Affinity of sulfamates and sulfamides to carbonic anhydrase II isoform: Experimental and molecular modeling approaches

Article abstract of DOI:10.1021/ci100112s

Sixteen aromatic and aliphatic sulfamides and sulfamates were synthesized and tested in their inhibition to carbonic anhydrase CAII activity. The weaker inhibition pattern shown by sulfamides as compared to sulfamates is interpreted in this research by means of molecular modeling techniques, including known inhibitors (topiramate and its sulfamide cognate) in the analysis. The results nicely explain the origin of the inhibitory activity, which is not only related to positive interactions of the ligand with the active site residues but also to the solvation pattern characteristic of each ligand.

Full text of DOI:10.1021/ci100112s

J. Chem. Inf. Model. 2010, 50, 1113–1122
1113
Affinity of Sulfamates and Sulfamides to Carbonic Anhydrase II Isoform: Experimental
and Molecular Modeling Approaches
Luciana Gavernet,*,† Jose L. Gonzalez Funes,† Luis Bruno Blanch,† Guillermina Estiu,‡
Alfonso Maresca,§ and Claudiu T. Supuran§
Medicinal Chemistry, Department of Biological Sciences, Faculty of Exact Sciences, National University of
La Plata, 47 and 115, La Plata B1900BJW, Argentina, Walther Cancer Research Center and Department of
Chemistry and Biochemistry, University of Notre Dame, Notre Dame, Indiana 46556-5670, and Universita`
degli Studi di Firenze, Polo Scientifico, Laboratorio di Chimica Bioinorganica, Rm. 188, Via della Lastruccia
3, 50019 Sesto Fiorentino (Florence), Italy
Received March 23, 2010
Sixteen aromatic and aliphatic sulfamides and sulfamates were synthesized and tested in their inhibition to
carbonic anhydrase CAII activity. The weaker inhibition pattern shown by sulfamides as compared to
sulfamates is interpreted in this research by means of molecular modeling techniques, including known
inhibitors (topiramate and its sulfamide cognate) in the analysis. The results nicely explain the origin of the
inhibitory activity, which is not only related to positive interactions of the ligand with the active site residues
but also to the solvation pattern characteristic of each ligand.
strategies have been used to search for new inhibitors.12-15
INTRODUCTION
Nevertheless, the most potent inhibitors used in clinic
Carbonic anhydrases (CAs) comprise a family of zinc
(acetazolamide, methazolamide, ethoxzolamide, dichlorophe-
metalloenzymes that catalyze the reversible hydration of
namide, dorzolamide, brinzolamide) still present a primary
carbon dioxide to bicarbonate ion.1-6 The assistance for the
sulfonamide group on a benzenoic or heterocyclic structure
rapid interconversion between these two species has an
that binds to the zinc ion through the deprotonated nitrogen
important effect on some vital physiological as well as
atom.4-7 Whereas this group is recognized as the most
pathologic processes. Most of the CA enzymes have been
frequent anchoring group in CAII, other closely related
characterized in detail, and classified in five distinct gene
functions have shown promising inhibitory action. This is
the case of the new generation anticonvulsant drug topiramate
families: R-, ꢀ-, γ-, δ-, and ꢁ-CA families.7
Sixteen R-CA isoforms have been isolated in mammals
(TOP, a sugar sulfamate, Figure 1) as well as several related
with different subcellular localization and tissue distribution.
compounds like its 4,5-cyclic sulfate congener (compound
In all cases, the catalytic site consists of a Zn (II) ion
3, Figure 1, TOP-like).16 TOP shows good activity in CA
tetrahedrally coordinated by the imidazole rings of three
inhibition, at variance with its sulfamide cognate (STOP,
histidine residues (His94, His96, and His119, PDB code
compound 2, Figure 1), which exhibits a weak inhibitory
2H15),7 with the fourth position occupied by a water
action.16 The different activity of sulfamates and sulfamides
molecule at an acidic pH (<8) and by a hydroxide ion at a
in CAII inhibition has been discussed in the literature16 as
higher pH.1-6 Due to the important role of R-CAs in higher
well as the larger potency of aryl than alkyl sulfamides.17
vertebrates, the study of these enzymes represents an
Nevertheless, the differences in the sulfamate/sulfamide CAII
attractive target for the design of inhibitors and activators
inhibitory activity reported for topiramate and related
with therapeutic value. Compounds possessing CA inhibitory
structures exceed the values characteristic of simpler alkyl
properties have been utilized as anticonvulsant, antiurolithic,
and aryl ones, pointing to an influence of the nonbinder
antiglaucoma, and anticancer agents.7 In terms of their
portions of the inhibitor in modulating the activity through
structural characteristics, most of the inhibitors possess a zinc
interactions with protein residues and/or indirectly impacting
binding function capable of substituting the nonprotein zinc
the pKa of the ligand.
ligand.7
Following previous research in this line, we decided to
investigate in some detail the different effects that can contribute
to the trend observed in the sulfamide/sulfamates CAII activity.
To this end, we synthesized and tested several N-substituted
and N,N′-disubstituted alkyl and aryl sulfamides (Figure 1) and
analyzed their interactions in the CAII active site using
molecular modeling techniques. From this set, representative
structures of alkyl and aryl sulfamides as well as their cognate
sulfamates were selected for further analysis. Extensive mo-
lecular dynamics simulations of N-butyl, N,N′-dibutyl, N-benzyl,
N,N′-dibenzyl, N-cyclohexyl and N,N′-dicyclohexyl-sulfamide,
Human CAII (hCAII) is the physiologically most relevant
and widespread CA isoform. It is also the most extensively
studied from structural and inhibitor design points of view
and is a benchmark in molecular dynamics simulations.8-11
Computational three-dimensional (3D) quantitative struc-
ture-activity relationship (QSAR) and virtual screening
* Corresponding author. Telephone: +54 0221-4235333. Fax: +54
02214223409. E-mail: lgavernet@biol.unlp.edu.ar.
† National University of La Plata.
‡ University of Notre Dame.
§ Universita` degli Studi di Firenze.
10.1021/ci100112s  2010 American Chemical Society
Published on Web 05/19/2010

1114 J. Chem. Inf. Model., Vol. 50, No. 6, 2010
GAVERNET ET AL.
Figure 1. Chemical structures of the compounds analyzed in this investigation. 1: topiramate, 2: topiramate sulfamide analogue, 3: topiramate
cyclic sulfate analogue. Sulfamides 4-12 and sulfamate 19 were synthesized and tested by some of us. Inhibitory activity of compounds
13-18 was taken from literature.
STOP, and their cognate sulfamates have been used to analyze
the interactions with residues and solvent waters of the active
site pocket that can modulate the pKa of the ligand and the
consequent strength of the binding.

SULFAMATES AND SULFAMIDES INHIBITORY ACTIVITY
J. Chem. Inf. Model., Vol. 50, No. 6, 2010 1115
EXPERIMENTAL SECTION
We used the default Autodock parameters for all the
variables but the charges of the ligands, for which AM1-
BCC charges were calculated using quacpac.27 We found
this performs better in the docking for this particular system
than the default Gasteiger charges.
The sulfamides/sulfamates were prepared as mentioned in
literature.18-20 Their synthesis, characterization, and iden-
tification have been already reported by some of us in
preceding investigations.18-20 Sulfamides 4-12 were pre-
pared by condensation of an excess of the corresponding
amine with sulfuryl chloride,18,19 whereas monosubstituted
sulfamide 15 was obtained by acidic hydrolysis of its
N-alkoxycarbonyl sulfamide derivative, the N-(benzyl)-N′-
(tert-butoxycarbonyl)sulfamide.19 This derivative was syn-
thesized using chlorosulfonyl isocyanate, tertbutanol, and the
benzylamine in the presence of triethylamine. Sulfamate 19
was prepared by adding formic acid to chlorosulfonyl
isocyanate to generate sulfamoyl chloride, which was then
allowed to react in situ with hexyl alcohol in the presence
of pyridine.
The structures were docked using the Lamarckian genetic
algorithm (LGA) in the “docking active site”, defined through
a grid centered on the N atom of the ASN67 residue of CAII
(coordinates: X ) 3.766, Y ) 1.766, Z ) -9.060), with 60,
50, and 60 grid point in X, Y, Z dimensions, respectively.
We use the default grid spacing (0.375 Å) and performed
50 docking runs, treating the docking active site as a rigid
molecule and the ligands as flexible, i.e., all nonring torsions
were considered active. The compounds were docked in their
deprotonated form when possible. Exceptions were com-
pounds 7, 8, and 11 bearing no H attached to the binding N
and compound 2 for which several protonation states were
analyzed.
Compounds 13-14 and 16-18 were not synthesized.
Their biological information was taken from literature.
CA Inhibition. An applied photophysics stopped-flow
instrument has been used for assaying the CA-catalyzed CO2
hydration activity.21 Phenol red (at a concentration of 0.2
µM) has been used as an indicator, working at the absorbance
maximum of 557 nm, with 20 mM of Hepes (pH 7.4) and
20 mM of NaBF4 (for maintaining constant the ionic
strength), following the initial rates of the CA-catalyzed CO2
hydration reaction for a period of 10-100 s. The CO2
concentrations ranged from 1.7 to 17 µM for the determi-
nation of the kinetic parameters and the inhibition constants.
For each inhibitor, at least six traces of the initial 5-10%
of the reaction have been used for determining the initial
velocity. The uncatalyzed rates were determined in the same
manner and subtracted from the total observed rates. Stock
solutions of inhibitor (10 µM) were prepared in distilled
deionized water, and dilutions up to 0.01 µM were done
thereafter with distilled deionized water. Inhibitor and
enzyme solutions were preincubated together for 15 min at
room temperature prior to assay, in order to allow for the
formation of the E-I complex. The inhibition constants were
obtained by nonlinear least-squares methods using PRISM
3, whereas the kinetic parameters for the uninhibited enzymes
were from Lineweaver-Burk plots, as reported earlier,22,23
and represent the mean from at least three different
determinations.
Computational. Docking. Binding of the compounds,
shown in Figure 1, was analyzed using AutoDockTools 1.5.0
and AutoDock 4.0 docking programs.24 The starting protein
was prepared from the 0.99 Å resolution crystal structure of
the CAII-sulfonamide complex deposited by Christianson
et al (PDB code 2FOU).25 We have chosen the highest-
resolution CAII crystal structure from the Protein Data Bank
(PDB code 2FOU) accepting that CAII possess a low degree
of movement upon the binding of different ligands.14 Only
for the cases of STOP, TOP, and TOP-like, the cocrystallized
structures were used, namely 2H15, 3HKU, and 1EOU,
respectively. In all cases, the crystallographic water mol-
ecules, the ligand, and any cocrystallized molecule/ion were
stripped. Hydrogen atoms were added using the leap module
of AMBER9.26 Special attention was given to the protonation
states of the His residues of the active site that were defined
as HID94, HID96, and HIE119.
For the particular case of STOP the conformational space
of the ligand was explored by a Macromodel28 conforma-
tional search using the Monte Carlo multiple search algorithm
with the OPLS-2005 force field,29 followed by further
structural clustering according to the heavy atoms root-mean-
squared (rms) value (XCluster).30
Molecular Dynamics. The conformations predicted by
Autodock were used as starting points for molecular dynam-
ics (MD) simulations, with the exception of 5 for which the
docking did not give any good pose, in agreement with the
low inhibitory activity observed for this molecule. For that
reason, we built the initial structure using, as a template,
the conformation obtained from the docking of 6. In case of
STOP, TOP, and TOP-like, we use their corresponding
cocrystallized structures as starting points.
MD simulations were carried out using the PMEMD
version included in the AMBER10 suite of programs,31 after
careful relaxation of the system using minimization and
equilibration protocols. The initial geometries were mini-
mized (1000 cycles for the water molecules followed by 2500
cycles for the entire systems). After a 20 ps NTV equilibra-
tion period with a weak restraint (10 kcal/mol Å2) for the
complex and a NTP 200 ps without restraint, production runs
larger than 12 ns were computed for each complex, for the
coordinates saved every 1000 time steps.
The ionizable residues were set to their normal ionization
states at pH 7, except for the HIS residues coordinating the
Zn metallocenter, which were modeled as HID94, HID96,
and HIE119. The protein atoms as well as all the water
molecules of the crystal structure were surrounded by a
periodic box of TIP3P32 water molecules that extended 10
Å from the protein. Na+ counterions were placed by LEaP26
to neutralize the system.
The ff03 version of the all-atom AMBER force field was
used to model the protein, and the GAFF force field was
used for the organic ligand.33 Atom-centered partial charges
were derived by using the AMBER antechamber program
(restrained electrostatic potential [RESP] methodology),34,35
after geometry optimization at the B3LYP/6-31G* level.36
Nonstandard force fields have been derived for the Zn active
site by means of geometry minimization (B3LYP/6-31G**,
G03),37 followed by calculation of the second derivatives
and RESP charges for the active site supermolecule defined
by three His residues, the Zn ion, and the sulfamide ligand.

1116 J. Chem. Inf. Model., Vol. 50, No. 6, 2010
GAVERNET ET AL.
Table 1. Biological Activity, Docking Scores, and Relevant
analogue of TOP, three different protonation states have been
considered: the neutral structure (STOPH), the deprotonated
form (STOPN), and a “zwitterion” conformation that in-
volves a negative coordinating N1 and a positive noncoor-
dinating N2 (STOPZ).
Geometric Data Derived from the Dockinga
compound
Ki (µΜ)
Ki ratio score d1 (Å) d2 (Å) rms fit
1
0.0116
–6.63 1.66
213.5 –4.83 1.43
213.5 –6.84 1.57
3.89
2.51
2.08
1.96
0.546
0.381
0.612
0.427
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
2 (STOPZ) 2.13516
2 (STOPN) 2.13516
3
Autodock is capable of predicting the binding disposition
of STOP, TOP, and TOP-like with rms fit errors from 0.4 to
0.6 (Table 1) relative to the respective crystallographic
structures (2H15 for STOP, 1EOU for TOP-like, and 3HKU
for TOP), which increase our confidence in the predicted
binding conformation of the other compounds (see Support-
ing Information, Figure S1). The error is smaller in the
STOPZ fit as compared to STOPN. On the other hand, no
good coordination was attained for the neutral structure
(STOPH). Figure 2 shows representative binding geometries
for N-substituted sulfamides, N,N′-disubtitued sulfamides and
sulfamates, as they result from the docking. STOP, TOP,
and TOP-like are shown in Figure 3.
The binding reproduces the X-ray determined interactions
of the zinc-binding function, stabilizing a H bond between
the THR199 OG and the H atom of the N1 of the sulfamide
group. In the single N-substituted sulfamides and sulfamates
(compounds 13-19, Figure 1), the interaction is reinforced
by a H bond comprising the backbone NH of the same
residue and the O of the sulfonyl group of the ligand. STOP
also displays similar H-bond interactions with THR199 and
THR200, adding interactions of the O-atoms of the organic
scaffold with ASN62 and GLN92. All of them are well
reproduced by the docking as well as the negative van der
Waals contacts between one methyl group of STOP and it
nearest residues (the side chain of ALA65 and the carboxa-
mide group of ASN67), previously discussed in the literature
as a possible reason for its low inhibitory activity.16 The
same holds for TOP-like and TOP, where the docking
perfectly resembles the interactions with THR199, THR200,
ASN62, and GLN92. The docking procedure also detects
the different orientation of these two related compounds in
the binding site. The molecules are rotated around the
anomeric C-CH2 bond relative to each other, and the 2,3
ring of one ligand has exchanged positions with the 4,5 ring
of the other ligand.
IC50 ) 3616
66.3
100.1
0.65
>200
>200
124
–5.76 1.58
–4.87 1.72
–
–4.86 1.70
–
–
–5.37 1.77
–
–
418
518
618
719
819
919
1019
1119
1219
1319
1419
1519
16
>200
135
>200
0.14817
0.4517
0.123
0.0743
0.0643
0.003443
2.82
–4.6
1.62
2.1 –5.43 1.60
7.5 –5.78 1.58
36.2 –6.64 1.59
–5.39 1.58
2.00
1.98
1.95
3.54
1.99
1.97
1.91
17
–6.3
1.58
NA
NA
NA
18
–6.25 1.59
–5.56 1.60
1920
a The references of the biological data published before as well
the references for the preparation of the compounds are indicated as
superscripts in the table. Ki ratio: Ki of sulfamides relative to the
Ki of its respective cognate sulfamate. NA: x-ray structure not
available. d1: Zn–N1 distances derived from the docking. c- d2:
Calculated N1H-OG1THR199 distances derived from docking.
RMSfit: root mean square for the superimposition of the exper-
imental and the docking conformation.
Different force fields were derived for mono- and disubsti-
tuted sulfamides, considering (His)3Zn-N1H1-SO2-NH2
and (His)3Zn-N1(CH1 )-SO2-NH2 coordination motifs. In
3
the MD simulation protocol, the time step was chosen to be
2 fs, and the SHAKE algorithm38 was used to constrain all
bonds involving hydrogen atoms. A nonbonded cutoff of 10.0
Å was used, and the nonbonded pair list was updated every
25 time steps. Langevin dynamics was used to control the
temperature (300 K) using a collision frequency of 1.0 ps-1,
with isotropic position scaling to maintain the pressure (1
atm).39 Periodic boundary conditions were applied to simu-
late a continuous system. To include the contributions of
long-range interactions, the particle mesh Ewald (PME)40,41
method was used with a grid spacing of ∼1 Å combined
with a fourth-order B-spline interpolation to compute the
potential and the forces in between grid points. The
trajectories were analyzed using the PTRAJ module of
AMBER.
No good coordination was achieved for the case of the
nonactive compounds, 5, 7, 8, 10, and 11, which bear sizable
substituents on N1. Substitution in N1 also impairs the activity
of 4 and 5 relative to the single-substituted analogs and is
reflected in a lower score. No H-bond interaction with the
THR199 OG can be established in this case, and only the
acceptor properties of the sulfamide sulfonyl O can add to
the stabilization (Figure 2C and D).
RESULTS
The different activity of sulfamides and sulfamates is not
well reproduced by the docking algorithms (compare com-
pounds 13 and 16, Table 1), as the calculation of the binding
energy value considers the coordination of the negative
species to the Zn active site, i.e., the different pKa is not
accounted for in the docking score. Even though the pKa
can be relevant in determining the different activity, it does
not seem to be enough to justify the Ki ratio of STOP/TOP,
two orders of magnitude larger than that of 13/16. Among
the different possible justifications discussed in the literature,
a steric clash of the methyl substituent of one of the dioxolane
rings of STOP with ALA65 has been mentioned. This group
is also present in TOP but has a slightly different orientation
Table 1 shows the biological activity determined for the
compounds shown in Figure 1.
The weaker inhibition pattern of sulfamides compared to
sulfamates is in agreement with previous results reported for
similar structures.17,42,43 The scores predicted by the Au-
todock docking are also given in Table 1, column 4, and
relevant calculated distances involved in Zn binding are
shown in columns 5 and 6. The N atoms in the sulfamide
function are referred to as N1 for the N atom coordinated to
Zn in the active site of CAII and N2 for the noncoordinating
N atom that bears the substitution (as exemplified for
compounds 2, 3, 13, and 16 in Figure 1). For the sulfamide

SULFAMATES AND SULFAMIDES INHIBITORY ACTIVITY
J. Chem. Inf. Model., Vol. 50, No. 6, 2010 1117
Figure 2. Binding geometries of compounds 4-19 in the CAII active site predicted by the Autodock docking algorithms. Histidine residues
of the active site are highlighted in green, and THR199 residue is highlighted in orange. Zn atom is represented as a nonbonded sphere in
gray. Only H atoms attached to N of the ligand are shown. Values of the relevant distances named as d1 and d2 are given in Table 1. For
clarity, carbon chains of the ligands are represented in different colors: (A) monosubstituted sulfamates 16 (green), 17 (orange), and 18
(light blue); (B) monosubstituted sulfamides 13 (green), 14 (light blue), and 15 (orange); (C) Disubstituted sulfamides 4 (violet), 9 (green),
and 12 (orange); (D) compound 6 (white); and (E) Compound 19 (blue).
that prevents this negative interaction. The compounds only
differ in the bioisosteric substitution of O by NH, which is
enough to trigger a different orientation of the dioxolane rings
as the S-X-C angle (X ) O, N) changes from 122 to 111
as X changes from O to NH. In this way, different intra-
and intermolecular interactions in the active site pocket can
be stabilized in both cases.
To assess the molecular basis of the differences observed
in the affinity of sulfamates and sulfamides toward hCA II,
12 ns MD simulations were carried out for butyl, benzyl,
and ciclohexyl-sulfamates and sulfamides as representative
of aliphatic, cyclic, and aromatic-substituted species. The
analysis was extended to the case of TOP and STOP as a
way of testing the credibility of the results through the
comparison with experimental data. TOP was modeled by
its analog TOP-like as no X-ray structure was available when
the simulations were run. For the case of STOP, different
protonation states have been considered (see Experimental
Section) to better account for the possible reasons for the
large fold difference (Table 1). The H-bond interaction
pattern connecting the ionized N1H group to THR199OG is
kept for both N-butyl sulfamide and N-butyl sulfamate over
the 12 ns MD runs with average distances of 2.16 and 2.05
Å, respectively. The binding is further stabilized by the
interaction of a sulfonyl O accepting a H from either the
THR200 HG or the THR199 H in a oscillating manner.
The aliphatic tail shows a similar orientation in both cases,
pointing toward the hydrophobic part of the active site
pocket; resembling the pattern shown in the crystallographic
structures of aliphatic sulfamates.44
The bioisosteric substitution of O by N2H does not impact
the interactions of the binder group in the CAII active site.
Nevertheless, striking differences become evident when the
interactions of the surrounding solvent water molecules are
considered. A solvent water molecule is always donating an
H atom to the N1H group of the sulfamide throughout the

1118 J. Chem. Inf. Model., Vol. 50, No. 6, 2010
GAVERNET ET AL.
Figure 3. Binding geometries of compounds 1-3 in the CAII active site predicted by the Autodock docking algorithms. HIS residues of
the active site are higtlighted in green, and THR residues are highlighted in orange. Zn atom is represented as a nonbonded sphere in gray.
Only H atoms attached to N of the ligand are shown. Values of the relevant distances named as d1 and d2 are given in Table 1. (A) STOP
in its “zwitterion” conformation (STOPZ). Other significant distances are: d3 ) 3.17, d4 ) 2.27, and d5 ) 3.10 Å. (B) STOP in its
deprotonated conformation (STOPN). Other significant distances are: d3 ) 3.39, d4 ) 2.66, and d5 ) 3.45 Å. (C) TOP. Other significant
distances are: d3 ) 3.03, d4 ) 2.49, and d5 ) 2.95 Å. (D) TOP-like. Other significant distances are: d3 ) 3.45 and d4 ) 3.45 Å.
