Carbonic anhydrase inhibitors: Inhibition of cytosolic isozymes I and II with sulfamide derivatives

Article abstract of DOI:10.1016/S0960-894X(03)00028-3

A novel class of effective CAIs has been identified, starting from a very weak carbonic anhydrase inhibitor (CAI), sulfamide, whose X-ray crystal structure in the adduct with hCA II has recently been reported. A series of N,N-disubstituted- and N-substitu

Full text of DOI:10.1016/S0960-894X(03)00028-3

Bioorganic & Medicinal Chemistry Letters 13 (2003) 837–840
Carbonic Anhydrase Inhibitors:
Inhibition of Cytosolic Isozymes I and II with Sulfamide
Derivatives
Angela Casini,a Jean-Yves Winum,b Jean-Louis Montero,b Andrea Scozzafavaa and
Claudiu T. Supurana,*
a
`
Universita degli Studi di Firenze, Dipartimento di Chimica, Laboratorio di Chimica Bioinorganica, Rm. 188,
Via della Lastruccia 3, I-50019 Sesto Fiorentino (Firenze), Italy
´ ´ ´
Universite Montpellier II, Laboratoire de Chimie Biomoleculaire, UMR 5032, Ecole Nationale Superieure
de Chimie de Montpellier, 8rue de l’Ecole Normale, 34296 Montpellier Cedex, France
b
Received 9 October 2002; revised 12 December 2002; accepted 20 December 2002
Abstract—A novel class of effective CAIs has been identified, starting from a very weak carbonic anhydrase inhibitor (CAI), sul-
famide, whose X-ray crystal structure in the adduct with hCA II has recently been reported. A series of N,N-disubstituted- and
N-substituted-sulfamides were prepared from the corresponding amines and N-(tert-butoxycarbonyl)-N-[4-(dimethylazaniumyli-
dene)-1,4-dihydropyridin-1-ylsulfonyl]azanide or the unstable N-(tert-butoxycarbonyl)sulfamoyl chloride. The disubstituted com-
pounds being too bulky, were ineffective as CAIs, whereas mono-substituted derivatives (incorporating aliphatic, cyclic and
aromatic moieties) as well as a bis-sulfamide, behaved as micro-nanomolar inhibitors of two cytosolic isozymes, hCA I and hCA II,
responsible for critical physiological processes in higher vertebrates. Aryl-sulfamides were more effective than aliphatic derivatives.
Low nanomolar inhibitors have been detected, which generally incorporated 4-substituted phenyl moieties in their molecule. This is
the first example of CAIs in which low nanomolar inhibitors were generated starting from a very ineffective lead molecule.
# 2003 Elsevier Science Ltd. All rights reserved.
Introduction
methazolamide MZA, ethoxzolamide EZA, dichloro-
phenamide DCP, dorzolamide DZA or brinzolamide
BRZ) contain a terminal sulfonamide as anchoring
group to coordinate the catalytic zinc.1ꢀ3 These sulfo-
namides are widely used clinically, mainly as anti-
glaucoma agents, but also for the therapy of other
diseases, for example increased intracranial pressure,
various neurological/neuromuscular pathologies, such
as epilepsy, genetic hemiplegic migraine and ataxia,
tardive diskinesia, hypokalemic periodic paralysis,
essential tremor and Parkinson’s disease, and mountain
sickness. Accordingly, drugs of this pharmacological
class are under constant development.2,3
Carbonic anhydrases (CAs, EC 4.2.1.1) catalyze a very
simple physiological reaction, the interconversion
between carbon dioxide and the bicarbonate ion, and
are involved in crucial physiological processes con-
nected with respiration and transport of CO2/bicarbo-
nate between metabolizing tissues and the lungs, pH
and CO2 homeostasis, electrolyte secretion in a variety
of tissues/organs, biosynthetic reactions (such as gluco-
neogenesis, lipogenesis and ureagenesis), bone resorp-
tion, calcification, tumorigenicity, and many other
physiologic or pathologic processes.1ꢀ3 Thus, it is not
surprising that many of their isozymes (14 are presently
known in higher vertebrates) have been discovered as
important targets for inhibitors with clinical applica-
tions.2,3 Almost all of the most potent inhibitors of CAs
(such as the clinically used acetazolamide AAZ,
Recently, the X-ray crystallographic structure of sulfa-
mide (H2NSO2NH2) bound to the physiologically rele-
vant isozyme hCA II has been reported by this group.4 It
was shown that sulfamide binds to the zinc ion of the
enzyme (Fig. 1) via its deprotonated amide group (Zn–N
distance of 1.76 A), and this nitrogen donates an H-bond
to Og of Thr 199 via its remaining hydrogen.4 Thus, Og
acts as an acceptor for the NH group of the inhibitor, but
*Corresponding author. Tel.: +39-055-457-3005; fax: +39-055-457-
3363; e-mail: claudiu.supuran@unifi.it
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00028-3

