Inhibition of human mitochondrial carbonic anhydrases VA and VB with para-(4-phenyltriazole-1-yl)-benzenesulfonamide derivatives

Article abstract of DOI:10.1016/j.bmcl.2008.07.010

A library of 10 novel benzenesulfonamides containing triazole-tethered phenyl 'tail' moieties were synthesized by a Cu(I) catalyzed 1,3-dipolar cycloaddition reaction (DCR) (i.e., click chemistry) between 4-azido benzenesulfonamide and a panel of variously substituted phenyl acetylenes. These compounds were very effective inhibitors (low nanomolar) of the human mitochondrial carbonic anhydrase isozymes VA and VB. Mitochondrial carbonic anhydrases are potential targets for anti-obesity therapies, acting to reduce lipogenesis through a novel mechanism of action. The inhibitors reported here should prove valuable as lead compounds to further investigate the potential of CA inhibition for this novel therapeutic application.

Full text of DOI:10.1016/j.bmcl.2008.07.010

Bioorganic & Medicinal Chemistry Letters 18 (2008) 4624–4627
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Inhibition of human mitochondrial carbonic anhydrases VA and VB
with para-(4-phenyltriazole-1-yl)-benzenesulfonamide derivatives
a
b
b
b,
*
Sally-Ann Poulsen a, , Brendan L. Wilkinson , Alessio Innocenti , Daniela Vullo , Claudiu T. Supuran
*
a Eskitis Institute for Cell and Molecular Therapies, Griffith University, Brisbane Innovation Park, Don Young Road, Nathan, Qld. 4111, Australia
b Polo Scientifico, Laboratorio di Chimica Bioinorganica, Rm. 188, Università degli Studi di Firenze, Via della Lastruccia 3, 50019 Sesto Fiorentino, Florence, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
A library of 10 novel benzenesulfonamides containing triazole-tethered phenyl ‘tail’ moieties were
synthesized by a Cu(I) catalyzed 1,3-dipolar cycloaddition reaction (DCR) (i.e., click chemistry) between
4-azido benzenesulfonamide and a panel of variously substituted phenyl acetylenes. These compounds
were very effective inhibitors (low nanomolar) of the human mitochondrial carbonic anhydrase isozymes
VA and VB. Mitochondrial carbonic anhydrases are potential targets for anti-obesity therapies, acting to
reduce lipogenesis through a novel mechanism of action. The inhibitors reported here should prove valu-
able as lead compounds to further investigate the potential of CA inhibition for this novel therapeutic
application.
Received 5 June 2008
Revised 2 July 2008
Accepted 7 July 2008
Available online 10 July 2008
Keywords:
Carbonic anhydrase
Glycoconjugates
Sulfonamide
Ó 2008 Elsevier Ltd. All rights reserved.
Anti-obesity
Click chemistry
An observed side effect of the widely used antiepileptic drugs
Zonisamide (ZNS) and Topiramate (TPM) is a significant loss of
body weight when administered to obese patients. These com-
pounds contain a sulfonamide (–SO2NH2 in ZNS) or sulfamate
(–OSO2NH2 in TPM) moiety, (Fig. 1).1 These functional groups are
both well characterized zinc binding functions (ZBFs) that have
prominent representation amongst known carbonic anhydrase
(CA) inhibitors, wherein a ZBF is a critical component of the CA
inhibition pharmacophore model. These facts have directed Supu-
ran and co-workers to commence an investigation of the possibil-
ity of a connection between CA inhibition by these compounds and
weight loss, in particular inhibition of cytosolic CA II and mito-
chondrial CA VA and VB. These studies have shown that both
ZNS and TPM are good inhibitors of CA II, VA and VB.2 The X-ray
crystal structures of both these drugs complexed with the human
CA II isozyme have also been reported.2 These 3-dimensional
enzyme-inhibitor structures reveal the molecular interactions that
readily account for the good inhibition observed for these com-
pounds at CA II.
O
O
S
O
O
NH2
S
O
O
O
O
NH2
O
N
O
O
ZNS
TPM
Figure 1. Zonisamide (ZNS) and Topiramate (TPM).
sion levels and tissue/organ localization.1a,3 Several hCA isozymes
are cytosolic (I–III, VII and XIII), four are membrane bound or trans-
membrane proteins (IV, IX, XII and XIV), one is secreted into the
saliva and milk (VI) and two are mitochondrial (VA and VB). The
activity of hCAs modulates cellular processes associated with res-
piration and transport of CO2HCO3À, provision of HCO3 as a sub-
À
strate for biosynthetic pathways, pH regulation, electrolyte and
fluid secretion, as well as bone resorption and calcification.1a,3 At
least 25 clinically used drugs have significant hCA inhibition activ-
ity, and the modulation of specific hCAs by inhibition is an avenue
for the treatment of a wide range of acquired and inherited
diseases.1a
Mitochondria are located in the cytoplasm of human cells
wherein they play significant roles in a number of cellular func-
tions. The dysfunction of mitochondria contributes to a range of
human diseases including obesity.4 The discovery and character-
ization of mitochondrial hCA isozymes VA and VB5 led to the
hypothesis that the manipulation of mitochondrial CA activity
The carbonic anhydrases (CA, EC 4.2.1.1) are a family of zinc(II)
metalloenzymes that catalyze the reversible hydration of carbon
dioxide (CO2) to give bicarbonate (HCO3À) and a proton (H+).
In humans 15 different CA isozymes and CA related proteins
(CARP) have been identified and characterized. These isozymes ex-
hibit variable kinetic parameters, subcellular localization, expres-
* Corresponding authors. Tel.: +61 7 3735 7825; fax: +61 7 3735 7656.
E-mail address: s.poulsen@griffith.edu.au (S.-A. Poulsen).
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.07.010

S.-A. Poulsen et al. / Bioorg. Med. Chem. Lett. 18 (2008) 4624–4627
4625
may provide an avenue for the development of anti-obesity thera-
peutics with a novel mechanism of action.1a CA isozymes are crit-
ical to fatty-acid biosynthesis. The mitochondrial CA isozymes
that is, ‘click chemistry’, has become a valued synthetic tool in
medicinal chemistry.7 Rather than being used as a passive linker
between two respective fragments of structural space, the 1,2,3-
triazole moiety can actively contribute to an enzyme-inhibitor
pharmacophore model. The triazole has moderate dipole character,
hydrogen bonding capability, rigidity and stability in the in vivo
environment.8 We have recently demonstrated the versatility of
the click chemistry methodology as an expedient method of gener-
ating both glycoconjugate and metallocene-based hCA inhibi-
tors.9,10 These compounds contain a benzenesulfonamide moiety
linked via a 1,2,3-triazole to either a sugar or a metallocene ‘tail’
moiety, and have so far been evaluated for their inhibition in vitro
of hCA I, II and IX; in some cases also for transmembrane CAs XII
and XIV (Fig. 2).9d
The impressive potency and selectivity of many of these inhib-
itors has prompted us to investigate this efficient chemistry to tar-
get small molecule CA inhibitors towards the mitochondrial CA
isozymes. The compounds described here have been designed with
a lipophilic tail moiety. The lipophilicity should enhance the mem-
brane permeability of these new compounds, an important physi-
cochemical property owing to the intracellular location of the
target CA isozymes.
À
provide the necessary mitochondrial supply of HCO3 substrate
(the mitochondrial membrane is impermeable to HCO3À) for bio-
synthetic processes dependant upon the mitochondrial enzyme
pyruvate carboxylase. Eventually this pathway leads to the forma-
tion of citrate from pyruvate. Citrate is translocated from mito-
chondria to the cytoplasm, then, through a biosynthetic pathway
again requiring HCO3À substrate (provided by cytosolic CA II), leads
to de novo lipogenesis.1a Isozymes hCA VA and VB have different
tissue distribution profiles, with isozyme VA found predominantly
in the liver, whilst isozyme VB has a much wider tissue distribu-
tion. The physiological impact of the tissue distribution profile dif-
ferences is as yet unknown. As well, the pharmacological
explanation for the weight loss observed with ZNS and TPM is still
under investigation. The results so far available from a number of
studies underpin a working hypothesis that inhibition of mito-
chondrial CA isozymes, possibly in conjunction with that of the
ubiquitous cytosolic isozyme CA II, may represent targets for novel
anti-obesity therapies, acting to reduce lipogenesis through inhibi-
tion of these CAs.1c,6
The Cu(I) catalyzed ligation of a terminal acetylene to an organ-
ic azide to form regiospecifically a 1,4-disubstituted-1,2,3-triazole,
A library of 10 benzenesulfonamides (12–21) containing tria-
zole-tethered, phenyl tail groups were synthesized by Cu(I) cata-
lyzed 1,3-DCR of the azido benzenesulfonamide fragment 1 with
a panel of variously substituted phenyl acetylenes (2–11) (Scheme
1).11 The members of this library represent a novel class of com-
pounds with no reported examples of 1,4-disubstituted-1,2,3-tria-
zoles wherein the triazole N-1 substituent is a benzenesulfonamide
and the triazole C-4 substituent derived from a phenyl acetylene
fragment. The variously substituted phenyl acetylene panel
encompassed commercially available compounds as well as 4-
ethynyl benzenesulfonamide (Scheme 1).