Figure 4. Representative snapshot that shows the interaction of butyl sulfamide with water molecules. The distances (in Å) correspond to
average values.
simulation, while the water O atom is the acceptor of a
H-bond coordination with the sulfamide N2H. THR200 is
also involved in this coordination alternating between
OG-water and HG-N2 interactions (Figure 4). Water
molecules also diffuse when the active site pocket is occupied
by a sulfamate, but the solvent donates its H to the sulfamate
O. Representative snapshots are shown in Figures 4 and 5.
The coordination of a water H atom compromises the lone
pair electrons of the Zn-N1 atom, significantly impairing
the coordination capability to the metal center.
Neither THR199 nor THR200 participates directly in
stabilizing the binding of the N,N′-disubstitued sulfamides.
The N2 atom can H bond a solvent water molecule and
partially interact through it with THR199. Another solvent
molecule can also reach deep into the pocket and H bond,
the N1 atom of N,N′-dibutylsulfamide. This interaction is not
as stable as in the single-substituted counterpart and is not
found in the simulation of the CAII-N,N′-dibenzylsulfamide
system. N,N′-disubstitution markedly impairs the activity as
it impedes important H-bond stabilizations of the binder
sulfamide group in the bottom of the active site pocket. The
effect is less evident for the case of the aromatic benzyl
substituent, as a partial π-stacking between the phenyl rings
reinforces the electronic density and, hence, a T-shape
interaction with PHE150 (see Supporting Information, Figure
S4).
Similar interactions with Zn, THR199, and THR200
stabilize the binding of N-benzyl sulfamide and N-benzyl
sulfamate, and a similar solvation pattern can be described
(see Supporting Information, Figure S2). There are no
differences in the solvation pattern or in the coordination
with THR199 and THR200 for the cyclohexyl substitution
(see Supporting Information, Figures S3).

SULFAMATES AND SULFAMIDES INHIBITORY ACTIVITY
J. Chem. Inf. Model., Vol. 50, No. 6, 2010 1119
Figure 7. Coordination of STOPZ (carbon atoms in white) in the
CAII active site resulting from conformational search. The X-ray
structure (2H15) is shown (carbon atoms in green) for comparison.
More details in the text.
Figure 5. Representative snapshot that shows the interaction of
butyl sulfamide with water molecules. The distances (in Å)
correspond to average values.
performed an exhaustive conformational search for the
different possible protonation states of STOP, neutral
(STOPH), deprotonated (STOPN), and the “zwitterion”
(STOPZ) that can be formed by proton transfer in the event
of the docking. We found the best agreement to the
experiment for the zwitterionic conformation (Figure 7). Only
this isomer reproduces the 111° value of the S-N-C angle
of STOP (110.2° calculated) which is compatible with a sp3
hybridization of the N atom. The resulting bound conforma-
tion is stabilized through the formation of two H bonds: a
N2-H-O bond with one dioxolane O6 acting as acceptor
and a second N2-H-OG reinforced by coordination of HG
to the STOP O3 (Figure 7). It also reproduces the interactions
of the sulfamide binder with THR199 and THR200 and the
H-bond interaction network of the cap with near neighbor
residues. As previously mentioned, no good coordination was
attained for the neutral structure (STOPH), and the rms
values obtained from superimposition of the docked con-
formations of STOPN and STOPZ (Table 1) with the
crystallographic conformation of STOP confirmed that
STOPZ better models the cocrystallized structure.
In order to better understand if the STOPZ conformation
can be formed through an assisted proton transfer between
the STOP N atoms, we repeated the conformational search
modeling an explicit water molecule close enough to STOP
to stabilize a STOP-water ensemble. The lowest conforma-
tion that results from the search features the water molecule
accepting a H from N1H2 and establishing a bifurcated H
bond with the second Stop N atom and the pyranose O3,
enabling the proton transfer that leads to STOPZ. The
superposition of the STOP-water supermolecule with 2H15
shows the water molecule playing the role of THR200 in
H-binding O3, suggesting a possible N1-OG-O3 H-bond
network (see Supporting Information, Figure S5). However,
the conformation of STOP in 2H15 features a long N1-OG
distance, not compatible with this interaction. The bifurcated
H-bond coordination prompted us to analyze a second
zwitterionic conformation (STOPZO, see Supporting Infor-
mation, Figure S6). The energies derived from the confor-
mational search in water solvent using the OPLS force field
suggests an equilibrium between both zwitterion structures
with the neutral conformer 60 kcal/mol higher in energy (see
Supporting Information, Figure S6).
Figure 6. Superimposition of STOP and TOP from the cocrystal-
lized structures 2H15 (STOP) and 3HKU (TOP).
There is no diffusion of solvent molecules into the CAII
active site pocket bound to bulky ligands as TOP-like or
STOP in the course of 12 ns MDs, which makes it even
more difficult to justify their Ki ratio. A thorough comparison
of the cocrystallized CAII-TOP (3HKU) and CAII-STOP
(2H15)7 has pointed out minor differences in the interactions
with the residues of the active site pocket, mainly focusing
in the H-bond interactions of the endocyclic sugar oxygens
and clashes of the methyl substituent of the dioxolane rings
with ALA65, which can be avoided in the case of TOP.
Although the cocrystallized ligands can be nicely superim-
posed in the binder and sugar cyclic moieties (cap), they
differ in the spatial orientation of the O/NH linker (Figure
6). The spatial orientation of the cap is, hence, more
conserved among similar ligands than the conformation of
the linker, showing that a different orientation would be
disfavored. Nevertheless, the different linker determines the
small differences in the fitting of the cap in the active site
pocket that have been pointed out as relevant in impairing
the activity.
Neither the O in TOP nor the N2H in STOP in the linker
region participates in any interaction with protein residues.
STOP N2 is, however, 2.76 Å from the endocyclic sugar O
(named as 3 in Figure 1) and 3.21 Å from the O6 atom of
one dioxolane ring, suggesting a H-bond coordination of N2
with at least one of these O atoms. In order to discern the
protonation state and possible intramolecular H-bond interac-
tions of the STOP molecule cocrystallized with CAII, we
On the knowledge that the sulfamide analogue binds to
CA II with the deprotonated sulfamide moiety coordinated
to Zn, we ran 12 ns MD simulations for STOPN and STOPZ,

1120 J. Chem. Inf. Model., Vol. 50, No. 6, 2010
GAVERNET ET AL.
of H-bond interactions in the binding conformation of STOP,
making STOPZ a better model that STOPN. The STOPZ
model features a ligand with a protonated N2 atom and a
negative N1. Accepting this model as closer to reality offers
a possible interpretation of the unpredictable low inhibitory
activity of STOP compared with other monosubstituted
sulfamides, as the lack of a negative net charge in the ligand
impairs its ability to interact with the positive Zn ion in the
enzyme metallocenter.
The MD simulation of TOP-like confirms the prevalence
of the positive interactions found in its X-ray structure. As
described for STOP, no water molecules diffuse to the active
site, and the cyclic sulfate O atoms of the cap are exposed
to the solvent and form H bonds as predicted before
considering crystallization waters16 (see Supporting Informa-
tion, Figure S7).
CONCLUSION
The discovery of novel chemical entities as potent inhibi-
tors and the origin of its activity at the molecular level are
major goals in medicinal chemistry. In this line, understand-
ing the effects that cause the lack of inhibition of new
possible candidates represents equally valuable information
for future inhibitor design. Through molecular modeling
techniques supported by experimental data, we provided in
this investigation an explanation for the different activity of
sulfamides compared with their cognate sulfamates.
The analysis showed that the nature and the position of
the sulfamide substituents as well the acidity and the ability
of solvation of the sulfamide function directly impact the
inhibitory capability. The results point to the importance of
an organic scaffold (N- or O-substituent_cap) that maximizes
interactions with hydrophobic and hydrophilic regions of the
enzyme active site. In the design of the cap, intramolecular
H bonds involving the N2H moiety should be avoided.
Disubstitution, although increasing the number of possible
interactions through the inclusion of a second substituent,
has a negative effect on the inhibitory activity, as the
substitution of the second sulfamide NH precludes the
interaction with THR199.
Our results analyze the influence of solvation on the zinc-
binding function of sulfamides and sulfamates and describe
an interesting effect of H-bond coordination on intramolecu-
lar interactions that results in a formal decrease of the
negative charge of N1. However, an appropriate substitution
that maximizes the positive interactions with the residues
and increases the acidity of the N1H can lead to interesting
sulfamide inhibitors. As examples, we can mention arylsul-
famides with an electron-withdrawing substituent in the
aromatic ring,17 where the 4-(trifluoromethyl)phenylsulfa-
mide (Ki ) 13 nm), pentaflluoromethylsulfamide (Ki ) 32
nm), and 4-cyanophenylsulfamide (Ki ) 16 nm) are more
active than their cognate sulfamates.
Figure 8. (A) Relevant interactions that stabilize STOPZ in the
active site of CAII during the MD simulations. Average values are
given. The superposition of one STOPZ representative MD snapshot
and the crystallographic ligand are shown in the lower right corner.
(B) Relevant interactions that stabilize STOPN in the active site of
CAII during the MD simulations. Average values are given. The
superposition of one STOPN representative MD snapshot and the
crystallographic ligand are shown in the lower right corner.
starting from the X-ray structure (PDB code 2H15) for
STOPN and modifying the ligand through protonation of the
N for STOPZ.
The main H-bond interactions derived from the docking
are conserved over the 12 ns MDs of CAII-STOPZ and
CAII-STOPN. As in X-ray structure, the binder sulfamide
function interacts with THR199 and THR 200, and the
organic scaffold makes a network of H bonds with residues
of the active site pocket in both cases (Figure 8A). In the
case of STOPZ, the intramolecular N2-O6 and N2-O3
interactions are kept throughout the simulation. Conversely,
the conformation of STOPN is only kept by the concerted
N2-H-OG/OG-HG-O H-bond network, but the N2-
endocyclic O6 distance is longer (3.54 Å) at the end of the
12 ns MD than in the X-ray structure (3.2 A), see Figure
8B. The slight relaxation of the structure impacts the
interaction of the cap with the surrounding residues. A
detailed comparison shows that only in the case of STOPZ
the experimental interactions of the dioxolane O (O5-GLN92
and O7-ASN62) are conserved. As mentioned before, no
water molecules were found in the active site of CAII during
the MD, and only the organic scaffold (cap) of the ligand
interacts with the solvent through the dioxolane O atoms.
The results obtained from docking, conformational analy-
sis, and MD simulations are indicative of a strong influence
ACKNOWLEDGMENT
L.E. Bruno Blanch is a member of the Facultad de Ciencias
Exactas, Universidad Nacional de La Plata, and L. Gavernet
is member of Consejo Nacional de Investigaciones Cient´ıficas
y Te´cnicas de la Repu´blica Argentina (CONICET). The work
described herein was done in part during a Postdoctoral leave
of Dr. L. Gavernet of the University of Notre Dame to whom

SULFAMATES AND SULFAMIDES INHIBITORY ACTIVITY
J. Chem. Inf. Model., Vol. 50, No. 6, 2010 1121
Inhibitors with Low Affinity for the Ubiquitous Isozyme II, Exempli-
fied by the Crystal Structure of the Topiramate Sulfamide Analogue.
J. Med. Chem. 2006, 49, 7024–7031.
she gratefully acknowledges. This research was supported
in part by the National Science Foundation (TG-CHE090124)
through TeraGrid resources provided by NICS and LONI
and by the Agencia de Promocio´n Cient´ıfica y Tecnolo´gica
(PICT 00339/2007), CONICET, the Universidad Nacional
de La Plata, Argentina, and the Walther Cancer Research
Center, University of Notre Dame. The computations were
performed on Kraken (a Cray XT5) at the National Institute
for Computational Sciences (http://www.nics.tennessee.edu/)
and on Queen Bee at the Louisiana Optical Network
Initiative. Generous allocation of computing resources by
the Center for Research Computing at the University of Notre
Dame is also acknowledged. Work from CTS lab has been
financed by an European Union grant of the seventh FP
programme (Metoxia project).