838
A. Casini et al. / Bioorg. Med. Chem. Lett. 13 (2003) 837–840
to the OH group of Thr 200 (3.20 A) and a third contact
has been found to another water molecule (3.26 A). In
summary, this very simple inhibitor, present as the
negatively charged (NH)SO2NHꢀ2 ion, shows a large
number of favorable contacts in the binding pocket of
CA, and may be used as a lead molecule, since sulfamide
itself is a weak inhibitor (KI of 35 mM against hCA I, and
of 82 mM against hCA II — for the esterase activity of
these enzymes)4,5 (Fig. 1). From data of Figure 1, it is
obvious that the second amino moiety of the inhibitor
points towards the exit of the active site, being thus
positioned in a very appropriate manner for derivatiza-
tion. This is the approach we report here, that is deriva-
tization of sulfamide at one of its NH2 moieties, which
led to nanomolar inhibitors of the two CA isozymes most
abundant in vertebrates, CA I and CA II.
Figure 1. The sulfamide–hCA II adduct: the Zn(II) ion of the enzyme,
its three histidine ligands (His 94, 96 and 119 CPK colors) and coor-
dinated sulfamidate are shown, together with active site residues/water
molecules (in red) involved in hydrogen bond networks (dotted, in
yellow) with the inhibitor molecule: Thr 199 and Thr 200, Glu 106 and
Arg 246 (figure generated from the PDB file described in ref. 4).
Chemistry
Sulfamides 1–26 have been prepared by an original
procedure recently described by Winum et al.,6 implying
the reaction of primary/secondary amines with either
N-(tert-butoxycarbonyl)-N-[4-(dimethylazaniumylidene)-
1,4-dihydropyridin-1-ylsulfonyl]azanide or the unstable
N-(tert-butoxycarbonyl)sulfamoyl chloride (obtained in
situ from chlorosulfonyl isocyanate and tert-butanol),
followed by removal of the protecting group (Scheme 1).
Some of these compounds were previously reported in
the literature,6ꢀ16 but they have not been tested as CA
in turn, its hydroxyl moiety donates an H-bond to one of
the terminal carboxylate oxygens of the adjacent Glu106.
Via its second oxygen, this carboxylate mediates a
hydrogen-bond network as acceptor to the backbone
NH of Arg246 and via a water molecule to the hydroxyl
oxygen of Tyr7. This extended extra-coordination results
in a distorted tetrahedral arrangement around the metal
ion, the remaining three ligands of zinc being His94,
His96 and His119 (as in the uninhibited enzyme).4 The
second oxygen of the SO2 group of the inhibitor was
shown to be close to the backbone NH of Thr199 (2.81
A), which accordingly accepts
a
H-bond from
Thr199NH. This oxygen replaces the ‘deep water’ in the
ligand-free enzyme.4 Finally, the second, uncharged NH2
group of sulfamide forms a hydrogen bond of 2.90 A to
an adjacent water molecule, which in turn is involved in
an H-bond network with three other water molecules
present in the binding pocket. In addition, this latter
sulfamide nitrogen forms a second weak hydrogen bond
Scheme 1.