The inhibition data of the parent 4-azido benzenesulfonamide
1; new sulfonamides 12–21; and clinically used compounds ZNS
and TPM against the human cytosolic isozymes CA I, CA II, and
the mitochondrial isozymes CA VA and CA VB are presented in Ta-
ble 1. Inhibition constants have been determined using the CA cat-
CA recognition
(zinc binding)
metallocene-triazole
'tail'
O
S
O
CA recognition
(zinc binding)
sugar-triazole
'tail'
NH2
N
N
N
OH
HO
HO
Ru
O
O
S
NH2
O
N
N
N
OH
hCA II Ki = 8.8 nM
hCA II Ki = 9.7 nM
Figure 2. Glycoconjugate and metallocene-based hCA inhibitors.9,10
X
O
O
(i)
Y
+
O
S
N3
X
O
S
N
N
N
Z
H2N
H2N
Z
Y
12-21
1
2-11
X
Y
H
Z
2,12
3,13
H
H
H
CH3
H
H
4,14
CH3
H
H
5,15
H
CH3
H
6,16
OCH3
H
7,17
SO2NH2
H
H
8,18
F
F
H
H
9, 19
10, 20
11, 21
CH3
H
H
OCH3
H
CH3
CF3
H
Scheme 1. Synthesis of benzenesulfonamides with triazole-tethered, variously substituted phenyl tail groups. Reagents and conditions: (i) azide 1 (0.2–0.5 M), arylacetylene
2–11 (1 equiv), CuSO4Á5H2O (0.1–0.2 equiv), sodium ascorbate (0.2–0.4 equiv), 1:1 t-BuOH/H2O, 45 °C, 1-h overnight, 55–88%.11

4626
S.-A. Poulsen et al. / Bioorg. Med. Chem. Lett. 18 (2008) 4624–4627
Table 1
acteristics. Nine of the ten derivatives had micromolar Kis in a com-
pact grouping with a value similar to that of the azido parent 1: Kis
ranged from 2200 to 6500 nM, the Ki of 1 was 3900 nM. The only
outlier in this library was the bis-sulfonamide 17, with two ZBFs
and an inhibition constant an order of magnitude better than for
all other compounds (Ki of 100 nM). This compound can bind to
the hCA active site zinc cation in two possible orientations, the ori-
entation of the triazole moiety being reversed in the two binding
modes. Thus, in the 1,4-diaryl-1,2,3-triazole, the triazole orienta-
tion appears to have a significant impact on hCA I binding that is
worthy of further investigation, and we intend to follow this up
in a future report.
Inhibition data for 1, new sulfonamides 12–21 and standard inhibitors ZNS and TPM
against human isozymes hCA I, II, VA and VB (h = human) 12
O
R
O
S
N
N
N
H2N
Compound
R
Ki (nM)a
CA VAc
CA Ib
CA IIb
CA VBc
ZNS
TPM
1
—
—
—
56
250
3900
35
10
47.0
20
63
56.0
6033
30
55.8
At isozyme hCA II the parent azido scaffold 1 had a Ki of 47 nM.
Seven of the phenyl triazoles (compounds 12, 15–17, 19, 20, 21)
were potent low nanomolar inhibitors of hCA II (Kis ranged from
7.7 to 11.7 nM). Compound 18 (para-fluoro derivative) had a Ki va-
lue similar to that of the azido parent 1 (Kis 40.3 nM), whilst those
for compounds 13 and 14 were in between (Kis 18.6 and 33.8 nM,
respectively). All triazole compounds were selective for hCA II over
hCA I, typically three orders of magnitude (except the bis-sulfon-
amide 17, which is two orders of magnitude selective). These re-
sults demonstrate, as did our earlier work, that the ‘tail’
approach for hCA inhibitors can readily discriminate selectivity be-
tween the physiologically dominant hCA isozymes I and II.9
At the mitochondrial CA isozyme VA the parent azido fragment
1 had a Ki of 56.0 nM. All aryl triazole inhibitors were stronger
inhibitors of hCA VA than ZNS, TPM or 1, with tightly clustered
low nanomolar Kis that ranged from 9.3 to 19.6 nM. The new com-
pounds are typically three orders of magnitude more potent inhib-
itors than that at cytosolic hCA I, and of similar potency for
inhibition of hCA II, with the exception of compounds 14 (meta-
methyl derivative) and 18 (para-fluoro derivative) which were bet-
ter inhibitors of hCA VA than hCA II. At hCA VB the parent azido
scaffold 1 had a Ki of 55.8 nM, equipotent with that at hCA VA.
