC 4.2.1.1) are zinc metal-
loenzymes that catalyze the reversible hydration of carbon
dioxide to give bicarbonate and a proton: CO2 þ H2O /
-
HCO3 þ Hþ. This reaction underpins cellular processes
-
associated with respiration and transport of CO2 and HCO3
,
-
the provision of HCO3 for biosynthesis, the regulation of
physiological pH, the secretion of electrolytes and fluids, as
well as bone resorption and calcification.1 In humans. 12
catalytically active CA isozymes (designated hCA, h =
human) have been characterized and all 12 share a conserved
active site topology that includes a tetrahedral Zn2þ cation
coordinated to the side chains of three histidine residues,
Figure 1. The Zn2þ core of CAs serves an essential function:
it is a strong Lewis acid that binds to and activates the
substrate water molecule. Hydration of CO2 does not proceed
at an appreciable rate under physiological conditions in the
absence of CA.1
The modulation of CA activity has been validated as an
avenue for the treatment of a range of acquired and inherited
diseases.1 Given the high degree of active site conservation
among isozymes of CA, the enzyme class does however
present a significant challenge as a drug discovery target.
hCA isozymes I and II are widely distributed and have been
*To whom correspondence should be addressed. Phone: þ61 7 3735
7825. Fax: þ61 7 3735 7656. E-mail: s.poulsen@griffith.edu.au.
a Abbreviations: CA, carbonic anhydrase; TPM, topiramate; Pe,
effective permeability; pHe, extracellular pH; pHi, intracellular pH;
PAMPA, parallel artificial membrane permeability assay; PVDF, poly-
vinylidene fluoride; DOPC, dioleoyl-3-phosphocholine; c LogP, calcu-
lated Log P; TPSA, total polar surface area; ELSD, evaporative light
scattering detection.
r
2009 American Chemical Society
Published on Web 09/30/2009
pubs.acs.org/jmc

6422 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 20
Lopez et al.
limit polypharmacology and result in therapeutics with im-
proved tolerability, safety, andefficacy.2 A successful example
of this approach was the pursuit and clinical launch of
aromatic sulfonamides as topically acting antiglaucoma
agents, namely dorzolamide (Merck & Co., 1995) and brin-
zolamide (Alcon Laboratories, 1998), Figure 3.3,4 These
compounds are potent inhibitors of hCA II within the ciliary
processes of the eye and are used clinically to reduce intra-
ocular pressure, which is the key physiological dysfunction
presenting in patients with glaucoma. Topical administration
avoids many of the side effects associated with earlier used
orally administered CA inhibitor-based drugs by selectively
targeting CAs in the eye tissue. The success of these com-
pounds in the market is testament to the principle that
efficacious CA inhibitors can provide safer and well tolerated
treatments when they act with specificity.
The historically broad pharmacological footprint of CA
inhibitors together with our modern day understanding of the
chemical biology of CA isozymes underpins opportunities for
the development of new CA-based therapies. An important
recent finding in CA research has been the validation of hCA
IX and XII as therapeutic targets for cancer chemotherapy in-
tervention.5,6 A phenotype of solid tumors is an acidic micro-
environment (low pHe) caused by heightened metabolism that
leads to an elevated level of both secreted lactic acid and CO2,
which when combined with poor clearance (owing to inade-
quate vasculature around the solid tumor) leads to increased
extracellular acidity.6,7 While low pHe promotes tumor growth
and development, a variation to intracellular pH (pHi) represents a
severe threat to cancer cell survival through disruption of critical
biological functions.6,7 Cancer cell survival thus depends on both
low pHe with a concomitant need for (normal) physiological pHi.
ꢁ
Pouyssegur and colleagues recently demonstrated that expression
of CA IX and XII grants hypoxic tumor cells a survival advantage,
as these enzymes regulate and maintain pHi levels that support
cellular functions.6 Transcriptional activation of the ca9 and ca12
genes in hypoxic tumor cells is regulated by hypoxia-induced
factor 1 (HIF-1).8 The results of in vivo studies in mice showed
that silencing of ca9 and ca12 genes significantly reduced the rate
of xenograft growth. A 40% reduction in xenograft tumor volume
occurred with ca9 silencing and a dramatic 85% reduction with
invalidation of both ca9 and ca12; this outcome is attributed to
slowed proliferation in the absence of extracellular CA.6 Several
research groups have provided validation that specific targeting of
CA IX (or IX and XII) may lead to an effective anticancer therapy,
particularly for the benefit of patients with hypoxic tumors.5-7 As
a candidate for drug intervention CAs possess two discriminating
attributes that provide opportunities for selective targeting with
small molecule inhibitors. First, hCA IX expression in hypoxic
tumors occurs within tissues that normally lack this isozyme such
as carcinomas of the lung, breast, esophagus, cervix, and kidney.9
Second, unlike the physiologically dominant cytosolic hCA I and
hCA II, hCA IX and XII are transmembrane proteins with an
extracellular enzyme active site. It has been demonstrated already
that hCA IX and XII display high affinity for primary aryl and
heteroaryl sulfonamides10,11 and that their inhibition can reverse
the acidification of the tumor microenvironment and retard tumor
growth.9 The validation of these extracellular domains as cancer
therapy targets has flagged an urgent need for the development of
cell impermeable CA inhibitors that specifically act on these
Figure 1. Schematic of the CA active site catalytic cycle that is
common to all catalytically active human CA isozymes. The zinc
cation is an excellent Lewis acid and is bound to three histidine
residues and the substrate water molecule. This Zn2þcoordinated
water has a pKa of ∼7.1 CA thus facilitates deprotonation of water
at physiological pH to generate the strongly basic hydroxide anion
Figure 3. Dorazolamide and brinzolamide are topically adminis-
tered CA inhibitors that are used clinically to lower intraocular
pressure in patients with glaucoma.
(OH-), which is the reactive species in the hydration of CO2 leading
-
to formation and release of HCO3
.
Figure 2. (a) Systemically administered CA inhibitors. All have a primary sulfonamide group as the zinc binding function. (b) CA active site
schematic of sulfonamide mediated inhibition (“canonical binding mode”).

Article
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 20 6423
membrane-associated CAs for use as tools or lead molecules for
cancer therapy and/or diagnostics discoveries.6,7
Results and Discussion
Chemistry. A set of 10 S-glycosyl primary sulfonamides
1-10 (Figure4) weresynthesized from per-O-acetylated sugar
derivatives as outlined in Scheme 1 and described by us
previously.21 Our general approach to the synthesis consists
of first introducing a thioacetate group at the anomeric center
of a per-O-acetylated sugar derivative. From this follows
formation of a NH-dimethoxybenzyl (DMB) glycosyl sulfe-
namide, oxidation of this sulfenamide to give a glycosyl
NH-DMB sulfonamide, and removal of the DMB sulfonam-
ide protecting group to yield a primary sulfonamide at the
anomeric center. Target sugars were prepared either as per-
O-acetylated (1-5) or fully deprotected (6-10) anomeric
sulfonamides. The chemistry was applied to a variety of
monosaccharides: D-glucose (f1 and 6), D-galactose (f2
and 7), and D-glucuronic acid methyl ester (f3 and 8) and
disaccharides lactose (f4 and 9) and maltose (f5 and 10).
Carbonic Anhydrase Inhibition. Enzyme inhibition data
was determined for physiologically dominant hCA I and II
and cancer-associated hCA IX and XII by assaying the CA
catalyzed hydration of CO2.23 Inhibition and isozyme selec-
tivity ratio data for the anomeric sulfonamides 1-10 as well
as for the standard CA inhibitor acetazolamide, the sugar
sulfamate drug topiramate, and the aliphatic sulfonamide
drug zonisamide are presented in Table 1. The structures for
topiramate and zonisamide are in Figure 5.
In the past few years, our group has developed a medicinal
chemistry program to differentiate inhibition of transmem-
brane CAs from cytosolic CAs.12-17 Our approach has been
to tether a carbohydrate moiety to the high-affinity aromatic
sulfonamide CA pharmacophore, leading to sulfonamide
glycoconjugates with
a sugar-aromatic-SO2NH2 motif.
Mounting evidence supports the premise that sugar-based
molecules have value as drug candidates, as they often present
desirable activity and safety profiles.18-20 The -SO2NH2 zinc
binding function of the glycoconjugates, as for other CA inhibi-
tors, facilitates a critical role in enzyme recognition and is essential
for efficacy. The role of the sugar is however to facilitate selective
or preferential inhibition of transmembrane CAs over cytosolic
CAs, as this hydrophilic moiety impairs the ability of the sulfon-
amide glycoconjugates to passively diffuse through lipid mem-
branes.12 The stereochemical and structural variability of the
carbohydrate moiety also provides the opportunity for interroga-
tion ofsubtledifferencesofactive sitearchitecture amongdifferent
CA isozymes. Our approach thus provides neutral and water-
soluble CA inhibitors that have excellent potential as isozyme
selective inhibitors as a consequence of tuning physicochemical
properties.12-17
Recently, one of our groups reported the synthesis of
S-glycosyl primary sulfonamides, a new class of compound
wherein a primary sulfonamide moiety is directly attached to
the anomeric center of a mono- or disaccharide fragment.21
Therein we disclosed novel synthetic methodology to access a
new chemical class that had so far proven elusive to chemical
synthesis. At the same time as our paper was published,
The majority of clinically relevant CA inhibitors possess a
flat aromatic moiety with a bland stereochemical profile.1
In this type of inhibitor, the sulfur atom of the primary
ꢁ
Somsak reported an alternate synthesis to similar com-
pounds.22 As new chemical entities that incorporate the
classical zinc binding primary sulfonamide function coupled
directly to a sterochemically rich carbohydrate scaffold
(that is, without an intervening aromatic moiety), we were
interested in investigating this novel sugar-SO2NH2 motif for
CA inhibition. Herein we present the inhibition profile of a
panel of anomeric sulfonamides (compounds 1-10) against
the cancer associated (hCA IX and XII) and physiologically
dominant (hCA I and II) carbonic anhydrase isozymes.
The novelty of the anomeric sulfonamide structural motif
together with the inhibition data acquired has led us also to
pursueprotein X-ray crystallography toidentify the structural
features of these molecules within the active site of hCA II
that are important for the observed enzyme inhibition
profile. We also investigate the membrane permeability
properties of these carbohydrate-based CA inhibitors to
evaluate the potential of the lipid membrane barrier
as a means toward selectively targeting small molecule CA
inhibitors to the extracellular cancer-associated isozymes IX
and XII.
Figure 4. Target anomeric sulfonamide compounds: per-O-acety-
lated derivatives (1-5) and fully deprotected derivatives (6-10).
Scheme 1. Synthesis of S-Glucosyl Primary Sulfonamides 1 and 6 from per-O-Acetylated D-glucosea
a Compounds 2-4 and 7-10 were synthesized similarly from per-O-acetylated sugar precursors.21

6424 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 20
Lopez et al.
Table 1. Inhibition and Isozyme Selectivity Ratio Data of hCA Isozymes I and II (Cytosolic), IX and XII (Tumor-Associated) with Anomeric
Sulfonamides 1-10, and the Established Drugs and CA Inhibitors Acetazolamide, Topiramate, and Zonisamide
Ki (μM)a
selectivity ratiosb
compd
hCA Ic
hCA IIc
hCA IXd
hCA XIId
hCA I/ hCA IX
10
4.3
11.2
1.16
hCA II/ hCA IX
acetazolamide
topiramate
zonisamide
0.25
0.25
0.056
4.53
3.92
4.34
4.68
4.12
3.90
3.93
3.71
3.97
4.15
0.012
0.005
0.035
4.50
4.73
4.35
4.51
4.98
4.91
4.55
3.90
3.92
4.10
0.025
0.058
0.005
3.91
4.09
4.15
4.03
4.08
4.05
4.19
3.95
4.19
4.22
0.0057
0.0038
11
0.48
0.09
7.0
1
2
3
4
5
6
7
8
9
10
4.66
4.75
4.71
4.60
4.61
4.69
4.80
4.55
4.77
4.84
1.15
1.16
1.05
1.12
1.22
1.21
1.09
0.99
0.94
0.97
0.96
1.05
1.16
1.01
0.96
0.94
0.94
0.95
0.98
a Errors in the range of (5-10% of the reported value from three determinations. b The Ki ratios are indicative of isozyme selectivity in vitro. c Human
(cloned) isozymes. d Catalytic domain of human (cloned) isozymes.
Acetazolamide and topiramate exhibit similar CA inhibi-
tion profiles. Both are low nanomolar inhibitors of hCA II
and XII (Kis 3.8-12 nM), slightly weaker inhibitors of hCA
IX (Kis 25-58 nM), and less potent inhibitors of hCA I (Kis
both 250 nM). Zonisamide is a very potent inhibitor of hCA
IX (Ki of 5.1 nM), a mid-potency inhibitor of hCA I and II
(Kis 35-56 nM), and a weak inhibitor of hCA XII (Ki of
Figure 5
11 μM). In contrast to standard CA inhibitors, the inhibition
profile for the anomeric sulfonamide compounds 1-10 is flat
sulfonamide zinc anchor group is directly bonded to a sp2
hybridized carbon atom, Figure 2a. Sulfonamide derivatives
such as the isosteric sulfamates and sulfamides are also well-
known CA inhibitors.24,25 In these compounds, the sulfon-
amide sulfur atom is attached to an electronegative atom:
oxygen for sulfamates (R-O-SO2NH2)26 and nitrogen for
sulfamides (R-NH-SO2NH2).27-29 Topiramate, a fructopyra-
nose sulfamate, is a billion dollar drug that is marketed world-
wide for the treatment of epilepsy and migraine.30,31 This
compound has been shown to be a good inhibitor of CAs in
vitro by a stopped flow technique monitoring the physiological
reaction catalyzed by these enzymes, i.e., CO2 hydration to
bicarbonate and protons.25 The chemical structure of this carbo-
hydrate-based sulfamate (Figure 5) is particularly notable from
the point of view that its structure is in stark contrast to the bland
sterochemical motif of many known CA inhibitors and hence
draws some interesting parallels with the new chemical entities
presented in this study, and we will discuss these further later in
this manuscript. There are also examples, albeit more limited, of
sulfonamides as CA inhibitors wherein the sulfonamide sulfur is
directly attached to a sp3 hybridized carbon to give a methylene
sulfonamide motif (R-CH2-SO2NH2). These compounds also
present as potent CA inhibitors, and an example of this inhibitor
class is the antiepileptic drug zonisamide, Figure 5.32 Both
topiramate and zonisamide have a complex biological effect,
which may or may not include CA inhibition in the mechanism of
action. Irrespective, these compounds are structurally novel CA
inhibitors when compared to the typical aromatic sulfonamides,
and as clinically used drugs they have passed the many hurdles
that small molecules face through the drug development pipeline,
and hence may serve as a valuable comparative chemical tools
for the current study of anomeric sulfonamides. To the best of our
knowledge, there are no compounds with a methine sulfonamide
motif (RR0-CH-SO2NH2) for which a CA inhibition response is
available;the anomeric sulfonamides herein belong in this
unreported class.
and there is no variation in the Ki values observed across all
CA isozymes investigated, leading to Ki selectivity ratios of
∼1.0 across isozymes, i.e., they are nonselective CA inhibi-
tors. Neither the stereochemistry presented by the differing
sugar moiety nor the nature of the sugar hydroxyl groups,
either as the polar free sugar (sugar-OH, 6-10) or less polar
and bulkier acetylated sugar (sugar-OAc, 1-5) impacted to
alter enzyme inhibition characteristics. At the physiologi-
cally dominant isozymes Kis were 3.71-4.68 μM at hCA I
and 3.90-4.98 μM at hCA II, while at the cancer-associated
isozymes, Kis were 3.91-4.22 μM at hCA IX and 4.55-
4.84 μM at hCA XII. This flat SAR and micromolar CA
inhibition represents an anomaly in our drug discovery
program wherein to date the “norm” has been to readily
prepare CA inhibitors that display nanomolar inhibition
constants similarly to many known sulfonamide-based CA
inhibitors. These atypical inhibition results prompted us to
turn our attention to a structure-based investigation of the
anomeric sulfonamides with hCA II. Using this structure-
based approach, we planned to discern the interactions of
our compounds with active site residues of hCA II so as to
identify the structural features of these ligands that are
responsible for the weaker micromolar inhibition observed.