A. Casini et al. / Bioorg. Med. Chem. Lett. 13 (2003) 837–840
839
inhibitors (CAIs) in earlier studies. A rather large num-
ber of substitution patterns (R and R’ moieties) have
been used in compounds 1–26 (aliphatic, aromatic,
substituted-aromatic, etc) with both mono- and N,N-
disubstituted compounds, together with a disulfamide
derivative (26) prepared, in order to detect best
substitution pattern(s) for efficient CA inhibition.17
aliphatic derivatives 1–3 and 9, 10, as well as benzylsul-
famide 4 behave as moderately active CAIs, with inhib-
ition constants in the range of 133–960 nM against hCA
I, and 123–890 nM against hCA II, respectively. In this
subseries, the bulkiest derivative (the 2-adamantyl-sub-
stituted compound 3) was the most ineffective whereas
the benzyl-substituted one (4) the most active inhibitor.
Very good inhibitors proved to be the aryl-substituted
sulfamides 11–25, which showed inhibition constants in
the range of 8–62 nM against hCA I, and 7–49 nM
against hCA II, respectively. The 4-substituted phe-
nylsulfamides showed better activity as compared to the
3-substituted compound 24, the naphthyl-substituted
derivative 25 or the pentafluorophenyl derivative 23
(which were the most ineffective in this subseries). For
the first type of such derivatives, best inhibition has
been observed for the 4-trifluoromethylphenyl-,
4-methoxyphenyl-, 4-methylphenyl-, 4-hydroxyphenyl-,
phenyl-, and 4-nitrophenyl-substituted sulfamides. A
special mention should be done regarding the aliphatic
bis-sulfamide 26, which behaved as an effective CA II
inhibitor and a moderate CA I inhibitor, with inhibition
constants of 27 nM against hCA II and 149 nM against
hCA I, respectively. Thus, many of the new CAIs
reported here possess potencies comparable or better
than those of the clinically used compounds acetazol-
amide, methazolamide, ethoxzolamide, dichloro-
Carbonic anhydrase inhibitory activity
The data of Table 1 show that most of the sulfamides 1–26
investigated here act as better inhibitors against iso-
zymes hCA I and hCA II as compared to the lead
molecule, sulfamide, which is a weak inhibitor.4,5 The
nature and the number of the groups substituting the
sulfamide moiety were the primary factors influencing
CA inhibitory properties in this class of derivatives.
Thus, the N,N-disubstituted derivatives 5–8 are inactive
as CAIs (except for 8 against CA II, which behaves as a
very weak inhibitors of this isozyme), probably because
of their very bulky nature and impossibility to accom-
modate within the enzyme active site. On the other
hand, all the remaining monosubstituted derivatives, as
well as the bis-sulfamide 26, are much more potent
CAIs as compared to the lead molecule, being effective
against both investigated isozymes. For example, the
Table 1. Inhibition data for derivatives 1–26 investigated in the present paper and standard sulfonamide CA inhibitors: RR0N-SO2NH2 (1–25);
NH2SO2-NH-RR0NH-SO2NH2 (26)
Inhibitor No.
R
R0
KI (nM)
hCA Ia
hCA IIa
Acetazolamide
Methazolamide
—
—
—
—
900
780
12
14
Ethoxzolamide
Dichlorophenamide
Dorzolamide
Brinzolamide
—
—
—
—
—
—
—
—
25
8
38
9
1200
>50,000
—
3
82,000
148
450
890
123
>100,000
>100,000
>100,000
647
148
131
12
Sulfamide (H2NSO2NH2)
1
2
3
4
5
6
7
8
—
n-Bu
Cyclohexyl
2-Adamantyl
PhCH2
i-Bu
i-Pr
Cyclohexyl
PhCH2
—
H
H
H
H
i-Bu
35,000
173
164
960
133
>100,000
>100,000
>100,000
>100,000
155
i-Pr
Cyclohexyl
PhCH2
—
—
H
9
ꢀ(CH2)5
ꢀ(CH2)6
Ph
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
163
13
15
8
19
23
18
14
16
18
26
20
17
34
62
4-Me-C6H4
4-CF3-C6H4
4-Cl-C6H4
4-Br-C6H4
4-I-C6H4
4-MeO-C6H4
4-HO-C6H4
4-O2N-C6H4
4-EtO2C-C6H4
4-NC-C6H4
4-Me2N-C6H4
C6F5
H
H
H
H
H
H
H
H
H
H
H
H
H
H
13
7
15
21
17
11
12
13
19
16
21
32
49
36
27
3-Benzoyl-C6H4
2-Naphthyl
(CH2)2–SS–(CH2)2
39
149
aHuman (cloned) isozymes, by the esterase method.18

840
A. Casini et al. / Bioorg. Med. Chem. Lett. 13 (2003) 837–840
phenamide and so on (Table 1). Isozyme hCA II was
more prone to inhibition by these sulfamides as com-
pared to isozyme hCA I, but the differences between
them are rather small, especially if compared to the
heterocyclic/aromatic sulfonamides used clinically.19 It
should also be mentioned that the lead molecule itself
had a higher affinity for isozyme I as compared to iso-
zyme II, behavior which unfortunately has not been
maintained in the derivatized sulfamides investigated
here. In fact, very few compounds with selectivity for
CA I over CA II have been reported up to now, and
sulfamide was one of them.
2. Supuran, C. T.; Scozzafava, A. Curr. Med. Chem. Imm.,
Endoc. Metab. Agents 2001, 1, 61.
3. Supuran, C. T.; Scozzafava, A. Exp. Opin. Ther. Pat. 2000,
10, 575.
4. Abbate, F.; Supuran, C. T.; Scozzafava, A.; Orioli, P.;
Stubbs, M.; Klebe, G. J. Med. Chem. 2002, 45, 3583.
5. Briganti, F.; Pierattelli, R.; Scozzafava, A.; Supuran, C. T.
Eur. J. Med. Chem. 1996, 31, 1001.
6. Winum, J.-Y.; Toupet, L.; Barragan, V.; Dewynter, G.;
Montero, J.-L. Organic Lett. 2002, 3, 2241.
7. Paquin, A. M. Angew. Chem. 1948, 60, 316.
8. McManus, J. M.; Farland, S. W.; Gerber, C. F.; McLa-
more, W. M.; Laubach, G. D. J. Med. Chem. 1965, 8, 766.
9. Graf, R. Chem. Ber. 1959, 92, 509.
10. Cherkasov, V. M.; Dashevskaya, T. A. Ukr. Khim. Zh.
1966, 32, 486; Chem. Abstr. 65, 5314b.
Conclusions
11. Davis, F. A.; McCauley, J. P.; Chattopadhyay, S.; Har-
akal, M. E.; Towson, J.-C.; Watson, W. H.; Tavanaiepour, I.
J. Am. Chem. Soc. 1987, 109, 3370.
We report here a novel class of effective CAIs, starting
from a very weak CAI, sulfamide, whose X-ray crystal
structure in the adduct with hCA II has recently been
reported. N,N-disubstituted sulfamides were too bulky,
and ineffective as CAIs, whereas mono-substituted deri-
vatives (incorporating aliphatic, cyclic and aromatic
moieties) as well as a bis-sulfamide, behaved as micro-
nanomolar inhibitors of two isozymes, hCA I and hCA
II, responsible for critical physiological processes in
higher vertebrates. Aryl-sulfamides were more effective
than aliphatic derivatives. In the first group of such
compounds, low nanomolar inhibitors have been
obtained, which generally incorporated 4-substituted
phenyl moieties in their molecule. This is the first exam-
ple of CAIs in which low nanomolar inhibitors were
generated starting from a very ineffective lead molecule.
12. Burke, P. O.; McDermott, S. D.; Hannigan, T. J.; Spillan,
W. J. J. Chem. Soc., Perkin Trans. 2 1984, 1851.
13. Catt, J. D.; Matier, W. L. J. Org. Chem. 1974, 39, 566.
14. Cohen, E.; Klarberg, B. J. Am. Chem. Soc. 1962, 84, 1994.
15. Snader, K. M.; Chakrin, L. W.; Cramer, R. D.; Gelernt,
M.; Miao, C. K.; Shah, D. H.; Venslavsky, J. W.; Willis, C. R.;
Sutton, B. M. J. Med. Chem. 1979, 22, 706.
16. Kirsanov, A. V.; Zolotov, Y. M. Zh. Obshch. Khim. 1958,
28, 343; Chem. Abstr. 52, 1:3664a.
17. An example of synthesis is illustrated below: compound
13: to a cold solution of 4-aminobenzotrifluoride (1 equiv) and
triethylamine (1.5 equiv) in methylene chloride was added a
solution of N-(tertbutoxycarbonyl)sulfamoyl chloride in
methylene chloride (prepared ab initio by adding chloro-
sulfonylisocyanate (1 equiv) on tert-butanol (1 equiv) at 0 ꢁC).
The mixture was stirred 1 h at room temperature and then
concentrated in vacuo. The residue was treated with a mixture
of ether- pentane and filtered. The precipitate was reacted with
a solution of 20% TFA in methylene chloride until complete
disapearance of starting material on TLC. Then the reaction
was concentrated, and the expected compound was pre-
cipitated in a mixture ether-pentane and filtered. Yield: 90%.
Colorless crystals, mp: 144–146 ꢁC; 1H NMR (200 MHz,
DMSO-d6): 10 (1H, s, NH), 7.6 (2H, d, J=8.5 Hz, Ar–H),
7.35 (2H, d, J=8.5 Hz, Ar–H), 7.3 (2H, s, NH2); MS ESI +30
eV: 263 [M+Na]+.
Acknowledgements
This research was financed by a grant from the Italian
CNR-target project Biotechnology and by CSGI. Thanks
are addressed to Dr. Alexander Weber (Marburg Uni-
versity, Germany) for generating Figure 1 from the
PDB file of the sulfamide–CA II adduct.
18. A stopped flow variant of the Poker and Stone spectro-
photometric method (Pocker, Y.; Stone, J. T. Biochemistry
1967, 6, 668) has been employed, using an SX.18MV-R
Applied Photophysics stopped flow instrument.
References and Notes
19. Casini, A.; Scozzafava, A.; Supuran, C. T. Med. Res. Rev.
2003, 23, 146.
1. Supuran, C. T.; Scozzafava, A. Exp. Opin. Ther. Pat. 2002,
12, 217.