At this isozyme the aryl triazoles had two distinct inhibition pro-
files—strongest inhibitors were compounds 12, 13, 15, 17, 18, 20,
21 (Ki range of 10.5–12.9 nM) and intermediate potency inhibitors
were compounds 14, 16, 19 (Ki range 51.8–54.2 nM). All com-
pounds were much better inhibitors than ZNS (Ki = 6033 nM),
whilst the strongest inhibitors were ꢀ3-fold better inhibitors than
TPM. As hCA VB has a wider tissue distribution this finding may
prove valuable when discriminating the mitochondrial hCA iso-
zymes is desirable.
Typically the chemical nature of the substituent on the aryl tail
moiety had less impact on the CA inhibition constants than did the
position of the substituent (ortho, meta or para). Compounds 12–15
allow the influence of a methyl group substitution to be assessed in
the ortho, meta and para positions of the phenyl tail. Compound 12
(where R = phenyl, that is, no methyl group substitution) has sim-
ilar Kis to the para- (13) and ortho- (15) methyl substituted phenyl
groups whilst the meta analogue (14) is weaker at hCA II, VA and
VB. Similarly, the other meta modified derivative (19) was also in
the weaker hCA VB inhibitor group described above. These Ki pro-
files indicate that meta substitution is in general detrimental to
hCA inhibition whilst ortho and para substitution patterns do not
effect inhibition constants greatly at either hCA II, VA or VB. There
is one exception to this general SAR: the para-methoxy compound
(16) had a weaker Ki at hCA VB than other para substituted
compounds.
12
13
5100
2100
7.9
17.8
12.8
10.6
10.6
18.6
CH3
14
4000
33.8
19.6
54.2
CH3
15
16
17
18
3000
2200
100
7.7
8.4
9.3
10.6
15.1
14.2
11.4
51.8
11.9
11.2
H3C
OCH3
8.1
SO2NH2
4600
40.3
F
F
19
20
6500
5800
5500
10.7
11.7
8.3
19.1
16.9
17.1
52.3
10.5
12.9
CH3
OCH3
H3C
F3C
21
a
Errors in the range of 5–10% of the reported value, from three determinations.
Human (cloned) isozymes, by the CO2 hydration method.
b
c
Human recombinant full length isozyme, by the CO2 hydration method.12
alyzed CO2 hydration assay, which measures the absorbance
change of a pH indicator and so is directly responsive to the endog-
enous CA catalyzed reaction.
ZNS inhibits isozymes I, II and VA with nanomolar Ki values,
whilst it inhibits more weakly isozyme VB (Ki = 6033 nM). TPM
inhibits all isozymes efficiently, with a Ki of 10 nM for isozyme II,
63 nM for isozyme VA and 30 nM for isozyme VB.
The parent azido benzenesulfonamide fragment (1) had similar
good efficacy against hCA II, VA and VB (Kis of 47.0, 56.0 and
55.8 nM, respectively). This azido compound was also a much
weaker inhibitor of hCA I (Ki = 3900 nM). The inhibition profile
for 1 compared to ZNS and TPM demonstrates that the classical
CA pharmacophore (i.e., an Ar–SO2NH2) is effective at targeting
the CA isozymes of interest in development of potential anti-obes-
ity therapies.
In conclusion, an inhibition study of the human cytosolic iso-
zymes CA I and II, and the mitochondrial isozymes CA and VB with
a novel library of benzenesulfonamides generated by click chemis-
try is presented. These compounds were low to mid nanomolar
inhibitors of hCA isozymes II, VA and VB, and weaker micromolar
inhibitors of hCA I. The inhibition profiles for these novel deriva-
tives should prove valuable lead work in the discovery of isozyme
At the physiologically dominant hCA I the variously substituted
phenyl moieties had minimal influence on enzyme inhibition char-

S.-A. Poulsen et al. / Bioorg. Med. Chem. Lett. 18 (2008) 4624–4627
4627
mechanism, distribution and physiological roles. In Carbonic Anhydrase: Its
Inhibitors and Activators; Supuran, C. T., Scozzafava, A., Conway, J., Eds.; CRC
Press: Boca Raton, FL, 2004; pp 1–24.
selective CA inhibitors targeting the mitochondrial hCA isozymes
VA and VB with potential use as anti-obesity agents.
