Discovery and characterization of novel selective inhibitors of carbonic anhydrase IX

Article abstract of DOI:10.1021/jm501003k

Human carbonic anhydrase IX (CA IX) is highly expressed in tumor tissues, and its selective inhibition provides a potential target for the treatment of numerous cancers. Development of potent, highly selective inhibitors against this target remains an unmet need in anticancer therapeutics. A series of fluorinated benzenesulfonamides with substituents on the benzene ring was designed and synthesized. Several of these exhibited a highly potent and selective inhibition profile against CA IX. Three fluorine atoms significantly increased the affinity by withdrawing electrons and lowering the pK_a of the benzenesulfonamide group. The bulky ortho substituents, such as cyclooctyl or even cyclododecyl groups, fit into the hydrophobic pocket in the active site of CA IX but not CA II, as shown by the compound's co-crystal structure with chimeric CA IX. The strongest inhibitor of recombinant human CA IX's catalytic domain in human cells achieved an affinity of 50 pM. However, the high affinity diminished the selectivity. The most selective compound for CA IX exhibited 10 nM affinity. The compound that showed the best balance between affinity and selectivity bound with 1 nM affinity. The inhibitors described in this work provide the basis for novel anticancer therapeutics targeting CA IX.

Full text of DOI:10.1021/jm501003k

Article
pubs.acs.org/jmc
Discovery and Characterization of Novel Selective Inhibitors of
Carbonic Anhydrase IX
Virginija Dudutiene,† Jurgita Matuliene,† Alexey Smirnov,† David D. Timm,† Asta Zubriene,†
̇
̇
̇
Lina Baranauskiene,† Vaida Morkunaite,† Joana Smirnoviene,† Vilma Michailoviene,† Vaida Juozapaitiene,†
̇
̇
̇
̇
̇
̅
Aurelija Mickeviciute,† Justina Kazokaite,† Sandra Baksyte,† Aiste Kasiliauskaite,† Jelena Jachno,†
̌
̌
̇
̇
̇
̇
̇
̅
Jurgita Revuckiene,† Migle Kisonaite,† Vilma Pilipuityte,† Egle Ivanauskaite,† Goda Milinaviciute,†
̌
̌
̇
̇
̇
̇
̇
̇
̇
̅
Vytautas Smirnovas,† Vilma Petrikaite,†,∥ Visvaldas Kairys,‡ Vytautas Petrauskas,† Povilas Norvaisas,†
̌
̇
†
†
†
†
†
̌
Darius Linge, Paulius Gibieza, Edita Capkauskaite, Audrius Zaksauskas, Egidijus Kazlauskas,
̌
̌
̇
̇
Elena Manakova,§ Saulius Grazulis,§ John E. Ladbury,⊥ and Daumantas Matulis*,†
̌
†Department of Biothermodynamics and Drug Design, Institute of Biotechnology, Department of Bioinformatics, Institute of
‡
Biotechnology, §Department of Protein−DNA Interactions, Institute of Biotechnology, Vilnius University, V. A. Graici
̌
uno 8, Vilnius
̅
LT-02241, Lithuania
∥Department of Drug Chemistry, Faculty of Pharmacy, Lithuanian University of Health Sciences, A. Mickevici
LT-44307, Lithuania
⊥Department of Biochemistry and Molecular Biology, University of Texas MD Anderson Cancer Center, 1515 Holcombe Boulevard,
Houston, Texas 77030, United States
̌
aus 9, Kaunas
S
* Supporting Information
ABSTRACT: Human carbonic anhydrase IX (CA IX) is highly expressed
in tumor tissues, and its selective inhibition provides a potential target for
the treatment of numerous cancers. Development of potent, highly
selective inhibitors against this target remains an unmet need in anticancer
therapeutics. A series of fluorinated benzenesulfonamides with sub-
stituents on the benzene ring was designed and synthesized. Several of
these exhibited a highly potent and selective inhibition profile against CA
IX. Three fluorine atoms significantly increased the affinity by withdrawing electrons and lowering the pKa of the
benzenesulfonamide group. The bulky ortho substituents, such as cyclooctyl or even cyclododecyl groups, fit into the
hydrophobic pocket in the active site of CA IX but not CA II, as shown by the compound’s co-crystal structure with chimeric CA
IX. The strongest inhibitor of recombinant human CA IX’s catalytic domain in human cells achieved an affinity of 50 pM.
However, the high affinity diminished the selectivity. The most selective compound for CA IX exhibited 10 nM affinity. The
compound that showed the best balance between affinity and selectivity bound with 1 nM affinity. The inhibitors described in
this work provide the basis for novel anticancer therapeutics targeting CA IX.
inhibition enhances the therapeutic effect of other drugs/
radiation.14−17
INTRODUCTION
■
Carbonic anhydrase IX (CA IX) is a transmembrane zinc
enzyme that catalyzes the conversion between carbon dioxide
and bicarbonate.1−3 CA IX is considered to be a good tumor
marker, as its expression is very limited in normal tissues
(almost exclusively expressed in gastrointestinal tract epithe-
lium) and it is highly expressed in various cancers.4−6 Inhibition
of CA IX is a promising therapeutic path for the treatment of
solid tumors where a hypoxic environment has developed.7−9
Hypoxia strongly induces the upregulation of CA IX. As
evidenced by in vivo tumor models using gene depletion or
overexpression strategies, CA IX is a functional mediator of
tumor growth and metastasis.10−12 CA IX activity is critical for
cancer cell survival and adaptation to hypoxic conditions.13 CA
IX also contributes to cancer progression by stimulating cancer
cell migration and invasion.7,12 CA IX activity is also linked to
increased tumor resistance to chemo- and radiotherapy, and its
Humans contain 15 highly homologous α-CA isoforms that
differ in their oligomeric structure (CAs VI, IX, and XII are
dimeric, whereas all other CAs are monomeric), catalytic
activity (CAs VIII, X, and XI are catalytically inactive), and
cellular localization. CAs I, II, III, VII, and XIII are cytosolic
proteins. Catalytic domains of CAs IV, IX, XII, and XIV are on
the cell surface; they either are bound to the cell membrane
through a GPI anchor (CA IV) or contain a transmembrane
domain (CAs IX, XII, and XIV), whereas CAs VA and VB are
located in the mitochondrial matrix; CA VI is the only secreted
isoform.2,18,19 The structure of the catalytic domain of human
α-CAs is composed of a twisted β-sheet surrounded by α-
Received: July 9, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jm501003k | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Table 1. Compound Dissociation Constants for All 12 Catalytically Active Human CA Isoforms
CA isoform
1 (VD12-09)
2 (PG7)
3 (VD11-4-2)
4 (VD10-13)
5 (VD10-35)
6 (AZM)
Kd Determined by the Fluorescent Thermal Shift Assay, nM (37 °C, pH 7.0, Pi Buffer)
CA I
50 000
1300
>200 000
1700
3300
210
>200 000
>200 000
>200 000
>200 000
>200 000
>200 000
>200 000
>200 000
9.5
710
60
0.11
6.7
0.20
17
1400
38
CA II
CA III
CA IV
40 000
25
29 000
590
330
1.3
33 000
160
290
22
40 000
100
1000
310
310
17
CA VA
CA VB
CA VI
2500
5.6
4300
330
95
200
46
67
CA VII
CA IX
chCA IX
CA XII
chCA XII
CA XIII
CA XIV
9.8
7.1
1.1
0.050
2.0
32
50
20
25
630
63
83
50
330
>200 000
>200 000
1700
3.3
220
310
8.3
250
250
29
130
330
50
500
6.7
140
3.6
26
4300
0.16
1.3
5.0
11
Kd Determined by the Isothermal Titration Calorimetry, nM (37 °C, pH 7.0, Pi or Tris Buffer)
CA I
>10 000
730
ND
ND
ND
ND
ND
ND
ND
1100
59
<10
<10
59
<10
<10
37
810
46
CA II
CA VII
CA IX
chCA IX
CA XII
CA XIII
1000
15
140
<10
7.3
20
63
57
51
22
22
12
14
14
230
94
140
32
130
60
420
42
18
Kd by Stopped-Flow Kinetic Inhibition Assay, Fitted by the Morrison eq, nM (25 °C, pH 7.5, Hepes Buffer)
CA I
46 000
910
ND
ND
ND
ND
ND
ND
560
<20
<5.0
25
ND
ND
ND
ND
ND
ND
290
10
CA II
29
CA IX
chCA IX
CA XII
CA XIII
<5.0
15
<10
6.7
<5.0
<5.0
110
<10
<5.0
25
170
<10
130
<10
<10
X-ray Crystallographic structure PDB IDs
CA II
4PYX
4Q06
ND
ND
ND
ND
ND
ND
4PYY
4Q07
4Q0L
4Q09
ND
ND
ND
ND
ND
ND
4PZH
ND
ND
ND
ND
ND
ND
chCA IX
CA XII
chCA XII
CA XIII
ND
4Q08
ND
ND
4HU1
a
Determined by FTSA, ITC, and stopped-flow kinetic inhibition assays. Available PDB IDs of X-ray crystallographic structures are listed.
helices. All catalytically active human CA isoforms contain a
zinc ion at the bottom of their deep active site cleft, which has
hydrophobic and hydrophilic surfaces.20,21 In addition to the
catalytic and transmembrane domains and a short intracellular
tail, CA IX has a proteoglycan-like domain, a unique feature
among other human CA isoforms. Its proposed function is to
act as an intrinsic buffer that increases CA IX adaptability to
acidic medium.22
Currently known CA inhibitors can be divided into two
groups: those that coordinate to an active site zinc and those
that do not. Among the zinc binders are sulfonamides/
sulfamates, ureates/hydroxamates, mercaptophenols, and metal
complexing anions.23 Aromatic sulfonamides are the most
widely studied CA inhibitors, with a clearly determined
mechanism of inhibition.24,25 Numerous sulfonamides are
being used in the clinic to treat pathologies such as edema,
glaucoma, epilepsy, altitude sickness, and others.23,26 Unfortu-
nately, current inhibitors not only bind to the targeted CA but
also to off-target CAs and thus often exhibit undesired side
effects.
are based on fluorinated benzenesulfonamides that have been
known to exhibit high affinity toward most CA isoforms.27−29
We characterized the binding and inhibition characteristics of
the inhibitors to all CAs using three complementary
techniques: isothermal titration calorimetry (ITC), the
fluorescent thermal shift assay (FTSA), and the stopped-flow
CO2 hydration assay. Structural characterization of compound
binding was achieved by determining compound co-crystal
structures with several selected CA isoforms (CAs I, II, and
XII) and several chimeric CAs (chCA IX and chCA XII), which
resemble CAs IX and XII. The compounds are good candidates
for further development as anticancer agents targeting CA IX.
RESULTS
■
Binding and Inhibition of CAs. CA IX is a membrane
protein containing a proteoglycan-like domain, a transmem-
brane region, and an intracellular region associated with
signaling. The protein dimerizes, exposing the catalytic domain
on the cell surface side of the membrane.1,30,31 The catalytic
domain of CA IX was prepared in mammalian cells and purified
by affinity chromatography. The measured catalytic activity of
CA IX and the inhibition profile by standard inhibitors such as
Here, we describe novel inhibitors that selectively and with
subnanomolar affinity bind and inhibit CA IX. The inhibitors
B
dx.doi.org/10.1021/jm501003k | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
acetazolamide (6, AZM) were similar to those previously
published (Table 1).
times weaker than that of CA IX. Thus, there is significant
selectivity of this compound to bind CA IX over other isoforms.
Compound 2 (PG7) bears an even bulkier cyclododecylamino
group and thus has quite low solubility. However, the solubility
was sufficient to measure its binding to all CA isoforms by
FTSA. Again, CA IX exhibited an affinity of approximately 10
nM, whereas almost all CA isoforms did not bind the
compound (the Kd was above the detection limit of the
FTSA under assay conditions of 200 μM), with the exception of
CAs XIII and XIV that bound 2 with submicromolar affinities
(Table 1).
To design CA IX-selective inhibitors, a series of fluorinated
benzenesulfonamides with substituents on the benzene ring was
synthesized that was designed to sterically fit in the active site
of CA IX but not in the active centers of other CAs. Figure 1
shows the chemical structures of several selected compounds
that are described in this study and exhibit subnanomolar
affinity and selectivity toward CA IX.
If the cyclooctylamino group substituent is instead in the
meta position (compound 3 (VD11-4-2), Table 1), then the
compound’s affinity toward CA IX increased, but it also bound
most other isoforms with nanomolar affinity. There was still
significant selectivity exhibited toward CA IX, but this was not
as profound as that of compounds in the ortho series.
Therefore, compounds like 1 may be better suited as drug
leads toward CA IX than 3.
Several compounds missing the second substituent and
bearing only the para tail (4 (VD10-13) and 5 (VD10-35),
Figure 1 and Table 1) were very potent binders of CAs.
However, they bound CA IX relatively weaker and exhibited
essentially no selectivity toward CA IX. These compounds
exhibited subnanomolar affinity toward CA I, an undesired
effect because CA I activity is deemed to be essential and its
inhibition could lead to toxic side effects. In addition, these
compounds are tight binders of CA VB.
Figure 1. Chemical structures of CA inhibitors discussed in this article.
Acetazolamide (6, AZM) is commonly used as a control inhibitor of
CAs.
Crystal Structures of Compounds Bound to CAs
Reveal the Basis of Specificity. The PDB codes of the
high-resolution X-ray crystallographic structures of CA−
compound complexes are listed at the bottom of Table 1. In
all 8 crystal structures determined in this work, the compound
is bound at the active site of the CA, and the amino group of
the sulfonamide is coordinated by the active site zinc atom. The
crystal structure of CA XIII with 5 (PDB ID 4HU1) has been
previously determined and described.27 Although we were able
to crystallize the compounds of interest (1, 3, and 5) with CAs
II and XII, crystallization of CA IX was elusive. As a result, we
adopted an innovative approach to provide structural insight
into small molecule binding. We made a chimeric version of CA
IX (chCA IX) that was based on using the CA II protein
(which we were able to crystallize) but for which the catalytic
site was modified to resemble CA IX. Both CAs II and IX have
essentially the same protein chain fold in the vicinity of the
active site of both proteins. Although the positions of most
amino acids are identical, some that are exposed on the active
site cleft are different. Therefore, to create a chimera of the
actual catalytic site, we made a multiple-residue mutant of CA
II that resembles the amino acids of CA IX. This approach is
similar to that for the previously designed mimic of CA IX.54
To confirm that such chimeric chCA IX could resemble the
behavior of actual human CA IX, an additional similar chimeric,
chCA XII, was also created. The compound binding
comparison between chCA XII and CA XII, and also between
chCA XII and CA II, showed that chimeric chCA XII
recognized and bound the compounds with similar affinities
to that of CA XII, but it was quite different from CA II. For
example, compound 3 bound CA XII with 3.3 nM and chCA
XII with 6.7 nM affinity. However, CA II bound the compound
with 60 nM affinity. Furthermore, compound 4 bound CA XII
with 220 nM and chCA XII with 310 nM affinities, but it bound
CA II with 6.7 nM affinity. Therefore, the six mutations that
The high affinity of the compounds was achieved by lowering
the pKa of the sulfonamide amino group by introducing
electron-withdrawing fluorine atoms to the conjugated benzene
ring system. Sulfonamide compounds bind and inhibit the CAs
in their deprotonated form that is present in a small
concentration in the absence of withdrawing groups. In our
compounds, a strong withdrawing effect is exhibited by the
three fluorine atoms on the benzene ring bound directly to the
sulfonamide headgroup.
A large number of techniques have been used to measure
inhibitor binding to CAs.24 The most common current method
for determining CA inhibition is the stopped-flow CO2
hydration inhibition assay.32−34 However, this method works
in the presence of substrate, whereas the direct binding assays
are performed in the absence of substrate. ITC,35−38 surface
plasmon resonance (SPR),39−41 and FTSA42−46 (often called
differential scanning fluorimetry47 or, in high-throughput,
ThermoFluor48−53) are frequently used techniques to charac-
terize binding in the absence of substrate.
Inhibitor affinities toward all 12 catalytically active human
CA isoforms were measured. Table 1 lists the observed affinities
of the compounds toward all 12 catalytically active CA
isoforms, as determined by FTSA. For comparison, the values
determined by the enzymatic inhibition assay and ITC are
shown for the most important isoforms for which selectivity is
to be confirmed, namely, CAs I, II, IX, XII, and XIII.
Compound 1 (VD12-09), bearing a bulky cyclooctylamino
group at the ortho position to the sulfonamide head, exhibited a
dissociation constant of 1.1 nM for CA IX. The remaining 12
catalytically active human CA isoforms bound the compound
with significantly lower affinity: CA XIV, 25 times weaker than
that of CA IX; CAs XIII, VB, VII, and XII, over 100 times
weaker than that of CA IX; CAs II, VA, and VI, over 1000 times
weaker than that of CA IX; and CAs I and III, about 100 000
C
dx.doi.org/10.1021/jm501003k | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 2. Compounds 1 (A; PDB ID 4Q06) and 3 (B; PDB ID 4Q07) bound to chCA IX, as determined by X-ray crystallography. Zn is shown as a
blue sphere, and the histidine residues holding the Zn atom are transparent. The amino acids of chCA IX are shown in gray. The terminal atoms of
amino acids that form the hydrophobic cavity are shown as CPK (light gray). Several atoms of the cyclooctyl group are also shown as CPK (dark
gray). Dashed lines connect the atoms that make hydrogen bonds or electron donor−acceptor interactions (with Zn). A water molecule is shown as
a red sphere. The compounds are shown in light steelblue.
were made to make chCA XII from CA II switched CA II into a
CA XII, resembling chCA XII.
disappears at these lower protein concentrations. Any ligands
that bind tighter to CA than Kd = 10 nM give a nonsymmetrical
dosing curve (Figure 4) that simply resembles the titration of
10 nM protein with a ligand that binds significantly tighter than
10 nM affinity. Such nonsymmetrical curves can be fit by using
the Morrison equation and should not be fit by a standard Hill
equation (see Experimental Section). Due to the above reasons,
of the three methods, only the FTSA method could be used to
determine the affinity of 3 to CA IX.
Compound 3 bound chCA IX quite similarly to that of 1
(Figure 3B). The overall position of the cyclooctyl group was
similar in both structures, but in 3, the hydrophobic group
bound deeper in the pocket, indicating that there was increased
contact with the protein.
It would be expected that chimeric CAs exhibit structurally
similar binding modes to that of the CAs they were meant to
resemble. Compound 3’s binding mode was determined bound
to both CA XII and chCA XII (Figure 3J). Both structures
showed a nearly identical binding mode of the compound to
actual CA XII and chimeric chCA XII. However, the structure
differed from CA II. Therefore, the six mutations introduced to
CA II in order to resemble CA XII have switched it into a new
protein that is more similar to CA XII than CA II both in terms
of its binding affinity and structure.
Interestingly, the structures of both 1 and 3 bound to chCA
IX did not contain any water molecules in the vicinity of the
active site. All water molecules were efficiently displaced by the
ligands. By way of contrast, only para-substituent-bearing
compound 5 exhibited average affinity to CA IX (50 nM),
but it bound to CA II significantly more tightly (17 nM) and
exhibited numerous water molecules in the vicinity of the
ligand (Figure 3C). This observation was true both for CA II
and CA XIII. Therefore, a bulky ortho or meta substituent is
needed to efficiently occupy the space in the active site and
displace water molecules. The two bulky hydrophobic groups in
compound 2 significantly reduced its water solubility, and no
crystal structure was determined for the compound.
Although the chimeric chCA IX shows similar trends in
binding data, it does not mimic those of wild-type CA IX to the
same extent that chCA XII resembles CA XII. For example,
compound 1 bound CA IX with 1.1 nM affinity, whereas it
bound chCA IX with 25 nM affinity. Still, this affinity is closer
to that of CA IX than to CA II (1300 nM). Compound 4
bound CA II with 6.7 nM affinity, whereas it bound CA IX with
32 nM affinity and chCA IX with 63 nM affinity. Thus, despite
the slight variation in binding data between chCA IX and CA
IX, we adopted chCA IX to provide structural insight as to how
the small molecules are binding.
Figure 2 shows the crystal structures of 1 and 3 bound to
chCA IX, emphasizing the nearly perfect filling of the
hydrophobic pocket by the cyclooctyl ring and the contacts
between the compounds and the protein. Figure 3 compares
the structures of 1 and 3 bound to chCA IX, CA II, CA XII, and
chCA XII. It is important to understand why compound 1
exhibited such profound selectivity toward CA IX. Figure 3A
compares 1 binding to CA II (green) and chCA IX
(aquamarine). chCA IX (as well as CA IX, as judged from
the only available crystal structure of CA IX in the literature,
PDB ID 3IAI) contains a deeper hydrophobic pocket than that
of CA II because the bulky Phe131 in CA II occupies part of
the pocket. Using this difference between CA II and CA IX, a
rather bulky cyclooctyl group has been designed to fit the
pocket. This is where it fit in chCA IX, but it could not fit in CA
II. Therefore, the ligand adopts the opposite orientation in CA
II, and the cyclooctyl group makes poor connections with the
protein. Therefore, a ligand such as 1 could be an example of
both a tight and selective CA IX-binding compound.
Compound 3, bearing the cyclooctyl group in the meta
position, was less selective toward CA IX than 1 but exhibited
some of the highest affinities of any CA inhibitor ever observed,
approximately 50 pM. Such affinity could be determined only
by FTSA and could not be determined by ITC or the stopped-
flow kinetic CO2 hydration assay because both ITC and the
stopped-flow kinetic CO2 hydration assays are limited by
protein concentration. The inhibition assay cannot be
performed at lower than 10 nM CA because the signal
Compound 3 bound CA II with 60 nM and CA IX with 50
pM affinities. Thus, there is 1000-fold selectivity toward CA IX.
The affinity of the chimeric chCA IX mutant that was used to
determine the structure was right in the middle of the range, 2.0
D
dx.doi.org/10.1021/jm501003k | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 3. X-ray crystallographic structures comparing the binding modes of inhibitors bound to the active sites of CA isoforms. Inhibitors are 1
(green colors), 3 (red colors), and 5 (yellow-orange colors). The surrounding amino acids of the CA isoforms are colored: CA II, yellow; chCA IX,
crystalline red; CA XII, crystalline pink; and chCA XII, crystalline green. The Zn atom is shown as a blue sphere, and the histidine residues holding
the Zn are transparent. (A) Compound 1 bound to CA II (green; PDB ID 4PYX) and chCA IX (aquamarine; PDB ID 4Q06). The cyclooctyl group
is pushed by Phe131 toward an opposite orientation in CA II as compared to that of Val131 in chCA IX, causing selectivity for chCA IX. Residue
labeling corresponds to chCA IX. (B) Comparison of the binding modes of compounds 1 (aquamarine; PDB ID 4Q06) and 3 (red; PDB ID 4Q07)
E
dx.doi.org/10.1021/jm501003k | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 3. continued
bound to chCA IX. The overall position of both compounds is similar. (C) Compound 5 bound to CA II (yellow; PDB ID 4PZH) and CA XIII
(orange; PDB ID 4HU1) exhibits a significantly different rotated position of the benzene ring. The active sites also contain numerous water
molecules bound deeply in the active site (shown as small spheres: yellow in CA II and orange in CA XIII). The presence of these water molecules
shows how much space is left unoccupied in the active site by compound 5. These water molecules are absent in the structures with compounds 1
and 3. (D) Compound 3 bound to CA II (there are two orientations shown in violet and violet-red; PDB ID 4PYY) and chCA IX (red; PDB ID
4Q07). The two orientations of the compound in CA II indicate the reason why the compound is bound much more weakly to CA II than to CA IX.
Residue labeling corresponds to chCA IX. (E) The binding mode of compounds 1 (green; PDB ID 4PYX) and 3 (two essentially opposite
orientations are shown in violet and violet-red; PDB ID 4PYY) bound to CA II. (F) Comparison of the two opposite orientations of the binding
positions of 1 (dark and light green; PDB ID 4Q08) and 3 (dark red; PDB ID 4Q09) to chCA XII. (G) The positions of compound 1 bound to
chCA IX (aquamarine green; PDB ID 4Q06) and chCA XII (two alternative orientations are shown in green-yellow and dark-green; PDB ID 4Q08).
Residue labeling corresponds to chCA IX. (H) The comparison of positions of 3 bound to chCA IX (red; PDB ID 4Q07) and chCA XII (dark-red;
PDB ID 4Q09). Residue labeling corresponds to chCA IX. (I) The positions of 3 bound to chCA IX (red; PDB ID 4Q07) and CA XII (pink; PDB
ID 4Q0L) showing essentially overlapping orientations. Residue labeling corresponds to chCA IX. (J) Compound 3 bound to CA XII (deep pink;
PDB ID 4Q0L) and chCA XII (dark-red; PDB ID 4Q09) in almost exactly the same orientation. Residue labeling corresponds to chCA XII.
nM. Therefore, these key amino acids caused the affinity to
change significantly. First, the compound bound CA II and
exhibited two oppositely facing positions. The electron density
(Figure 5E) clearly could not be fit with a single pose of the
ligand. The presence of two rather different positions indicates
that their binding energies are quite similar and that the affinity
is not high, as shown by the binding measurements.
The structures of compounds 1 and 3 bound to chCA XII
and CA XII were also solved for comparison. However, their
binding affinities or selectivities are significantly lower and
would be less suitable to inhibit CA XII. The present
compounds are thus more suitable to be developed as CA IX
inhibitors rather than CA XII.
There is only one report on the crystallization of human CA
IX30 and thus inhibitor development for CA IX has been
hindered by unsuccessful co-crystallization of compounds with
CA IX. Therefore, a mimic of CA IX, a multiple-residue mutant
in the vicinity of the active site, has been proposed and
crystallized with some inhibitors.54 Our designed chimeric CA
IX (chCA IX) slightly differed from this by the selection of
mutations that were supposed to better resemble the
environment of the active site of CA IX. Unfortunately, such
an approach cannot be considered as fully replacing the
crystallization of CA IX itself. The binding data of chCA IX did
not precisely resemble the binding to CA IX. However, this
approach is currently the best available to enable structural
visualization of the mode of binding of various CA IX
inhibitors. Furthermore, this approach worked excellently for
CA XII, where we determined the crystal structures of both
chimeric chCA XII and CA XII itself. Both the binding data and
the crystal structures confirmed that chCA XII is an excellent
model of CA XII.
Several groups of CA inhibitors have shown promising
selectivity in binding CA IX, such as nitroimidazole
sulfonamides,55 glycosidic sulfonamides,56 diarylpyrazole ben-
zenesulfonamides,57 and technecium-containing arylsulfona-
mides.58 Some nonsulfonamide compounds, namely, coumar-
ins, were also reported to be selective CA IX inhibitors.59−61
Furthermore, previously reported polyfluorinated compounds
have shown selectivity toward CA IX.62 However, their affinity
and selectivity were relatively low because the compounds were
fluorinated on the benzene ring that was not directly attached
to the sulfonamide headgroup. The studies of these inhibitors
were based on following only enzyme inhibition, which is often
insufficient for complete characterization of the binding
reaction. Furthermore, the reported selectivity was often
limited and arguably of insufficient magnitude for an inhibitor
against the most common CA I and CA II isoforms, i.e., the
appropriate inhibition constant should be at least 1000-fold
tighter than that for CA I and CA II.
DISCUSSION
■
The discovery of high-affinity and -selectivity compounds
toward human CA IX over all 11 remaining active human CAs
has been described and reconciled both in terms of affinities
and the structure of the complexes. The high affinity of the
compounds was achieved by lowering the pKa of the
sulfonamide amino group by introducing electron-withdrawing
fluorine atoms to the conjugated benzene ring system.
Sulfonamide compounds bound and inhibited the CAs in the
deprotonated form that was present in a small concentration
without the presence of the withdrawing groups. For example,
the pKa of benzenesulfonamide is approximately 11, whereas
for pentafluorobenzenesulfonamide, it is about 8. Therefore,
the observed dissociation constant for the fluorinated
sulfonamide would be approximately 1000-fold stronger due
to the higher fraction of the active, deprotonated form. In
compounds 1−5, a strong withdrawing effect was exhibited by
the fluorine atoms on the benzene ring.
The strong selectivity of the compounds was ensured by the
good fit of the cyclooctyl group of both compounds to the
hydrophobic pocket of CA IX. The compounds bound nicely
into the pocket, with numerous hydrophobic contacts between
the protein and ligand. This was not possible for CA II, where
the ligand bound in an opposite orientation and made weaker
contacts with the protein. Similarly, CA XII was also weakly
affected by the compounds. The cyclooctyl group did not fit as
in CA IX, and the compound was bound in the opposite
orientation, which was significantly weaker in its binding
energy. There are no structures with the remaining CAs.
Therefore, we do not know the exact structural reasons why the
compounds are poorer binders of these isoforms than to CA
IX.
It seems that in order to achieve highly selective compounds
it is insufficient to have a single para substituent on the
benzenesulfonamide. There are no steric constraints, and such
compounds bind quite easily to most CA isoforms.
Furthermore, it is insufficient to have a single ortho substituent
because such compounds are also quite easily accommodated in
the active site of most CAs. In order to achieve significant
selectivity, it is necessary to have at least two substituents, and
the most promising positions are ortho and para. Compounds
that have substituents farther away from the sulfonamide group,
F
dx.doi.org/10.1021/jm501003k | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 4. Compounds 1 (left) and 3 (right) binding and inhibition of CAs. (A) Binding of compounds, as determined by the thermal shift assay.
Data points show the ΔTm as a function of the total added compound concentration, and the lines are simulated according to ref 42. Red filled
squares, CA IX; magenta open squares, chCA IX; black filled triangles, CA II; and blue filled circles, CA I. The largest ΔTm shift for similar proteins
corresponds to strongest binding Kd. The inset graphs show normalized raw fluorescence data as a function of temperature at zero (filled red
diamonds) and 50 μM (open red triangles) of total added compound concentration. The melting midpoints correspond to the Tm. (B) Binding of
the compounds, as determined by isothermal titration calorimetry. Colors and symbols for CA isoforms are the same as those in panel A. The ITC
curve fitting Kd’s are listed in Table 1. Insets show the raw ITC curves of the respective compound binding to CA IX. (C) The inhibition of CA
isoforms, as determined by the stopped-flow kinetic CO2 hydration assay. Colors and symbols for CA isoforms are the same as those in panel A. Data
points correspond to the percent inhibition of a CA as a function of the total added compound concentration. The lines are fit according to the
Morrison equation as explained in the Experimental Section. Insets show raw activity curves (drop in absorbance/pH due to acidification by the CA
IX) at various added compound concentrations: magenta, 0 nM; cyan, 15.6 nM; violet, 31.3 nM; and green, spontaneous CO2 hydration in the
absence of CA IX. The CA IX concentration was 20 nM. All three methods conclusively indicate that both compounds 1 and 3 bound and inhibited
CA IX significantly stronger than that for CA I and CA II. Furthermore, compound 3 bound tighter to most CA isoforms than that of 1. However,
compound 1 exhibited a greater selectivity ratio toward CA IX than that of 3.
e.g., on consecutive rings, are expected to bear lower selectivity
toward a desired isoform.
Compound 3 possessed an even higher affinity for CA IX,
approximately 50 pM. However, this compound bound more
strongly not only to the target CA IX but also to nearly all other
CAs. Therefore, despite greater affinity, it seems that 3 has
lower potential to be developed further as a drug candidate
than that of 1. The affinity of 3 for CA IX is extremely high.
Such affinities are rarely observed for any protein−ligand
system.
Several compounds have been designed and produced that
strongly and selectively bind CA IX, as shown by three
independent techniques. Crystal structures show the structural
arrangement of the ligands in the CA IX active site, as modeled
by chimeric CA IX. The compounds had to possess a highly
The three compounds, 1−3, are good potential lead
compound candidates. However, further development of the
compounds may prove to be beneficial for improving affinity
and selectivity. The three compounds are quite different. 2 is
the most selective toward CA IX, but it has the lowest affinity
(10 nM). 1 possesses 1 nM affinity toward CA IX, and, under
the assumption that it would affect the other CAs in the human
body only at 0.1 μM, it would affect only CA XIV in addition to
CA IX. At 1 μM concentration, it would affect CAs VB, VII,
XII, XIII, and XIV. However, the most abundant CAs, I and II,
still would not be affected.
G
dx.doi.org/10.1021/jm501003k | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 5. Electron densities of compounds bound to the active sites of CA isoforms. (A) 1 bound to CA II (PDB ID 4PYX). (B) 1 bound to chCA
IX (PDB ID 4Q06). (C) 1 bound to chCA XII (two alternative positions are visible; PDB ID 4Q08). (D) 5 bound to CA II (PDB ID 4PZH). (E) 3
bound to CA II (two alternative positions are visible; PDB ID 4PYY). (F) 3 bound to chCA IX (PDB ID 4Q07). (G) 3 bound to chCA XII (PDB
ID 4Q09). (H) 3 bound to CA XII (PDB ID 4Q0L). Modeled compounds are shown in the same colors as those in previous images: 1, green; 3,
red-violet; and 5, yellow. The catalytic Zn2+ is shown as a blue sphere. The electron density map |Fobs − Fcalc| calculated in the absence of ligand is
contoured at 3.0σ (A, H) and at 2.5−2.7σ (B−G).
pounds 4 and 5 has been previously described (4 is 3c and 5 is 3d
described in ref 27).
hydrophobic group at an ortho position relative to the aryl
sulfonamide in order to exhibit high selectivity. Extremely high
selectivity diminished affinity because of the strain that the
compound−protein structure exhibits.
Synthesis of 2-(Cyclooctylamino)-3,5,6-trifluoro-4-[(2-
hydroxyethyl)thio]benzenesulfonamide (1). A mixture of 2,3,5,6-
tetrafluoro-4-[(2-hydroxyethyl)thio]benzenesulfonamide (4) (0.17 g,
0.55 mmol), Et3N (0.08 mL, 0.57 mmol), DMSO (1 mL), and
cyclooctylamine (0.07 g, 0.57 mmol) was stirred at 60 °C for 16 h.
The mixture was then diluted with H2O (20 mL) and extracted with
EtOAc (3 × 10 mL). The combined organic phase was dried over
MgSO4 and evaporated under reduced pressure.
EXPERIMENTAL SECTION
■
Chemical Compounds. All starting materials and reagents were
commercial products that were used without further purification.
Melting points of the compounds were determined in open capillaries
on a Thermo Scientific 9100 Series and are uncorrected. Column
chromatography was performed using silica gel 60 (0.040−0.063 mm,
1
Merck). H and 13C NMR spectra were recorded on a Varian Unity
Inova spectrometer (300 and 75 MHz, respectively) with TMS as an
internal standard, and proton and carbon chemical shifts are expressed
in parts per million (ppm) in the indicated solvent. 19F NMR spectra
were recorded on a Varian Unity Inova spectrometer (282 MHz) with
CFCl3 as an internal standard, and fluorine chemical shifts are
expressed in parts per million (ppm) in the indicated solvent.
Multiplicity was defined as s (singlet), d (doublet), t (triplet), q
(quartet), dd (double doublet), ddd (double double doublet), m
(multiplet), br s (broad singlet), br d (broad doublet), or br t (broad
triplet). TLC was performed with silica gel 60 F254 aluminum plates
(Merck) and visualized with UV light. High-resolution mass spectra
(HRMS) were recorded on a Dual-ESI Accurate-Mass Q-TOF LC/
MS 6520 mass spectrometer (Agilent Technologies). The purity of
final compounds was verified by HPLC to be >95% using the Agilent
1290 Infinity instrument with a Poroshell 120 SB-C18 (2.1 mm × 100
mm, 2.7 μm) reversed-phase column. Analytes were eluted using a
linear gradient of water/methanol (20 mM ammonium formate in
both phases) from 60:40 to 30:70 over 12 min, from 30:70 to 20:80
over 1 min, and then 20:80 over 5 min at a flow rate of 0.2 mL/min.
UV detection was at 254 nm.
The product was purified by chromatography on a column of silica
gel (0.040−0.063 mm) with EtOAc/CHCl3 (1:1), Rf = 0.59. Yield:
0.15 g, 56%, mp 68−69 °C. 1H NMR (300 MHz, CDCl3): 1.40−1.75
(12H, m, cyclooctane), 1.80−1.95 (2H, m, cyclooctane), 2.54 (1H, br
s, OH), 3.14 (2H, t, J = 5.7 Hz, SCH2CH2), 3.74 (2H, t, J = 5.7 Hz,
SCH2CH2), 3.75−3.85 (1H, m, CH of cyclooctane, signal overlaps
with signal of SCH2CH2), 5.77 (2H, s, SO2NH2), 6.16 (1H, br s, NH).
13C NMR (75 MHz, CDCl3): 23.7 (cyclooctane), 25.8 (cyclooctane),
27.5 (cyclooctane), 33.0 (cyclooctane), 37.5 (SCH2CH2, br t), 56.4
(CH of cyclooctane, d, J (19F−13C) = 11 Hz), 61.2 (SCH2CH2), 117.9
1
2
(C1, dd, J (19F−13C) = 12 Hz, J (19F−13C) = 5 Hz), 118.3 (C4, t, J
(19F−13C) = 21 Hz), 132.7 (C2, d, J (19F−13C) = 15 Hz), 142.1 (C5,
1
2
3
ddd, J (19F−13C) = 240 Hz, J (19F−13C) = 16 Hz, J (19F−13C) = 5
Hz), 145.1 (C6, ddd, 1J (19F−13C) = 247 Hz, 2J (19F−13C) = 16 Hz, 3J
(19F−13C) = 4 Hz), 149.1 (C3, d, J (19F−13C) = 243 Hz). 19F NMR
(282 MHz, CDCl3): −124.5 (C3−F, d, J = 11 Hz), −143.0 (C5−F,
dd, 1J = 27 Hz, 2J = 12 Hz), −149.0 (C6−F, d, J = 26 Hz). HRMS for
C16H23F3N2O3S2 [(M+H)+] calcd, 413.1175; found, 413.1175.
Instant JChem was used for compound structure database
management, search, and prediction (Instant JChem 6.1.3, 2013,
ChemAxon; http://www.chemaxon.com). The synthesis of com-
H
dx.doi.org/10.1021/jm501003k | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Table 2. X-ray Crystallographic Data Collection and Refinement Statistics
protein−
compound
CA II−1
CA II−3
CA XII−3
CA II−5
chCA IX−1
chCA XII−1
chCA XII−3
chCA IX−3
resolution (Å)
25.22−1.80
21 036
23.76−1.75
24 612
22.09−2.00
44 882
40.82−1.06
100 624
40.74−1.15
77 998
40.66−1.07
98 666
69.03−1.20
72 417
69.86−1.15
77 927
Nref (unique)
Rmerge (outer
shell)
0.127 (0.422)
0.060 (0.133)
0.068 (0.269)
0.051 (0.376)
0.063 (0.189)
0.032 (0.339)
0.042 (0.346)
0.065 (0.204)
I/σ (outer shell)
9.9 (2.3)
10.8 (6.1)
3.6 (3.5)
12.2 (2.8)
2.9 (3.0)
15.9 (4.1)
6.7 (6.0)
14.7 (7.2)
7.0 (6.7)
14.0 (3.0)
3.1 (2.3)
21.1 (4.4)
6.7 (6.4)
8.6 (4.1)
3.5 (3.3)
multiplicity (outer 3.7 (3.7)
shell)
completeness (%) 95.1 (92.0)
(outer shell)
99.9 (99.8)
74.5 (64.9)
93.5 (77.6)
91.4 (72.0)
93.7 (75.0)
98.8 (98.2)
92.0 (73.6)
Natoms
2284
0.178
0.23
2465
0.149
0.2
8896
2527
2568
2663
2502
3103
0.131
0.159
14.353
0.023
2.45
Rwork
0.264
0.328
20.194
0.014
1.739
4Q0L
0.135
0.162
16.549
0.026
2.588
4PZH
0.137
0.169
17.922
0.024
2.448
4Q06
0.128
0.156
14.474
0.024
2.492
4Q08
0.139
0.178
17.478
0.024
2.333
4Q09
Rfree
Baverage
RMSbonds
RMSangles
PDB ID
18.949
0.02
15.231
0.02
2.137
4PYX
2.138
4PYY
4Q07
a
All datasets were collected at 100 K; test set size was 10%.
The synthesis of 2 and 3 is described in the Supporting Information.
Protein Preparation. Expression and purification of CAs I, II, VII,
XII and XIII were previously described: CA I, in ref 63; CA II, in ref
43; CA VI, in ref 63; CAs VII and XIII, in ref 64; and CA XII, in ref 65.
Preparation of CAs III, VA, VB, XIV, chCA IX, and chCA XII is
described in the Supporting Information.
Isothermal Titration Calorimetry. ITC experiments were per-
formed using ITC200 or VP-ITC instruments (MicroCal, Inc.,
Northampton, MA, USA) with 5−10 μM protein solution in the
cell and 50−100 μM of the ligand solution in the syringe. A typical
experiment consisted of 18 or 25 injections (2 or 10 μL each) within 2
or 3 min intervals. Experiments were carried out at 37 °C in 50 mM
phosphate or Tris buffer containing 100 mM NaCl at pH 7.0, with a
final DMSO concentration of 2%. The affinities by ITC were
determined from the slope of the ITC integrated curve, and in order
for the Wiseman factor, c, to not exceed 1000 at 10 μM protein
concentration, the Kd should be weaker than 10 nM. The dissociation
constants stronger than 10 nM could not be determined by ITC and
thus are shown as <10 nM in Table 1.
Stopped-Flow Kinetic CO2 Hydration Assay. The carbon dioxide
hydration activity of recombinant human CA was measured at 25 °C
using an Applied Photophysics SX.18MV-R stopped-flow spectrom-
eter according to ref 33. Reaction velocities were measured by
recording the absorbance of phenol-red indicator (30 μM, λ = 557
nm). The sample consisted of carbonic anhydrase, 0−100 μM
inhibitor (in ≤0.2% DMSO), and 25 mM Hepes reaction buffer
containing 25 mM NaCl, pH 7.5. Saturated CO2 solution was
prepared by bubbling the gas in Milli-Q water at 25 °C for 1 h. CA
concentration was chosen according to its activity: 300−500 nM, CA
I; 5−20 nM, CA II; 20 nM, chCA IX; 20 nM, CA IX; 40 nM, CA XII;
and 100−150 nM, CA XIII. The Kd value was determined using
Morrison equation derived by Morrison and co-workers67,68 and well-
explained in the book by Copeland.69 According to the Morrison
equation, the fraction of protein bound with ligand (equivalent to the
fraction of enzyme inhibited by the inhibitor) can be expressed as
Preparation of CAs IV and IX in Mammalian Cells. The cDNAs of
human CA IX and CA IV were purchased from RZPD Deutsches
Ressourcenzentrum fur Genomforschung GmbH (Germany). Ex-
̈
pression of CAs IX and IV in mammalian cells was carried out using
the pCEP4dS vector designed for the secretion of recombinant
mammalian proteins.66
For the construction of pCEP4dS-CAIX plasmids, the DNA
fragments, corresponding to the catalytic domain of CA IX (amino
acids 38−414) were inserted into a multicloning site of the pCEP4dS
vector. For the construction of the pCEP4dS-CAIV plasmid, a
nucleotide sequence encoding CA IV’s catalytic domain (amino acids
19−284) was inserted into a multicloning site of the pCEP4dS vector.
Because of the linker located between the secretion signal and the
coding sequences of the CAs, expressed CA IX had an additional 5
amino acids (DAAHM) and CA IV had an additional 8 amino acids
(DAAHMKLM) located at the N terminus.
Expression of CAs IX and IV was carried out using the FreeStyle
Max 293 expression system (Invitrogen, Life Technologies). FreeStyle
293-F suspension cell culture was maintained in 125−1000 mL
Erlenmeyer flasks containing 30−240 mL of FreeStyle medium in a 37
°C incubator with a humidified atmosphere of 8% CO2, on an orbital
shaker platform rotating at 135 rpm. FreeStyle cells were transiently
transfected with the purified pCEP4dS-CAIX or pCEP4dS-CAIV
plasmids according to the manufacturer’s recommendations. After 5−7
days, the cell culture was centrifuged at 6000g for 20 min, and the
secreted recombinant proteins were purified from the supernatant
using a CA-affinity column containing p-aminomethylbenzenesulfona-
mide-agarose (Sigma Life Science). The eluted CA IX was dialyzed
into 50 mM sodium phosphate buffer, pH 7.0, containing 100 mM
NaCl. CA IV was dialyzed into storage buffer containing 20 mM
HEPES, 50 mM NaCl, pH 7.5. All purified proteins were stored at
−80 °C.
fb = 1 − [CA]T + [I]T + Kiapp
(
2
−
([CA]T + [I]T + Kiapp
)
− 4[CA]T[I]T /2[CA]T
)
where [CA]T is the total concentration of the enzyme CA, [I]T is the
total added inhibitor concentration, and Kai pp is the apparent inhibition
constant equal to the protein−ligand binding constant under the
assumption of one ligand/inhibitor binding to one protein molecule.
Structure Determination by X-ray Crystallography. The
protein (CA) stock solutions (20 mM sodium Hepes buffer, pH 7.5,
50 mM NaCl) were concentrated by ultrafiltration to 20−60 mg/mL.
Crystallization by the sitting-drop vapor-diffusion method was started
by mixing equal volumes (2−3 μL) of protein solution with the
corresponding reservoir buffer. Crystals were grown at 20 °C for
several weeks. Crystallization buffers, space groups, and cell parameters
of CA crystals are described in Supporting Information Table S1. The
complexes of ligands with CA isoforms were prepared by soaking a CA
Binding and Inhibition Methods. Fluorescent Thermal Shift
Assay. FTSA experiments were performed in a Corbett Rotor-Gene
6000 (Qiagen Rotor-Gene Q) instrument using the blue channel
(excitation, 365
20 nm; detection, 460
15 nm). Samples
contained 5−10 μM protein, 0−200 μM ligand, 50 μM solvatochromic
dye ANS (8-anilino-1-naphthalenesulfonate), and 50 mM phosphate
buffer containing 100 mM NaCl at pH 7.0, with the final DMSO
concentration at 2%. The applied heating rate was 1 °C/min. Data
analysis was performed as previously described.42
I
dx.doi.org/10.1021/jm501003k | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
crystal with a 0.5 mM solution of ligand prepared by mixing a 50 mM
stock solution of ligand in DMSO with the corresponding reservoir
solution. The soaked crystals were measured after several days.
Diffraction data from complexes of 5 with CA II and of 1 and 3 with
chCA IX and chCA XII were collected at EMBL beamlines P14 and
P13 at the storage ring PETRAIII (DESY, Hamburg). Data from CA II
crystals soaked with 1 and 3 and 3 with CA XII were collected using a
MicroMax 007 HF (Rigaku, Japan) X-ray diffractometer at the
Institute of Biotechnology, Vilnius University (Lithuania). Data
collection and refinement statistics are listed in Table 2.
(4) Ivanov, S.; Liao, S. Y.; Ivanova, A.; Danilkovitch-Miagkova, A.;
Tarasova, N.; Weirich, G.; Merrill, M. J.; Proescholdt, M. A.; Oldfield,
E. H.; Lee, J.; Zavada, J.; Waheed, A.; Sly, W.; Lerman, M. I.;
Stanbridge, E. J. Expression of hypoxia-inducible cell-surface trans-
membrane carbonic anhydrases in human cancer. Am. J. Pathol. 2001,
158, 905−919.
(5) McDonald, P. C.; Winum, J.-Y.; Supuran, C. T.; Dedhar, S.
Recent developments in targeting carbonic anhydrase IX for cancer
therapeutics. Oncotarget 2012, 3, 84−97.
́
́
(6) Pastorekova, S.; Parkkila, S.; Parkkila, A. K.; Opavsky, R.; Zelník,
Data sets collected at the synchrotron were processed using XDS.70
MOSFLM71 was used to process the data sets collected with the
MicroMax 007 HF diffractometer. All structures were solved by
molecular replacement using MOLREP.72 Initial phases for the
structures of CA II, chCA IX, and chCA XII were obtained using
PDB ID 3HLJ. 1JD0 was used for the CA XII crystal structures. A
single protein chain stripped of all ligands was used as initial model in
all molecular replacement procedures. Inhibitor 3D models were
created using AVOGADRO,73 and molecule geometry description was
generated using LIBREFMAC.74 Protein models were refined and
manually remodeled using REFMAC75 and COOT.76 All graphic
representations were made with MOLSCRIPT,77 BOBSCRIPT,78 and
RASTER3D.79 Coordinates and structure factors have been submitted
to the RCSB Protein Databank, and their accession codes are given in
Table 2.
V.; Saarnio, J.; Pastorek, J. Carbonic anhydrase IX, MN/CA IX:
analysis of stomach complementary DNA sequence and expression in
human and rat alimentary tracts. Gastroenterology 1997, 112, 398−408.
(7) Benej, M.; Pastorekova, S.; Pastorek, J. Carbonic anhydrase IX:
regulation and role in cancer. Subcell. Biochem. 2014, 75, 199−219.
(8) Supuran, C. T. Inhibition of carbonic anhydrase IX as a novel
anticancer mechanism. World J. Clin. Oncol. 2012, 3, 98−103.
(9) Tafreshi, N. K.; Lloyd, M. C.; Bui, M. M.; Gillies, R. J.; Morse, D.
L. Carbonic anhydrase IX as an imaging and therapeutic target for
tumors and metastases. Subcell. Biochem. 2014, 75, 221−254.
(10) Gieling, R. G.; Williams, K. J. Carbonic anhydrase IX as a target
for metastatic disease. Bioorg. Med. Chem. 2013, 21, 1470−1476.
(11) Kivela, A. J.; Knuuttila, A.; Rasanen, J.; Sihvo, E.; Salmenkivi, K.;
̈
̈
̈
Saarnio, J.; Pastorekova, S.; Pastorek, J.; Waheed, A.; Sly, W. S.; Salo, J.
A.; Parkkila, S. Carbonic anhydrase IX in malignant pleural
mesotheliomas: a potential target for anti-cancer therapy. Bioorg.
Med. Chem. 2013, 21, 1483−1488.
ASSOCIATED CONTENT
* Supporting Information
■
S
(12) McDonald, P. C.; Dedhar, S. Carbonic anhydrase IX (CAIX) as
a mediator of hypoxia-induced stress response in cancer cells. Subcell.
Biochem. 2014, 75, 255−269.
(13) Sedlakova, O.; Svastova, E.; Takacova, M.; Kopacek, J.; Pastorek,
J.; Pastorekova, S. Carbonic anhydrase IX, a hypoxia-induced catalytic
component of the pH regulating machinery in tumors. Front. Physiol.
2014, 4, 400.
Synthesis of compounds 2 and 3; preparation of recombinant
CA isoforms; and crystallization buffers, space groups, and cell
parameters of CA crystals (Table S1). This material is available
free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*Tel.: +370-5-269-1884. Fax. +370-5-260-2116. E-mail:
matulis@ibt.lt, daumantas.matulis@bti.vu.lt.
■
́ ́
(14) Doyen, J.; Parks, S. K.; Marcie, S.; Pouyssegur, J.; Chiche, J.
Knock-down of hypoxia-induced carbonic anhydrases IX and XII
radiosensitizes tumor cells by increasing intracellular acidosis. Front.
Oncol. 2012, 2, 199.
(15) Dubois, L.; Peeters, S.; Lieuwes, N. G.; Geusens, N.; Thiry, A.;
Wigfield, S.; Carta, F.; McIntyre, A.; Scozzafava, A.; Dogne, J.-M.;
́
Supuran, C. T.; Harris, A. L.; Masereel, B.; Lambin, P. Specific
inhibition of carbonic anhydrase IX activity enhances the in vivo
therapeutic effect of tumor irradiation. Radiother. Oncol. 2011, 99,
424−431.
(16) Dubois, L.; Peeters, S. G. J. A.; van Kuijk, S. J. A.; Yaromina, A.;
Lieuwes, N. G.; Saraya, R.; Biemans, R.; Rami, M.; Parvathaneni, N. K.;
Vullo, D.; Vooijs, M.; Supuran, C. T.; Winum, J. Y.; Lambin, P.
Targeting carbonic anhydrase IX by nitroimidazole based sulfamides
enhances the therapeutic effect of tumor irradiation: a new concept of
dual targeting drugs. Radiother. Oncol. 2013, 108, 523−528.
(17) Ilardi, G.; Zambrano, N.; Merolla, F.; Siano, M.; Varricchio, S.;
Vecchione, M.; De Rosa, G.; Mascolo, M.; Staibano, S. Histopatho-
logical determinants of tumor resistance: a special look to the
immunohistochemical expression of carbonic anhydrase IX in human
cancers. Curr. Med. Chem. 2014, 21, 1569−1582.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This research was funded by a grant from the Research Council
of Lithuania (LIG-09/2012). The authors are grateful for grant
FP7-REGPOT-2009-1 “MoBiLi”, agreement no. 245721, and
COST projects TD0905 and CM0804. The authors are grateful
to Prof. Claudiu T. Supuran and Daniella Wullo for teaching to
perform the stopped-flow CO2 hydration assay
ABBREVIATIONS USED
■
CA, carbonic anhydrase; CA IX, carbonic anhydrase isoform
IX; chCA IX, chimeric carbonic anhydrase isoform IX; FTSA,
fluorescent thermal shift assay; ITC, isothermal titration
calorimetry
(18) Aspatwar, A.; Tolvanen, M. E. E.; Ortutay, C.; Parkkila, S.
Carbonic anhydrase related proteins: molecular biology and evolution.
Subcell. Biochem. 2014, 75, 135−156.
REFERENCES
■
(19) Supuran, C. T. Carbonic anhydrasesan overview. Curr. Pharm.
(1) De Simone, G.; Supuran, C. T. Carbonic anhydrase IX:
biochemical and crystallographic characterization of a novel antitumor
target. Biochim. Biophys. Acta 2010, 1804, 404−409.
Des. 2008, 14, 603−14.
(20) Aggarwal, M.; Boone, C. D.; Kondeti, B.; McKenna, R.
Structural annotation of human carbonic anhydrases. J. Enzyme Inhib.
Med. Chem. 2013, 28, 267−277.
(2) Frost, S. C. Physiological functions of the alpha class of carbonic
anhydrases. Subcell. Biochem. 2014, 75, 9−30.
(21) De Simone, G.; Alterio, V.; Supuran, C. T. Exploiting the
hydrophobic and hydrophilic binding sites for designing carbonic
anhydrase inhibitors. Expert Opin. Drug Discovery 2013, 8, 793−810.
(22) Innocenti, A.; Pastorekova, S.; Pastorek, J.; Scozzafava, A.; De
Simone, G.; Supuran, C. T. The proteoglycan region of the tumor-
associated carbonic anhydrase isoform IX acts as an intrinsic buffer
(3) Pastorek, J.; Pastorekova,
́
S.; Callebaut, I.; Mornon, J. P.; Zelník,
M.; Liao, S.; Portetelle, D.; Stanbridge, E.
V.; Opavsky, R.; Zat’ovicova,
́
́
J. Cloning and characterization of MN, a human tumor-associated
protein with a domain homologous to carbonic anhydrase and a
putative helix-loop-helix DNA binding segment. Oncogene 1994, 9,
2877−2888.
J
dx.doi.org/10.1021/jm501003k | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
optimizing CO2 hydration at acidic pH values characteristic of solid
tumors. Bioorg. Med. Chem. Lett. 2009, 19, 5825−5828.
(23) Alterio, V.; Di Fiore, A.; D’Ambrosio, K.; Supuran, C. T.; De
Simone, G. Multiple binding modes of inhibitors to carbonic
anhydrases: how to design specific drugs targeting 15 different
isoforms? Chem. Rev. 2012, 112, 4421−4468.
(41) Rogez-Florent, T.; Duhamel, L.; Goossens, L.; Six, P.; Drucbert,
A.-S.; Depreux, P.; Danze, P.-M.; Landy, D.; Goossens, J.-F.; Foulon,
́
C. Label-free characterization of carbonic anhydrase-novel inhibitor
interactions using surface plasmon resonance, isothermal titration
calorimetry and fluorescence-based thermal shift assays. J. Mol.
Recognit. 2014, 27, 46−56.
(42) Baranauskiene, L.; Hilvo, M.; Matuliene, J.; Golovenko, D.;
̇
̇
(24) Krishnamurthy, V. M.; Kaufman, G. K.; Urbach, A. R.; Gitlin, I.;
Gudiksen, K. L.; Weibel, D. B.; Whitesides, G. M. Carbonic anhydrase
as a model for biophysical and physical-organic studies of proteins and
protein−ligand binding. Chem. Rev. 2008, 108, 946−1051.
(25) Miller, W. H.; Dessert, A. M.; Roblin, R. O. Heterocyclic
sulfonamides as carbonic anhydrase inhibitors. J. Am. Chem. Soc. 1950,
72, 4893−4896.
Manakova, E.; Dudutiene, V.; Michailoviene, V.; Torresan, J.; Jachno,
̇
̇
J.; Parkkila, S.; Maresca, A.; Supuran, C. T.; Grazulis, S.; Matulis, D.
̌
Inhibition and binding studies of carbonic anhydrase isozymes I, II and
IX with benzimidazo[1,2-c][1,2,3]thiadiazole-7-sulphonamides. J.
Enzyme Inhib. Med. Chem. 2010, 25, 863−870.
̌
(43) Cimmperman, P.; Baranauskiene, L.; Jachimoviciute, S.; Jachno,
̇
̅
̇
J.; Torresan, J.; Michailoviene, V.; Matuliene, J.; Sereikaite, J.; Bumelis,
̇
̇
̇
(26) Swenson, E. R. Safety of carbonic anhydrase inhibitors. Expert
Opin. Drug Saf. 2014, 13, 459−472.
V.; Matulis, D. A quantitative model of thermal stabilization and
destabilization of proteins by ligands. Biophys. J. 2008, 95, 3222−3231.
(44) Cimmperman, P.; Matulis, D. Protein thermal denaturation
measurements via a fluorescent dye. In Biophysical Approaches
Determining Ligand Binding to Biomolecular Targets; Alberto
Podjarny, A. D., Kieffer, B., Eds.; RSC Publishing: Cambridge, 2011;
pp 247−274.
(27) Dudutiene, V.; Zubriene, A.; Smirnov, A.; Gylyte, J.; Timm, D.;
̇
̇
̇
Manakova, E.; Grazulis, S.; Matulis, D. 4-Substituted-2,3,5,6-
̌
tetrafluorobenzenesulfonamides as inhibitors of carbonic anhydrases
I, II, VII, XII, and XIII. Bioorg. Med. Chem. 2013, 21, 2093−2106.
(28) Krishnamurthy, V. M.; Bohall, B. R.; Kim, C.-Y.; Moustakas, D.
T.; Christianson, D. W.; Whitesides, G. M. Thermodynamic
parameters for the association of fluorinated benzenesulfonamides
with bovine carbonic anhydrase II. Chem.−Asian J. 2007, 2, 94−105.
(29) Winum, J. Y.; Innocenti, A.; Vullo, D.; Montero, J. L.; Supuran,
C. T. Carbonic anhydrase inhibitors; fluorinated phenyl sulfamates
show strong inhibitory activity and selectivity for the inhibition of the
tumor-associated isozymes IX and XII over the cytosolic ones I and II.
Bioorg. Med. Chem. Lett. 2009, 19, 5082−5085.
̌
̌
(45) Zubriene, A.; Capkauskaite, E.; Gylyte, J.; Kisonaite, M.;
̇
̇
̇
̇
̌
Tumkevicius, S.; Matulis, D. Benzenesulfonamides with benzimidazole
moieties as inhibitors of carbonic anhydrases I, II, VII, XII and XIII. J.
Enzyme Inhib. Med. Chem. 2014, 29, 124−131.
(46) Zubriene, A.; Kazlauskas, E.; Baranauskiene, L.; Petrauskas, V.;
̇
̇
Matulis, D. Isothermal titration calorimetry and thermal shift assay in
drug design. Eur. Pharm. Rev. 2011, 16, 56−59.
(47) Niesen, F. H.; Berglund, H.; Vedadi, M. The use of differential
scanning fluorimetry to detect ligand interactions that promote protein
stability. Nat. Protoc. 2007, 2, 2212−2221.
(30) Alterio, V.; Hilvo, M.; Di Fiore, A.; Supuran, C. T.; Pan, P.;
Parkkila, S.; Scaloni, A.; Pastorek, J.; Pastorekova, S.; Pedone, C.;
Scozzafava, A.; Monti, S. M.; De Simone, G. Crystal structure of the
catalytic domain of the tumor-associated human carbonic anhydrase
IX. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 16233−16238.
(48) Cummings, M. D.; Farnum, M. A.; Nelen, M. I. Universal
screening methods and applications of ThermoFluor. J. Biomol. Screen.
2006, 11, 854−863.
(31) Hilvo, M.; Baranauskiene, L.; Salzano, A. M.; Scaloni, A.;
̇
(49) Klinger, A. L.; McComsey, D. F.; Smith-Swintosky, V.; Shank, R.
P.; Maryanoff, B. E. Inhibition of carbonic anhydrase-II by sulfamate
and sulfamide groups: an investigation involving direct thermodynamic
binding measurements. J. Med. Chem. 2006, 49, 3496−500.
(50) Matulis, D.; Kranz, J. K.; Salemme, F. R.; Todd, M. J.
Thermodynamic stability of carbonic anhydrase: measurements of
binding affinity and stoichiometry using ThermoFluor. Biochemistry
2005, 44, 5258−5266.
(51) Mezzasalma, T. M.; Kranz, J. K.; Chan, W.; Struble, G. T.;
Schalk-Hihi, C.; Deckman, I. C.; Springer, B. A.; Todd, M. J.
Enhancing recombinant protein quality and yield by protein stability
profiling. J. Biomol. Screening 2007, 12, 418−428.
(52) Nettleship, J. E.; Brown, J.; Groves, M. R.; Geerlof, A. Methods
for protein characterization by mass spectrometry, thermal shift
(ThermoFluor) assay, and multiangle or static light scattering. Methods
Mol. Biol. 2008, 426, 299−318.
(53) Pantoliano, M. W.; Petrella, E. C.; Kwasnoski, J. D.; Lobanov, V.
S.; Myslik, J.; Graf, E.; Carver, T.; Asel, E.; Springer, B. A.; Lane, P.;
Salemme, F. R. High-density miniaturized thermal shift assays as a
general strategy for drug discovery. J. Biomol. Screening 2001, 6, 429−
440.
(54) Pinard, M. A.; Boone, C. D.; Rife, B. D.; Supuran, C. T.;
McKenna, R. Structural study of interaction between brinzolamide and
dorzolamide inhibition of human carbonic anhydrases. Bioorg. Med.
Chem. 2013, 21, 7210−7215.
(55) Rami, M.; Dubois, L.; Parvathaneni, N.-K.; Alterio, V.; van
Kuijk, S. J. A.; Monti, S. M.; Lambin, P.; De Simone, G.; Supuran, C.
T.; Winum, J.-Y. Hypoxia-targeting carbonic anhydrase IX inhibitors
by a new series of nitroimidazole-sulfonamides/sulfamides/sulfamates.
J. Med. Chem. 2013, 56, 8512−8520.
Matulis, D.; Innocenti, A.; Scozzafava, A.; Monti, S. M.; Di Fiore, A.;
De Simone, G.; Lindfors, M.; Janis, J.; Valjakka, J.; Pastorekova, S.;
Pastorek, J.; Kulomaa, M. S.; Nordlund, H. R.; Supuran, C. T.;
Parkkila, S. Biochemical characterization of CA IX, one of the most
active carbonic anhydrase isozymes. J. Biol. Chem. 2008, 283, 27799−
27809.
