Intrinsic thermodynamics of high affinity inhibitor binding to recombinant human carbonic anhydrase IV

Article abstract of DOI:10.1007/s00249-017-1256-0

Membrane-associated carbonic anhydrase (CA) isoform IV participates in carbon metabolism and pH homeostasis and is implicated in the development of eye diseases such as retinitis pigmentosa and glaucoma. A series of substituted benzenesulfonamides were de

Full text of DOI:10.1007/s00249-017-1256-0

Eur Biophys J
DOI 10.1007/s00249-017-1256-0
ORIGINAL ARTICLE
Intrinsic thermodynamics of high afnity inhibitor binding
to recombinant human carbonic anhydrase IV
Aurelija Mickevičiūtė1 · David D. Timm1 · Marius Gedgaudas1 · Vaida Linkuvienė1 · Zhiwei Chen2 ·
Abdul Waheed2 · Vilma Michailovienė1 · Asta Zubrienė1 · Alexey Smirnov1 · Edita Čapkauskaitė1 ·
Lina Baranauskienė1 · Jelena Jachno1 · Jurgita Revuckienė1 · Elena Manakova3 · Saulius Gražulis3 ·
Jurgita Matulienė1 · Enrico Di Cera2 · William S. Sly2 · Daumantas Matulis1
Received: 5 July 2017 / Revised: 7 September 2017 / Accepted: 15 September 2017
© European Biophysical Societies’ Association 2017
Abstract Membrane-associated carbonic anhydrase
(CA) isoform IV participates in carbon metabolism and pH
homeostasis and is implicated in the development of eye
diseases such as retinitis pigmentosa and glaucoma. A series
of substituted benzenesulfonamides were designed and their
binding afnity to CA IV was determined by fuorescent
thermal shift assay and isothermal titration calorimetry
(ITC). Compound [(4-chloro-2-phenylsulfanyl-5-sulfamoyl-
benzoyl)amino]propyl acetate (19) bound CA IV with the
Kd of 1.0 nM and exhibited signifcant selectivity over the
remaining 11 human CA isoforms. The compound could
be developed as a drug targeting CA IV. Various forms of
recombinant CA IV were produced in Escherichia coli and
mammalian cell cultures. Comparison of their temperature
stability in various bufers and salt solutions demonstrated
that CA IV is most stable at slightly alkaline conditions and
at elevated sodium sulfate concentrations. High-resolution
X-ray crystallographic structures of ortho-Cl and meta-
thiazole-substituted benzene sulfonamide in complex with
CA IV revealed the position of and interactions between
the ligand and the protein. Sulfonamide inhibitor binding
to CA IV is linked to several reactions—the deprotona-
tion of the sulfonamide amino group, the protonation of
CA–Zn(II)-bound hydroxide at the active site of CA IV, and
the compensating reactions of the bufer. The dissection of
binding-linked reactions yielded the intrinsic thermody-
namic parameters, characterizing the interaction between
CA IV and the sulfonamides in the binding-able protonation
forms, including Gibbs energy, enthalpy, and entropy, that
could be used for the characterization of binding to any CA
in the process of drug design.
Keywords Carbonic anhydrase · Isothermal titration
calorimetry · Fluorescent thermal shift assay · Diferential
scanning fuorimetry · Intrinsic thermodynamics of
protein–ligand binding · X-ray crystallography
Abbreviations
AZM
Acetazolamide (also abbreviated as AAZ,
ACTAZ)
BSA
CA
Benzenesulfonamide
Carbonic anhydrase
CA IV
Isozyme 4 of the human carbonic anhydrase
protein family
EZA
FTSA
ITC
Ethoxzolamide
Fluorescent thermal shift assay
Isothermal titration calorimetry
Methazolamide (also METHZ)
MZM
* Daumantas Matulis
PAMBS Para-aminomethyl-benzenesulfonamide
matulis@ibt.lt; daumantas.matulis@bti.vu.lt
TFS
Trifuoromethane sulfonamide (also TFMSA)
1
Department of Biothermodynamics and Drug Design,
TPM
Topiramate
Institute of Biotechnology, Life Sciences Center, Vilnius
University, Saulėtekio 7, 10257 Vilnius, Lithuania
2
Department of Biochemistry and Molecular Biology, Saint
Introduction
Louis University School of Medicine, St. Louis, MO 63104,
USA
Carbonic anhydrases (CAs) are metalloenzymes that cata-
lyze the reversible hydration of CO2 into bicarbonate and
protons and thus are essential to maintaining intracellular
3
Department of Protein-DNA Interactions, Institute
of Biotechnology, Life Sciences Center, Vilnius University,
Saulėtekio 7, 10257 Vilnius, Lithuania
Vol.:(0123456789)
1 3

Eur Biophys J
and extracellular pH. CAs are found in all three domains of
life and belong to several diferent families: α, β, γ, δ, ζ, η
and θ (Zolfaghari Emameh et al. 2016a, b; Kikutani et al.
2016). α-CAs are found in many organisms, including fungi
(together with β-CAs) (Elleuche and Pöggeler 2010). β-CAs
are also found in plants, archaea, and both eukaryotes and
prokaryotes (Supuran 2004). γ-CAs are found in bacteria
(Ferry 2010). δ-CAs and ζ-CAs arefound in certain marine
planktonic species (Lane et al. 2005). η-CAs are found in
Plasmodium falciparum, malaria causative agent (Del Prete
et al. 2014). θ-CAs are found in marine diatom Phaeodac‑
tylum tricornutum (Kikutani et al. 2016). Most CAs have
zinc(II) ions in their active center, though some can have
cadmium, iron or cobalt (Lane et al. 2005, Ferry 2010).
Human CAs belong to the α-CA family and are zinc-con-
taining metalloenzymes. Humans have 15 CA isoforms but
only 12 of them contain Zn(II) and are catalytically active.
Each isoform has somewhat diferent kinetic properties and
inhibitor afnity profles, and may be located in diferent tis-
sues. In the cell, diferent CAs can be located in the cytosol
(CA I, CA II, CA III, CA VII, and CA XIII), in mitochondria
(CA VA and CA VB), or on the cell membrane (CA IV, CA
IX, CA XII, and CA XIV), or be secreted in saliva and milk
(CA VI) (Purkerson and Schwartz 2007; Liu et al. 2012).
These enzymes are important not only for pH maintenance,
but also for signal transduction, bone resorption, calcifca-
tion, renal acidifcation, gluconeogenesis, gastric acid for-
mation, metabolism, adaptation to cellular stress, biosyn-
thesis, and other processes (Breton 2001; Kivelä et al. 2005;
Krishnamurthy et al. 2008).
malfunction of this metabolon correlates with some diseases
(McMurtrie et al. 2004). CA IV catalyses CO2 exchange in
lungs and the hydration of CO2 that is produced in muscle
during exercise (Waheed et al. 1996). Together with CA
XIV, CA IV is the main CA in the brain extracellular space
that maintains pH homeostasis. CA XIV and CA IV also
maintain intracellular pH in the hippocampal neurons (Svi-
char et al. 2009). Together with CA II, CA IV participates
in compacting of myelin membranes, but CA IV alone is not
sufcient to keep myelin compact (Cammer et al. 1995). CA
I–IV are found in esophageal epithelium where, among other
functions, they protect against gastric acid refux (Christie
et al. 1997).
CA IV plays a role in the development of several dis-
eases, such as retinitis pigmentosa and glaucoma. Retinitis
pigmentosa, an inherited progressive eye disease, results in
blindness at the age of 40–50 years. Retinitis pigmentosa
afects one in 3000–7000 people, both men and women
(Köhn et al. 2008; Datta et al. 2009; Ferrari et al. 2011).
In these patients, CA IV carries the R14W mutation and
the mutated protein cannot maintain the pH homeostasis,
causing the photoreceptors to degenerate due to apoptosis
(Yang et al. 2005). Recently it was shown that CA IV plays
a very important role in suppressing colorectal cancer and
in wound healing (Zhang et al. 2016). After injury, CA IV
expression is induced by infammation and, together with
CA IX, it generates an acidic micro-environment, crucial
for neutrophil survival and keratynocyte migration (Barker
et al. 2017). CA IV and CA III functions might be altered in
rheumatoid arthritis, systemic lupus erythematosus, type 1
and 2 diabetic nephropathy, type 1 and 2 diabetes, and heart
failure, since high titers of anti–CA III and/or anti–CA IV
antibodies were found in the serum of people afected by
these diseases (Liu et al. 2012). There are also antibodies to
CA IV in the serum of people diagnosed with autoimmune
pancreatitis and Sjögren’s syndrome (Nishimori et al. 2005).
CA IV is one of the most enzymatically active CA iso-
forms. Its kcat for CO2 hydration is 1.1 × 106 s−1 while
the CA II kcat is 8 × 105 s−1. CA IV is also more active in
HCO3− dehydration than CA II (Baird et al. 1997). CA IV is
inhibited by bromophenol compounds (Balaydin et al. 2012).
It is also inhibited by acetazolamide (AZM), ethoxzola-
mide (EZA), and bromosulfanilamide. Celecoxib and some
halogenated sulfanilamides are also inhibitors of medium
potency. CA IV is located outside the cell. Therefore, it can
be selectively inhibited by positively charged sulfonamides
that cannot cross the plasma membrane (Supuran 2008).
Since CA IV is involved in many processes, selective inhibi-
tors might have clinical importance (Innocenti et al. 2005).
The crystal structure of the secretory form of human CA
IV was frst determined by Stams et al. (Stams et al. 1996).
A unique loop on the opposite side of the active site, the
Val-131-Asp-136 segment, was found to be specifc to this
CA IV is the most widely distributed of all membrane-
associated CA isoforms and has a unique glycosylphosphati-
dylinositol anchor that attaches it to the membrane in the
outer surface (Purkerson and Schwartz 2007). This anchor
may also activate the protein. It is also known that CA IV
can be activated by small concentrations (< 20 mM) of chlo-
ride, bromide, and phosphate (Baird et al. 1997) and it was
the frst discovered membrane–associated CA (Datta et al.
2010). The CA IV propeptide length is 312 amino acids
and the MW is about 35 kDa (Kivelä et al. 2005). Human
CA IV is physiologically stabilized by two disulphide bonds
between Cys24–Cys36 and Cys46–Cys229 (Waheed et al.
1996). These bonds make CA IV stable, for example, in the
presence of 5% SDS, while CA II is deactivated in these
conditions (Baird et al. 1997). CA IV is widely distributed
in the human body, including kidneys, lungs, colon, pancreas
cell plasma membranes, eye and brain capillaries, nasal
mucosa, esophageal epithelium, salivary glands, and heart
muscle (Pastorekova et al. 2004; Supuran 2004; Purkerson
and Schwartz 2007; Datta et al. 2009). CA IV is the most
important protein for HCO3− resorption in kidneys (Sterling
et al. 2002). CA II and CA IV are part of the bicarbonate
transport metabolon. Maintaining pH is very important and
1 3

Eur Biophys J
isozyme instead of the alpha-helix conformation found in
other isozymes.
most potent binder of CA IV, with an extraordinary Kd-obs
of 1 nM.
We have designed a series of 35 novel substituted ben-
zenesulfonamides and have discovered a compound 19
(3-[(4-chloro-2-phenylsulfanyl-5-sulfamoyl-benzoyl)amino]
propyl acetate) that bound CA IV with the Kd of 1.0 nM and
had signifcant selectivity towards this isoform over other
human CA isoforms. The high-resolution crystal structure
for compound 20 in complex with CA IV revealed the most
important interactions between the ligand and protein.
A detailed thermodynamic analysis of EZA and trif-
luormethanesulfonamide (TFS) binding to CA IV was per-
formed by isothermal titration calorimetry (ITC) and fuo-
rescent thermal shift assay (FTSA) to determine the intrinsic
thermodynamics of compound binding. Intrinsic values do
not depend on pH or bufer because they take into account
the binding-linked protonation reactions, while direct deter-
minations of binding or inhibition by any enzymatic or bio-
physical techniques always yield the observed dissociation
constants that should not be directly used to correlate the
structure with thermodynamics. The pH dependence of the
binding afnities and the observed enthalpies of binding
enabled us to obtain the pKa and the enthalpy of protona-
tion of the Zn bound water at the active site of CA IV. These
values should help other researchers to determine the intrin‑
sic thermodynamics of binding to CA IV and develop drugs
targeting CA IV.
For analysis of the selectivity, we tested the binding of
fve selected compounds toward all CA isoforms and com-
pared the data with BSA and AZM (Table 2). Compound 5
was a selective and high-afnity binder of CA I and CA VB
(Kd-obs in the low nanomolar range), and had a lower afn-
ity to CA IV and CA VA (Kd-obs in the range 170–100 nM).
Compounds 18, 19, and 26 bound CA IV 28, 4000, and
23-times stronger, respectively, than CA I, one of the most
abundant isoforms found in the human organism. However,
their binding afnity to CA VII, CA IX, CA XII, CA XIII,
and CA XIV was higher or similar to CA IV. Compound
19 was selective toward CA IV, bound CA IV 4000 times
stronger than CA I and 10 times stronger than CA II. This
compound, or compounds with similar structures, may be
developed as drugs targeting CA IV.
Crystal structure of CA IV in complex with inhibitors
The crystal structures of human CA IV in complex with the
synthesized compound 20 and fve commercially available
CA inhibitors, namely EZA, AZM, methazolamide (MZM),
4-aminomethylbenzenesulfonamide chloride (PAMBS) and
topiramate (TPM), were determined by X-ray crystallogra-
phy. Data collection and refnement statistics are shown in
Table 3. The quality of electron densities for the complexes
of compound 20 and fve other inhibitors is shown in Fig. 3,
parts b1–b6. For the complex with compound 20, the asym-
metric unit of the crystal structure contains four subunits.
Three sets of electron densities attributable to the inhibitor
were found in protein chains A, B, and C. The location of
the compound is nearly identical in all three protein chains
with small diferences in the positions of the second ring.
The location of the 2-chlorobenzenesulfonamide ring in
the active site of several human CA isoforms (CA XII, CA
XIII, and CA II) described earlier (Čapkauskaitė et al. 2013)
is very similar. The chlorine atom is anchored in a hydro-
phobic pocket formed by the residues conserved between
human CA isoforms (in CA II Val121, Leu141, Val143, and
Leu198). On the other hand, the active site of CA IV is more
hydrophilic (Fig. 3c1), but the hydrophilic residues do not
interact with inhibitor 20. The dissociation constants of com-
pound 20 with CA IV and CA II are very similar (200 and
290 nM, respectively, Table 2). The dissociation constants
of 20 with CA XII and CA XIII are also very similar (330
and 100 nM accordingly, Table 2). These values can refect
similar interaction of 20 with these CA isoforms (Fig. 3a1),
since a relatively simple compound interacts with the most
conserved part of the CA binding site.
Results and discussion
Observed thermodynamics of ligand binding to CA IV
The binding afnity of a novel series of substituted benze-
nesulfonamides 1–35 (Fig. 1) to CA IV was determined by
FTSA. The compounds that bind to the active site of CA
IV stabilized the protein and shifted the protein melting
temperature (Tm) upward (Fig. 2). The observed dissocia-
tion constants (Kd-obs) were obtained after the simulation
or regression of the ligand-dosing curve according to the
model (Cimmperman et al. 2008). According to the data
in Table 1, many compounds showed better afnities for
CA IV than the non-substituted compound benzenesul-
fonamide (BSA). Tetrafuorbenzenesulfonamide deriva-
tives 1–10 bearing substitutions at the para position bound
CA IV with Kd-obs ranging from 50 to 830 nM. The intro-
duction of ortho substituent (compounds 11–15) resulted
in much weaker binding to CA IV (Kd-obs in the range of
400–4000 nM). However, meta substituted derivatives
17–18 bearing a short hydroxyethylthio group at the para
position had higher binding afnities (Kd-obs 17–18 nM).
For the series of 19–35 compounds, compound 19 was the
The buried surface area (in Å2) calculated by PISA v1.06
(Krissinel and Henrick 2007) values of the side chains that
interact with 20 for protein chain A are given in Table 4. The
1 3

Eur Biophys J
SO2NH2
F
SO2NH2
F
SO2NH2
F
SO2NH2
SO2NH2
F
SO2NH2
F
SO2NH2
F
SO2NH2
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
S
S
S
NH
VD12-13
S
SO2
VD10-35
F
VD10-13
VD10-9
1
VD12-22
2
VD12-05
VD11-09
VD10-18
3
5
4
7
8
6
HO
HO
SO2NH2
SO2NH2
F
SO2NH2
H
N
F
F
F
F
F
F
SO2NH2
SO2NH2
HO
H
N
H
N
F
F
F
F
SO2NH2
F
F
F
VD10-28
F
H
N
F
F
S
S
S
O
O
F
F
VD11-71
11
N
VD10-50
10
S
S
S
N
N
N
9
VD12-09
VD12-17
F
N
12
13
VD12-36
14
HO
HO
SO2NH2
SO2NH2
F
SO2NH2
SO2NH2
SO2NH2
F
H
N
F
F
F
Cl
F
F
F
F
F
H
N
O
F
F
N
H
N
H
N
H
VD12-37
SO2
S
SO2
SO2
O
O
O
O
VD11-4-2
17
VD11-54-2
16
15
EA4-2c
VD12-34
18
19
HO
HO
SO2NH2
SO2NH2
SO2NH2
SO2NH2
SO2NH2
SO2NH2
Cl
Cl
Cl
Cl
O
O
N
O
N
N
OH
OH
S
S
NH2
O
O
S
N
NH
N
N
E11-27
21
N
NH
E52
22
E35
20
E11-51
24
E6
25
E2
23
N
N
O
O
SO2NH2
SO2NH2
SO2NH2
SO2NH2
SO2NH2
SO2NH2
SO2NH2
Cl
Cl
Cl
Cl
S
S
O
HO
E40
27
O
O
S
O
O
O
N
N
N
O
NH
E46
26
S
OH
OH
E67
28
N
N
N
S
N
O
E99
E11-26
29
E86
30
E11-50
31
N
N
NH
32
NH
O
O
O
SO2NH2
SO2NH2
SO2NH2
SO2NH2
SO2NH2
S
SO2NH2
SO2NH2
S
SO2NH2
SO2NH2
O
O
F
F
N
S
O
N
F
N
N
N
TFS
O
N
NH
S
O
S
O
O
O
O
O
H2N
S
PAMBS
MZM
TPM
AZM
E95
33
VD12-10
34
VD11-31
35
EZA
N
HO
NH
O
Fig. 1 The structures of synthesized compounds 1–35 and six conventional CA inhibitors EZA, TFS, AZM, MZM, PAMBS, and TPM, that
were used in this study
largest buried surface area values contain residues Asn69,
Met74, Gln96, His98, Leu207, and Thr209. Most of these
residues are present both in CA IV and CA II isoforms,
except for Phe131 in CA II, which does not have a structural
counterpart in the crystal structures of CA IV, and Met74
in CA IV, which coincides spatially with Asn67 in CA II.
In the protein chain A of CA IV, an electron density was
found, corresponding to residues Lys128–Gln140, that was
absent in other protein chains of PDB ID 5IPZ as well as
in other CA IV crystal structures (except PDB ID 1ZNC,
2.8 Å) (Stams and Christianson 2000; Vernier et al. 2010).
Despite the poor quality of the electron map, we were able to
put most of the residues into the protein model. The fexible
loop interacts with the surface of the symmetric equivalent
of the protein chain B, and it is likely that such a conforma-
tion refects an artifact of the crystal packing. Interestingly,
the residues that correspond to this loop are quite well fxed
in CAII (residues 127–137) (Fig. 3a2) and form part of the
1 3

Eur Biophys J
(a)
(b)
(a)
(b)
pH 5.5
CA I
40
20
40
0
0
20
15
70
200
pH 7.0
CA IV
8
6
4
0
0
10
70
70
200
50
60
T (oC)
80
50
60
70
80
T (oC)
75
70
65
60
55
75
70
65
60
55
(c)
(c)
Kd(pH 7.0) = 36 nM
Kd(CA IV) = 29 nM
Kd( pH 5.5) = 67 nM
Kd(CA I) = 670 nM
10-5
10-4
10-3
-6
-6
10-5
10-4
0
0
10
10
Lt (M)
Lt (M)
Fig. 2 Fluorescent thermal shift assay data of EZA (left) binding to
CA IV and compound 26 (right) binding to CA I and CA IV. Left
panels a-b show the thermal melting fuorescence curves of CA IV
(10 µM) in the universal bufer, pH 5.5 and 7.0, in the presence of
various EZA concentrations, 0 µM (flled square), 4.4 µM (open
square), 8.8 µM (flled triangle), 17.5 µM (open triangle), 35 µM
(flled circle), and 70 µM (open circle). Right panels a, b show the
thermal melting fuorescence curves of CA I (5 µM) and CA IV
(10 µM) in phosphate bufer, pH 7.0, in the presence of various com-
pound 26 concentrations, 0 µM (flled square), 12 µM (open square),
50 µM (flled triangle), 100 µM (open triangle), and 200 µM (open
circle). Panels c show the ligand dosing curves plotting the melting
temperatures (Tm) from data in panels a, b as a function of added
compound concentration. The lines were regressed according to
(Cimmperman et al. 2008)
active site in CA II. Phe131, which is often involved in the
interaction with sulfonamide inhibitors, is also located in
this region.
the active site pocket in CA II. The structures of residues
126–138 in human CA IV were very fexible and there were
four diferent conformations (including the disordered one)
in the segment found so far. Although the conformations in
this region were diferent, all of them made the active site
pocket bigger and allowed larger inhibitors to interact with
the active site.
For the complexes with AZM, EZA, TPM, PAMBS, and
MZM, crystals were grown in space group p212121, with
almost identical unit cells containing four molecules in the
asymmetric unit (Table 3). The structures of the CA mol-
ecules A, B, C, and D including inhibitors were very similar,
with the rms deviation of ~0.2 Å for these fve complexes.
However, residues 126–138 (using the numbering of CA
II) in molecules A and B were disordered while residues
126–138 in molecule C and D were ordered and visible in
the electron density map. Comparing the structures of mol-
ecule C in the complex of CA IV and AZM, molecule A of
PDB ID 1ZNC, the frst crystal structure of human CA IV
where the residues 126–138 in molecule B were disordered
(Stams et al. 1996), molecule A in complex of human CA
IV and compound 20, and human CA II (PDB ID 4HT0)
(Fig. 3a2), showed that the orientation of residues 126–138
in CA IV difers greatly from CA II. The structure of resi-
dues 126–138 in CA II is very similar with that of another
CA family, in which residue 131 was an important residue
located in a short helix and interacted with inhibitors in most
isoforms of the CA family. This helix acts as a lid and closes
The active site structure in the complex of human CA
IV and inhibitor consisted of a metal binding site, protein
ligands from the nitrogen atoms of His94, His96, and His119
as well as the nitrogen atom from the inhibitor, a hydropho-
bic pocket containing the residues Val121,Val143, Ile141,
Val207, Trp209, and Leu198, and a hydrophilic pocket
consisting of Tyr7, Asn62, His64, Ser65, Met67, Gln92,
Thr199, and Thr200. Figure 3, parts c1, c2, c3, c4, c5, and
c6, shows the interactions between inhibitors and the protein
in the active site areas. All inhibitors formed two hydrogen
bonds with the nitrogen atom from the main chain and the
oxygen atom from the side chain of Thr199, which is a cata-
lytic residue in CA IV. Residue His64 from the hydrophilic
pocket is a proton shuttle in catalysis (Stams et al. 1996).
Gly-63- → Gln substitution in rodent CA IV revealed the
importance of residue His64 as a proton shuttle (Tamai et al.
1996). In fact, the complex structure of CA IV and AZM in
1 3

Eur Biophys J
Table 1 The tested compound afnities to CA IV, determined by the
fuorescent thermal shift assay at pH 7.0 and 25 °C
also observed that there were two conformations for the side
chains of His64 in some crystal structures of CA II and CA
VII. For the active site structure for compound 20 in com-
plex with CA IV, there were no other interactions between
the inhibitors and protein for the structures with EZA and
PAMBS in CA IV (Fig. 3c3, c5), which revealed the most
important hydrophobic interactions between the ligand and
the protein. However, in the active site structure of CA IV
with TPM, there were six hydrogen bonds in total between
TPM and the hydrophilic residues. This bonding was attrib-
uted to the more hydrophilic oxygen atoms in TPM and pro-
vided a guide for designing new inhibitors with improved
binding selectivity towards CA IV. Unfortunately, the crystal
structure of 19 with CA IV was not obtained. This structure
could have revealed the structural features of the lead com-
pound, but the crystallization procedure was unsuccessful.
Compound Compound lab name pKa Kd_obs (nM) Kd_intr (nM)
1
VD10-9
VD12-22
VD10-13
VD10-35
VD12-05
VD10-18
VD11-9
VD12-13
VD10-28
VD10-50
VD11-71
VD12-09
VD12-17
VD12-36
VD12-37
VD11-54-2
VD11-4-2
VD12-34
EA4-2c
E35
E11-27
E52
E2
E11-51
E6
E46
E40
E67
E11-26
E86
E11-50
E99
8.3 260
8.4 250
8.3 170
7.4 830
8.3 170
8.0 100
8.2 220
8.8 110
7.9 50
8.2 770
8.8 3300
8.8 400
8.8 3300
8.9 400
8.8 4000
8.1 670
8.2 17
4.5
3.4
2.9
85
2.9
3.3
4.7
0.62
2.0
16
19
2.2
19
1.8
22
18
0.36
0.59
0.0036
0.89
8.9
0.64
0.51
0.064
1.5
0.13
0.18
0.19
2.0
0.95
2.9
0.023
0.18
0.057
0.15
2.0
0.40
35
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
Determination of pKa of Zn‑bound water molecule
in the active site of CA IV
The pKa and enthalpy of protonation values of Zn(II)-bound
hydroxide in CA cannot be determined directly by pH titra-
tion or ITC because there are many overlapping ionizing
groups in proteins. Therefore, the pKa is determined indi-
rectly by modeling the U-shape dependence of the afnity
on pH and the X-shape observed enthalpy dependence on
pH (Fig. 4). The enthalpy of the hydroxide protonation is
also determined from the model of the observed enthalpy
dependence on pH in two bufers of diferent protonation
enthalpy (Fig. 4).
8.0 18
9.0 1.0
8.9 200
9.4 6300
8.9 140
9.0 140
9.5 50
9.0 440
8.9 29
8.9 40
9.4 133
8.9 460
9.1 330
8.6 330
9.4 16
9.4 125
10.5 500
10.2 670
10.0 5600
8.0 16.2
6.3 118
7.3 50
The binding of EZA, measured by ITC in two bufers,
phosphate (Pi) and TRIS, demonstrated the dependence of
the observed binding enthalpy ∆bHobs on the bufer (Fig. 4a).
The binding parameters strongly depended on the bufer pH as
well, both for EZA and for TFS binding (Fig. 4b, c). Fitting of
the observed experimental data obtained in TRIS and Pi buf-
ers by applying Eq. (4) yielded the intrinsic binding enthalpy
(horizontal dashed line). Fitting of the data enabled the deter-
mination of the pKa of CA IV, which was equal to 7.0.
Another useful technique for the estimation of pKa of
the Zn(II)-bound water form of CA IV is FTSA. Figure 4d
shows the dependence of the observed binding afnity on
pH. The resulting U-shape of pH dependence can be approx-
imated by Eq. (1). The value of pKa for the Zn(II)-bound
water form of CA IV used for ftting the U-shape was 6.5,
a value which is slightly lower than that obtained by ITC
technique. The intrinsic Gibbs energy for EZA binding was
equal to −53.6 kJ/mol.
E95
VD12-10
VD11-31-2
BSA
EZA
TFS
AZM
6.0
MZM
PAMBS
TPM
7.1 143
10.1 3300
8.7 45
23
1.0
0.32
The observed values were determined experimentally while the
intrinsic afnities were calculated according to Eq. (1)
Characterization of CA IV stability by FTSA
molecule A showed two conformations for the side chain of
His64; one of them directed toward the active site and fur-
ther certifed the importance for this residue in CA IV. It was
FTSA was used to identify bufer conditions (pH ranges
from 5 to 10) that stabilized CA IV. Diferent additives (salt
1 3

Eur Biophys J
Table 2 The observed dissociation constants of selected compounds for all catalytically active human CA isoforms, determined by the fuores-
cent thermal shift assay at pH 7.0 and 25 °C
Compound
Kd_obs (nM)
CA I
CA II
CA III
CA IV
CA VA
CA VB
CA VI
CA VII
CA IX
CA XII
CA XIII
CA XIV
5
0.017
500
4000
5000
670
2860
714
0.83
10
10
290
33
1110
13
6670
170
18
1.0
200
29
5560
50
100
1250
1050
670
40
2860
500
0.020
2.5
40
630
250
12,500
25
50
33
590
250
100
5000
83
0.45
1.7
3.6
67
29
2860
8.3
2.0
0.050
2.0
130
25
500
14
17
0.63
2.5
130
100
56
3570
25
2.0
1.1
2.5
20
4.0
3330
22
18
11,100
33,300
100,000
2200
200,000
16,700
2.5
3.8
330
77
5560
71
19
20
26
BSA
AZM
The Kd for compounds 5, 18, and 26 binding to CA I, II, VII, XII, and XIII are taken from our previously published data, recalculated to 25 °C
(Dudutienė et al. 2013, 2015; Čapkauskaitė et al. 2013; Zubrienė et al. 2017). Binding data for 19 and 20 are reported here for the frst time
substances, glycerol, DMSO) and EZA inhibitor concen-
trations were used to examine a stabilizing or destabilizing
efect on CA IV. The stability of CA IV was afected by pH,
bufer, and the presence of various salts (Fig. 5). The protein
was most stable at pH 8.0–8.5 (Fig. 5a, b). CA IV at pH 8.0
was more stable in Pi and HEPES bufers than in TRIS and
carbonate bufers. Destabilization of CA IV by carbonate at
pH 8–10 and by citrate bufer at pH 4.5–6.0 was observed.
These bufers may have to be avoided in CA IV purifcation
or storage.
(Kišonaitė et al. 2014), thus the pKa for meta substituted
compounds 20 and 27 with chlorine atoms were 8.9, and
for para-substituted benzenesulfonamide 32 was 9.4. The
thermodynamic parameters of protonation of the depro-
tonated inhibitor sulfonamide group and the CA-Zn(II)-
bound hydroxide anion, determined by ITC at 25 °C, are
listed in Table 5.
The intrinsic Kd-intr for synthesized compounds 1–35
and for reference compounds are listed in Table 1. The
intrinsic Kd-intr for all compounds was signifcantly higher
than the observed dissociation constants. It is evident that
the lower the value of pKa for sulfonamide, the smaller the
diference between the intrinsic and the observed dissocia-
tion constants (Fig. 6a, b). The largest diferences between
the observed and intrinsic Kd values were observed for
compounds 34, 35, and BSA, where the binding to CA IV
difered more than 8300-fold. The most tightly binding
inhibitor, according to both the observed and intrinsic Kds,
was compound 19, where the intrinsic Kd reached 3.6 pM,
whereas the observed Kd was equal to 1.0 nM. Since the
pKa of TFS was equal to 6.3, the lowest among the tested
compounds, the intrinsic Kd value for TFS was closest to
the observed (Kd-intr/Kd-obs = 3.3).
Regular salts such as sodium chloride, sodium sulfate,
and ammonium sulfate had diferent infuences on the sta-
bility of CA IV. Sodium sulfate exhibited the greatest sta-
bilization efect on the protein—the addition of 1 M sulfate
increased the protein Tm by about 5 °C. Sodium chloride
had a minor efect on CA IV stability, whereas the addi-
tion of ammonium sulfate slightly decreased protein Tm,
primarily because it diminished the pH of the protein solu-
tion (Fig. 5c).
Both glycerol and DMSO had a minor destabilizing efect
for CA IV (Fig. 5d). At 5–10% (v/v) of glycerol, the protein
stability decreased by about 1 °C. The addition of DMSO at
these concentrations destabilized the protein by about 2 °C.
The enthalpy-entropy compensation graph (Fig. 6c)
shows that the intrinsic Gibbs energy of binding for the
tested series of compounds (listed in Table 6) spreads over
24 kJ/mol, whereas the ΔΔbHint varies by 38 kJ/mol and
TΔΔbSintr by 50 kJ/mol. There is no clear enthalpy-entropy
compensation and the enthalpy and entropy contribution
to binding afnity is favorable, except for AZM, for which
the large negative enthalpy (ΔbHint = −50.6 kJ/mol) is
slightly opposed by the negligible entropy contribution
(TΔbSintr = −3.7 kJ/mol). EZA, TPM, AZM, MTZ and
compound 26 bound to CA IV with the dominant enthalpy
contribution, whereas compounds 5, 18, 19 and 20—with
the dominant favorable entropy. Favorable entropy and
Intrinsic thermodynamics of ligand binding
In order to calculate the intrinsic binding parameters by
Eq. (1), the protonation parameters of both protein and sul-
fonamide are necessary. The pKas of most tested sulfona-
mides were previously determined (Zubrienė et al. 2015;
Linkuviene et al. 2016). The pKa of the sulfonamide group
for compound 19 was measured spectrophotometrically by
measuring absorption at diferent pHs. The pKas of 20, 27,
and 32 were determined by analogy with other meta- and
para-substituted benzenesulfonamides. We have recently
shown that only the head-groups but not the tail groups on
benzenesulfonamides infuenced the ionization constants
1 3

Eur Biophys J
Table 3 X-ray crystallographic data collection and refnement statistics
Names of inhibi-
tors:
Compound 20
Acetazolamide
Ethoxzolamide
Topiramate
4-Aminometh- Methazolamide
ylbenzenesul-
fonamide
Abbreviation:
AZM
20
EZA
3
TPM
14
PAMBS
MZM
20
Concentration
(mM):
0.4
20
Bufer/salt
100 mM MES,
pH 6.5/200 mM
(NH4)2SO4
100 mM Na
acetate, pH
4.6/200 mM
(NH4)2SO4
100 mM Na
acetate, pH
4.6/200 mM
(NH4)2SO4
100 mM Na
acetate, pH
4.6/200 mM
(NH4)2SO4
100 mM Na
acetate, pH
4.6/200 mM
(NH4)2SO4
100 mM Na acetate,
pH 4.6/200 mM
(NH4)2SO4
PEG (%)
PDB ID
2000(20)
5IPZ
4000(21)
5JN8
3350(16)
5JN9
4000(21)
5JNA
3350(16)
5JNC
4000(23)
5KU6
Data collection
Wavelength (Å)
Space group
0.83
P21
1.54
P212121
a = 64.9
b = 123.5
c = 151.2
1.54
P212121
a = 64.8
b = 124.0
c = 151.3
1.54
P212121
a = 64.9
b = 123.8
c = 151.4
1.54
P212121
a = 64.8
b = 123.8
c = 151.4
1.54
P212121
a = 65.0
b = 123.7
c = 151.4
Unit cell dimen- a = 48.7
sions (Å)
b = 256.4
c = 48.6
β = 117.3°
Molecules/asym-
metric unit
4
4
4
4
4
4
Resolution range 40–2.1
40–1.85
40–2.1
40–2.0
40–2.0
40–1.8
(Å)
Observations
466,257
102,801
225,442
68,006
375,127
80,516
268,928
80,221
484,979
110,310
Unique observa- 51,979
tions
Completeness
(%)
Rsym (%)
I/σ(I)
85.0 (77.0)
98.5 (96.8)
94.5 (93.4)
96.8 (92.7)
96.5 (90.9)
97.3 (94.2)
4.9 (41.7)
10.5 (2.1)
11.3 (47.4)
10.2 (2.0)
10.8 (46.4)
8.1 (1.9)
9.6 (42.4)
12.1 (2.4)
11.4 (43.4)
8.0 (2.0)
8.9 (42.2)
11.4 (2.1)
Refnement
Resolution (Å)
Rcryst, Rfree
40–2.1
0.225, 0.306
46,821/5105
40–1.85
0.185, 0.228
97,258/5127
40–2.1
0.170, 0.220
64,559/3413
40–2.0
0.172, 0.218
76,157/4002
40–2.0
0.174, 0.218
76,144/4039
40–1.8
0.184, 0.216
104,356/5481
Refections
(working/test)
Protein atoms
8252
8391
4/4
8376
4/4
8392
4/4
8376
4/4
8376
4/4
Zn++ ions/inhibi- 4/3
tors
Solvent mol-
120
962
916
888
937
915
ecules
RMSD bond
lengthsa (Å)
0.012
1.7
0.010
0.013
0.010
0.012
1.6
0.009
1.4
RMSD anglesa
(°)
1.3
1.8
1.3
RMSD ΔB (Å2) 3.43/3.23/3.70
1.49/1.54/2.25
20.8
1.90/2.23/2.96
28.9
1.44/1.46/2.34
26.5
1.62/1.88/2.51 1.36/1.48/2.27
(mm/ms/ss)
<B > protein
50.5
24.0
21.0
(Å2)
<B > Zn++
/
32.3/47.2
41.8
12.2/27.0
32.6
19.6/27.2
38.5
16.9/22.2
38.8
15.5/34.2
36.2
12.9/31.3
33.8
inhibitors (Å2)
<B > solvent
(Å2)
Ramachandran plot
Most favored (%) 90
100
100.0
100
100
100
1 3

Eur Biophys J
Table 3 (continued)
Names of inhibi-
tors:
Compound 20
Acetazolamide
Ethoxzolamide
Topiramate
4-Aminometh- Methazolamide
ylbenzenesul-
fonamide
Abbreviation:
AZM
20
EZA
3
TPM
14
PAMBS
MZM
20
Concentration
(mM):
0.4
20
Generously
allowed (%)
Disallowed (%)
9.8
0.2
0
0
0
0
0
0
0
0
0
0
mm main chain–main chain, ms main chain-side chain, ss side chain-side chain
aRoot-mean-squared deviation (RMSD) from ideal bond lengths and angles and RMSD in B-factors of bonded atoms
enthalpy contributions to binding afnity, in almost equal
magnitudes, were observed for TFS.
intrinsic enthalpy contributed favorably to the Gibbs energy
of EZA binding, whereas TFS binding was driven by entropy
and enthalpy in equal parts. CA IV was most stable at pH
8.0–8.5. High sodium sulfate concentrations signifcantly
stabilized the protein. Sodium chloride had insignifcant
efect on protein stability. High-resolution crystal structures
for compound 20, EZA, and PAMBS in complex with CA
IV revealed the most important hydrophobic interactions
between the ligand and the protein. The structure of residues
126–138 in human CA IV was a loop conformation and was
very fexible. There were four diferent conformations in the
segment. Although the conformations in this segment were
diferent, all of them made the active site pocket bigger and
allowed larger inhibitors to interact with the active site. This
conformation is a unique feature of CA IV in the CA family.
The results presented here show that binding of CA IV to
sulfonamides was signifcantly infuenced by the composi-
tion of the bufer and its pH. The pKa and ionization enthalpy
of the Zn(II)-bound water molecule in the active site of CA
IV were determined due to the EZA and TFS titrations at
various pHs of TRIS and Pi bufers. Measurement of the
binding afnity at various pHs by FTSA helped validate the
pKa values as well. When the pKa and the enthalpy of proto-
nation of sulfonamide and protein, as well as the enthalpy of
bufer protonation, are known, the intrinsic thermodynamic
parameters can be calculated and correlated with the high-
resolution protein–ligand crystal structures. It was necessary
to determine the pKa and enthalpy of protontion of the water
molecule bound to Zn only once. When this has been accom-
plished here, future researchers can use these values in the
calculation of the intrinsic parameters of any other sulfona-
mide ligand binding. A single determination at a single pH
in a known bufer will be sufcient to calculate the intrinsic
parameters. Such information may be valuable for further
drug design of more efective and selective compounds.
These known values enabled us to calculate the intrinsic
binding values even from a single ITC titration experiment
in a bufer with known protonation enthalpy and at a known
pH (Table 6).
Experimental section
Chemical compounds
EZA, AZM, TPM, MZM, TFS, and PAMBS were purchased
from Sigma-Aldrich (St. Louis, MO, USA). The synthesis,
chemical structural characterization, and the purity of com-
pounds 1–35 has been previously described (Dudutienė et al.
2013, 2015; Čapkauskaitė et al. 2013; Zubrienė et al. 2014),
with the exception of 19, 20, 27, and 32.
Conclusions
Synthesis of 19, 20, 27, and 32
The observed binding afnity and the binding enthalpy of
EZA to CA IV are pH-dependent. The pH dependence of
the observed thermodynamic parameters shows the con-
tributions of linked protonation-deprotonation reactions of
both ligand and protein. The dissection of binding linked
reactions yielded the intrinsic thermodynamic parameters,
describing the reaction between the Zn-bound water form of
CA IV and the anionic form of sulfonamide. The dominant
All starting materials and reagents were commercial prod-
ucts and were used without further purifcation. The melting
points of the synthesized compounds were determined in
open capillaries on a Thermo Scientifc 9100 Series appa-
ratus without further correction. IR spectra were obtained
on a Perkin-Elmer FT-IR spectrophotometer Spectrum BX
II in KBr. 1H and 13C NMR spectra were recorded on a Var-
ian Unity Inova (300 and 75 MHz, respectively) or Bruker
1 3

Eur Biophys J
(400 and 100 MHz, respectively) spectrometers in DMSO-d6
using residual DMSO signals (2.52 and 40.21 ppm for 1H
and 13C NMR spectra, respectively) as the internal standard.
TLC was performed with silica gel 60 F254 aluminum plates
(Merck, Kenilworth, NJ, USA) and visualized with UV light.
High-resolution mass spectra (HRMS) were recorded on a
Dual-ESI Q-TOF 6520 mass spectrometer (Agilent Tech-
nologies, Santa Clara, CA, USA). The purity of fnal com-
pounds was verifed by HPLC to be > 95% using the Agi-
lent 1290 Infnity 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/metha-
nol (20 mM ammonium formate in both phases) from 60:40
to 30:70 over 12 min, then from 30:70 to 20:80 over 1 min,
and then 20:80 over 5 min at a fow rate of 0.2 mL/min. UV
detection was at 254 nm.
Compound 19a, 2,4-dichloro-5-sulfamoyl-benzoic acid,
was converted to chloroanhydride, then subjected to ami-
dation of the acid choride with 3-aminopropan-1-ol to pro-
duce the amide, compound 19b. Compound 19b was then
treated to esterifcation with ethylacetate to produce com-
pound 19c, 3-[(2,4-dichloro-5-sulfamoyl-benzoyl)amino]
1 3

Eur Biophys J
◂Fig. 3 X-ray crystallographic structures of compounds in com-
plex with CA IV. (a1) The overlay of the binding sites of CA IV and
CA II. The residues important for 2-chlorobenzene ring binding are
colored yellow (CA IV) and orange (CA II). The residues that inter-
act with compound 20 in CA IV are colored green and the spatially
corresponding residues of CA II are colored dark green. The red
and dark green labels belong to CA IV and CA II residues, respec-
tively. (a2) Comparison for the structures: molecule C in complex
with AZM (green), molecule A in the PDB entry 1ZNC, frst crystal
structure of CA IV at 2.8 Å resolution (blue), molecule A in complex
with compound 20 (red), and human CA II (PDB ID: 4HTO, yellow,
residue 131 is shown). The orientation of residues 126–138 in CA
IV difers from CA II. The structure of residues 126–138 in CA II
is very similar to other members of the CA family, in which residue
131 was an important residue located in a short helix and interacted
with inhibitors in most CA family. This helix acts as a lid and closes
the active site pocket in CA II. The structures of residues 126-138 in
human CA IV were fexible and there were four diferent conforma-
tions (including a disordered one) in the four observed molecules.
Although the conformations in this region were diferent, all of them
made the active site pocket bigger and allowed larger inhibitors to
interact with the active site. (b1) The electron density of compound
20 in the active site of human CA IV (expressed in E. coli). The elec-
tron density map |Fobs-Fcalc| was calculated after removing the coor-
dinates of compound 20 from the model and contoured at 3.0 σ. (b2)
Electron density map |Fo-Fc| after removing the coordinates of AZM
from the model, contoured at 2.2 σ. (b3) The electron density map
|Fo-Fc| after removing the coordinates of EZA from the model, con-
toured at 3.0 σ. (b4) The electron density map |Fo-Fc| after remov-
ing the coordinates of TPM from the model, contoured at 3.0 σ. (b5)
The electron density map |Fo-Fc| after removing the coordinates of
PAMBS from the model, contoured at 3.0 σ. (b6) The electron den-
sity map |Fo-Fc| after removing the coordinates of MZM from the
model, contoured at 2.5 σ. (c) The binding mode of compounds in
the active site of CA IV. The Zn(II) atom is shown as a pink sphere.
The residues forming the hydrophilic area are colored aquamarine.
(c1) Interactions between compound 20 and CA IV. The chlorine
atom of chlorobenzene in 20 is located in the hydrophobic area of the
CA IV active site (residues colored yellow). This hydrophobic area is
conservative for human CA active sites (Alterio et al. 2012) and the
binding mode of the chlorobenzene ring is common for CA XII, XIII,
and II. Selected other residues of active site are shown in green. (c2)
Interactions between AZM and protein (molecule D), hydrogen bonds
DMF in toluene (5 ml) was refuxed for 4 h. Excess SOCl2
and toluene were removed by distillation under reduced
pressure, and the crude acid chloride was used directly in the
next step. The solution of 2,4-dichloro-5-sulfamoylbenzoyl
chloride in THF (30 ml) was added dropwise to a solution
of 3-aminopropan-1-ol (2.25 g, 30.0 mmol) in THF (20 ml)
at 0 °C and allowed to stirr for 1 h. The mixture was warmed
to room temperature and stirred for another 2 h. THF was
removed under reduced pressure. Water was added to the
residue and product was extracted with EtOAc. The organic
layer was washed with 5% HCl(aq), dried over anhydrous
MgSO4, filtered, and concentrated. Yield was 60%, mp
153–155 °C (H2O). 1H NMR (400 MHz) δ ppm: 1.67 (2H,
quint, J = 6.4 Hz, CH2), 3.30 (2H, q, J = 6.4 Hz, NHCH2),
3.48 (2H, t, J = 6.4 Hz, CH2OH), 4.49 (1H, br s, OH), 7.83
(2H, s, SO2NH2), 7.92 (1H, s, C3-H), 7.94 (1H, s, C6-H),
8.65 (1H, t, J = 5.6 Hz, NH). 13C NMR (100 MHz) δ ppm:
32.6, 36.9, 58.9, 129.2, 132.1, 132.5, 134.4, 136.4, 140.4,
164.9. HRMS calculated for C10H12Cl2N2O4S [(M + H)+]:
326.9968, found: 326.9971.
3‑[(2,4‑Dichloro‑5‑sulfamoyl‑benzoyl)amino]propyl
acetate (19c)
2,4-Dichloro-N-(3-hydroxypropyl)-5-sulfamoyl-benzamide
(19b) (327 mg, 1.00 mmol) was refuxed in EtOAc (7 mL)
with 3 drops of concentrated H2SO4 for 2 h. The reaction
mixture was concentrated under reduced pressure and the
resultant precipitate was washed with H2O. Yield: 288 mg,
1
78%, mp 149-151 °C. H NMR (400 MHz) δ ppm: 1.83
(2H, quint, J = 6.4 Hz, CH2), 2.02 (3H, s, CH3), 3.30 (2H,
q, J = 6.8 Hz, NHCH2), 4.08 (2H, t, J = 6.4 Hz, CH2O),
7.81 (2H, s, SO2NH2), 7.93 (1H, s, C3-H), 7.95 (1H, s,
C6-H), 8.71 (1H, t, J = 5.6 Hz, NH). 13C NMR (100 MHz)
δ ppm: 21.2, 28.5, 36.5, 62.1, 129.1, 132.2, 132.5,
134.4, 136.1, 140.2, 164.9, 170.9. HRMS calculated for
C12H14Cl2N2O5S[(M + H)+]: 369.0073, found: 369.0071.
are shown as dashed lines and Zn++
is shown as a pink sphere. (c3)
Interactions between EZA and protein (molecule D), hydrogen bonds
are shown as dashed lines and Zn(II) is shown as a pink sphere. (c4)
Interactions between TPM and protein (molecule B), hydrogen bonds
are shown as dashed lines and Zn(II) is shown as a pink sphere. (c5)
Interactions between PAMBS and protein (molecule C), hydrogen
bonds are shown as dashed lines and Zn(II) is shown as a pink sphere.
(c6) Interactions between MZM and protein (molecule A), hydrogen
bonds are shown as dashed lines and Zn++ is shown as a pink sphere
3‑[(4‑Chloro‑2‑phenylsulfanyl‑5‑sulfamoyl‑benzoyl)amino]
propyl acetate (19)
The mixture of 3-[(2,4-dichloro-5-sulfamoyl-benzoyl)
amino]propyl acetate (19c) (250 mg, 0.677 mmol), MeOH
(5 mL), thiophenol (82.1 mg, 0.745 mmol), and Et3N
(82.2 mg, 0.813 mmol) was refuxed for 3 h. MeOH was
evaporated under reduced pressure and the resultant pre-
cipitate was washed with H2O. Yield 39%, mp 168–171 °C
propyl acetate. Finally, compound 19c was converted to
compound 19, 3-[(4-chloro-2-phenylsulfanyl-5-sulfamoyl-
benzoyl)amino]propyl acetate, via nucleophilic substitu-
tion of the choride with thiophenol (Scheme 1).
1
(toluene:2-PrOH (8:1)). H NMR (400 MHz) δ ppm: 1.86
2,4‑Dichloro‑N‑(3‑hydroxypropyl)‑5‑sulfamoyl‑benzamide
(19b)
(2H, quint, J = 6.4 Hz, CH2), 2.03 (3H, s, CH3), 3.32–3.36
(2H, m, NHCH2), 4.10 (2H, t, J = 6.4 Hz, CH2O), 6.82 (1H,
s, C3-H), 7.55–7.56 (5H, m, Ph-H), 7.64 (2H, s, SO2NH2),
8.01 (1H, s, C6-H), 8.81 (1H, t, J = 5.2 Hz, NH). 13C NMR
(100 MHz) δ ppm: 21.2, 28.6, 36.7, 62.3, 128.6, 129.6,
The mixture of 2,4-dichloro-5-sulfamoylbenzoic acid (19a)
(2.70 g, 10.0 mmol), SOCl2 (3.57 g, 30.0 mmol), and 1 drop
1 3

Eur Biophys J
130.4, 130.8, 131.4, 132.3, 133.6, 135.2, 138.0, 144.6, 166.0,
170.9. HRMS calculated for C18H19ClN2O5S2[(M + H)+]:
443.0497, found: 443.0495.
5-(2-Aminothiazol-4-yl)-2-chloro-benzenesulfona-
mide (20) and 3-(4-chloro-3-methyl-phenyl)-5,6-dihydro-
2H-imidazo[2,1-b]thiazol-3-ol (27) were obtained using the
same reaction conditions from 5-(2-bromoacetyl)-2-chloro-
benzenesulfonamide (20a) and thioureas (Scheme 2).
S-Alkylation of phenylmethanethiol with
5-(bromoacetyl)-2-chlorobenzenesulfonamide (32a) to give
4-(2-benzylsulfanylacetyl)benzenesulfonamide (32) was car-
ried out in the presence of sodium acetate in tetrahydrofuran
at room temperature (Scheme 3).
Synthesis of (20a) was accomplished from commercially
available 1-(4-chloro-3-nitrophenyl)ethanone as reported
previously (Oelschlager 1961), Fujikura et al. 1982). 4-(bro-
moacetyl)benzenesulfonamide (32a) was synthesized from
commercially available 1-(4-aminophenyl)ethanone as
described in (Fujikura et al. 1982).
General procedure for the synthesis of 20, 27, and 32
A mixture of corresponding bromoacetophenone 20a, 32a
(0.360 mmol), appropriate thioureas or phenylmethanethiol
(0.704 mmol), and NaOAc (63.0 mg, 0.768 mmol) in THF
(3 ml) was stirred at room temperature for 24 h. The reaction
mixture was poured into water. The precipitate was fltered
of, washed with water, and then with Et2O.
5‑(2‑Aminothiazol‑4‑yl)‑2‑chloro‑benzenesulfonamide (20)
Yield 82%, mp 235-237 °C. IR ν cm−1: 3444, 3353, 3257
1
(NH2). H NMR (300 MHz) δ, ppm: 7.20 (3H, s, NH2,
C5′-H), 7.62 (2H, s, SO2NH2), 7.63 (1H, d, J = 8.1 Hz,
C3-H), 7.99 (1H, dd, J = 8.4 Hz, J = 2.1 Hz, C4-H), 8.46
(1H, d, J = 2.1 Hz, C6-H). 13C NMR (75 MHz) δ, ppm:
104.3, 126.8, 129.2, 130.2, 132.3, 134.7, 141.9, 148.2,
169.3. HRMS calculated for (C9H8ClN3O2S2): [(M + H)+]:
289.9819, found: 289.9819.
3‑(4‑Chloro‑3‑methyl‑phenyl)‑5,6‑dihy‑
dro‑2H‑imidazo[2,1‑b]thiazol‑3‑ol (27)
Yield 80%, mp 158–160 °C. 1H NMR (300 MHz) δ, ppm:
2.96–3.19 (2H, m, NCH2), 3.69 (1H, d, J = 11.7 Hz,
SCH2), 3.79 (1H, d, J = 11.7 Hz, SCH2), 3.91–4.12 (2H,
m, NCH2), 7.10 (1H, br s, OH), 7.69–7.72 (3H, m, C3-H,
NH2), 7.77 (1H, dd, J = 1.8 Hz, J = 8.1 Hz, C4-H), 8.20
(1H, d, J = 1.8 Hz, C6-H). 13C NMR (75 MHz) δ, ppm: 42.9,
49.8, 61.4, 87.8, 127.5, 131.0, 131.8, 132.5, 141.7, 142.0,
170.4. HRMS calculated for (C11H12ClN3O3S2): [(M + H)+]:
334.0081, found: 334.0084.
1 3

Eur Biophys J
-20
-30
-40
-50
-60
-70
0
-10
-20
-30
-40
-50
-60
EZA
(a)
EZA
(b)
Pi
20.5
20.0
19.5
Pi
∆
Tris
Tris
0
20
40
Time (min)
0.0
0.5
1.0
1.5
2.0
5
6
7
8
9
pH
Molar ratio
0
-10
-20
-30
-40
-50
ITC (Pi)
ITC (Tris)
FTSA
(d)
EZA
TFS
(c)
-30
-40
-50
observed
Pi
Tris
∆
intrinsic
5
6
7
8
9
5
6
7
8
9
10
pH
pH
Fig. 4 The dependence of sulfonamide inhibitor binding to CA IV
as a function of pH. a The observed enthalpies of EZA binding to
CA IV in sodium phosphate (flled square) and TRIS chloride (open
square) bufer, both at pH 7.0, 25 °C, determined by the ITC. Panels
b, c show the observed enthalpies of EZA, b and TFS c binding to
CA IV in sodium phosphate (flled square) and TRIS chloride (open
square) bufers (at 25 °C, as a function of pH). Data points represent
the observed enthalpies obtained by ITC, while the curves were ft
according to Eq. (4). d The observed and intrinsic Gibbs energies of
EZA binding to CA IV as a function of pH. Datapoints show ∆bGobs
obtained by ITC in phosphate (flled square) and TRIS (open square)
bufers and by FTSA in the universal bufer (flled triangle). Solid
bent line is the ft according to Eq. (1) recalculated to Gibbs energies.
Dashed lines show the contributions of the fractions of deprotonated
EZA and protonated CA IV. The straight horizontal line shows the
intrinsic Gibbs free energy of EZA binding that is independent of pH
70
70
(a)
(b)
65
60
55
50
60
PIPES
TRIS
Gly
HEPES
NaPi
NaAcetate
NaCitrate
NaiCO3
MES
50
5
6
7
8
9
10
4
5
6
7
8
9
10
pH
pH
70
65
60
64
62
60
Na2SO4
(c)
(d)
NaCl
(NH4 )2SO4
DMSO
Glycerol
0
250
500
750
1000
0
5
10
cw (%)
L (mM)
t
Fig. 5 Thermal stability of CA IV in various bufers, pHs, and salts.
a CA IV melting temperature (Tm) dependence on pH of univer-
sal bufer. b CA IV melting temperature (Tm) dependence on pH in
various bufers (100 mM): sodium acetate (NaAcetate, open square),
sodium citrate (NaCitrate, asterisk), sodium phosphate (NaPi, open
triangle), sodium carbonate (NaCO3, open down triangle), glycine
(Gly, flled circle), TRIS (flled triangle), MES (multiplication),
HEPES (flled triangle), and PIPES (open circle). c The melting
temperature of CA IV as a function of various salt concentrations:
(NH4)2SO4 (flled triangle), Na2SO4 (flled circle), and NaCl (flled
square). d CA IV stability at various added concentrations of DMSO
(flled square) and glycerol (flled triangle)
1 3

Eur Biophys J
Table 5 The thermodynamic
parameters of protonation of
the deprotonated inhibitor
Inhibitor/protein
PKa
ΔGprot (kJ/mol)
ΔHprot (kJ/mol)
TΔSprot (kJ/mol)
5a
8.3
8.0
9.0
8.9
8.9
8.0
6.3
7.3
7.1
10.1
8.7f
6.8
−47.4
−45.7
−51.4
−50.8
−50.8
−45.7
−36.0
−41.7
−40.5
−57.7
−49.7
−38.8
−28.9
−26.8
−30.1
−26.8
−26.8
−29.5
−22.4
−23.0
−25.1
−23.0
−29.3
−33.0
18.5
18.9
21.3
24.0
24.0
16.2
13.6
18.7
15.4
34.6
20.4
5.8
sulfonamide group and the CA–
Zn(II)-bound hydroxide anion,
determined by ITC at 25 °C
18b
19
20
26c
EZAd
TFSd
AZMe
MZMb
PAMBSb
TPM
CA IV–Zn–H2O
b
c
d
aData taken from (Zubrienė et al. 2015), (Zubrienė et al. 2017), (Kišonaitė et al. 2014), (Jogaitė et al.
2013), e(Matulis and Todd 2004), f(Maryanof et al. 1998)
10000
(c)
(a)
(b)
1011
109
107
105
∆
Kb-obs Kb-intr
0
-20
-40
G
i
nt
1000
100
10
= -41.6 kJ/mol
r
6.3 7.3 8.0 9.0 10.2
∆
G
intr
= -65.4 kJ/mo
l
1
6
7
8
pKa
9
10
TFS AZM
18
19
35
-40
-20
0
H
(kJ/mol)
∆
intr
Fig. 6 Comparison between the observed binding constants and the
intrinsic afnities. Only the intrinsic values should be used for the
drawing of compound structure–afnity correlations. a Dependence
of the Kb-int/Kb-obs ratio on the compound pKa. The ratio of binding
constants increases with the pKa. b The intrinsic (black column) and
observed (white column) Kb values of compounds with diferent pKa
values 6.3, 7.3, 8.0, 9.0, and 10.2. c The intrinsic enthalpy–entropy
compensation plot for compound binding to CA IV. Diagonal lines
represent the Gibbs energies (solid upper line −41.6 kJ/mol, equiva-
lent to 100 nM Kd, solid bottom line −65.4 kJ/mol equivalent to
0.01 nM Kd)
Table 6 The intrinsic
Inhibitor
ΔbHintr (kJ/mol)
Kb_intr (M−1
)
ΔbGintr (kJ/mol)
TΔbSintr (kJ/mol)
thermodynamic parameters of
compound binding to CA IV,
determined by ITC at 25 °C
5
−12.9
−12.5
−19.1
−17.3
−30.9
−46.0
−21.5
−50.6
−30.6
ND
2.5 × 107
2.8 × 108
6.7 × 109
2.1 × 107
1.7 × 108
2.5 × 109
2.8 × 107
1.7 × 108
4.4 × 107
1.1 × 109
3.1 × 109
−48.8
−52.7
−65.3
−51.6
−56.5
−53.6
−42.5
−46.9
−43.6
−51.5
−54.2
35.9
40.1
46.2
34.3
25.6
7.6
21.0
−3.7
13.0
ND
18
19
20
26
EZA
TFS
AZM
MZM
PAMBS
TPM
−29.9
24.3
1 3

Eur Biophys J
Cl
O
Cl
O
N
H
OH
OH
1. SOCl2, ∆
EtOAc, ∆
2. HR, THF, 0°C
Cl
H+
Cl
NH2
NH2
SO2
SO2
19a
19b
Cl
O
O
S
O
O
N
H
O
CH3
N
H
H
3
O
C
PhSH, Et3N
MeOH
Cl
Cl
NH2
SO2
NH2
SO2
19c
19
Scheme 1 Synthesis of compound 19
S
N
NH2
S
H2 N
NH2
Cl
NaOAc, THF
O
r.t., 24h
NH2
SO2
Br
20
S
S
Cl
N
NH
HN
N
NH2
SO2
OH
20a
Cl
NaOAc, THF
r.t., 24h
NH2
SO2
27
Scheme 2 Synthesis of compounds 20 and 27
1 3

Eur Biophys J
O
O
S
Br
BnSH
NaOAc, THF
H
2NO2S
H
2NO2S
r.t., 24h
32a
32
Scheme 3 Synthesis of the compound 32
4‑(2‑Benzylsulfanylacetyl)benzenesulfonamide (32)
sorbent. Elution bufer contained 0.1 M sodium acetate,
0.5 M NaClO4, pH 5.6. Eluted protein was dialyzed into
a bufer containing 20 mM HEPES, 20 mM NaCl, pH 8.
After dialysis, CA IV (19–284 aa) was further purifed using
an anion exchange column with Q SepharoseTM Fast Flow
(GE Healthcare, Little Chalfont, United Kingdom). Protein
was eluted with 20 mM HEPES, 1 M NaCl, pH 8. Eluted
protein was dialyzed into a storage bufer containing 20 mM
HEPES, 50 mM NaCl, pH 7.5, and stored at −80 °C.
CA IV (19–284 aa) purity was confirmed using
SDS–PAGE. Concentration was measured by the Brad-
ford method. The molecular weight of the purifed CA IV
(19–284 aa) was measured by HRMS and was 30,320.74 Da,
nearly equal to the theoretical molecular weight calculated
from the amino acid sequence minus Met without water
(30,323.4 Da). The diference in weight of 2 Da occurs due
to the possible reduction–oxidation of Cys found in CA IV
(19–284 aa).
Yield 82%, mp 155-157 °C, IR ν cm−1: 3382, (NH2), 1690
1
(CO). H NMR δ, ppm: 3.74 (2H, s, SCH2), 3.96 (2H, s,
CH2CO), 7.24–7.29 (1H, m, C4′-H), 7.33–7.40 (4H, m,
C2′,3′,5′,6′-H), 7.59 (2H, s, NH2), 7.96 (2H, d, J = 8.4 Hz,
C2,6-H), 8.15 (2H, d, J = 8.7 Hz, C3,5-H). 13C NMR δ,
ppm: 35.9, 37.3, 126.6, 127.8, 129.2 (2C), 129.8, 130.0,
138.3, 148.5, 194.6. HRMS calculated for (C15H15NO3S2):
[(M + H)+]: 322.0566, found: 322.0562.
Production of recombinant CA IV
The cDNA of CA IV was purchased from RZPD Deutsches
Ressourcenzentrum für Genomforschung (Berlin, Germany).
The gene encoding the catalytic domain of human CA IV
(19–284 aa) was expressed in Escherichia coli OrigamiTM
B (DE3) (Novagen, EMD Millipore, Billerica, MA, USA)
cells. The transformed colony was transferred to BHI (Brain
Heart Infusion) medium containing 60 μM ZnSO4 and the
Human CA IV was purifed using CA-inhibitor afn-
ity resins as described (Waheed et al. 1996). Post afnity
enzyme was dialyzed against 10 mM Tris–Sulfate pH 7.5 or
10 mM ammonium bicarbonate pH 8.0. Dialyzed enzyme
was lyophilized for storage at 4 °C. Quality of the enzyme
was assessed by SDS–PAGE and by measuring the specifc
activities. This CA IV preparation was used to obtain the
crystal structures with AZM, MZM, PAMBS, EZA, and
TPM. Both CA IV preparations gave specifc activity of
3000–4000 EU/mg.
cells were cultivated at 37 °C until reaching 0.6–0.7 OD600
.
Then 0.5 mM ZnSO4 was added and protein expression was
induced with 1 mM isopropyl β-d-1-thiogalactopyranoside
(IPTG). The cells were grown for 20 h at 16 °C. The bio-
mass was harvested by centrifugation at 4000×g for 20 min
at 4 °C.
Mammalian CA IV (19–284 aa) was produced as previ-
ously described (Dudutiene et al. 2014).
The pellet was suspended in the lysis bufer (20 mM
HEPES, 0.15 M NaCl, pH 8.0) containing 1× protease
inhibitor cocktail (Roche Applied Science, Indianapolis, IN,
USA) and 1 mM PMSF (SERVA, Heidelberg, Germany).
The cells were incubated by mixing at 4 °C for 60 min and
then sonicated for 15 min, 70% amplitude, using 1/1 min
intervals. The supernatant was separated by centrifugation
at 30,000×g for 25 min at 4 °C. Supernatant was mixed
with p-aminomethylbenzene sulfonamide-agarose (Sigma-
Life Science Aldrich) and left to incubate overnight slightly
shaken at 4 °C. The next day, CA IV (19–284 aa) was puri-
fed using a CA-afnity column containing the mentioned
Compound binding to CA IV by isothermal titration
calorimetry (ITC)
ITC measurements were performed using the iTC200 instru-
ment (Microcal, Inc., Northampton, MA, USA, now part
of Malvern Instruments Ltd., UK). Protein solution (10 µM
in the cell, volume 200 µl) was titrated with the compound
solution in the syringe (100 µM, 40 µl). A typical experi-
ment consisted of 19 injections, 2 µL each, with 2–3 min
intervals between injections. Experiments were performed at
25 °C in 50 mM TRIS chloride or sodium phosphate bufers
1 3

Eur Biophys J
containing 100 mM NaCl and 0.5-2% DMSO concentrations
both in the cell and syringe. Protein stock solutions were
dialyzed against the bufers that were used to prepare the
ligand solutions. ITC data were analyzed using MicroCal
Origin software. The frst point obtained using the 0.5 µL
injection in the integrated data graph was deleted. The bind-
ing constants and the enthalpies and entropies of binding
were estimated after ftting the data with the single binding
site model.
Each of these reactions contributes to the observed bind-
ing parameters, the binding constant, and the enthalpy of
binding.
The intrinsic binding constant Kb_intr is equal to the
observed binding constant Kb_obs divided by the available
fractions of deprotonated inhibitor and protonated Zn(II)-
bound water form of CA
Kb_obs
Kb_intr
=
.
(1)
−
(fRSO NH fCAZnH O
The enthalpy of TPM deprotonation was measured using
VP-ITC instrument. Inhibitor (0.25 mM) was deprotonated
by adding 1.3 equivalent of NaOH and titrated with 2.5 mM
HNO3. The experiment consisted of 58 injections (5 μl each)
added at 4 min intervals. Experiment was performed at
25 °C with 0.5% of DMSO, equal in the syringe and the cell.
)
2
2
The fractions of the deprotonated inhibitor and the Zn-
bound water form of CA can be calculated if both pKa values
of the sulfonamide (pKa_sulf) and the CA (pKa_ZnH2O) are
known:
10pH−pK
a_sulf
Compound binding to CA IV by the fuorescent thermal
−
fRSO NH
=
,
a_sulf
(2)
2
1 + 10pH−pK
shift assay (FTSA)
10pH−pK
a_ZnH
O
During the FTSA experiments, the protein solution was
heated from room temperature to 99 °C while following
the extrinsic fuorescence of a solvatochromic probe, such
as 1,8-anilinonaphthalene sulfonate (ANS), that strongly
changes its fuorescence upon protein unfolding (Stryer
1965; Slavik et al. 1982; Matulis and Lovrien 1998; Matu-
lis et al. 1999). The thermal melting temperatures (Tm) are
strongly dependent on the ligand binding afnity and con-
centration as previously described (Brandts and Lin 1990;
Matulis et al. 2005; Cimmperman et al. 2008). FTSA experi-
ments were performed with a Corbett Rotor-Gene 6000
(QIAGEN Rotor-Gene Q) instrument using the blue chan-
nel (excitation 365 20, detection 460 15 nm), applying
the heating rate of 1 °C/min. Usually the samples contained
20 µL of 10 µM protein, 0–400 µM ligand, 50 µM 8-ani-
lino-1-naphthalene sulfonate, 50 mM sodium phosphate at
pH 7.0, 50 mM NaCl, and 2% DMSO. The pH dependence
of the observed binding constant (Kb) was determined in
the bufer containing 50 mM sodium phosphate, 50 mM
sodium acetate, 25 mM sodium borate, and 50 mM NaCl,
pH 5.0–10.0. The FTSA data was ft and analyzed as previ-
ously described (Cimmperman et al. 2008).
2
fCAZnH O = 1 −
.
O
(3)
1 + 10pH−pK
2
a_ZnH
2
The observed enthalpy is the sum of all protonation
events and the intrinsic binding reaction itself. Linked reac-
tions that have to be subtracted are the protonation of the
zinc-bound hydroxide ion in the active center of CA, depro-
tonation of sulfonamide, and the protonation/deprotonation
of the bufer:
ΔbHintr = ΔbHobs − nsulfΔb_proton_sulf
H
− nCAΔb_proton_CAH + nbufΔb_proton_bufH,
(4)
where ∆bHobs is the observed binding enthalpy,
−
nsulf = fRSO NH is the number of protons binding to the
2
inhibitor, ∆b_proton_sulf H is the enthalpy of inhibitor proto-
nation, nCA = 1 − fCAZnH O is the number of protons bound
2
to the Zn-hydroxide, ∆b_proton_CA H is the enthalpy of CA
protonation, nbuf = nsulf + nCA, is the net sum of taken up or
released protons, and ∆b_proton_buf H is the bufer protona-
tion enthalpy. The enthalpy of TRIS protonation is equal to
−47.45 kJ/mol and the enthalpy of phosphate bufer protona-
tion is equal to −5.1 kJ/mol at 25 °C (Goldberg and Lennen
2002).
Intrinsic thermodynamics of compound binding to CA
IV
Crystallization, X‑ray difraction data collection,
and structure determination
Several linked reactions occur during sulfonamide inhibitor
binding to CA that prevent direct determination of the intrin-
sic afnity, namely, the zinc-bound hydroxide ion protona-
tion in the active center of the CA, the ligand sulfonamide
group deprotonation, a coordination bond formation between
the protein and deprotonated sulfonamide amine, and fnally,
the compensating protonation–deprotonation of the bufer.
The crystallization bufer contained 0.2 M ammonium sul-
fate, 0.1 M sodium MES (pH 6.5) and 20% PEG2000 meth-
ylmonoester, and crystals were obtained by crystallization of
a mixture of CA IV (25 mg/ml) with 0.4 mM of compound
20 for the complex of CA IV and compound 20. For the
other fve complexes of CA IV and inhibitors, 8 mg/ml CA
1 3

Eur Biophys J
IV in 50 mM Tris sulfate, pH 8.0 and 2 mM benzamidine
were mixed with 3–20 mM inhibitors (Table 1) before crys-
tallization. Crystals were grown by the hanging drop vapor
difusion technique using an Art Robbins Instruments (Sun-
nyvale, CA, USA) Phoenix liquid handling robot in a solu-
tion of 100 mM Na acetate, pH 4.6, 200 mM ammonium sul-
fate, and 16–23% PEG 4000 or PEG 3350 (Table 1). Crystals
were grown in 1–2 weeks at 277 K and were cryoprotected
in a solution containing mother liquid and 15% glycerol
prior to fash freezing. Data were collected at the EMBL P14
beamline at the PETRAIII storage ring, DESY, Hamburg
(Germany) for the complex with compound 20. Data for
other complexes were collected with a home source (Rigaku
1.2 kw MMX007 generator with VHF optics) Rigaku Raxis
IV++ detector. All data were collected at 100 K.
will release the atomic coordinates and experimental data
upon article publication.
Acknowledgements This research was funded by the Grant SEN-
04/2015 from the Research Council of Lithuania to J.M. and by
National Institutes of Health Grants HL049413 and HL073813 to
E.D.C. The authors also acknowledge the COST projects CM1406,
CM1407, CA15126, and CA15135. The authors are grateful to Gleb
Burenkov for assistance with data collection at EMBL P14 beamline
at the PETRA III storage ring, DESY Hamburg. The authors also
thank Tracey Baird for editorial assistance in the preparation of this
manuscript.
References
Alterio V, Fiore AD, D’Ambrosio K, Supuran CT, Simone GD (2012)
Multiple binding modes of inhibitors to carbonic anhydrases: how
to design specifc drugs targeting 15 diferent isoforms? Chem
Rev 112:4421–4468
Datasets for the complex with compound 20 were pro-
cessed by MOSFLM (Leslie 2006) and other datasets were
indexed, integrated, and scaled with the HKL2000 software
package (Otwinowski and Minor 1997). All structures were
solved by molecular replacement using the MOLREP pro-
gram (Vagin and Teplyakov 1997). The initial phases were
received from PDB entry 3F7B for the complex with com-
pound 20 and PDB entry 1ZNC for other complexes. For the
complex with compound 20, 3D models of inhibitors were
created using the molecule editor AVOGADRO (Hanwell
et al. 2012). The chemical and geometric descriptions of
inhibitors used in refnement were generated by LIBCHECK
CCP4 (Collaborative Computational Project, N. 4 1994).
For other complexes, the models of inhibitors were obtained
from Protein Data Bank. A model of PAMBS was obtained
from RCS (Triscarbonyl-cyclopentadienyl-4-aminomethylb-
enzene sulfonamide) and the Triscarbonyl-cyclopentadienyl
was removed. The chemical and geometric descriptions for
these inhibitors were generated automatically by REFMAC
(Murshudov et al. 1997). Refnement and electron density
generation were performed with REFMAC (Murshudov
et al. 1997). Ten and 5% of the refections were randomly
selected as test sets for cross-validation for the complex with
compound 20 and other complexes, respectively. Model
building and analysis of the structures were carried out using
COOT (Emsley and Cowtan 2004). In the fnal refnement
stage, TLS tensors modeling rigid-body anisotropic tempera-
ture factors were calculated and applied to the models except
for the complexes with compound 20, EZA, and PAMBS.
Ramachandran plots were calculated using PROCHECK
(Morris et al. 1992). Statistics for data collection and refne-
ment are summarized in Table 3. Atomic coordinates and
structure factors have been deposited in Protein Data Bank
(accession codes 5IPZ, 5JN8, 5JN9, 5JNA, 5JNC, and 5KU6
for the complexes with compound 20, AZM, EZA, TPM,
PAMBS, and MZM, respectively).
Baird JTT, Waheed A, Okuyama T, Sly WS, Fierke CA (1997) Cataly-
sis and inhibition of human carbonic anhydrase IV. Biochemistry
36:2669–2678
Balaydin HT, Senturk M, Menzek A (2012) Synthesis and carbonic
anhydrase inhibitory properties of novel cyclohexanonyl bromo-
phenol derivatives. Bioorg Med Chem Lett 22:1352–1357
Barker H, Aaltonen M, Pan P, Vähätupa M, Kaipiainen P, May U,
Prince S, Uusitalo-Järvinen H, Waheed A, Pastoreková S, Sly WS,
Parkkila S, Järvinen TA (2017) Role of carbonic anhydrases in
skin wound healing. Exp Mol Med 49:e334
Brandts JF, Lin LN (1990) Study of strong to ultratight protein inter-
actions using diferential scanning calorimetry. Biochemistry
29:6927–6940
Breton S (2001) The cellular physiology of carbonic anhydrases. JOP
2:159–164
Cammer W, Zhang H, Tansey F (1995) Efects of carbonic anhydrase
II (CAII) defciency on CNS structure and function in the myelin-
defcient CAII-defcient double mutant mouse. J Neurosci Res
40:451–457
Čapkauskaitė E, Zubrienė A, Smirnov A, Torresan J, Kišonaitė M,
Kazokaitė J, Gylytė J, Michailovienė V, Jogaitė V, Manakova
E, Gražulis S, Tumkevičius S, Matulis D (2013) Benzenesul-
fonamides with pyrimidine moiety as inhibitors of human car-
bonic anhydrases I, II, VI, VII, XII, and XIII. Bioorg Med Chem
21:6937–6947
Christie KN, Thomson C, Xue L, Lucocq JM, Hopwood D (1997)
Carbonic anhydrase isoenzymes I, II, III, and IV are present
in human esophageal epithelium. J Histochem Cytochem
45:35–40
Cimmperman P, Baranauskienė L, Jachimovičiūtė S, Jachno J, Torresan
J, Michailovienė V, Matulienė J, Sereikaitė J, Bumelis V, Matu-
lis D (2008) A quantitative model of thermal stabilization and
destabilization of proteins by ligands. Biophys J 95:3222–3231
Collaborative Computational Project, N. 4 (1994) The CCP4 suite:
programs for protein crystallography. Acta Cryst D 50:760–763
Datta R, Waheed A, Bonapace G, Shah GN, Sly WS (2009) Pathogen-
esis of retinitis pigmentosa associated with apoptosis-inducing
mutations in carbonic anhydrase IV. Proc Natl Acad Sci USA
106:3437–3442
Datta R, Shah GN, Rubbelke TS, Waheed A, Rauchman M, Goodman
AG, Katze MG, Sly WS (2010) Progressive renal injury from
transgenic expression of human carbonic anhydrase IV folding
mutants is enhanced by defciency of p58IPK. Proc Natl Acad
Sci USA 107:6448–6452
PDB IDs: 5IPZ—20, 5JN8—AZM, 5JN9—EZA,
5JNA—TPM, 5JNC—PAMBS, and 5KU6—MZM. Authors
1 3

Eur Biophys J
Krissinel E, Henrick K (2007) Inference of macromolecular assemblies
from crystalline state. J Mol Biol 372:774–797
Del Prete S, Vullo D, