1,3-Dipolar Cycloaddition, HPLC Enantioseparation, and Docking Studies of Saccharin/Isoxazole and Saccharin/Isoxazoline Derivatives as Selective Carbonic Anhydrase IX and XII Inhibitors

Article abstract of DOI:10.1021/acs.jmedchem.9b01434

Two series of saccharin/isoxazole and saccharin/isoxazoline hybrids were synthesized by 1,3-dipolar cycloaddition. The new compounds showed to be endowed with potent and selective inhibitory activity against the cancer-related human carbonic anhydrase (hCA) IX and XII isoforms in the nanomolar range, while no affinity was encountered for off-targets, such as hCA I and II. Successive enantioseparation on a milligram scale of the most representative compounds led to the discovery that (S)-isomers were more potent than their corresponding (R)-enantiomers. Lastly, molecular modeling studies were conducted to define those structural requirements that were responsible for the discrimination among selected human isoforms of carbonic anhydrases. Two nanomolar hCA IX and XII inhibitors were also screened for their selective toxicity against non tumoral primary cells (fibroblasts) and against a breast adenocarcinoma cell line (MCF7) in hypoxic environment. The efficacious combination of these compounds with doxorubicin on MCF7 cells was demonstrated after 72 h of treatment.

Full text of DOI:10.1021/acs.jmedchem.9b01434

Subscriber access provided by ECU Libraries
Article
1,3-Dipolar cycloaddition, HPLC enantioseparation and docking
studies of saccharin/isoxazole and saccharin/isoxazoline
derivatives as selective carbonic anhydrase IX and XII inhibitors
Melissa D'Ascenzio, Daniela Secci, Simone Carradori, Susi Zara, Paolo Guglielmi, Roberto Cirilli,
Marco Pierini, Giulio Poli, Tiziano Tuccinardi, Andrea Angeli, and Claudiu Trandafir Supuran
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b01434 • Publication Date (Web): 23 Jan 2020
Downloaded from pubs.acs.org on January 25, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the
course of their duties.

Page 1 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1,3-Dipolar cycloaddition, HPLC enantioseparation and docking
studies of saccharin/isoxazole and saccharin/isoxazoline derivatives as
selective carbonic anhydrase IX and XII inhibitors
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Melissa D’Ascenzioa§, Daniela Seccia,*, Simone Carradorib, Susi Zarab, Paolo Guglielmia, Roberto
Cirillic, Marco Pierinia, Giulio Polid, Tiziano Tuccinardid, Andrea Angelie,f, Claudiu T. Supurane
aDipartimento di Chimica e Tecnologie del Farmaco, Sapienza University of Rome, P.le A. Moro 5,
00185 Rome, Italy
bDepartment of Pharmacy, “G. d’Annunzio” University of Chieti-Pescara, Via dei Vestini 31,
66100 Chieti, Italy
cCentro nazionale per il controllo e la valutazione dei farmaci, Istituto Superiore di Sanità, Viale
Regina Elena 299, 00161 Rome, Italy
dDepartment of Pharmacy, University of Pisa, Via Bonanno 6, 56126 Pisa, Italy
eNeurofarba Department, Section of Pharmaceutical and Nutraceutical Sciences, University of
Florence, Via U. Schiff 6, 50019 Sesto Fiorentino (Florence), Italy
fCentre of Advanced Research in Bionanoconjugates and Biopolymers Department, “Petru Poni”
Institute of Macromolecular Chemistry, 41A Grigore Ghica-Voda Alley, 700487 Iasi, Romania
*Corresponding author: Daniela Secci, Dipartimento di Chimica e Tecnologie del Farmaco,
Sapienza University of Rome, P.le A. Moro 5, 00185 Rome, Italy. E-mail:
daniela.secci@uniroma1.it
§Current affiliation: D'Arcy Thompson Unit, School of Life Sciences, University of Dundee,
Dundee (Scotland). E-mail: m.dascenzio@dundee.ac.uk
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 56
1
2
3
4
Abstract
5
6
7
8
Two series of saccharin/isoxazole and saccharin/isoxazoline hybrids were synthesised by 1,3-
dipolar cycloaddition. The new compounds showed to be endowed with potent and selective
inhibitory activity against the cancer-related human carbonic anhydrase (hCA) IX and XII isoforms
in the nanomolar range, while no affinity was encountered for off-targets such as hCA I and II.
Successive enantioseparation on a mg-scale of the most representative compounds led to the
discovery that (S)-isomers were more potent than their corresponding (R)-enantiomers. Lastly,
molecular modelling studies were conducted in order to define those structural requirements that
were responsible for the discrimination among selected human isoforms of carbonic anhydrases.
Two nanomolar hCA IX and XII inhibitors were also screened for their selective toxicity against
non tumoral primary cells (fibroblasts) and against a breast adenocarcinoma cell line (MCF7) in
hypoxic environment. The efficacious combination of these compounds with doxorubicin on MCF7
cells was demonstrated after 72 h of treatment.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Keywords: Saccharin, isoxazole, isoxazoline, chiral HPLC, docking, carbonic anhydrase inhibitors.
Introduction
Carbonic anhydrases (CAs, EC 4.2.1.1) are ubiquitous metalloenzymes that catalyse the
interconversion of carbon dioxide and water into bicarbonate and hydrogen ions.1 Among the seven
CA families, αCA are further categorized in 15 diverse human (h) isoforms playing pivotal roles in
a multitude of physiological functions and pathological conditions.2 hCA I and II are cytosolic
proteins with high enzymatic efficiency that are constitutively expressed in all tissues. Conversely,
hCA IX and XII are membrane-associated proteins that can be overexpressed in the hypoxic tumour
environment, with CA IX not significantly present in the majority of healthy tissues.3 Owing to the
peculiar role of these multidomain enzymes in many solid tumours, compounds that are able to act
against the extracellular facing carbonic anhydrases IX and XII are currently under investigation.4,5
ACS Paragon Plus Environment

Page 3 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
In fact, recent research efforts have led to the identification of a large plethora of CAs inhibitors
characterized (or not) by a zinc binding group, as the presence of a metal ion in the active site
constitutes one of the most conserved features within this family of enzymes.6-11
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In particular, saccharin derivatives have been reported as good hCAs inhibitors by various research
groups that, in the last few years, focused their attention toward the development of new derivatives
based on this scaffold.12-17 Analysing the structure and substitution pattern of previously published
inhibitors possessing a saccharin core, it was highlighted that two general approaches had been
exploited for the design of new compounds (Figure 1): the first strategy relied on the substitution of
the benzene ring of saccharin while maintaining the cyclic secondary sulfonamide functional group
in its unsubstituted form (Figure 1, I and II). This approach produced inhibitors endowed with
activity in the nanomolar range against isoforms IX and XII, although residual activity against the
off-targets hCA I and hCA II was unfortunately retained.15,18
1st approach
R
N
O
S
O
O
S
O
O
S
O
NH
N
HN
R
NH
N
NH
O
O
O
O
Saccharin
I
II
2nd approach
H
H
O
S
O
S
O
S
O
N
N
N
R
R
O
O
OH
R
O
OH
O
III
IV
V
Figure 1. The two different approaches used for the synthesis of saccharin-based hCAs inhibitors.
The second approach was based on the substitution of the cyclic secondary sulfonamide with
different groups in order to obtain N-substituted saccharins (Figure 1, III). These molecules were
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 56
1
2
3
4
5
6
7
8
9
effective inhibitors of hCA IX and XII, even if a few of them retained activity against off-target
isoforms.14,17 Successive efforts focused on obtaining the “open” form of these compounds using
two synthetic strategies: the reductive opening of the heterocyclic ring to give a secondary
sulfonamide and benzylic alcohol (Figure 1, IV),13 or the hydrolysis of the cyclic sulfonamide
under basic conditions to generate a carboxylic acid and a secondary sulfonamide (Figure 1, V).19
Our research group contributed mostly to the development of N-substituted saccharins, which meant
that a compound obtained by 1,3-dipolar cycloaddition between N-propargyl substituted saccharin
and N-hydroxy-4-nitrobenzimidoyl chloride was already reported in one of our first works,13,14,17
among others (Figure 2).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
O
insertion of
isoxazole
"linker"
evaluation of
isoxazole/isoxazoline
ring as "linker"
O
O
O
S
O
N
S
S
N
N
O
O
O
O
Ar/Het
O
N
N
R
NO2
(nM) hCAI/II > 50.000
(nM) hCAIX = 310
(nM) hCAXII = 240
KI
KI
KI
Figure 2. Design of new saccharin/isoxazoline and saccharin/isoxazole derivatives.
The first compound endowed with a heterocyclic “linker” between the methylene and the phenyl
ring showed activity in the nanomolar range only against tumor-related isoforms, while being
ineffective against the two main off-targets. In light of the above, we designed two series of new
inhibitors obtained by inserting two different heterocyclic systems as linker groups: the isoxazole
and isoxazoline ring, respectively. The chosen linkers possess some common characteristics, e.g.
they are both five-membered rings containing nitrogen and oxygen, they display identical
connectivity and position of substituents. However, the isoxazole ring is a flat-aromatic system that
can give п-stacking interactions inside the active site; on the other hand, isoxazoline - the 4,5-
dihydrogenated form of isoxazole – is not aromatic and possesses two tetrahedral carbon atoms that
ACS Paragon Plus Environment

Page 5 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
alter the flat conformation (although п-interactions could still be established using the double bond
in the ring). Furthermore, the presence of four different substituents on one of the two tetrahedral
carbon (C5 of isoxazoline) accounts for the existence of a stereogenic center. Therefore, all the
saccharin/isoxazoline compounds can exist as a pair of enantiomers that can be resolved and tested
in order to determine the configuration preferred by the enzymes and, in conclusion, the eutomer.
The inhibition data obtained from the biological assays were further corroborated through molecular
modelling studies including docking and molecular dynamics (MD) simulations. In particular, we
focused our attention on the most interesting compounds in order to determine the interactions
established inside the active site of hCAs that could help rationalise activity and selectivity. Two
potent hCA IX and XII inhibitors were selected to further assess their effects on viability and
cytotoxicity in an in vitro model of normal human primary cells (fibroblasts), and, at the same time,
the antiproliferative effect of hCA IX and XII inhibitors in combination with doxorubicin, on an
adenocarcinoma breast cancer cell line (MCF7 cells).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Chemistry
The synthesis of this small library of new saccharin derivatives followed a multi-step approach,20,21
which led to the preparation of the appropriate reagents for the final 1,3-dipolar cycloaddition, i.e.
allyl/propargyl N-substituted saccharins (S1-S2) and hydroxyiminoyl acid chlorides (chloro-oximes
C1-C9). Final compounds (1a-9a and 1b-9b) were obtained in moderate yields (Scheme 1).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 56
1
2
3
O
H
4
5
6
7
O
Ar/Het
O
S
NH2OH . HCl
NH
(b)
MeOH, Et3N,
80 °C
O
8
9
N
H
Allyl/propargyl bromide
DMF, K2CO3, 80 °C
HO
(a)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Ar/Het
(O1-O9)
NCS, HCl(g)
DMF, 50 °C
O
O
(c)
S
N
N
Cl
O
HO
S1 (N-allyl) S2 (N-propargyl)
Ar/Het
(C1 C9)
-
-
(d)
1,3 dipolar cycloaddition
Et3N, anhydrous EtOAc
O
O
S
N
Ar/Het
O
N
O
O
(1a 9a 1b 9b)
-
;
-
N
Cl
NCS =
Ar = phenyl, 3-/4-NO2-phenyl, 2-/3-/4-OCH3-phenyl
Het = thiophen-2-yl, pyridin-2/3-yl
O
Scheme 1. Synthetic approach for the synthesis of new derivatives 1a-9a and 1b-9b.
N-alkylated saccharins (S1-S2) were obtained following our previously reported procedure,17 which
involved a nucleophilic substitution reaction between saccharin and allyl/propargyl bromide
(Scheme 1, a). The hydroximinoyl acid chlorides (chloro-oximes) were synthesised in two steps
from the corresponding benzaldehydes.22 The first step (Scheme 1, b) required the reaction between
the appropriate aldehyde with hydroxylamine hydrochloride in the presence of triethylamine in
methanol (80 °C) to give the corresponding oximes (O1-O9). At a later time, the freshly
synthesised oximes (O1-O9) were reacted with N-chlorosuccinimide (NCS) in N,N-
dimethylformamide (DMF) in the presence of catalytic amounts of gaseous hydrogen chloride
(HCl(g)) (Scheme 1, c). In fact, Howe et al. observed that the reaction of NCS with benzaldoximes
exhibited an induction period that could be reduced through the addition of small amounts of
gaseous HCl. In this way, we avoided the fairly exothermic behaviour observed when the reaction
ACS Paragon Plus Environment