(17) Casini, A.; Winum, J.-Y.; Montero, J.-L.; Scozzafava, A.; Supuran,
C. T. Carbonic anhydrase inhibitors: inhibition of cytosolic isozymes
I and II with sulfamide derivatives. Bioorg. Med. Chem. Lett. 2003,
13, 837–840.
(18) Gavernet, L.; Dominguez Cabrera, J.; Bruno Blanch, L.; Estiu´, G. 3D-
QSAR design of novel antiepileptic sulfamides. Bioorg. Med. Chem.
2007, 15, 1556–1567.
(19) Gavernet, L.; Barrios, I.; Sella Cravero, M.; Bruno Blanch, L. Design,
synthesis, and anticonvulsant activity of some sulfamides. Bioorg. Med.
Chem. 2007, 15, 5604–5614.
(20) Barrios, I. A.; Rocha Arrieta, L.; Bruno-Blanch, L. E. Unpublished
results.
(21) Khalifah, R. G. The carbon dioxide hydration activity of carbonic
anhydrase. I. Stop-flow kinetic studies on the native human isoenzymes
B and C. J. Biol. Chem. 1971, 246, 2561–2573.
(22) Innocenti, A.; Vullo, D.; Pastorek, J.; Scozzafava, A.; Pastorekova,
S.; Nishimori, I.; Supuran, C. T. Carbonic anhydrase inhibitors.
Inhibition of transmembrane isozymes XII (cancer-associated) and XIV
with anions. Bioorg. Med. Chem. Lett. 2007, 17, 1532–1537.
(23) Supuran, C. T. Carbonic anhydrases: novel therapeutic applications
for inhibitors and activators. Nat. ReV. Drug DiscoVery 2008, 7, 168–
181.
(24) Morris, G. M.; Goodsell, D. S.; Halliday, R. S.; Huey, R.; Hart, W. E.;
Belew, R. K.; Olson, A. J. Automated Docking Using a Lamarckian
Genetic Algorithm and and Empirical Binding Free Energy Function.
J. Comput. Chem. 1998, 1639–1662.
(25) Jude, K. M.; Banerjee, A. L.; Haldar, M. K.; Manokaran, S.; Roy, B.;
Mallik, S.; Srivastava, D. K.; Christianson, D. W. Ultrahigh Resolution
Crystal Structures of Human Carbonic Anhydrases I and II Complexed
with “Two-Prong” Inhibitors Reveal the Molecular Basis of High
Affinity. J. Am. Chem. Soc. 2006, 128, 3011–3018.
Supporting Information Available: Additional figures
that describe the interactions of the ligands in the CAII active
site. This information is available free of charge via the
Internet at http://pubs.acs.org
REFERENCES AND NOTES
(1) Lindskog, S.; Henderson, L. E.; Kannan, K. K.; Liljas, A.; Nyman,
P. O.; Strandberg, B. Carbonic anhydrase. In The Enzymes, 3rd ed.;
Boyer, P. D., Ed.; Academic Press: New York, 1971; Vol. 3, pp 587-
665.
(2) Pocker, Y.; Sarkanen, S. Carbonic anhydrase: structure, catalytic
versatility, and inhibition. AdV. Enzymol. Relat. Areas Mol. Biol. 1978,
47, 149–274.
(3) Sly, W. S.; Hu, P. Y. Human carbonic anhydrases and carbonic
anhydrase deficiencies. Annu. ReV. Biochem. 1995, 64, 375–401.
(4) Supuran, C. T.; Scozzafava, A. Carbonic anhydrase inhibitors and their
therapeutic potential. Expert Opin. Ther. Pat. 2000, 10, 575–600.
(5) Supuran, C. T.; Scozzafava, A.; Cassini, A. Carbonic anhydrase
inhibitors. Med. Res. ReV. 2003, 23, 146–189.
(6) Pastorekova, S.; Parkkila, S.; Pastorek, J.; Supuran, C. T. Carbonic
anhydrase: current state of the art, therapeutic applications and future
prospects. J. Enzyme Inhib. Med. Chem. 2004, 19, 199–229.
(7) Supuran, C. T.; Scozzafava, A. Carbonic anhydrases as targets for
medicinal chemistry. Bioorg. Med. Chem. 2007, 15, 4336–4350, and
references cited therein.
(8) Hartsough, D. S.; Merz, K. M., Jr. Dynamic Force Field Models:
Molecular Dynamics Simulations of Human Carbonic Anhydrase II
Using a Quantum Mechanical/Molecular Mechanical Coupled Poten-
tial. J. Phys. Chem. 1995, 99, 11266–11275.
(9) Jackman, J. E.; Merz, K. M.; Fierke, C. A. Disruption of the active
site solvent network in carbonic anhydrase II decreases the efficiency
of proton transfer. Biochemistry 1996, 35, 16421–16428.
(10) Peng, Z.; Merz, K. M.; Banci, L. Binding of cyanide, cyanate, and
thiocyanate to human carbonic anhydrase II. Proteins 1993, 17, 203–
216.
(11) Rossi, K. A.; Merz, K. M.; Smith, G. M.; Baldwin, J. J. Application
of the free energy perturbation method to human carbonic anhydrase
II inhibitors. J. Med. Chem. 1995, 38, 2061–2069.
(12) Oltulu, O.; Yasar, M. M. Eroglu, E.A QSAR study on relationship
between structure of sulfonamides and their carbonic anhydrase
inhibitory activity using the eigenvalue (EVA) method. Eur. J. Med.
Chem. 2009, 44, 3439–3444.
(26) Schafmeister, C.; Ross, W. S.; Romanovski, V. LEaP; University of
California: San Francisco, 1995.
(27) Quality Atomic Charges, Proton Assignment and Canonicalization,
QuACPAC; OpenEye Scientific Software, Inc.: Santa Fe, NM, 2007.
