X-ray crystallography-promoted drug design of carbonic anhydrase inhibitors

Article abstract of DOI:10.1039/c5cc01854d

1-N-Alkylated-6-sulfamoyl saccharin derivatives were prepared and assayed as carbonic anhydrase inhibitors (CAIs). During X-ray crystallographic experiments an unexpected hydrolysis of the isothiazole ring was evidenced which allowed us to prepare highly

Full text of DOI:10.1039/c5cc01854d

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
X-ray crystallography-promoted drug design of
carbonic anhydrase inhibitors†
Cite this: Chem. Commun., 2015,
51, 7108
Jekaterina Ivanova,‡a Janis Leitans,‡b Muhammet Tanc,‡cd Andris Kazaks,b
Raivis Zalubovskis,*a Claudiu T. Supuran*cd and Kaspars Tarsbe
Received 4th March 2015,
Accepted 20th March 2015
DOI: 10.1039/c5cc01854d
www.rsc.org/chemcomm
1-N-Alkylated-6-sulfamoyl saccharin derivatives were prepared SLC-0111, are also known, this compound being in Phase I
and assayed as carbonic anhydrase inhibitors (CAIs). During X-ray clinical trials for the treatment of patients with advanced solid,
crystallographic experiments an unexpected hydrolysis of the iso- metastatic tumors overexpressing CA IX/XII.3
thiazole ring was evidenced which allowed us to prepare highly
Despite promising achievements in selective inhibition of
potent enzyme inhibitors with selectivity for some isoforms with hCA IX and hCA XII there is still a demand for more effective
medical applications.
and selective inhibitors of various CA isoforms, such as CA II,
VA, VB, IX, etc.2
The artificial sweetener saccharin (SAC) (Fig. 1) was previously
The mechanism of CA inhibition by SAC is rather different
reported as an efficient inhibitor of several isoforms of the compared to that of primary sulfonamides, the most investigated
human metalloenzymes carbonic anhydrases (CAs, EC 4.2.1.1) class of CA inhibitors (CAIs) including those used clinically (AAZ).
with promising selectivity towards the cancer associated iso- Even though in both cases the binding to the Zn ion within the
forms hCA IX and hCA XII,1 both of which have been recently active site of CA takes place by the deprotonated nitrogen of the
validated as drug targets for anti-cancer therapy or imaging sulfonamide group, the SAC binding significantly differs from
of hypoxic tumors.2 It should be noted that CAs are efficiently that of primary sulfonamides. The presence of the acyl group
but indiscriminately inhibited by most sulfonamides such as incorporated into the isothiazole ring and the absence of a
acetazolamide (AAZ) but hCA IX selective inhibitors, such as proton on the nitrogen exhibit a rather different binding pattern
of SAC to the enzyme compared to primary sulfonamides.2
Such different interactions directly reflect the inhibition
profile of SAC, which efficiently inhibits only the cytosolic
isoform hCA VII and the tumor associated one hCA IX compared
to primary sulfonamides such as AAZ, which is a highly efficient
inhibitor of 14 out of the 15 hCAs known to date.1,2 For this purpose
Fig. 1 Chemical structures of known CAIs.
SAC was extensively used as a lead molecule for obtaining novel CAIs
ultimately.4–6 For example, we synthesized 6-sulfamoylsaccharin 1
and its 1-substituted derivatives 2 (Scheme 1)4 where the opportunity
to investigate competition of binding of the primary and secondary
sulfonamide to the enzyme (in the case of 1) emerged.7
Indeed, recently we reported the high resolution X-ray crystal
structure of the adduct of hCA II with 1, which proved that only
the primary sulfonamide participates in the interaction with the
metal ion.7 Thus a series of N-substituted saccharin derivatives
2a–2i appeared to be of interest for being prepared by reacting 1
with alkyl/aralkyl bromides in DMF (Scheme 1, see the ESI† for
details). We investigated the inhibitory properties of these
compounds and their binding to the enzyme by means of kinetic
experiments and X-ray crystallography.
a Latvian Institute of Organic Synthesis, Aizkraukles 21, LV-1006 Riga, Latvia.