(32) Biswas, S.; McKenna, R.; Supuran, C. T. Effect of incorporating
a thiophene tail in the scaffold of acetazolamide on the inhibition of
human carbonic anhydrase isoforms I, II, IX and XII. Bioorg. Med.
Chem. Lett. 2013, 23, 5646−5649.
(33) Khalifah, R. G. The carbon dioxide hydration activity of carbonic
anhydrase. I. Stop-flow kinetic studies on the native human
isoenzymes B and C. J. Biol. Chem. 1971, 246, 2561−2573.
(34) McKenna, R.; Supuran, C. T. Carbonic anhydrase inhibitors
drug design. Subcell. Biochem. 2014, 75, 291−323.
(35) Biela, A.; Betz, M.; Heine, A.; Klebe, G. Water makes the
difference: rearrangement of water solvation layer triggers non-
additivity of functional group contributions in protein−ligand binding.
ChemMedChem. 2012, 7, 1423−1434.
(36) Holdgate, G. A. Making cool drugs hot: isothermal titration
calorimetry as a tool to study binding energetics. BioTechniques 2001,
31, 164−184.
(37) Ladbury, J. E. Isothermal titration calorimetry: application to
structure-based drug design. Thermochim. Acta 2001, 380, 209−215.
(38) Wiseman, T.; Williston, S.; Brandts, J. F.; Lin, L. N. Rapid
measurement of binding constants and heats of binding using a new
titration calorimeter. Anal. Biochem. 1989, 179, 131−137.
(39) Babon, J. J. Quantitative analysis of JAK binding using
isothermal titration calorimetry and surface plasmon resonance.
Methods Mol. Biol. 2013, 967, 57−67.
(40) Jecklin, M. C.; Schauer, S.; Dumelin, C. E.; Zenobi, R. Label-free
determination of protein-ligand binding constants using mass
spectrometry and validation using surface plasmon resonance and
isothermal titration calorimetry. J. Mol. Recognit. 2009, 22, 319−329.
(56) Winum, J.-Y.; Colinas, P. A.; Supuran, C. T. Glycosidic carbonic
anhydrase IX inhibitors: a sweet approach against cancer. Bioorg. Med.
Chem. 2013, 21, 1419−1426.
(57) Rogez-Florent, T.; Meignan, S.; Foulon, C.; Six, P.; Gros, A.;
́ ́
Bal-Mahieu, C.; Supuran, C. T.; Scozzafava, A.; Frederick, R.;
K
dx.doi.org/10.1021/jm501003k | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Masereel, B.; Depreux, P.; Lansiaux, A.; Goossens, J.-F.; Gluszok, S.;
Goossens, L. New selective carbonic anhydrase IX inhibitors: synthesis
and pharmacological evaluation of diarylpyrazole-benzenesulfona-
mides. Bioorg. Med. Chem. 2013, 21, 1451−1464.
(58) Can, D.; Spingler, B.; Schmutz, P.; Mendes, F.; Raposinho, P.;
Fernandes, C.; Carta, F.; Innocenti, A.; Santos, I.; Supuran, C. T.;
Alberto, R. [(Cp-R)M(CO)3] (M=Re or 99mTc) Arylsulfonamide,
arylsulfamide, and arylsulfamate conjugates for selective targeting of
human carbonic anhydrase IX. Angew. Chem., Int. Ed. 2012, 51, 3354−
3357.
(59) Carradori, S. Selective carbonic anhydrase IX inhibitors based
on coumarin scaffold as promising antimetastatic agents:
WO2012070024. Expert Opin. Ther. Pat. 2013, 23, 751−756.
(60) Sharma, A.; Tiwari, M.; Supuran, C. T. Novel coumarins and
benzocoumarins acting as isoform-selective inhibitors against the
tumor-associated carbonic anhydrase IX. J. Enzyme Inhib. Med. Chem.
2014, 29, 292−296.
organization of prior chemical knowledge and guidelines for its use.
Acta Crystallogr., Sect. D: Biol. Crystallogr. 2004, 60, 2184−2195.
(75) Murshudov, G. N.; Vagin, A. A.; Dodson, E. J. Refinement of
macromolecular structures by the maximum-likelihood method. Acta
Crystallogr., Sect. D: Biol. Crystallogr. 1997, 53, 240−55.
(76) Emsley, P.; Cowtan, K. Coot: model-building tools for
molecular graphics. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2004,
60, 2126−2132.
(77) Kraulis, P. J. MOLSCRIPT: a program to produce both detailed
and schematic plots of protein structures. J. Appl. Crystallogr. 1991, 24,
946−950.
(78) Esnouf, R. M. Further additions to MolScript version 1.4,
including reading and contouring of electron-density maps. Acta
Crystallogr., Sect. D: Biol. Crystallogr. 1999, 55, 938−940.
(79) Merritt, E. A.; Bacon, D. J. Raster3D: photorealistic molecular
graphics. Methods Enzymol. 1997, 277, 505−524.
(61) Touisni, N.; Maresca, A.; McDonald, P. C.; Lou, Y.; Scozzafava,
A.; Dedhar, S.; Winum, J.-Y.; Supuran, C. T. Glycosyl coumarin
carbonic anhydrase IX and XII inhibitors strongly attenuate the growth
of primary breast tumors. J. Med. Chem. 2011, 54, 8271−8277.
(62) Pastorekova, S.; Vullo, D.; Casini, A.; Scozzafava, A.; Pastorek,
J.; Nishimori, I.; Supuran, C. T. Carbonic anhydrase inhibitors:
inhibition of the tumor-associated isozymes IX and XII with
polyfluorinated aromatic/heterocyclic sulfonamides. J. Enzyme Inhib.
Med. Chem. 2005, 20, 211−217.
̌
(63) Capkauskaite, E.; Zubriene, A.; Smirnov, A.; Torresan, J.;
Kiso
̇
̇
̌
naite, M.; Kazokaite, J.; Gylyte, J.; Michailoviene, V.; Jogaite, V.;
̇ ̇ ̇ ̇ ̇
Manakova, E.; Grazulis, S.; Tumkevici
̌
us, S.; Matulis, D. Benzene-
̌
sulfonamides with pyrimidine moiety as inhibitors of human carbonic
anhydrases I, II, VI, VII, XII, and XIII. Bioorg. Med. Chem. 2013, 21,
6937−6947.
(64) Sudzius, J.; Baranauskiene, L.; Golovenko, D.; Matuliene, J.;
̌
̇
̇
̅
Michailoviene, V.; Torresan, J.; Jachno, J.; Sukackaite, R.; Manakova,
̇
̇
E.; Grazulis, S.; Tumkevici
̌
us, S.; Matulis, D. 4-[N-(Substituted 4-
̌
pyrimidinyl)amino]benzenesulfonamides as inhibitors of carbonic
anhydrase isozymes I, II, VII, and XIII. Bioorg. Med. Chem. 2010, 18,
7413−7421.
(65) Jogaite, V.; Zubriene, A.; Michailoviene, V.; Gylyte, J.;
̇
̇
̇
̇
Morkunaite, V.; Matulis, D. Characterization of human carbonic
̇
̅
anhydrase XII stability and inhibitor binding. Bioorg. Med. Chem. 2013,
21, 1431−1436.
̌
̌
(66) Dekaminaviciute, D.; Kairys, V.; Zilnyte, M.; Petrikaite, V.;
̇ ̇ ̇
̅
̌
Jogaite, V.; Matuliene, J.; Gudlevici
̌
ene, Z.; Vullo, D.; Supuran, C. T.;
̇
̇
̇
̌
Zvirbliene, A. Monoclonal antibodies raised against 167−180 aa
̇
sequence of human carbonic anhydrase XII inhibit its enzymatic
activity. J. Enzyme Inhib. Med. Chem. 2014, DOI: 10.3109/
14756366.2013.856424.
(67) Morrison, J. F. Kinetics of the reversible inhibition of enzyme-
catalysed reactions by tight-binding inhibitors. Biochim. Biophys. Acta
1969, 185, 269−286.
(68) Williams, J. W.; Morrison, J. F.; Duggleby, R. G. Methotrexate, a
high-affinity pseudosubstrate of dihydrofolate reductase. Biochemistry
1979, 18, 2567−2573.
(69) Copeland, R. A. Evaluation of Enzyme Inhibitors in Drug
Discovery: A Guide for Medicinal Chemists and Pharmacologists; John
Wiley & Sons, Inc.: Hoboken, NJ, 2013.
(70) Kabsch, W. Integration, scaling, space-group assignment and
post-refinement. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2010, 66,
133−144.
(71) Leslie, A. G. W. The integration of macromolecular diffraction
data. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2006, 62, 48−57.
(72) Vagin, A.; Teplyakov, A. MOLREP: an automated program for
molecular replacement. J. Appl. Crystallogr. 1997, 30, 1022−1025.
(73) Avogadro: an open-source molecular builder and visualization
tool, version 1.0.0, 2009; http://avogadro.openmolecules.net/.
(74) Vagin, A. A.; Steiner, R. A.; Lebedev, A. A.; Potterton, L.;
McNicholas, S.; Long, F.; Murshudov, G. N. REFMAC5 dictionary:
L
dx.doi.org/10.1021/jm501003k | J. Med. Chem. XXXX, XXX, XXX−XXX