(28) Mohamadi, F.; Richards, N. G.; Guida, W. C.; Liskamp, R.; Lipton,
M.; Caufiel, C.; Chang, G.; Hendrickson, T.; Still, W. C. Macromodel:
An Integrated Software System for Modeling Organic and Bioorganic
Molecules Using Molecular Mechanics. J. Comput. Chem. 1990, 11,
440–467.
(29) Jorgensen, W. L.; Tirado-Rives, J. Potential energy functions for
atomic-level simulations of water and organic and biomolecular
systems. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 6665–6670.
(30) Shenkin, P. S.; McDonald, D. Q. Cluster Analysis of Molecular
Conformations. J. Comput. Chem. 1994, 15, 899–916.
(31) Case, D. A.; Darden, T. A.; Cheatham, I. T. E.; Simmerling, C. L.;
Wang, J.; Duke, R. E.; Luo, R.; Crowley, M.; Walker, R. C.; Zhang,
W.; Merz, K. M. J.; Wang, B.; Hayik, S.; Roitberg, A.; Seabra, G.;
Kolossvary, I.; Wong, K. F.; Paesani, F.; Vanicek, J.; Wu, X.; Brozell,
S.; Steinbrecher, H.; Gohlke, H.; Yang, L.; Tan, C.; Mongan, J.;
Hornak, V.; Cui, G.; Mathews, D. H.; Seetin, M. G.; Sagui, C.; Babin,
V.; Kollman, P. A. AMBER 10; University of California: San
Francisco, CA, 2008.
(32) Jorgensen, W. L. C. J.; Madura, J.; Impey, R. W.; Klein, M. L.
Comparison of simple potential functions for the simulation of liquid
water. J. Chem. Phys. 1983, 79, 926–935.
(33) Wang, J.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.; Case, D. A.
Development and testing of a general AMBER force field. J. Comput.
Chem. 2004, 25, 1157–1174.
(34) Bayly, C. A.; Cieplak, P.; Cornell, W. D.; Kollman, P. A. A well
behaved elestrostatic potential based method using charge restraints
for deriving atomic charges: The RESP model. J. Phys. Chem. 1993,
97, 10269–10280.
(35) Fox, T.; Kollman, P. A. Application of RESP Methodology in the
Parametrization of Organic Solvents. J. Phys. Chem. B 1998, 102,
8070–8079.
(13) Eroglu, E.; Turkmen, H. A DFT-based quantum theoretic QSAR study
of aromatic and heterocyclic sulfonamides as carbonic anhydrase
inhibitors. J. Mol. Graphics Modell. 2007, 26, 701–708.
(14) Tuccinardi, T.; Nuti, E.; Ortore, G.; Supuran, C. T.; Rossello, A.;
Martinelli, A. Analysis of Human Carbonic Anhydrase II: Docking
Reliability and Receptor-Based 3D-QSAR Study. J. Chem. Inf. Model.
2007, 47, 515–525.
(15) Singh, J.; Shaik, B.; Singh, S.; Sikhima, S.; Agrawal, V. K.;
Khadikar, P. V.; Supuran, C. T. QSAR studies on the activation of
the human carbonic anhydrase cytosolic isoforms I and II and
secretory isozyme VI with amino acids and amines. Bioorg. Med.
Chem. 2007, 15, 6501–9.
(36) Becke, A. D.; Yarkony, D. R. In Modern Electronic Structure Theory
Part II; World Scientific: Singapore, 1995.
(37) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin,
K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone,
V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa,
J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene,
M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.;
Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev,
O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala,
P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.;
Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas,
O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.;
(16) 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

1122 J. Chem. Inf. Model., Vol. 50, No. 6, 2010
GAVERNET ET AL.
Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.;
Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen,
W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, revision
C.02; Gaussian, Inc.: Wallingford, CT, 2004.
(42) Winum, J. Y.; Vullo, D.; Casini, A.; Montero, J. L.; Scozzafava, A.;
Supuran, C. T. Carbonic Anhydrase Inhibitors. Inhibition of Cytosolic
Isozymes I and II and Transmembrane, Tumor-Associated Isozyme
IX with Sulfamates Including EMATE Also Acting as Steroid
Sulfatase Inhibitors. J. Med. Chem. 2003, 46, 2197–2204.
(43) Winum, J. Y.; Vullo, D.; Casini, A.; Montero, J. L.; Scozzafava, A.;
Supuran, C. T. Carbonic Anhydrase Inhibitors: Inhibition of Trans-
membrane, Tumor-Associated Isozyme IX, and Cytosolic Isozymes I
and II with Aliphatic Sulfamates. J. Med. Chem. 2003, 46, 5471–
5477.
(38) Ryckaert, J. P.; Ciccotti, G.; Berendsen, H. J. C. Numerical integration
of the cartesian equations of motion of a system with constraints:
Molecular dynamics of n-alkanes. J. Comput. Phys. 1977, 23, 327–
341.
(39) Pastor, R. W.; Brooks, B. R.; Szabo, A. An analysis of the accuracy
of Langevin and molecular dynamics algorithms. Mol. Phys. 1988,
65, 1409–1419.
(40) Petersen, H. G. Accuracy and efficiency of the particle-mesh-ewald
method. J. Chem. Phys. 1995, 103, 3668–3679.
(44) Vitale, R. M.; Alterio, V.; Innocenti, A.; Winum, J. Y.; Monti, S. M.;
De Simone, G.; Supuran, C. T. Carbonic Anhydrase Inhibitors.
Comparison of Aliphatic Sulfamate/Bis-sulfamate Adducts with
Isozymes II and IX as a Platform for Designing Tight-Binding, More
Isoform-Selective Inhibitors. J. Med. Chem. 2009, 52, 5990–5998.
(41) Essman, V.; Perera, L.; Berkowitz, M. L.; Darden, T.; Lee, H.;
Pedersen, L. G. A smooth particle-mesh-Ewald method. J. Chem. Phys.
1995, 103, 8577–8593.
CI100112S