E-mail: raivis@osi.lv; Fax: +371-67550338; Tel: +371-67014826
b Biomedical Research and Study Center, Ratsupites 1, LV-1067 Riga, Latvia
c
`
Universita degli Studi di Firenze, NEUROFARBA Department, Section of
Pharmaceutical Chemistry, Via Ugo Schiff 6, 50019 Sesto Fiorentino, Florence,
Italy. E-mail: claudiu.supuran@unifi.it; Fax: +39-055-4573385;
Tel: +39-055-4573005
d
`
Universita degli Studi di Firenze, Polo Scientifico, Laboratorio di Chimica
Bioinorganica, Rm. 188, Via dellaLastruccia 3, 50019 Sesto Fiorentino, Florence,
Italy
e University of Latvia, Faculty of Biology, Department of Molecular Biology,
Kronvalda bulv. 4, LV-1010 Riga, Latvia
† Electronic supplementary information (ESI) available: A detailed description of
the synthesis and characterization of compounds 1–4, as well as the enzyme
inhibitory assays. See DOI: 10.1039/c5cc01854d
In order to visualize the binding mode of saccharin sulfon-
amide derivatives 2 to hCA II, we solved the high resolution
‡ These authors contributed equally to this study.
7108 | Chem. Commun., 2015, 51, 7108--7111
This journal is ©The Royal Society of Chemistry 2015

View Article Online
Communication
ChemComm
Fig. 3 Different binding modes of SAC (grey carbons, thin sticks) and
N-substituted ‘open’ saccharin 3e (black carbons, thick sticks). Both
compounds are coordinating active site Zn ions (grey sphere), but 3e is
bound to the metal ion by its primary sulfonamide whereas SAC by the
secondary, acylated sulfonamide.
Scheme 1 Synthesis of 1-N-substituted 6-sulfamoylsaccharins 2 and
their hydrolysis products 3 and 4.
we thereafter prepared all the corresponding open forms of the
6-sulfamoyl saccharins 1 and 2, obtaining the bissulfamoyl
carboxylic acids 3 and 4 (Scheme 1), under alkaline hydrolytic
conditions. Even though we expected that these carboxylic acids
3 and 4 might undergo a ring closure under neutral or acidic
conditions, we did not observe the isothiazole ring closure even
by storing compounds 3 and 4 for a prolonged period at ambient
temperatures.
All compounds obtained were subjected to CA inhibition
studies summarized in Table 1.
Fig. 2 Comparison of binding modes of compounds 2i, 2e and 2d within the
hCA II active site. Compound 2i/3i is shown in panel A, 2e/3e is shown in panel
B and 2d/3d is shown in panel C. The zinc ion is the gray sphere and its
coordinating residues (His94, 96 and 119) are shown in green. Residues 67, 92,
131, 135, 198, 200 and 202 participating in hydrogen bonding, hydrophobic and
van der Waals contacts with inhibitors are also indicated. For the sake of clarity,
Fo À Fc OMIT electron density is shown only for ligands and contoured at 3s.
The figure was prepared by using Pymol (DeLano ThePyMOL Molecular
Graphics System San Carlos, CA, USA, DeLano Scientific).
Four hCA isoforms were included in this study: two cytosolic
ones hCA I and hCA II, and two tumor-associated transmembrane
isoforms hCA IX and hCA XII, all of which are drug targets for
various applications of their inhibitors.2,3 Data of Table 1 show the
following interesting findings.
Against the slow cytosolic isoform hCA I the activity range of
compounds 1–4 was between 2.6 and 451 nM. Almost all
compounds showed a better inhibition compared to the non-
crystal structures of hCA II in complex with compounds 2i, 2e selective compound AAZ. Only compound 4 showed low inhibition
and 2d reported here. The electron density was interpretable for all against this isoform, whereas derivatives 1, 2a and 3g had a
inhibitors (Fig. 2) and surprisingly revealed that the isothiazole ring comparable inhibition profile to that of AAZ. However the most
was open in all of them. Even though the isothiazole ring opening interesting observation was that in pairs of closed/open ring
occurs most probably by alkaline hydrolysis due to the relatively high derivatives 1/4 and 2/3, the net reduction of the inhibitory activity
pH (of 9.0) in the crystallization buffer it unexpectedly and clearly for the open forms 4 and 3 (1.7 to up to 43 times) compared to the
revealed a new possibility of designing CAIs. One should mention corresponding closed form ones 1 and 2 occurred. The most
that initially we explored the possibility that the enzyme itself significant reduction of activity, by two orders of magnitude, was
hydrolyzed the amide bond from derivatives 2, but this did not observed for compounds 3e and 3g, which were 27 and 43 times,
occur (data not provided). Indeed, although the CAs have esterase respectively, less inhibitory compared to the corresponding benz-
and thioesterase activity,8,9 they do not possess peptidase activity. isothiazoles 2e and 2g. The only exceptions to this rule were the
Notably all three compounds were bound in a very similar fashion, pairs 2a/3a and 2h/3h, for which the closed form was less inhibitory
coordinating to the zinc ion with their primary sulfonamide, than the open ones (Table 1).