Page 7 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
initiates after a considerable portion of NCS has been added.22 The resulting chloro-oximes were
not contaminated by side-products (i.e. chlorinated ring products) and could be used without
purification in the next 1,3-dipolar cycloaddition step. In this reaction, the 1,3-dipole was generated
in situ by dehydrohalogenation of the corresponding chloro-oxime (C1-C9) and then reacted with
the dipolarophile (N-allyl/propargyl saccharins) in dry ethyl acetate (Scheme 1, d). The
regioselectivity of the isoxazoline ring formation is usually controlled by the energy of the
interacting orbitals (frontier molecular orbitals, FMO), although in some cases the electronically
preferred orientation might be disfavoured due to steric effects.23 However, the reaction of nitrile
oxides with monosubstituted olefins is quite exclusive, giving only (or predominantly) 5-substituted
isoxazolines, regardless of the nature of the substituent on the dipolarophile. Similarly, the
cycloaddition of nitrile oxides to acetylenic dipolarophiles leads to 5-substituted isoxazoles.21,23
The purity of all final compounds was >95% as determined by HPLC analysis, employing two
different chromatographic methods for isoxazoline and isoxazole derivatives. The compounds were
analysed using gradient elution with a binary mobile phase composed by water and acetonitrile, and
were shown to be >95% HPLC pure (Figures S4-S10; S15-S23). The compounds 8a and 9a were
evaluated for their purity after chiral resolution through the analytical checks of the enantiomeric
purity of each enantiomer (Figures S11-S14).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Two-dimensional (2D) Nuclear Overhauser Enhancement Spectroscopy (NOESY) NMR
The regioselectivity of the cycloaddition step was studied using compound 3b as a model.
Compound 3b was synthesised by the reaction of N-hydroxy-pyridine-2-carbimidoyl chloride and
N-propargyl saccharin. 2D-NMR (NOESY) analysis conducted on compound 3b demonstrated a
spatial relationship between proximal protons that could be highlighted in 5-substituted isoxazoles
(Figure 3, a) but not in their 4-substituted analogues (Figure 3, b), thus confirming the overall
tendency of this type of 1,3-dipolar cycloadditions to produce 5-substituted rings.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 56
1
2
3
4
5
6
7
8
9
b
a
O
O
N
H
H
O
O
S
5
N
H
S
4
N
O
N
N
N
O
O
O
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Structural isomers originated by the 1,3-dipolar cycloaddition of N-hydroxy-pyridine-2-
carbimidoyl chloride and N-propargyl saccharin: (a) compound 3b; (b) alternative structure
proposed for compound 3b.
Chiral resolution of compounds 8a and 9a
The reaction between N-allyl saccharin and chloro-oximes C1-C9 determined the generation of a
chiral centre (carbon C5) on the isoxazoline ring. In an attempt to determine whether hCA isoforms
would show any preference for one enantiomer over the other, compounds 8a and 9a were
submitted to enantioselective HPLC separation. Pure enantiomers of each compound were obtained
in mg-amount and submitted for further assays (Figure 4).
O
O
O
chiral
HPLC
O
O
O
S
S
S

+
N
N
N


Ar/Het
Ar/Het
Ar/Het
O
O
O
N
N
N
O
O
O
Figure 4. Resolution of saccharin/isoxazoline compounds by chiral HPLC.
All the synthesized saccharin/isoxazoline and saccharin/isoxazole derivatives were then tested to
evaluate their inhibitory activity against the ubiquitous off-target isoforms, hCA I and II, as well as
the cancer-related ones, hCA IX and XII, using a stopped-flow CO2 hydrase assay method.24 The
hCAs inhibition data (KI) are summarized in Table 1. Ki ratios between target and off-target
isoforms were also reported in square brackets as a measure of compound selectivity. The hCA
ACS Paragon Plus Environment

Page 9 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
tested in this work were in line with those reported in previous works, for better comparison and to
improve the SAR within the saccharin-based scaffold.
7
8
Results and discussion
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Two-dimensional Nuclear Overhauser Enhancement NMR
The interaction between the two protons highlighted in red in Figure 3 was highlighted in the
NMR-NOESY spectrum reported in Figure 5 below.
Figure 5. NMR-NOESY of compound 3b.
The interaction between protons with spatial proximity was confirmed by the presence of the
“green” circles along the ideal line that connects proton C4(H) of the isoxazole ring (singlet at
~7.06 ppm) and the pyridine signals between 7.90 and 8.10 ppm, which contain the proton at the C3
position of the pyridine ring. This outcome confirmed the regioselectivity of the reactions in our
experimental conditions.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 56
1
2
3
4
Chiral resolution of compounds 8a and 9a
5
6
7
8
The semipreparative enantioseparation of 8a and 9a was carried out on coated amylose-based
Chiralpak AS-H chiral stationary phase using pure ethanol as a mobile phase. In Figure 6 we
reported the mg-scale enantioseparation and analytical checks of the enantiomeric purity of the
collected enantiomers. Both enantiomeric separations gave yields of about 85% and enantiomeric
excess of >98% for both enantiomers (S11-S14).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(R) (S)
(R) (S)
8a
9a
(a)
(b)
0
5
10
15
20
0
5
10
15
20
25
(c)
(d)
3
6
9
12
15
3
6
9
12
15
Elution time (min)
Elution time (min)
Figure 6. Traces (a): typical chromatogram illustrating the semipreparative enantioseparation of 3
mg of 8a and 2 mg of 9a on a 250 x 10 mm i.d. AS-H column. Mobile phase: ethanol; flow rate: 2.5
mL/min; temperature: 40 °C; detector: UV at 280 nm. Traces (b): analytical resolution of 8a and 9a.
Traces (c) and (d): purity check of the enantiomers collected at a semipreparative scale. Mobile
phase: ethanol; flow rate: 0.8 mL/min; temperature: 40 °C; detector: UV at 254 nm.
Assignment of the absolute configuration to compounds 8a and 9a
ACS Paragon Plus Environment

Page 11 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The absolute configuration of both enantiomers of 8a and 9a was determined by comparison of the
experimental and computed electronic circular dichroism (ECD) spectra (Figure 7). A detailed
description of the adopted procedure is provided in the Experimental Section. In brief, the obtained
spectra allowed us to assign the (S)-configuration to the more retained species and the
(R)-configuration to the first eluted enantiomers, as shown in Figure 7 for both compounds 8a and
9a.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. Experimental (full line) and calculated (dashed line) ECD spectra of the enantiomers of
8a (left side) and 9a (right side) in ethanol. In both graphs the first enantiomer eluted from the
chiral HPLC column is shown as a green line, while the second one is shown as a red line. The red
dashed traces refer to the ECD spectra calculated for the enantiomer (S)-8a and (S)-9a, respectively.
Inhibition of hCA I, II, IX, and XII
Comparing the activity of compounds 1a-9a and 1b-9b with the N-allyl/propargyl saccharin parent
drugs (S1 and S2), it was possible to assess that the insertion of an isoxazoline/isoxazole ring
bearing aromatic or heteroaromatic substituents improved both activity and isoform selectivity. In
particular, the effects of these changes were more pronounced with regards to selectivity. As a
matter of fact, data reported in Table 1 showed that, except for derivatives (R)-9a and (S)-9a, all the
synthesised compounds were selective inhibitors of hCA IX and hCA XII. Moreover, a preferential
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 56
1
2
3
4
5
6
7
8
9
affinity for isoform XII was observed, with KI spanning from 1.0 nM ((S)-9a) to 240 nM (derivative
2b). The nature of the linker (isoxazoline rather than isoxazole) or the substituents on the phenyl
ring, as well as the heterocyclic system, influenced the inhibitory activity against the tested
isoforms. Due to the relatively small size of the two series, it was difficult to establish a clear
structure-activity trend. However, a few interesting observations could be made: compounds 1a and
1b, characterised by the presence of a NO2 group on the meta position of the phenyl substituent,
showed similar inhibitory activity against isoform XII (1a, KI hCA XII = 76.5 nM; 1b, KI hCA XII
= 57.3 nM), while compound 1a was more effective than 1b against isoform IX (1a, KI hCA IX =
102.9 nM; 1b, KI hCA IX = 230.2 nM). Compounds 2a and 2b, both possessing a p-NO2 substituted
phenyl ring, displayed a similar trend in hCA IX inhibition, while compound 2b (containing the
isoxazole linker) showed impaired activity against isoform XII (2a, KI hCA XII = 77 nM; 2b, KI
hCA XII = 240 nM). In the case of these four derivatives (1a-2a, 1b-2b) it could be noted that
starting from 1a, containing a m-NO2 substituted phenyl ring bound to the isoxazoline ring, to
compound 2b endowed with p-NO2 substituted phenyl ring connected to the isoxazole system, there
was a decrease of activity against hCA IX (Table 1). This trend was opposite compared to
derivatives spanning from 6a to 7b, which were characterised by the presence of a methoxy
substituent (OMe). Indeed, compound 7b - endowed with an isoxazole linker and p-OMe
substituted phenyl ring - showed good inhibitory activity against isoform IX (KI hCA IX = 22.1
nM). All the other derivatives possessing a methoxy group were found to be inactive (5a, 5b, 6a),
or at least less effective (6b and 7a) against isoform IX, when compared to compound 7b. On the
other side, they were good inhibitors of hCA XII (7.0 < KI hCA XII (nM) < 9.4), with the best
inhibitory activity exhibited by compound 6b. These data seemed to suggest that the nature of the
linker did not favour one isoform over the other and did not affect enzyme inhibition in a specific
manner. In fact, the transition from an isoxazoline ring in 1a to the isoxazole ring of 1b increased
the activity against hCA XII, while reduced the inhibitory effect against hCA IX. In the case of
compounds 2a-2b, we observed a reduction of activity against both tumour-related isoforms when
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 13 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
passing from the isoxazoline linker (2a) to an isoxazole ring (2b). The change of linker type for
pairs 6a-6b and 7a-7b only influenced the inhibitory activity against hCA IX, while hCA XII was
not affected, showing almost comparable inhibition data. Albeit further substitution patterns still
need to be evaluated, these preliminary results showed that the nature of the heteroaromatic linker,
together with the type and position of aromatic substituent, could have a synergistic effect on the
inhibitory activity of the corresponding derivatives, rather than being independent features.
Bioisosteric heterocyclic systems bound to the linker moiety were also evaluated. For example,
compounds 3a and 3b were endowed with a pyridin-2-yl moiety. These derivatives showed similar
activity against hCA IX (3a, KI hCA IX = 258.2 nM; 3b, KI hCA IX = 245.4 nM), whereas
compound 3a exhibited better activity against hCA XII compared to its isoxazole containing
analogue (3a, KI hCA XII = 9.7 nM; 3b, KI hCA XII = 66.5 nM). Derivatives 4a and 4b, which
were characterised by the presence of a pyridin-3-yl system, showed comparable activity against
isoform XII, while they differed for activity on hCA IX (4a, KI hCA IX = 273.5 nM; 4b, KI hCA IX
= 86.2 nM). These data suggested that the influence of the linker is not unequivocal. Compounds
containing phenyl (8b) or thiophene (9b) substituents bound to the isoxazole ring showed low
nanomolar inhibition of hCA XII (8b, KI hCA XII = 7.5 nM; 9b KI hCA XII = 8.2 nM), while
derivative 9b displayed better affinity than 8b with regards to hCA IX (8b, KI hCA IX = 50.9 nM;
9b, KI hCA IX = 7.5 nM). Therefore, the five-membered heterocyclic system seemed to be the best
tolerated group bound to the linker. Finally, the evaluation of potential chiral recognition operated
by specific hCAs isoforms was evaluated by subjecting previously purified enantiomers of
compounds 8a and 9a to biological testing. The enantiomers of compound 8a were ineffective
towards off-targets hCA I and hCA II, while their inhibitory activity against hCA IX and XII was
deemed to be comparable. Therefore, we could conclude that in this case the enzymes did not
exhibit a “chiral preference” between the two enantiomers (Table 1). A different outcome was
obtained for the enantiomers of compound 9a. The (S)-enantiomer of derivative 9a lacked that
selectivity for the tumour-related isoforms that was recorded when both enantiomers were tested as
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 56
1
2
3
4
5
6
7
8
9
a racemic mixture, as it inhibited both the off-targets hCA I and II (KI hCA I = 365 nM; KI hCA II =
13.1 nM). However, it also exhibited a very low nanomolar affinity for hCA XII ((S)-9a, KI hCA
XII = 1.0 nM) and hCA IX ((S)-9a, KI hCA IX = 2.4 nM), resulting in the most potent inhibitor of
the series. The (R)-enantiomer of compound 9a was inactive against hCA II, showing poor residual
activity towards isoform I ((R)-9a, KI hCA I = 358.0 nM), while the activity against the two tumour
related isoforms remained in the low nanomolar range ((R)-9a, KI hCA IX = 26.4 nM; KI hCA XII =
8.6 nM), although inferior to that of the (S) enantiomer. In the light of the above, we could conclude
that due to the chiral properties of compound (S)-9a, this is the eutomer of the enantiomeric pair.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1. Inhibition data of selected human CA isoforms (hCA I, II, IX and XII) with compounds
1a-9a and 1b-9b reported here and the standard sulfonamide inhibitor acetazolamide (AAZ) by a
stopped flow CO2 hydrase assay.
O
O
S
KI (nM) a
N
R
Compd
O
hCA I [hCA I/hCA hCA II [hCA hCA IX [hCA IX
R
hCA XII
XII]b
II/hCA XII]b
/hCA XII] b
S117
S217
>10000 [>106.4]
86 [0.9]
140 [1.5]
94
>10000 [>17.5]
58 [0.1]
210 [0.4]
570
NO2
NO2
1a
1b
>10000 [>130.7] >10000 [>130.7]
102.9 [1.3]
76.5
57.3
O
O
N
N
>10000 [>174.5] >10000 [>174.5]
>10000 [>129.9] >10000 [>129.9]
230.2 [4.0]
NO2
2a
2b
3a
235.1 [3.1]
310 [1.3]
77.0
240
9.7
O
N
>10000 [>41.7]
>10000 [>41.7]
NO2
O
N
O
N
>10000
>10000 [>1030.9]
258.2 [26.6]
[>1030.9]
N
ACS Paragon Plus Environment

Page 15 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
N
3b
4a
4b
>10000 [>150.4] >10000 [>150.4]
245.4 [3.7]
273.5 [29.7]
86.2 [10.8]
66.5
9.2
O
O
O
N
N
N
N
N
>10000 [>1087]
>10000 [>1250]
>10000 [>1087]
>10000 [>1250]
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
8.0
H3CO
5a
5b
6a
6b
>10000 [>1063]
>10000 [>1075]
>10000 [>1315.8]
>10000 [>1428.6]
>10000 [>1063] >10000 [>1063]
>10000 [>1075] >10000 [>1075]
9.4
9.3
7.6
7.0
O
O
N
H3CO
N
OCH3
OCH3
>10000
>10000 [>1315.8]
[>1315.8]
O
O
N
>10000
247.1 [35.3]
[>1428.6]
N
>10000
7a
7b
>10000 [>1234.6]
>10000 [>1250]
139.1 [17.2]
[>1234.6]
8.1
8.0
OCH3
O
O
N
OCH3
>10000 [>1250]
22.1 [2.8]
N
(R)-8a
(S)-8a
>10000 [>204.5] >10000 [>204.5]
30.0 [0.6]
29.3 [0.8]
48.9
36.5
O
N
N
>10000 [>274]
>10000 [>274]
>10000
8b
>10000 [>1333.3]
50.9 [6.8]
7.5
O
[>1333.3]
>10000
[>1162.8]
13.1 [13.1]
(R)-9a
(S)-9a
358 [41.6]
365 [365]
26.4 [3.1]
2.4 [2.4]
8.6
1.0
O
N
S
>10000
9b
>10000 [>1219.5]
7.5 [0.9]
8.2
O
[>1219.5]
N
S
AAZ
250 [43.8]
12 [2.1]
25 [4.4]
5.7
57
58
59
60
aMean from 3 different assay, by a stopped flow technique (errors were in the range of  5-10% of
the reported values). bKIs ratio for the indicated enzyme isoforms.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 56
1
2
3
4
5
6
Molecular modelling studies
7
8
9
Molecular modelling studies, including docking and molecular dynamics simulations, were
performed with the purpose of identifying a reliable binding mode into hCA XII and hCA IX for
this novel series of saccharin ligands, in an attempt to rationalise their activity and selectivity profile
(Figure 8). Compound 8b, which showed to be one of the most potent and selective hCA XII and
hCA IX inhibitor, was used as a reference compound of the series and docked into the X-ray
structures of the two enzyme isoforms (PDB codes 1JD0 and 3IAI for hCA XII and hCA IX,
respectively). The software GOLD with GoldScore fitness function was used for this study. The
ligand-protein complexes predicted by docking were then refined and analysed through MD
simulation studies. As shown in Figure 8, according to the binding mode predicted by the
computational protocol, compound 8b interacted with hCA XII by coordinating the prosthetic zinc
ion in the active site of the enzyme with one of the two sulfonamide oxygen atoms belonging to the
saccharin scaffold, while the other oxygen of the sulfonamide fragment forms an H-bond with the
hydroxyl group of T227 that is maintained for most of the MD simulation. This latter oxygen shows
an additional H-bond with the backbone oxygen of T227. A further H-bond is then observed
between the carbonyl oxygen of the ligand and the amide group of Q117 side chain. The phenyl
ring of the ligand saccharin core is placed in a mainly hydrophobic pocket constituted by W32,
N92, H94, H121 and H119, forming a stable π-π stacking with the imidazole ring of this latter
residue and lipophilic interactions with the other residues. Finally, the phenylisoxazole group of
compound 8b interacts with the residues belonging to the lipophilic side of the enzyme catalytic
site, forming hydrophobic contacts with V147, S158, S161, L167, L225, and P229. The binding
mode predicted for the ligand into hCA IX was similar to that generated for hCA XII. In fact,
compound 8b showed the same orientation into the hCA IX binding cavity, with the
phenylisoxazole moiety laying on the hydrophobic wall of the catalytic site and the saccharin
scaffold placed in proximity of the zinc ion, with the fused phenyl ring located into the small pocket
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 17 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
delimited by W141, N198, H200, Q203, H226 and H228. The ligand forms an H-bond with Q224
as observed in hCA XII (with the homologue residue Q117) through the carbonyl oxygen and
coordinates the zinc ion with a sulfonamide oxygen. Moreover, the compound forms the two H-
bonds with the backbone nitrogen of T333 (homologue of T227 in hCA XII) and the hydroxyl
group of the same residue.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 8. Minimized average structures of compound 8b (green) in complex with (A) hCA XII
(PBD code: 1JD0) and (B) hCA IX (PBD code: 3IAI). The protein surface surrounding the ligand is
shown in grey.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 56
1
2
3
4
5
6
7
8
9
In order to get insightful information on the selectivity profile of this series of ligands, the same
docking protocol applied on hCA XII and hCA IX was also used to evaluate the potential binding
mode of compound 8b into hCA II and hCA I. The docking results suggested that the binding
disposition described for the ligand into hCA XII and hCA IX could not be assumed into the other
two CA isoforms, due to the steric hindrance of non-conserved residues. Precisely, the presence of
F130 on the hydrophobic side of hCA II binding pocket, with its benzyl side chain oriented toward
the centre of the catalytic site, would hamper the proper disposition of the ligand phenylisoxazole
group, thus preventing the compound to assume the binding mode predicted into the tumour-related
hCA isoforms (Figure S1). Differently, the presence of few non-conserved residues within the
catalytic site of hCA I would totally occlude the small pocket where the phenyl ring of the ligand
saccharin scaffold should be placed according to the binding mode predicted into hCA XII and hCA
IX. In particular, the side chain of H201 of hCA I, that replaces the threonine residues T227 and
T333 of hCA XII and hCA IX, respectively, is directed inside the pocket, thus preventing the
binding of the ligand (Figure S2). These results and considerations could provide a possible
explanation for rationalizing the selectivity profile of compound 8b and the related series of
compounds. Finally, the binding dispositions into hCA XII and hCA IX of the two enantiomers (R)-
8a and (S)-8a, obtained by replacing the isoxazole ring of 8b with an isoxazoline linker, were
evaluated using the same computational protocol comprising docking and MD studies. The results
obtained were consistent with the biological activities determines for the two compounds, since they
showed a very similar binding mode into both hCA XII and hCA IX, which was also comparable to
that predicted for their analogue 8b into the two enzyme isoforms. As shown in Figure S3, the two
8a enantiomers coordinate the zinc ion with their sulfonamide moiety and place the phenyl ring of
the saccharin scaffold into the lipophilic pocket formed by W32, N92, H94, H119 and H121 in hCA
XII and W141, N198, H200, Q203, H226 and H228 in hCA IX. The phenylisoxazoline group of
both ligands is oriented toward the hydrophobic wall of the two enzymes and the same pattern of H-
bond interactions predicted for 8b can be observed, namely, a couple of H-bonds with the hydroxyl
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 19 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
group and the backbone nitrogen of T227/T333 (of hCA XII/IX) and a third H-bond with the side
chain of Q117/Q224 (of hCA XII/IX). At present, we could not identify a satisfying rationale for
explaining the reduced selectivity for hCA XII and hCA IX of compound (S)-9a, nor its increased
activity against the two CA isoforms with respect to either its enantiomer (R)-9a or its close
analogues 9b, 8b and (R/S)-8a. Further in-depth computational analyses will be carried out in order
to investigate and rationalize this experimental evidence.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Biological evaluation on normal cells
According to the hCA inhibitory data reported in Table 1 and keeping in mind the possibility for
nitro-containing molecules to be considered as “red flags” in Medicinal Chemistry, two compounds,
2b and 7b, were selected among the most potent as nanomolar hCA IX and XII inhibitors. More in
detail, compounds 2b and 7b are characterized by a structurally similar skeleton (saccharin-
isoxazoles), whereas they only differ for the presence of an electron-withdrawing moiety (NO2 in
2b) or an electron-donating group (OCH3 in 7b) at the para position of the aryl ring. They were
tested in in vitro models, represented by both normal and tumoral cells, in order to evaluate the
antiproliferative and cytotoxic effect of the two above mentioned compounds.
Firstly, compounds 2b and 7b were screened and tested on normal human cells represented by
gingival fibroblasts (HGFs). Doses ranging from to 0.05 to 200 µM were selected administrating
compounds 2b and 7b up to 72 h of culture. To evaluate the cytotoxicity, a Lactate Dehydrogenase
(LDH) assay, which quantitatively measures LDH released upon cell lysis in the medium, was
performed; 24, 48 and 72 h of culture were chosen as experimental times (Figure S24). Regarding
compound 2b (nitro-containing molecule), at all experimental times there are not statistically
significant differences between the tested doses (Figure S24 A), conversely, when compound 7b
(methoxy-containing molecule) is administered, after 24 h of culture, a statistically significant
reduction in LDH released % is recorded in HGFs treated with compound 7b 200 µM and 100 µM
with respect to control sample (DMSO) and to all other tested concentrations. After 48 and 72 h of
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 56
1
2
3
4
5
6
treatment with compound 7b no significant differences are detected among the tested concentrations
(Figure S24, B).
7
8
9
Secondly, after 24, 48 and 72 h of culture, a metabolic activity test (MTT assay), which is
descriptive of cell viability rate, was performed on HGFs treated with compounds 2b and 7b. Again
for compound 2b no significant differences among the tested concentrations are recorded at all
experimental times (Figure S25, A). Conversely, when HGFs are treated with compound 7b a
statistically significant decrease in viable cells % could be clearly evidenced after 24 and 48 h with
200, 100 and 50 µM doses with respect to cells treated with 5, 0.5, 0.05 µM and DMSO. After 72 h
of culture, a decrease of viable cells % is detected only for cells treated with 200 and 100 µM
compared to HGFs treated with 5, 0.5, 0.05 µM and to DMSO sample (Figure S25, B).
Taken together, these results clearly evidence that, in terms of both cytotoxicity and cell viability,
the nitro-containing compound 2b does not affect the pivotal biological processes occurring within
the cells; in fact, even when administered at very high concentrations, the proliferation rate and the
cytotoxic effect are comparable with control samples, thus indicating a good tolerability of
molecule 2b in normal cells. Concerning compound 7b, it can be argued that an appreciable
tolerability is recorded for doses lower than 50 µM, however, even when compound 7b is
administered at very high doses (100 and 200 µM), the cell viability never goes below 50% thus
underlining a high safety profile despite the substitution pattern.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Biological evaluation on breast adenocarcinoma cells (MCF7)
Compounds 2b and 7b were then tested in a tumoral biological in vitro model represented by MCF7
breast adenocarcinoma cells. It is commonly recognized that breast cancer is as a chemoresistant
tumor25 and this often represents the main reason for inadequate therapeutic strategies. Considering
that doxorubicin is a first-line drug for breast cancer therapy, we aimed at evaluating the co-
treatment of compounds 2b and 7b with doxorubicin on MCF7 cells for 72 h. hCA IX and XII
inhibitors were shown to counteract the overexpression of these cancer-related CA isozymes
involved in tumor progression, metastasis development and angiogenesis in cancer cells adapted to
ACS Paragon Plus Environment

Page 21 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a hypoxic environment. For this reason, before administering compounds 2b and 7b in combination
with doxorubicin, a pretreatment with CoCl2 was carried out to induce hypoxia for 48 h. This
approach is based on the well-established chemical induction of HIF-1α exerted by cobalt(II)
chloride hexahydrate in order to overexpress CA IX and XII isozymes in MCF7 cells.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A dose–response evaluation, obtained by a MTT test and using doxorubicin doses ranging from
0.05 to 200 µM, was firstly performed aiming to select a subtoxic dose to be applied in combination
with molecules 2b and 7b (Figure 9A). Given that our results show that, after 72 h of treatment,
doxorubicin IC50 can be included between 50 and 5 µM, we chose 2.5 µM as a suitable subtoxic
dose to be combined with 2b and 7b concentrations ranging from 0.05 to 200 µM.
MTT test, performed on MCF7 treated with compound 2b + doxorubicin 2.5 µM for 72 h, discloses
no statistically significant differences among the tested concentrations (Figure 9B). Whereas when
MCF7 are treated with compound 7B in combination with doxorubicin 2.5 µM for 72 h, a
statistically significant reduction of viable cell percentage is recorded in 200 µM sample respect to
all other experimental conditions. A similar trend is detectable for cells treated with 100 µM
concentration with respect to 5, 0.5, 0.05 µM and control (CoCl2 + DMSO) samples (Figure 9C).
These data let us to hypothesize that probably MCF7 cells are chemoresistant to treatment
performed with compound 2b, even when combined with subtoxic doses of doxorubicin while the
combination at high doses of compound 7b with doxorubicin 2.5 µM could represent a
chemosensitizer strategy to treat breast adenocarcinoma cells. To elucidate this hypothesis a
morphological analysis, by means of phase contrast light microscope, was also carried out. The
obtained results, shown in Figure 10, are quite surprising as taken pictures clearly show that MCF7
cells treated with compound 2b, at both 100 and 200 µM, are round shaped, close to detachment,
similarly to cells treated with compound 7b. This could be explained admitting that MCF7 cells
treated with compound 2b, after 72 h, are still viable, as recognized by the viability assay (Figure
9), but they seem to be early affected by the compound. This led us to assume that compound 2b, to
affect the viability of tumoral cells, probably needs a longer treatment respect to compound 7b.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 56
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 9. MTT cell viability assay on MCF7 cell line. A Histogram represents the viability dose-
response of MCF7 cells exposed to different concentrations of doxorubicin (from 0.05 to 200 μM)
for 72 h. B and C Histograms represent the viability dose-response of MCF7 cells exposed to
different concentrations of compounds 2b and 7b (from 0.05 to 200 μM), respectively, in
combination with doxorubicin 2.5 μM (subtoxic dose) for 72 h. Proliferation was assessed using
MTT assay and normalized to control cells treated with DMSO (0.2% as final concentration) +
CoCl2 + doxorubicin 2.5 μM. The most representative of three separate experiments is shown. Data
are presented as the mean ± standard deviation.
* p<0.005; **p<0.0005; ***p<0.0002.
ACS Paragon Plus Environment

Page 23 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 10. Phase contrast microscopy images of MCF7 cells treated with doxorubicin, compounds
2b and 7b for 72 h. A: DMSO, B: DMSO + CoCl2 , C: Doxorubicin 2.5 µM, D: Doxorubicin 5 µM;
E: 2b 100 µM + doxorubicin 2.5 µM, F: 2b 200 µM + doxorubicin 2.5 µM, G: 7b 100 µM +
doxorubicin 2.5 µM, H: 7b 200 µM + doxorubicin 2.5 µM. Adherent and viable cells are detectable
in samples A and B, whereas a high number of detached and floating cells is appreciable in all other
experimental points, particularly in samples G and H. Magnification ×20.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 56
1
2
3
4
5
6
7
8
9
Considering that doxorubicin is very efficient in vitro, when given as monotherapy, but can be
cardiotoxic in vivo at therapeutic levels, our results confirmed the putative role of hCA IX and XII
inhibitors as chemosensitizers or coadjuvants to reduce antitumoral drug concentration and thus
limiting the possibility to develop harmful side effects. In addition, there is a strong need to develop
combined therapies to define synergistic or additive actions improving the therapeutic index of the
therapy or decrease adaptive resistance.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Conclusions
In this paper we report on the development of two series of potent and selective hCAs inhibitors
based on the well-established saccharin scaffold. On the basis of the previously reported “tail
approach”, we explored the introduction of isoxazole and isoxazoline linkers for the decoration of
saccharin cyclic sulfonamide. In fact, in silico studies suggested that a phenyl-isoxazole moiety
could be important for the establishment of pivotal interactions with the lipophilic side of the
catalytic site, while the sulfonamide fragment was predicted to be responsible of both zinc
coordination and H-bonds formation with key anchoring residues. In vitro results evidenced that
compounds belonging to both series were characterised by strong affinity for hCA IX and XII and
marked selectivity over hCA I and II. Moreover, two of the newly synthesised compounds were
submitted to HPLC enantioseparation to assess the enzyme ability to discriminate between the two
enantiomers.
It is also worth noting that two of the most potent nanomolar hCA IX and XII inhibitors did not
displayed any cytotoxic effect up to 200 μM on primary human fibroblasts despite the substitution
pattern (nitro or methoxy moiety), whereas compound 7b was shown to act as chemosensitizer and
coadjuvant in combination with subtoxic doses of doxorubicin on MCF7 breast cancer cell line.
Overall, these in vitro activity data are encouraging for further future development of this scaffold.
Experimental protocols
ACS Paragon Plus Environment

Page 25 of 56
Journal of Medicinal Chemistry
1
2
3
Chemistry
4
5
6
7
8
Unless otherwise indicated, all reactions were carried out under a positive pressure of nitrogen
(balloon pressure) in washed and oven-dried glassware. Solvents were used as supplied without
further purification. Where mixtures of solvents are specified, the stated ratios are volume:volume.
Reagents for the chemical and biological experiments were used directly as supplied by Sigma-
Aldrich® Italy, unless otherwise stated. All melting points were measured on a Stuart® melting point
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
13
apparatus SMP1, and are uncorrected. Temperatures are reported in °C. H and C NMR spectra
were recorded at 400 and 101 MHz, respectively, on a Bruker spectrometer using CDCl3, DMSO-
d6, CD3OD and CD3CN, as the solvents at room temperature. The samples were analysed with a
final concentration of ~30 mg/mL. Chemical shifts are expressed as δ units (parts per millions)
1
relative to the solvent signal. H spectra are described as follows: δH (spectrometer frequency,
solvent): chemical shift/ppm (multiplicity, J-coupling constant(s), number of protons, assignment).
13C spectra are described as follows: δC (spectrometer frequency, solvent): chemical shift/ppm
(assignment). Multiplets are abbreviated as follows: br – broad; s – singlet; d – doublet; t – triplet; q
– quartet; m – multiplet. Coupling constants J are valued in Hertz (Hz). Infra-red spectra were
recorded on a Bruker Tensor 27 FTIR spectrometer equipped with an attenuated total reflectance
attachment with internal calibration. Absorption maxima (νmax) are reported in wavenumbers (cm-1).
Column chromatography was carried out using Sigma-Aldrich® silica gel (high purity grade, pore
size 60 Å, 230-400 mesh particle size). All the purifications and reactions were carried out by TLC
performed on 0.2 mm thick silica gel-aluminium backed plates (60 F254, Merck). Visualization was
performed under ultra-violet irradiation (254 and 365 nm). Where given, systematic compound
names are those generated by ChemBioDraw Ultra 12.0 following IUPAC conventions.
The purity of all final compounds was >95% as determined by HPLC analyses performed using the
Shimadzu Prominence-i LC-2030C 3D system, that integrates autosampler, binary pump, column
oven and a photodiode array detector (PDA). The separation was performed using the column
Kinetex Evo C18, 5 µm, 100 Å (Phenomenex, Bologna, Italy) at a constant flow of 1.25 mL min-1,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 56
1
2
3
4
5
6
7
8
9
employing a binary mobile phase elution gradient. The two eluents used were water (solvent A) and
acetonitrile (solvent B) mixed in different elution gradients for isoxazole and isoxazoline
derivatives. For the compounds endowed with an isoxazole moiety the initial eluent composition
was 50% of solvent B. This concentration has been kept constant for 3 min, then raised up to 90%
in 7 min and maintained at this value for 1 min. At the end of this time the concentration returned at
the initial value for the column reconditioning (4 min), to a total run time of 15 min
(chromatograms are reported as Supporting Information).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For the derivatives bearing the isoxazole moiety the chromatographic run started with 30% of
solvent B, that also in this case was kept constant for 3 min. The concentration was then raised up to
90% in 7 min, and maintained at this value for 1 min. After this time the concentration returned at
the initial value for the column reconditioning (4 min), to a total run time of 15 min. With the
exception of compound 2a which exhibited scarce solubility, all the analyte solutions were prepared
in acetonitrile at the concentration of about 1 mg mL-1, and 2 µL were directly injected for the
HPLC analysis. All compounds reported were > 95% HPLC pure. The solvents used in the HPLC
analysis were acetonitrile, purchased from Carlo Erba Reagents and mQ water 18 MΩ cm, obtained
from Millipore’s Direct-Q3 system.
Synthesis and characterization data for saccharin derivatives S1-S2
2-Allylbenzo[d]isothiazol-3(2H)-one 1,1-dioxide (S1) and 2-(prop-2-yn-1-yl)-1,2-benzothiazol-
3(2H)-one 1,1-dioxide S2. See reference [17] for synthesis procedure and characterization data.
Synthesis and characterization data for oximes O1-O9
3-Nitrobenzaldehyde oxime O1. In an oven-dried flask, hydroxylamine hydrochloride (1.0 g, 0.014
mol) and triethylamine (3.0 mL, 0.021 mol) were dissolved in 10 mL of anhydrous methanol and
stirred for 5 min at room temperature. 3-Nitrobenzaldehyde (1.45 g, 0.0096 mol) was added and the
reaction mixture stirred at 80 °C until complete consumption of the starting material. The reaction
was cooled down to room temperature and the solvent removed in vacuo. 20 mL of ice-cold water
were added to the residue and the resulting suspension was kept at 0-4 °C overnight in order to
ACS Paragon Plus Environment

Page 27 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
obtain oxime precipitate, which was filtered and washed with a mixture of petroleum ether/diethyl
ether, obtaining the title compound as a light yellow solid (1.35 g, 85% yield). See reference [26]
for characterization data.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4-Nitrobenzaldehyde oxime O2. See reference [17] for synthesis procedure and characterization
data.
Pyridine-2-aldoxime O3. In an oven-dried flask, hydroxylamine hydrochloride (1.0 g, 0.014 mol)
and triethylamine (3.0 mL, 0.021 mol) were dissolved in 10 mL of anhydrous methanol and stirred
for 5 min at room temperature. Pyridine-2-carboxaldehyde (0.92 mL, 0.0096 mol) was added and
the reaction mixture stirred at 80 °C until complete consumption of the starting material. The
reaction was cooled down to room temperature and the solvent removed in vacuo. 20 mL of ice-
cold water were added to the residue and the resulting suspension was kept at 0-4 °C overnight in
order to obtain oxime precipitate, which was filtered and washed with a mixture of petroleum
ether/diethyl ether, obtaining the title compound as a white solid (0.82 g, 70% yield). See reference
[27] for characterization data.
Pyridine-3-aldoxime O4. In an oven-dried, flask hydroxylamine hydrochloride (1.0 g, 0.014 mol)
and triethylamine (3.0 mL, 0.021 mol) were dissolved in 10 mL of anhydrous methanol and stirred
for 5 min at room temperature. Pyridine-3-carboxaldehyde (0.90 mL, 0.0096 mol) was added and
the reaction mixture stirred at 80 °C until complete consumption of the starting material. The
reaction was cooled down to room temperature and the solvent removed in vacuo. 20 mL of ice-
cold water were added to the residue and the resulting suspension was kept at 0-4 °C overnight in
order to obtain oxime precipitate, which was filtered and washed with a mixture of petroleum
ether/diethyl ether, obtaining the title compound as a white solid (1.07 g, 92% yield). See reference
[28] for characterization data.
2-Methoxybenzaldehyde oxime O5. In an oven-dried, flask hydroxylamine hydrochloride (1.0 g,
0.014 mol) and triethylamine (3.0 mL, 0.021 mol) were dissolved in 10 mL of anhydrous methanol
and stirred for 5 min at room temperature. 2-Methoxybenzaldehyde (1.30 g, 0.0096 mol) was added
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 56
1
2
3
4
5
6
7
8
9
and the reaction mixture stirred at 80 °C until complete consumption of the starting material. The
reaction was cooled down to room temperature and the solvent removed in vacuo. 20 mL of ice-
cold water were added to the residue and the resulting suspension was kept at 0-4 °C overnight in
order to obtain oxime precipitate, which was filtered and washed with a mixture of petroleum
ether/diethyl ether, obtaining the title compound as a light yellow solid (1.29 g, 89% yield). See
reference [26] for characterization data.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3-Methoxybenzaldehyde oxime O6. In an oven-dried, flask hydroxylamine hydrochloride (1.0 g,
0.014 mol) and triethylamine (3.0 mL, 0.021 mol) were dissolved in 10 mL of anhydrous methanol
and stirred for 5 min at room temperature. 3-Methoxybenzaldehyde (1.30 g, 0.0096 mol) was added
and the reaction mixture stirred at 80 °C until complete consumption of the starting material. The
reaction was cooled down to room temperature and the solvent removed in vacuo. 20 mL of ice-
cold water were added to the residue and the resulting suspension was extracted with ethyl acetate
(3 x 20 mL). The organics were reunited, dried over sodium sulphate and evaporated in vacuo to
give the title compound as thick oil (1.27 g, 88% yield). See reference [26] for characterization data.
4-Methoxybenzaldehyde oxime O7. In an oven-dried flask, hydroxylamine hydrochloride (1.0 g,
0.014 mol) and triethylamine (3.0 mL, 0.021 mol) were dissolved in 10 mL of anhydrous methanol
and stirred for 5 min at room temperature. 4-Methoxybenzaldehyde (1.30 g, 0.0096 mol) was added
and the reaction mixture stirred at 80 °C until complete consumption of the starting material. The
reaction was cooled down to room temperature and the solvent removed in vacuo. 20 mL of ice-
cold water were added to the residue and the resulting suspension was kept at 0-4 °C overnight in
order to obtain oxime precipitate, which was filtered and washed with a mixture of petroleum
ether/diethyl ether, obtaining the title compound as a white solid (1.29 g, 89% yield). See reference
[26] for characterization data.
Benzaldehyde oxime O8. In an oven-dried flask, hydroxylamine hydrochloride (1.0 g, 0.014 mol)
and triethylamine (3.0 mL, 0.021 mol) were dissolved in 10 mL of anhydrous methanol and stirred
for 5 min at room temperature. Benzaldehyde (0.97 mL, 0.0096 mol) was added and the reaction
ACS Paragon Plus Environment

Page 29 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
mixture stirred at 80 °C until complete consumption of the starting materials. The reaction was
cooled down to room temperature and the solvent removed in vacuo. 20 mL of ice-cold water were
added to the residue and the resulting suspension was extracted with ethyl acetate (3 x 20 mL). The
organics were reunited, dried over sodium sulphate and evaporated in vacuo to give the title
compound as colourless oil (0.87 g, 75% yield). See reference [26] for characterization data.
Thiophene-2-carbaldehyde oxime O9. In an oven-dried flask, hydroxylamine hydrochloride (1.0 g,
0.014 mol) and triethylamine (3.0 mL, 0.021 mol) were dissolved in 10 mL of anhydrous methanol
and stirred for 5 min at room temperature. Thiophene-2-carbaldehyde (1.07 g, 0.0096 mol) was
added and the reaction mixture stirred at 80 °C until complete consumption of the starting materials.
The reaction was cooled down to room temperature and the solvent removed in vacuo. 20 mL of
ice-cold water were added to the residue and the resulting suspension was kept at 0-4 °C overnight
in order to obtain oxime precipitate, which was filtered and washed with a mixture of petroleum
ether/diethyl ether, obtaining the title compound as a light brown solid (0.77 g, 63% yield). See
reference [27] for characterization data.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis and characterization data for chloro oximes C1-C9
N-hydroxy-3-nitrobenzimidoyl chloride C1. In an oven-dried flask, m-nitrobenzaldoxime O1 (1 g,
0.006 mol) was dissolved in 10 mL of anhydrous DMF. N-chlorosuccinimide (0.16 g, 0.0012 mol)
was added and the reaction stirred for 10 min at room temperature. The solution was purged with 20
mL of HCl(g), stirred for 10 additional min and then heated up to 50 °C. N-chlorosuccinimide was
added portionwise (0.16 g or 0.0012 mol at the time, up to 1.20 g or 0.009 mol) over half an hour.
The progression of the reaction was monitored using iodine starch paper. When the iodine starch
paper was not turning brown anymore, the reaction was quenched with 4 volumes of ice-cold water.
The aqueous phase was extracted with diethyl ether (3 x 20 mL), the organics reunited were dried
over sodium sulphate and evaporated in vacuo to give the title compound as a yellow-orange solid
(0.83 g, 69% yield). See reference [29] for characterization data.
ACS Paragon Plus Environment