4. Murphy, M. P.; Smith, R. A. J. Adv. Drug Deliv. Rev. 2000, 41, 235.
5. (a) Dodgson, S. J.; Forster, R. E.; Storey, B. T.; Mela, L. PNAS 1980, 77, 5562; (b)
Fujikawa-Adachi, K.; Nishimori, I.; Taguchi, T.; Onishi, S. J. Biol. Chem. 1999, 274,
21228.
Acknowledgments
This work was financed in part by the Australian Research
Council (Grant No. DP0877554) and the Eskitis Institute for Cell
and Molecular Therapies (Griffith University, S.-A.P.) and an EU
grant of the 6th framework programme (DeZnIT project) (Univer-
sity of Florence, C.T.S.).
6. Supuran, C. T. Expert Opin. Ther. Pat. 2003, 13, 1545.
7. (a) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. Angew. Chem., Int.
Ed. 2002, 41, 2596; (b) TornØe, C. W.; Christensen, C.; Meldal, M. J. Org. Chem.
2002, 67, 3057; (c) Sharpless, K. B.; Manetsch, R. Expert Opin. Drug Discov. 2006,
1, 525.
8. Kolb, H. C.; Sharpless, K. B. Drug Discov. Today 2003, 8, 1128.
9. (a) Wilkinson, B. L.; Bornaghi, L. F.; Houston, T. A.; Innocenti, A.; Supuran, C. T.;
Poulsen, S.-A. J. Med. Chem. 2006, 49, 6539; (b) Wilkinson, B. L.; Bornaghi, L. F.;
Houston, T. A.; Innocenti, A.; Vullo, D.; Supuran, C. T.; Poulsen, S.-A. J. Med.
Chem. 2007, 50, 1651; (c) Wilkinson, B. L.; Innocenti, A.; Vullo, D.; Supuran, C.
T.; Poulsen, S.-A. J. Med. Chem. 2008, 51, 1945; (d) Wilkinson, B. L.; Bornaghi, L.
F.; Houston, T. A.; Innocenti, A.; Vullo, D.; Supuran, C. T.; Poulsen, S.-A. Bioorg.
Med. Chem. Lett. 2007, 17, 987.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2008.07.010.
10. Salmon, A. J.; Williams, M. L.; Innocenti, A.; Vullo, Da.; Supuran, C. T.; Poulsen,
S.-A. Bioorg. Med. Chem. Lett. 2007, 17, 5032.
References
11. General procedure for synthesis of phenyl triazole benzenesulfonamides (12–
21). A mixture of azide (1, 1.0 equiv) and acetylene (2–11, 1.0 equiv) was
suspended in tert-butyl alcohol and distilled water (1:1, 0.2–0.5 M final
concentration). A solution of sodium ascorbate (0.2 equiv) in water, followed
by a solution of CuSO4Á5H2O (0.1 equiv) in water was successively added. The
bright yellow heterogeneous mixture was stirred vigorously at 40 °C until TLC
indicated reaction completion. The mixture was evaporated under reduced
pressure and the resulting residue was purified by flash chromatography. Full
synthesis details and spectroscopic data for 12–21 can be found in the
Supplementary Material.
1. (a) Supuran, C. T. Nat. Rev. Drug Discov. 2008, 7, 168; (b) Picard, F.; Deshaies, Y.;
Lalonde, J.; Samson, P.; Richard, D. Obesity Res. 2000, 8, 656; (c) Gadde, K. M.;
Franciscy, D. M.; Wagner, R. H.; Krishnan, K. R. K. J. Am. Med. Assoc. 2003, 289,
1820.
2. (a) Casini, A.; Antel, J.; Abbate, F.; Scozzafava, A.; David, S.; Waldeck, H.;
Schäfer, S.; Supuran, C. T. Bioorg. Med. Chem. Lett. 2003, 13, 841; (b) De Simone,
G.; Di Fiore, A.; Menchise, V.; Pedone, C.; Antel, J.; Casini, A.; Scozzafava, A.;
Wurl, M.; Supuran, C. T. Bioorg. Med. Chem. Lett. 2005, 15, 2315.
3. (a) Supuran, C. T.; Scozzafava, A. Bioorg. Med. Chem. 2007, 15, 4336; (b)
Pastorekova, S.; Parkkila, S.; Pastorek, J.; Supuran, C. T. J. Enz. Inhib. Med. Chem.
2004, 19, 199; (c) Supuran, C. T. Carbonic anhydrases: catalytic and inhibition
12. The CO2 hydration assay procedure is described in the Supplementary
Material.