whereas oxygens of carboxyl and sulfone groups made H-bonds
For the rapid isoform hCA II a similar inhibition pattern was
with the Asn67 and Gln92 side chains. The R moieties occupied a observed as for hCA I discussed above. All compounds except 4
hydrophobic pocket, formed by the side chains of residues Phe131, showed an excellent, better inhibitory activity than AAZ, with
Val135, Leu198, Leu204 and Pro202 (Fig. 2).
Kis in the range of 0.2–8.4 nM. A similar reduction of the
The observed binding mode of all three compounds is inhibitory activity of the open versus the closed forms was also
substantially different from that previously reported for unsub- observed with most compounds, but already many of the closed
stituted saccharin (PDB code 2Q38),1 or derivative 1,7 when the ones were low nanomolar hCA II inhibitors and thus, this
two inhibitors were not hydrolysed. Thus, the binding observed for reduction seems to be less relevant for this isoform than for
compounds 3, obtained by hydrolysis of derivatives 2 reported here, hCA I. The nature of the R group also influenced the inhibition
is indeed very different compared to other saccharin based CAIs pattern of these derivatives significantly. Thus, an increase of the
reported so far4–7 (Fig. 3). Inspired by these crystallographic results aliphatic chain from C2 to C4 led to an increase of the hCA II
This journal is ©The Royal Society of Chemistry 2015
Chem. Commun., 2015, 51, 7108--7111 | 7109

View Article Online
ChemComm
Communication
Table 1 CA inhibition data of isoforms hCA I, II, IX and XII with saccharin
derivatives 1–2 and the corresponding open forms 3–4 reported in this
communication, by a CO2 hydrase stopped-flow assay12
synthetic procedures, and as shown above, they possess notable
inhibitory properties, with a profile quite different from that of the
structurally related, closed form (or the primary sulfonamide AAZ).
In fact all sulfonamides 3a–3i were highly effective, CA II-selective
inhibitors, and this type of profile is very rare or even absent among
the many sulfonamide CAIs reported so far.10 Furthermore, the
crystallographic experiments (Fig. 2 and 3) also showed that the R
moiety present in these compounds may adopt a variety of
orientations within the CA II active site, which may explain their
very high affinity for this isoform and the relatively lower ones for
other isoforms such as hCA I, IX and XII (Table 1). As hCA II is the
main target for designing anti-glaucoma CAIs (in clinical use for
decades but with many side effects due to inhibition of other
isoforms),11 these findings may lead to the design of water-soluble
(due to the presence of the COOH moiety, which may form sodium
salts), highly effective and selective hCA II inhibitors belonging to a
novel chemical space.