The sugar sulfamate topiramate has a structural motif that is
related to that of the anomeric sulfonamides 1-10, as each
are based on natural product sugar scaffold, a zinc binding
function, and lack an aromatic moiety. Topiramate is a good
hCA II inhibitor (Ki of 5 nM), whereas for the anomeric
sulfonamides 1-10, Kis were 3.90-4.98 μM, i.e., approxi-
mately 3 orders of magnitude weaker inhibitors than topir-
amate. We specifically targeted protein X-ray crystallo-
graphy to enable comparison of the active site binding
attributes of the new ligands with those of topiramate in
the hCA II active site. We also attempted to find clues as to
potential ligand interactions with hCA IX using a homology
model of the catalytic domain of this membrane-bound

Article
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 20 6425
Table 2. Data Collection and Structure Refinement Statistics of Ligand-Bound hCA II Crystal Structures
ligand
data set
4
hCA II:4
7
hCA II:7
10
hCA II:10
topiramate
hCA II:TPM
Data Collection
AS-PX1
0.9568
X-ray source
wavelength (A)
AS-PX1
0.9568
P21
in-house
1.5418
P21
in-house
1.5418
P21
˚
space group
cell parameters (A; deg)
˚
resolution (A)
P21
˚
42.4, 41.7, 72.5; 104.5
1.8
7.5 (7.3)
42.5, 41.7, 72.5; 104.5
1.8
7.5 (7.5)
42.4, 41.7, 72.5; 104.9
2.4
3.5 (3.1)
42.3, 41.4, 72.4; 104.6
1.8
3.3 (3.2)
multiplicity
completeness (%)
Rmerge
99.5 (96.9)
0.055 (0.287)
99.9 (100)
0.062 (0.558)
99.3 (95.5)
0.053 (0.160)
99.8 (98.7)
0.033 (0.058)
Structure Refinement
no. of water molecules
2
average B actor (A )
93
23.8
87
24.2
164
14.9
354
16.6
˚
˚
rms bond lengths (A)
rms bond angles (deg)
0.005
1.436
2.9
0.005
1.340
2.5
0.018
1.782
1.6
0.013
1.368
1.5
2
˚
rms bonded B factors (A )
Ramachandran plot (%)a
R factor
88.4, 11.1, 0.5, 0
0.196
0.220
88.0, 11.6, 0.05, 0
0.200
0.220
84.9, 14.2, 0.9, 0
0.154
0.247
88.0, 11.5, 0.5, 0
0.153
0.198
Rfree
Ligand
28.2
n/a
2
average B factor (A )
˚
59.5
49.9/69.5
33.5
22.3/34.1
12.2
n/a
2
hexose ring 1/ring 2 (A )
˚
a Residues in most favored regions, additional allowed regions, generously allowed regions, disallowed regions.
Table 3. Comparison of Active Site Residue Configuration of Apo- and
Ligand-Bound Structures of hCA IIa
enzyme. Because of the high degree of conservation of
residues within the active site, no significant differences in
ligand-protein interactions could be identified. Currently,
there is no crystal structure available for hCA IX, however
recently a structure of a hCA IX active site mimic that is a
double mutant of hCA II was reported.33
Structures of hCA II:Ligand Complexes. To obtain insights
into ligand-protein interactions at the atomic level, we
attempted determination of crystal structures of human
recombinant CA II in complex with ligands 1-10 as well
as topiramate. Using cocrystallization as well as soaking
techniques, not all ligands were found to be present in the
crystals examined. With cocrystallization, crystal structures
of hCA II in complex with one monosaccharide (7), two
disaccharides (4 and 10), and topiramate were obtained.
Data collection and refinement statistics are presented in
Table 2.
The vast majority of publications on drug discovery
targeting hCA II annotate residue numbers in the active site
that are offset by one relative to the amino acid sequence of
hCA II. This may be a consequence of the numbering scheme
used in the early hCA II crystal structures (PDB codes 1CA2,
2CBB), which skip residue number 126. In this study, we
have applied a residue numbering scheme that is in accor-
dance with the amino acid sequence of hCA II.
As expected, the binding mode of the sulfonamide (or
sulfamate) moiety to the catalytic zinc ion is invariant, and
all four ligands share the binding mode depicted in Figure 2b.
The crystal structure of hCA II:topiramate is in excellent
agreement with the previously published structure of this
ligand in complex with hCA II.25 All ligands investigated in
this study occupy space in the active site cleft of hCA II
without altering the configuration of residues lining the
active site (see Table 3, Figure 6). Notably, even topiramate
with its rather oblate shape is not causing a significant
change in configuration of active site residues.
reference structure
apo-hCA II (PDB: 1CA2)
˚
positional rms (A)
hCA II:TPM
positional rms (A)
˚
comparison structure
hCA II:4
hCA II:7
0.151
0.133
0.177
0.244
0.266
0.238
0.233
hCA II:10
hCA II:TPM
a The positional root mean square deviation of 22 residues lining the
active site was calculated for all 184 atoms using the program Align.34
the side chains of Asn62, Gln92, and Thr199 through direct
hydrogen bonds, while Ala65 and Phe130 provide hydro-
phobic contacts to the aliphatic parts of the two isopropy-
lidene moieties, Figure 7, Table 4. The crystal structures of
the three anomeric sulfonamide ligands reveal the conserva-
tion of a hydrogen bond with the side chain of Thr199 and a
hydrophobic contact with Leu197, Figure 7 and Table 4.
Despite the complexity of disaccharides 4 and 10, these
ligands are not able to form as many direct interactions with
protein residues as topiramate. Compounds 10 and 7 show
indirect interactions of their free hydroxyl groups with
protein side chain residues through a network of ordered
water molecules. Poor electron density was observed for the
distal hexose ring of disaccharides 4 and 10 as well as for
several of the acetyl moieties of 4, indicating that the ligands
are not well ordered in the active site cleft. This observation is
also consistent with the thermal displacement factors of the
ligand atoms of data sets hCA II:4 and hCA II:10, Table 2.
For the data sets reported here, there was no ambiguity as to
the presence of ligands in the active site cleft because the
initial electron density clearly allowed us to distinguish
between cases where water, sulfate anion, or sulfonamide
was bound to the catalytic zinc ion, Figure 8.
Both topiramate and the anomeric sulfonamides of
the kind presented in this study possess a central carbohy-
drate skeleton and present various functional groups that
In addition to the canonical sulfamate interactions
(Figure 2b), the topiramate sugar moiety also interacts with

6426 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 20
Lopez et al.
Figure 6. Comparison of ligand and active site residue configurations. The four ligand-bound structures of hCA II determined in this study (4,
green; 10, pale green; 7, blue; topiramate, red) are superimposed on the structure of apo-hCA II (PDB: 1CA2; orange). The 22 residues lining the
active site of hCA II are shown as wireframe, ligands are drawn as stick models. The stereo figure was prepared with PyMOL.35
Figure 7. Schematic view of ligand interactions with protein residues in hCA II:ligand crystal structures. The canonical sulfonamide and
sulfamate interactions (Figure 2b) are not shown for the purpose of clarity.
Table 4. Summary of Ligand Interactionswith Protein Residues in hCA
II:Ligand Complex Crystal Structuresa
ligand
4
10
7
topiramate
direct hydrogen bonds
H2O hydrogen bonds
vdW interactions
1
0
3
1
1
3
1
2
1
4
0
2
a The canonical sulfonamide and sulfamate interactions are not
included.
potentially could interact with protein residues. A key struc-
tural difference we have identified between the anomeric
sulfonamides and topiramate is the overall shape of these
molecules. Topiramate presents a “T” shaped ligand with the
sugar moiety and its two isopropylidine hydroxyl protecting
groups neatly plugging the hCA II active site anchored by the
Figure 8. Initial difference electron density of hCA II:7 contoured
at 1.8σ. The density features allowed unambiguous placement of the
hexose ring. Stereo figure prepared with PyMOL.35

Article
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 20 6427
Figure 9. Comparisonofligandshapes. Shape complementarity between ligands and activesiteofhCA II. The ligandatoms are shown asspheres
and the protein surface iscolored byelectrostatic potential. Upper panel figures preparedwith AstexViewer.36 All ligands oflower panel are shown
in their superimposed, protein-bound conformations and rendered as sphere models. Lower panel figures prepared with PyMOL.35
zinc binding sulfamate group. The diisopropylidine sugar
moiety effectively spans the diameter of the conical shaped
˚
active site cleft at a height of about 5 A above the catalytic
edge about passive diffusion across biological membrane
37-41
barriers is crucial. We13-17 and others
have adopted a
strategy to deliberately manipulate physicochemical properties
so as to impart membrane impermeability into the CA small
molecule inhibitor. So far, few studies have interrogated the
effectiveness of this design strategy with a quantitative assess-
ment of membrane permeability.38,39
zinc ion, enabling a total of six direct interactions with the
protein (Table 4, Figure 9). In contrast, the anomeric sulf-
onamide compounds tested in this study present an elon-
gated rather than “T” shaped ligand. The sugar moieties of
these ligands do not provide sufficient transverse bulk to
span the active site cleft as for topiramate, Figure 9. This
shape difference leads to a high degree of flexibility of ligand
conformations in the protein-bound state and to less than
optimal active site occupancy, exemplified through weaker
and fewer interactions of the inhibitor with the enzyme active
site, Table 4 and Figure 9.
Comparison with hCA IX Active Site. Using a homology
model of hCA IX (90-341), potential interactions of com-
pounds 4, 7, 10, and topiramate were assessed. The constitu-
tion of the active site of hCA IX is highly similar to that of
hCA II as far as amino acid sequence identity is concerned.
Comparing the two active sites, only eight residues are sub-
stituted, mostly in a conservative fashion: Glu69/Thr155,
Leu60/Arg146, Ile91/Leu173, Ala65/Ser151, Asn67/Gln153,
Phe130/Val212, Val134/Leu216, Leu203/Ala287. Out of
these, only Ala65/Ser151 and Asn67/Gln153 are of practical
interest because the other substitutions occur in the periph-
ery and at the top of conical active site cleft.33 The substitu-
tion of Asn67 with Gln153 in case of hCA IX does not
introduce new functionality in this area but may allow
interactions with a ligand situated in the center of the cleft
due to spacer arm being longer by one methylene unit.
Ser151, substituting for Ala65 of hCA II, may be more
within the interaction radius of the core structure of a ligand.
When comparing the structures of 4, 7, and 10 bound to hCA
II with the homology model of hCA IX by means of super-
position, it is obvious that none of the compounds explores
the active site space to an extent that would utilize interac-
tions with Ser151 or Gln153. Interestingly, when superim-
posing our hCA II:topiramate structure with the hCA IX
homology model, one of the isopropylidene units of this
ligand clashes with Ser151, which may explain the lower
inhibition of topiramate for hCA IX (see Table 1).
In general, passive diffusion across biological membrane
barriers is related to the lipophilicity of the small molecule
and permeability generally decreases with increasing polarity
or capacity for hydrogen bonding. Log P represents intrinsic
lipophilicity and compounds with Log P < 0 typically have
good solubility but poor lipid bilayer permeability (although
paracellular permeability may be a contributor for very small
hydrophilic molecules).42 In this study, we have calculated
the theoretical cLog P for the anomeric sulfonamides as well
as the standard CA inhibitors acetazolamide, topiramate,
and zonisamide, Table 5. The anomeric sulfonamides 6-10
are hydrophilic, with cLog Ps that range from -2.8 for the
monosaccharides (four free hydroxyl groups) to -5.0 (disac-
charides with seven free hydroxyl groups). The cLog P values
for the acetylated sugars 1-5 are consistent with the incorpo-
rated acetate groups, decreasing the polarity of the resulting
sugar moiety, with cLog Ps that range from -0.6 for the
monosaccharides (four acetate groups) to -0.2 for disac-
charides (seven acetate groups). Although compounds 1-5
are several log units less negative than 6-10, their cLog P
values still fall within the range indicative of molecules with
poor membrane permeability, i.e., Log P < 0.
Another small molecule property that has been shown to
be a descriptor for characterizing lipid membrane barrier
diffusion is topological polar surface area (TPSA). Mole-
2
˚
cules with a TPSA greater than approximately 140 A are
likely to have a low capacity for penetrating cell membranes,
2
while those with PSA e 60 A typically have good passive
˚
permeability properties.43 The calculated TPSA for the free
2
sugar disaccharides 9 and 10 is 229 A , which is well above
˚
the upper limit for good passive membrane diffusion, while
for monosaccharides 6-8, the calculated values are some-
2
˚
what lower (150-167 A ) but are also above the upper
2
˚
140 A limit, Table 4. For the acetylated compounds 1-5,
the calculated TPSAs range from 174 A (monosaccharide
2
˚
Passive Membrane Permeability. In the development of
anticancer compounds that target hCA IX and XII knowl-
2
sulfonamide) to 271 A (disaccharide sulfonamide), again
˚

6428 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 20
Lopez et al.
Table 5. Physicochemical Properties Relevant to Passive Membrane Diffusion with Anomeric Sulfonamides 1-10 and the Standard CA Inhibitors
Acetazolamide and Topiramate
a cLog P and TPSA data calculated using ChemBioDraw Ultra 11.0. b Value is represented as mean ( SD (n = 4 wells). c The concentration of test
compound in the acceptor chamber was lower than the LLQ. Upper limits for the Pe values were estimated using the respective analytical LLQ values as
the maximum concentration in the acceptor chamber. d During the course of the experiment, some degradation (∼35%) of the compound occurred.
well above the threshold for good passive diffusion,
Table 5.
apparent in vitro effective permeability (Pe); high perme-
ability compounds typically have Pe values in excess of 3 ꢀ
10-6 cm s-1, while low permeability compounds typically
Permeability measurements through artificial lipid mem-
brane biolayers have been extensively used to classify small
molecules for their passive membrane permeability charac-
teristics, thereby providing an experimental indicator for
drug permeability that complements cLog P and TPSA
values.44-47 The technique provides a measurement for
have Pe values less than 3 ꢀ 10-6 cm s
.
-1 48,49 The Pe of per-O-
acetylated anomeric sulfonamides (1 and 5), fully deprotected
anomeric sulfonamides (6 and 10), topiramate, acetazolamide,
and control compounds were determined using a parallel
artificial membrane permeability assay (PAMPA).45-47 The

Article
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 20 6429
experimentally determined Pe values for control compounds,
naldolol and verapamil, representing low and high membrane
permeability markers, respectively, were used to confirm suitable
permeability properties of the membranes. The experimental
values of Pe at pH 7.4 using the PAMPA assay are given in
Table 5. The only compound for which a measurable concentra-
tion was observed in the acceptor compartment was topiramate,
with a Pe of 0.54 ((0.04) ꢀ 10-6 cm s-1. On the basis of the
permeability classification criteria, the permeability of topira-
mate across the PAMPA lipid membrane is low or poor. For the
other compounds tested, including acetazolamide and the test
compounds, the concentrations in the acceptor chamber were
below the analytical lower limits of quantitation (LLQ), suggest-
ing that these compounds have very low permeability in this
assay. A theoretical estimate of the maximum Pe value for each
compound was calculated using the respective LLQ values,
Table 4. The PAMPA results for 1, 5, 6, and 10 as representative
anomeric sulfonamides (that include an example of monosac-
charide, disaccharide, sugar-OAc and sugar-OH) confirm that
this class of compound would be expected to have poor passive
membrane permeability. This characteristic of the molecules is
expected to lead to preferential inhibition of the transmembrane
CAs IX and XII over cytosolic CAs, however further in vivo
studies are required to confirm this hypothesis.
and results have provided an extensive insight into the ligand-
active site molecular recognition attributes that will guide the
future development of this class of compound as CA inhibi-
tors. For example, the attachment of appropriate moieties to
modify the overall shape of the ligands to enable more direct
interactions with active site protein residues and enhance
enzyme inhibition while retaining the membrane imperme-
ability and drug lead-like properties of the compounds pre-
sented will be undertaken.
Experimental Section
Chemistry. The syntheses of compounds 1-10 were prepared
as previously described by ourselves.21 All reagents were pur-
chased from Sigma-Aldrich Chemical Company with the excep-
tion of methyl 1,2,3,4-tetra-O-acetyl-β-D-glucopyranuronate,
which was synthesized as described in the literature.50 All
reactions were monitored by TLC using Merck F60254 silica
plates visualized with UV light (λ = 254 nm), sulphuric acid
stain (5% H2SO4 in ethanol), and/or orcinol stain (1 g of orcinol
monohydrate in a mixture of EtOH:H2O:H2SO4 72.5:22.5:5).
Silica gel flash chromatography was performed on Davisil silica
˚
gel 60 A (230-400 mesh). Reverse phase chromatography was
performed using Phenomenex C-18 silica prepacked cartridges
and eluted with a gradient of H2O-MeOH (from 100:0 to
0:100). NMR (1H, 13C {1H}, gCOSY, and HSQC) spectra were
recorded on a Varian 500 MHz spectrometer at room tempera-
ture. For 1H and 13C NMR run in CDCl3, chemical shifts (δ) are
reported in ppm relative to the solvent residual peak: proton
(δ = 7.27 ppm) and carbon (δ = 77.2 ppm). Chemical shifts for
1H and 13C NMR run in DMSO-d6 are reported in ppm relative
to residual solvent proton (δ = 2.50 ppm) and carbon (δ = 39.5
ppm) signals, respectively. For 1H NMR run in D2O, chemical
shifts are reported in ppm relative to the solvent residual peak:
proton (δ = 4.80 ppm). Multiplicity is indicated as follows: s
(singlet), d (doublet), t (triplet), m (multiplet), dd (doublet of
doublet), ddd (doublet of doublet of doublet), br s (broad
singlet). Coupling constants are reported in hertz (Hz). Melting
points are uncorrected. Low resolution mass spectra were
acquired in positive ion mode on an Applied Biosystems Pty
Ltd. Mariner ESI-TOF mass spectrometer. High resolution
mass spectra were performed in positive ion mode on an Apex
III Bruker Daltonics 4.7T Fourier transform mass spectrometer
(FTMS) fitted with an Apollo electrospray ionization source.
All MS analysis samples were prepared as solutions in methanol.
Optical rotations were measured on a Jasco P-1020 polarimeter
at 25 °C with Na-589 nm wavelength and reported as an average
of 10 measurements. The purities of isolated products were
determined by HPLC obtained on an LCMS instrument (MS-
ZQ Waters; HPLC-Alliance Waters) using a HPLC column
(Ascentis, C18 3 μm, 5 cm ꢀ 4.6 mm). Elution was performed
with a gradient of water:methanol (containing 1% formic acid)
from 95:5 to 0:100 for 10 min at a flow rate of 1 mL/min. UV
(200 to 400 nm) and evaporative light scattering detection
(ELSD-Alltech) detection were used. Purity of all compounds
proved to be g95%.
Conclusions
It has been widely acknowledged that physicochemical
properties must be addressed at the onset of a drug discovery
campaign so that early candidates have enhanced drug-like
profiles. New oncology drugs are desperately needed, and
medicinal chemistry must capitalize on any small molecule
design avenues available so as to develop safe and efficacious
therapeutics that selectively target and kill cancer cells. To
maximize the benefits of future chemotherapies involving CA
inhibition as part of a cancer treatment regime, it will be
essential to develop compounds that specifically target the
extracellular domains of the relevant CAs. Selective inhibition
among isozymes is a challenging hurdle in the drug discovery
process, and for the CA enzyme family, the highly conserved
activesitestructure andtopology has madeitdifficult totarget
subtle isozyme differences by rational drug design. Herein we
have explored an alternate strategy toward CA isozyme
selectivity that takes into account the relevant biological
landscape for both normal and cancerous cells or tissues
associated with CAs. Our approach has been to target the
extracellular domains of CA enzymes involved in cancer by
the design of compound structures and properties that impart
poor membrane permeability. Such compounds may prove to
be desirable for chemotherapy if delivered through a route of
drug administration other than oral.
Specifically, we have presented a new class of CA inhibitor
comprising of S-glycosyl primary sulfonamides (1-10). The
aromatic moiety of the classical zinc binding sulfonamide CA
inhibitors is absent from these compounds and instead they
incorporate a hydrophilic mono- or disaccharide fragment
directly attached to the sulfonamide group. We have investi-
gated the membrane permeability properties of the carbohy-
drate-based CA inhibitors and confirm that these compounds
have poor passive membrane permeability and hence are
positioned to selectively target extracellular CA active sites
in the in vivo context. Protein X-ray crystallography struc-
tures for three anomeric sulfonamides as well as the sugar
sulfamate, topiramate, wereobtained incomplex with hCAII,
Carbonic Anhydrase Catalytic Inhibition Assay. An SX.
18MV-R Applied Photophysics stopped-flow instrument has
been used for assaying the CA I, II, IX, and XII CO2 hydration
activity.23 Phenol red (at a concentration of 0.2 mM) has been
used as indicator, working at the absorbance maximum of 557
nm, with 10 mM Hepes (pH 7.5) as buffer, 0.1 M NaClO4 (for
maintaining constant the ionic strength-this anion is not
inhibitory), following the CA-catalyzed CO2 hydration reaction
for a period of 10-100 s. Saturated CO2 solutions in water at 20
°C were used as substrate. Stock solutions of inhibitors were
prepared at a concentration of 10-50 mM (in the assay buffer)
and dilutions up to 1 nM done with the assay buffer mentioned
above. Inhibitor and enzyme solutions were preincubated to-
gether for 15 min at room temperature prior to assay in order to

6430 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 20
Lopez et al.
allow for the formation of the E-I complex. The inhibition
constants were obtained by nonlinear least-squares methods
using PRISM 3. The curve-fitting algorithm allowed us to
obtain the IC50 values, working at the lowest concentration of
substrate of 1.7 mM), from which Ki values were calculated by
using the Cheng-Prusoff equation. The catalytic activity (in the
absence of inhibitors) of these enzymes was calculated from
Lineweaver-Burk plots and represent the mean from at least
three different determinations. Enzyme concentrations were
10.3 nM for CA I and CA II, 12 nM for CA IX, and 15 nM
for hCA XII. Kinetic parameters and inhibition constants
were calculated as described previously.11,51 Enzymes used here
were recombinant ones, prepared and purified as described
earlier.11,51
Protein X-ray Crystallography. Human recombinant CA II
was expressed in Escherichia coli BL21(DE3) using autoinduc-
tion52 as implemented in our laboratory. Soluble protein was
purified by anion exchange chromatography using Q-Sepharose
resin. The concentrated protein was subjected to cocrystalliza-
tion with ligands, using a selection of ammonium sulfate con-
taining conditions from our in-house factorial collection (2.6 or
2.9 M (NH4)2SO4 buffered with either 0.1 M TRIS pH 8.0, 0.1 M
TRIS pH 8.5, or 0.1 M glycine pH 9.5). Crystallization droplets
were constituted by mixing 3 μL of protein stock solution (18 mg
mL-1 in 100 mM NaCl, 20 mM HEPES, pH 8.0), with 2 μL
reservoir, and 1 μL of ligand solution (60 mM in MeOH)
solution. The final concentration of protein in the droplets
was 9 mg mL-1; the molar ratio of ligand to protein was 33.
X-ray diffraction data was obtained under cryoconditions at the
in-house diffractometer (Rigaku MM007-HF with an R-Axis
IVþþ detector), as well as the Australian Synchrotron beamline
PX1 equipped with a Quantum ADSC CCD detector. Data
were processed using Mosflm53 or XDS54 and the CCP4 suite.55
All crystals belonged to the monoclinic space group P21, and CA
II:ligand complex structures were determined by difference
Fourier techniques using the structure of human CA II (PDB
code 1CA2). The ligand-bound structures were refined using
CNS56 with ligand topologies generated by PRODRG.57 Man-
ual model building and visual inspection was performed using
the programs O58 and Coot.59 Refined structures were checked
for acceptable geometry using PROCHECK,60 and the Rama-
chandran plot showed no residues in the disallowed regions.
Coordinates and structure factors have been deposited with the
PDB (accession codes: 3HKN, 3HKQ, 3HKT, and 3HKU for
hCA II complexes with 4, 7, 10, and topiramate).
assay. The acceptor compartment was filled with 300 μL of
blank transport buffer. Compounds were allowed to permeate
across the artificial membrane for 4 h at 37 °C with 200 rpm
orbital shaking, after which the sandwich was separated and
samples from the donor and acceptor wells transferred into a
fresh 96-well plate for analysis. Control compounds, nadolol
and verapamil, representing low and high membrane perme-
ability markers, respectively, were included in the PAMPA
plate. To monitor for membrane integrity, 20 μM nadolol (a
low permeability marker) was added to all wells concurrently
with the test compound. Compound stability in transport buffer
at 37 °C (4 h) was also assessed. All samples were quenched with
an equal volume of acetonitrile, and the concentrations of test
and control compounds were quantitated by LC-MS using ESI-
positive ion mode. The lower limits of quantitation (LLQ) were
found to be 0.005 μM (1, 5, topiramate), 0.05 μM
(acetazolamide), and 1 μM (6, 10). The effective permeability
coefficient (Pe) of each compound was calculated as described
previously.46
Acknowledgment. This work was financed in part by the
AustralianResearchCouncil(grant no. DP0877554toS.-A.P.
and S.C.), Griffith University, an EU grant of the 6th frame-
work programme (DeZnIT project to C.T.S.), and an EU
grant of the 7th framework programme (METOXIA project
to C.T.S.). We thank A/Prof. Jean-Yves Winum for the
sample of topiramate used in this study. cDNA for hCA II
was kindly provided by Prof. Carol Fierke, University of
Michigan. A.H. acknowledges the award of beam time by
the Australian Synchrotron and on-site support by Trevor
Huyton.
References
(1) For recent reviews of carbonic anhydrases see: Supuran, C. T.
Carbonic anhydrases: novel therapeutic applications for inhi-
bitors and activators. Nat. Rev. Drug Discovery 2008, 7, 168–
181. (b) Krishnamurthy, V. M.; Kaufman, G. K.; Urbach, A. R.; Gitlin,
I.; Gudiksen, K. L.; Weibel, D. B.; Whitesides, G. M. Carbonic
Anhydrase as
a Model for Biophysical and Physical-Organic
Studies of Proteins and Protein Ligand Binding. Chem. Rev. 2008,
108, 946–1051.
(2) Lipinski, C.; Hopkins, A. Navigating chemical space for biology
and medicine. Nature 2004, 855–861.
(3) Marquis, R. E.; Whitson, J. T. Management of glaucoma: focus on
pharmacological therapy. Drugs Aging 2005, 22, 1–21.
(4) Sugrue, M. F. Pharmacological and ocular hypotensive properties
of topical carbonic anhydrase inhibitors. Prog. Retinal Eye Res.
2000, 19, 87–112.
Homology Modeling. Secondary structure elements of hCA
IX were predicted using PSIPRED,61 and a structure-based
sequence alignment of the catalytic subunit of hCA IX with
hCA XII (PDB: 1JD0) was generated manually. Twenty inde-
pendent homology models were computed with MODELER,62
and that with the lowest energy was selected. The overall
geometry was scrutinized with PROCHECK,60 and models
were visually inspected with O.58
(5) Svastova, E.; Hulikova, A.; Rafajova, M.; Zatovicova, M.;
Gibadulinova, A.; Casini, A.; Cecchi, A.; Scozzafava, A.; Supuran,
ꢁ
C. T.; Pastorek, J.; Pastorekova, S. Hypoxia activates the capacity
of tumor-associated carbonic anhydrase IX to acidify extracellular
pH. FEBS Lett. 2004, 577, 439–445.
ꢁ
(6) Chiche, J.; Ilc, K.; Laferriere, J.; Trottier, E.; Dayan, F.; Mazure,
N. M.; Brahimi-Horn, M. C.; Pouyssegur, J. Hypoxia-inducible
ꢁ
Membrane Permeability Studies. The apparent in vitro effec-
tive permeability (Pe) of per-O-acetylated anomeric sulfona-
mides (1 and 5), fully deprotected anomeric sulfonamides (6 and
10), topiramate, and acetazolamide were determined using a
parallel artificial membrane permeability assay (PAMPA).45-47
Standard 96-well PVDF filter plates were coated with a solution
of lipid material (dioleoyl-3-phosphocholine (DOPC)) to pre-
pare the artificial membrane. The filter plate was sandwiched
with a 96-well acceptor plate filled with transport buffer (Hanks
balanced salt solution containing 20 mM HEPES, pH 7.4). The
solubility of each test compound in the transport buffer at the
donor concentration was confirmed prior to the experiment.
For topiramate, acetazolamide, 1, and 5, donor solutions (20
μM) were prepared in transport buffer and quadruplicate
aliquots (200 μL) transferred to the donor wells. For 6 and 10,
a donor concentration of 200 μM was used owing to the poor
analytical sensitivities for these compounds using the LCMS
carbonic anhydrase IX and XII promote tumor cell growth by
counteracting acidosis through the regulation of the intracellular
pH. Cancer Res. 2009, 69, 358–368.
(7) Swietach, P.; Vaughan-Jones, R. D.; Harris, A. L. Regulation of
tumor pH and the role of carbonic anhydrase 9. Cancer Metastasis
Rev. 2007, 26, 299–310.
(8) Wykoff, C. C.; Beasley, N. J. P.; Watson, P. H.; Turner, K. J.;
Pastorek, J.; Sibtain, A.; Wilson, G. D.; Turley, H.; Talks, K. L.;
Maxwell, P. H.; Pugh, C. W.; Ratcliffe, P. J.; Harris, A. L.
Hypoxia-inducible regulation of tumor-associated carbonic anhy-
drases. Cancer Res. 2000, 60, 7075–7083.
ꢁ
(9) Thiry, A.; Dogne, J. M.; Masereel, B.; Supuran, C. T. Targeting
tumor-associated carbonic anhydrase IX in cancer therapy. Trends
Pharmacol. Sci. 2006, 27, 566–573.
(10) Vullo, D.; Franchi, M.; Gallori, E.; Pastorek, J.; Scozzafava, A.;
Pastorekova, S.; Supuran, C. T. Carbonic anhydrase inhibitors.
Inhibition of the tumor-associated isozyme IX witharomatic and
heterocyclic sulfonamides. Bioorg. Med. Chem. Lett. 2003, 13,
1005–1009.

Article
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 20 6431
(11) Vullo, D.; Innocenti, A.; Nishimori, I.; Pastorek, J.; Scozzafava,
ꢁ
cytosolic isozyme II and mitochondrial isozyme V: solution and X-
ray crystallographic studies. Bioorg. Med. Chem. Lett. 2005, 15,
2315–2320.
A.; Pastorekova, S.; Supuran, C. T. Carbonic anhydrase inhibitors.
Inhibition of the transmembrane isozyme XII with sulfonamides;
a new target for the design of antitumor and antiglaucoma drugs?
Bioorg. Med. Chem. Lett. 2005, 15, 963–969.
(33) Genis, C.; Sippel, K. H.; Case, N.; Cao, W.; Avvaru, B. S.;
Tartaglia, L. J.; Govindasamy, L.; Tu, C.; Agbandje-McKenna,
M.; Silverman, D. N.; Rosser, C. J.; McKenna, R. Design of a
carbonic anhydrase IX active-site mimic to screen inhibitors for
possible anticancer properties. Biochemistry 2009, 48, 1322–
1331.
(34) Hofmann, A.; Wlodawer, A. PCSB;a program collection for
structural biology and biophysical chemistry. Bioinformatics
2002, 18, 209–210.
(35) DeLano, W. The PyMOL Molecular Graphics System; 2002; http://
www.pymol.org.
(36) Hartshorn, M. J. AstexViewer: A visualisation aid for structure-
based drug design. J. Comput.-Aided. Mol. Des. 2002, 16, 871–
881.
(37) Scozzafava, A.; Menabuoni, L.; Mincione, F.; Supuran, C. T.
Carbonic anhydrase inhibitors. A general approach for the pre-
paration of water-soluble sulfonamides incorporating polyamino-
polycarboxylate tails and of their metal complexes possessing long-
lasting, topical intraocular pressure-lowering properties. J. Med.
Chem. 2002, 45, 1466–1476.
(38) Vullo, D.; Steffansen, B.; Brodin, B.; Supuran, C. T.; Scozzafava,
A.; Nielsen, C. U. Carbonic anhydrase inhibitors: transepithelial
transport of thioureido sulfonamide inhibitors of the cancer-asso-
ciated isozyme IX is dependent on efflux transporters. Bioorg. Med.
Chem. 2006, 14, 2418–2427.
(39) Scozzafava, A.; Briganti, F.; Ilies, M. A.; Supuran, C. T. Carbonic
anhydrase inhibitors: synthesis of membrane-impermeant low
molecular weight sulfonamides possessing in vivo selectivity for
the membrane-bound versus cytosolic isozymes. J. Med. Chem.
2000, 43, 292–300.
(40) Pastorekova, S.; Casini, A.; Scozzafava, A.; Vullo, D.; Pastorek, J.;
Supuran, C. T. Carbonic anhydrase inhibitors: The first selective,
membrane;impermeant inhibitors targeting the tumor-associated
isozyme IX. Bioorg. Med. Chem. Lett. 2004, 14, 869–873.
(41) Rami, M.; Cecchi, A.; Montero, J.-L.; Innocenti, A.; Vullo, D.;
Scozzafava, A.; Winum, J.-Y.; Supuran, C. T. Carbonic anhydrase
inhibitors: design of membrane-impermeant copper(II) complexes
of DTPA-, DOTA-, and TETA-tailed sulfonamides targeting the
tumor-associated transmembrane isoform IX. ChemMedChem
2008, 3, 1780–1788.
(12) Winum, J.-Y.; Poulsen, S.-A.; Supuran, C. T. Therapeutic applica-
tions of glycosidic carbonic anhydrase inhibitors. Med. Res. Rev.
2009, 29, 419–435.
(13) Singer, M.; Lopez, M.; Bornaghi, L. F.; Innocenti, A.; Vullo, D.;
Supuran, C. T.; Poulsen, S.-A. Inhibition of carbonic anhydrase
isozymes with benzene sulfonamides incorporating thio, sulfinyl
and sulfonyl glycoside moieties. Bioorg. Med. Chem. Lett. 2009, 19,
2273–2276.
(14) Wilkinson, B. L.; Innocenti, A.; Vullo, D.; Supuran, C. T.; Poulsen,
S.-A. Inhibition of carbonic anhydrases with glycosyltriazole
benzene sulfonamides. J. Med. Chem. 2008, 51, 1945–1953.
(15) Wilkinson, B. L.; Bornaghi, L. F.; Houston, T. A.; Innocenti, A.;
Vullo, D.; Supuran, C. T.; Poulsen, S.-A. Carbonic anhydrase
inhibitors: inhibition of isozymes I, II and IX with triazole-linked
O-glycosides of benzene sulfonamides. J. Med. Chem. 2007, 50,
1651–1657.
(16) Wilkinson, B. L.; Bornaghi, L. F.; Houston, T. A.; Innocenti, A.;
Vullo, D.; Supuran, C. T.; Poulsen, S.-A. Inhibition of membrane-
associated carbonic anhydrase isozymes IX, XII and XIV with a
library of glycoconjugate benzenesulfonamides. Bioorg. Med.
Chem. Lett. 2007, 17, 987–992.
(17) Wilkinson, B. L.; Bornaghi, L. F.; Houston, T. A.; Innocenti, A.;
Supuran, C. T.; Poulsen, S.-A. A novel class of carbonic anhydrase
inhibitors: glycoconjugate benzene sulfonamides prepared by
“click-tailing”. J. Med. Chem. 2006, 49, 6539–6548.
(18) Aich, U.; Campbell, C. T.; Elmouelhi, N.; Weier, C. A.;
Sampathkumar, S.-G.; Choi, S. S.; Yarema, K. J. Regioisomeric
SCFA attachment to hexosamines separates metabolic flux from
cytotoxicity and MUC 1 suppression. ACS Chem. Biol. 2008, 3,
230–240.
(19) Elmouelhi, N.; Aich, U.; Paruchuri, V. D. P.; Meledeo, M. A.;
Campbell, C. T.; Wang, J. J.; Srinivas, R.; Khanna, H. S.; Yarema,
K. J. Hexosamin template. A platform for modulating gene
expression and for sugar-based drug discovery. J. Med. Chem.
2009, 52, 2515–2530.
(20) Meutermans, W.; Le, G. T.; Becker, B. Carbohydrates as scaffolds
in drug discovery. ChemMedChem 2006, 1, 1164–1194.
(21) Lopez, M.; Drillaud, N.; Bornaghi, L. B.; Poulsen, S.-A. Synthesis
of S-glycosyl primary sulfonamides. J. Org. Chem. 2009, 74, 2811–
2816.
(42) Kerns, E. H.; Di, L. Drug-Like Properties: Concepts, Structure
Design and Methods: From ADME to Toxicity Optimization, 1st ed.;
Academic Press: London, 2008; pp 43-47.
(43) Palm, K.; Stenberg, P.; Luthman, K.; Artursson, P. Polar molec-
ular surface properties predict the intestinal absorption of drugs in
humans. Pharm. Res. 1997, 14, 568–571.
(44) Kerns, E. H.; Di, L. Drug-Like Properties: Concepts, Structure
Design and Methods:From ADME to Toxicity Optimization; 1st ed.;
Academic Press: London, 2008; pp 287-298.
(45) Kansy, M.; Senner, F.; Gubernator, K. Physicochemical high
throughput screening: parallel artificial membrane permeation
assay in the description of passive absorption processes. J. Med.
Chem. 1998, 41, 1007–1010.
ꢁ
ꢁ
(22) Czifrak, K.; Somsak, L. Synthesis of anomeric sulfonamides and
their behavior under radical-mediated bromination conditions.
Carbohydr. Res. 2009, 344, 269–277.
(23) Khalifah, R. G. T. The carbon dioxide hydration activity of
carbonic anhydrase. J. Biol. Chem. 1971, 246, 2561–2573.
(24) Winum, J.-Y.; Scozzafava, A.; Montero, J.-L.; Supuran, C. T. New
zinc binding motifs in the design of selective carbonic anhydrase
inhibitors. Mini Rev. Med. Chem. 2006, 6, 921–936.
(25) Casini, A.; Antel, J.; Abbate, F.; Scozzafava, A.; David, S.;
Waldeck, H.; Schafer, S.; Supuran, C. T. Carbonic anhydrase
inhibitors: SAR and X-ray crystallographic study for the interac-
tion of sugar sulfamates/sulfamides with isozymes I, II and IV.
Bioorg. Med. Chem. Lett. 2003, 13, 841–845.
(46) Avdeef, A. Physicochemical profiling (solubility, permeability and
charge state). Curr. Top. Med. Chem. 2001, 1, 277–351.
(47) Wohnsland, F.; Faller, B. High-throughput permeability pH pro-
file and high-throughput alkane/water log P with artificial mem-
branes. J. Med. Chem. 2001, 44, 923–930.
(26) Winum, J.-Y.; Scozzafava, A.; Montero, J.-L.; Supuran, C. T.
Sulfamates and their therapeutic potential. Med. Res. Rev. 2005,
25, 186–228.
(27) Winum, J.-Y.; Scozzafava, A.; Montero, J.-L.; Supuran, C. T. The
sulfamide motif in the design of enzyme inhibitors. Expert Opin.
Ther. Pat. 2006, 16, 27–47.
(28) Winum, J.-Y.; Scozzafava, A.; Montero, J.-L.; Supuran, C. T.
Therapeutic potential of sulfamides as enzyme inhibitors. Med.
Res. Rev. 2006, 26, 767–792.
(29) Winum, J.-Y.; Temperini, C.; El Cheikh, K.; Innocenti, A.; Vullo,
D.; Ciattini, S.; Montero, J.-L.; Scozzafava, A.; Supuran, C. T.
Carbonic anhydrase inhibitors: clash with Ala65 as a means for
designing inhibitors with low affinity for the ubiquitous isozyme II,
exemplified by the crystal structure of the topiramate sulfamide
analogue. J. Med. Chem. 2006, 49, 7024–7031.
(30) Shank, R. P.; Maryanoff, B. E. Molecular pharmacodynamics,
clinical yherapeutics, and pharmacokinetics of topiramate. CNS
Neurosci. Ther. 2008, 14, 120–142.
(48) Zhu, C.; Jiang, L.; Chen, T.-M.; Hwang, K.-K. A comparative
study of artificial membrane permeability assay for high through-
put profiling of drug absorption potential. Eur. J. Med. Chem.
2002, 37, 399–407.
(49) FDA Guidance for Industry 2000: Waiver of In Vivo Bioavailability
and Bioequivalence Studies for Immediate-Release Solid Oral Do-
sage Forms Based on a Biopharmaceutics Classification System; U.S.
Food and Drug Adminstration, 2000.
(50) Nakajima, R.; Ono, M.; Aiso, S.; Akita, H. Synthesis of methyl-1-
O-(4-hydroxymethamphetaminyl)-R-D-glucopyranuronate. Chem.
Pharm. Bull 2005, 53, 684.
(51) Winum, J.-Y.; Vullo, D.; Casini, A.; Montero, J.-L.; Scozzafava,
A.; Supuran, C. T. Carbonic anhydrase inhibitors: Inhibition of
transmembrane, tumor-associated isozyme IX, and cytosolic iso-
zymes I and II with aliphatic sulfamates. J. Med. Chem. 2003, 46,
5471.
(31) Maryanoff, B. E. Pharmacutical “gold” from neurostabilizing
agents: Topiramate and successor molecules. J. Med. Chem.
2009, 52, 3431–3440.
(32) De Simone, G.; Di Fiore, A.; Menchise, V.; Pedone, C.; Antel, J.;
Casini, A.; Scozzafava, A.; Wurl, M.; Supuran, C. T. Carbonic
anhydrase inhibitors. Zonisamide is an effective inhibitor of the
(52) Studier, F. Protein production by auto-induction in high density
shaking cultures. Protein Expr. Purif. 2005, 41, 207–234.
(53) Leslie, A. Recent changes to the MOSFLM package for processing
film and image plate data. Joint CCP4 þ ESF-EAMCB Newsletter
on Protein Crystallography 1992, 26.

6432 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 20
Lopez et al.
(54) Kabsch, W. Automatic processing of rotation diffraction data
from crystals of initially unknown symmetry and cell constants.
J. Appl. Crystallogr. 1993, 26, 795–800.
(55) Collaborative Computational Project Number 4. The CCP4 suite:
programs for protein crystallography. Acta Crystallogr., Sect. D:
Biol. Crystallogr. 1994, 50, 760-763.
(56) Brunger, A.; Adams, P.; Clore, G.; Delano, W.; Gros, P.; Grosse-
Kunstleve, R.; Jiang, J.; Kuszewski, J.; Nilges, N.; Pannu, N.;
Read, R.; Rice, L.; Simonson, T.; Warren, G. Crystallography and
NMR system (CNS): a new software system for macromolecular
structure determination. Acta Crystallogr., Sect. D: Biol. Crystal-
logr. 1998, 54, 905–921.
(57) Schuettelkopf, A. W.; van Aalten, D. M. F. PRODRG;a tool for
high-throughput crystallography of protein-ligand complexes.
Acta Crystallogr., Sect. D: Biol. Crystallogr. 2004, 60, 1355–1363.
(58) Jones, T. A.; Zou, J. Y.; Cowan, S.; Kjeldgaard, M. Improved
methods for building protein models in electron density maps and
location of errors in these models. Acta Crystallogr., Sect. A:
Found. Crystallogr. 1991, 47, 110–119.
(59) Emsley, P.; Cowton, K. Coot: Model-Building Tools for Molecular
Graphics. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2004, 60,
2126–2132.
(60) Laskowski, R.; MacArthur, M.; Moss, D.; Thornton, J. PRO-
CHECK: a program to check the stereochemical quality of protein
structures. J. Appl. Crystallogr. 1993, 26, 283–291.
(61) Bryson, K.; McGuffin, L. J.; Marsden, R. L.; Ward, J. J.; Sodhi, J.
S.; Jones, D. T. Protein structure prediction servers at University
College London. Nucleic Acids Res. 2005, 33, W36–W38.
(62) Sali, A.; Blundell, T. Comparative protein modelling by satisfac-
tion of spatial restraints. J. Mol. Biol. 1993, 234, 779–815.