In conclusion we report here new CAIs obtained by a ‘side
reaction’ which occurred during an X-ray crystallographic study
of sulfonamide–CA adducts. We have demonstrated the high
potential of the newly obtained compounds (open/closed forms
of 1-N-substituted saccharins or the unsymmetrically substituted
bissulfamoyl benzoic acids), possessing an improved selectivity
towards some CA isoforms with medical applications. Considering
the chemical simplicity and good water solubility of the newly
Ki (nM)
Compound
R
hCA I hCA II hCA IX hCA XII
1
—
Et
nPr
nBu
251
257
49
4.8
4.3
4.6
2.6
8.4
1.0
0.6
0.6
3.5
0.4
0.4
0.2
0.8
0.4
1.7
0.2
1.1
6.0
0.7
1.5
2.2
0.3
2.8
31.7
12
337
452
278
51.2
380
271
51.8
52.9
321
378
92.8
89.5
130
333
78.1
67.4
73.6
70.5
71.6
42.1
25
52.9
7.2
5.9
5.7
5.2
9.5
5.8
7.9
14.3
6.7
77.7
63.6
58.7
68.3
67.8
50.5
195
27.0
48.6
63.3
5.7
2a
2b
2c
2d
2e
2f
2g
2h
2i
3a
3b
3c
3d
3e
3f
3g
3h
3i
nC5H11
CH2CHCHMe
Bn
CH2C6H4(4-NO2) 4.9
CH2C6H4(4-Br)
CH2CH2Ph
Et
nPr
nBu
nC5H11
CH2CHCHMe
Bn
CH2C6H4(4-NO2) 213
CH2C6H4(4-Br)
CH2CH2Ph
—
—
57.8
9.1
66.2
81.4
41.4
29.9
125
64.0
38.7
59.0
451
250
4
AAZa
a
Acetazolamide (AAZ) was used as a standard inhibitor for all CAs
investigated in this communication.
inhibitory properties but a further increase to C5 was detrimental obtained CAIs, their scaffold may find applications in the develop-
for the inhibitory activity (compare 2d to 2a–c, Table 1). However, ment of new types of CAIs, probably by modulating the nature of the
unsaturated or aralkyl chains (as in 2e–2i) led again to highly moieties substituting in position 1 the saccharin derivatives (the R
effective, subnanomolar CAIs, for all the substitution patterns of moiety). Indeed, in this communication we explored few substitution
compounds 2e–2i, i.e., benzyl, 4-substituted benzyl moieties or patterns which are aliphatic, alkenyl and aralkyl groups. By extend-
phenethyl.
ing the type and nature of these moieties, which as shown in the
An opposite inhibition pattern was observed in the case of the crystal structures, interact with amino acid residues critical for the
tumor-associated transmembrane isoform hCA IX with compounds binding of inhibitors, compounds with improved potency and
2–4 reported here. Even though none of the compounds was selectivity may presumably be obtained.
superior to AAZ, the inhibitory activity increased going from the
This research was in part financed by two FP7 EU projects
closed to the open forms for the compound pair 1/4 and most of (InnovaBalt and Dynano).
the pairs 2/3. As shown in Table 1, the highest increase, more than
4 times, was observed for compounds 3a, 3h, 3i and 4 with Kis in
the range of 42.1–92.9 nM, which are effective inhibitors of this
tumor-associated isoform.
Notes and references
¨
1 K. Kohler, A. Hillebrecht, J. Schulze Wischeler, A. Innocenti,
A. Heine, C. T. Supuran and G. Klebe, Angew. Chem., Int. Ed. Engl.,
2007, 46, 7697.
For the second transmembrane isoform, hCA XII, the inhibi-
tion pattern was similar to those of the cytosolic isoforms hCA I
and hCA II discussed above. All closed forms except 1 and 2h
exhibited comparable inhibitory activity with AAZ, whereas the
open forms 3a–3f, 3h–4 were around one order of magnitude less
inhibitory compared to AAZ. Overall, many low nanomolar hCA
XII inhibitors were detected such as for instance 2a–2i, which
had inhibition constants ranging between 5.2 and 14.3 nM, in
the same range as the classical sulfonamide inhibitor AAZ.
The most interesting finding of this communication is
however the fact that our drug design has been guided by the
crystallographic work, which evidenced a hydrolytic process
taking place during the crystallization experiments. Unexpectedly,
the hydrolysis afforded compounds possessing a free COOH moiety
in addition to the primary and secondary sulfamoyl moieties. This
type of sulfonamide was in fact not synthesized so far using other
2 (a) V. Alterio, A. Di Fiore, K. D’Ambrosio, C. T. Supuran and G. De
Simone, Chem. Rev., 2012, 112, 4421; (b) C. T. Supuran, Nat. Rev.
Drug Discovery, 2008, 7, 168.
3 (a) See more at ClinicalTrails.gov: Safety Study of SLC-0111 in
Subjects With Advanced Solid Tumours-ClinicalTrials_gov.mht;
(b) F. Pacchiano, F. Carta, P. C. McDonald, Y. Lou, D. Vullo,
A. Scozzafava, S. Dedhar and C. T. Supuran, J. Med. Chem., 2011,
54, 1896; (c) C. T. Supuran, J. Enzyme Inhib. Med. Chem., 2012,
27, 759; (d) C. T. Supuran, J. Enzyme Inhib. Med. Chem., 2013, 28, 229.
4 E. M. Ivanova, E. Y. Simin, I. V. Vozny, P. Trapencieris and
ˇ
R. Zalubovskis, Chem. Heterocycl. Compd., 2012, 47, 1561.
5 M. D’Ascenzio, S. Carradori, C. De Monte, D. Secci, M. Ceruso and
C. T. Supuran, Bioorg. Med. Chem., 2014, 22, 1821.
6 (a) J. Moeker, T. S. Peat, L. F. Bornaghi, D. Vullo, C. T. Supuran and
S. A. Poulsen, J. Med. Chem., 2014, 57, 3522; (b) B. P. Mahon,
A. M. Hendon, J. M. Driscoll, G. M. Rankin, S. A. Poulsen,
C. T. Supuran and R. McKenna, Bioorg. Med. Chem., 2015, 23, 849.
7 V. Alterio, M. Tanc, J. Ivanova, R. Zalubovskis, I. Vozny, S. M. Monti,
A. Di Fiore, G. De Simone and C. T. Supuran, Org. Biomol. Chem.,
2015, 13, 4064.
7110 | Chem. Commun., 2015, 51, 7108--7111
This journal is ©The Royal Society of Chemistry 2015

View Article Online
Communication
ChemComm
8 (a) Y. Pocker and J. T. Stone, J. Am. Chem. Soc., 1965, 87, 5497;
(b) A. Innocenti, A. Scozzafava, S. Parkkila, L. Puccetti, G. De Simone
and C. T. Supuran, Bioorg. Med. Chem. Lett., 2008, 18, 226; (c) H. Çavdar,
D. Vullo, A. Maresca, F. Carta, A. Scozzafava and C. T. Supuran,
Chem. Commun., 2012, 48, 3551; (d) J. Y. Winum, A. Maresca,
F. Carta, A. Scozzafava and C. T. Supuran, Chem. Commun., 2012,
˘
´
D. Ekinci, O. Talaz, N. Saraçoglu, M. -Sentu¨rk and C. T. Supuran, J. Enzyme
48, 8177; (e) B. Metayer, A. Mingot, D. Vullo, C. T. Supuran and
˘
Inhib. Med. Chem., 2012, 27, 148; (d) E. A. Kazancıoglu, M. Gu¨ney,
S. Thibaudeau, Chem. Commun., 2013, 49, 6015.
M. -Sentu¨rk and C. T. Supuran, J. Enzyme Inhib. Med. Chem., 2012, 27, 880. 11 (a) A. Maresca, F. Carta, D. Vullo and C. T. Supuran, J. Enzyme Inhib.
9 M. Tanc, F. Carta, A. Scozzafava and C. T. Supuran, ACS Med. Chem.
Lett., 2015, 6, 292.
10 (a) F. Pacchiano, M. Aggarwal, B. S. Avvaru, A. H. Robbins,
A. Scozzafava, R. McKenna and C. T. Supuran, Chem. Commun.,
2010, 46, 8371; (b) A. Di Fiore, A. Maresca, V. Alterio, C. T. Supuran
Med. Chem., 2013, 28, 407; (b) S. M. Monti, A. Maresca, F. Carta,
G. De Simone, F. A. Mu¨hlschlegel, A. Scozzafava and C. T. Supuran,
Bioorg. Med. Chem. Lett., 2012, 22, 859; (c) F. Carta, M. Aggarwal,
A. Maresca, A. Scozzafava, R. McKenna, E. Masini and C. T. Supuran,
J. Med. Chem., 2012, 55, 1721.
and G. De Simone, Chem. Commun., 2011, 47, 11636; (c) S. Parkkila, 12 R. J. Khalifah, J. Biol. Chem., 1971, 246, 2561.
This journal is ©The Royal Society of Chemistry 2015
Chem. Commun., 2015, 51, 7108--7111 | 7111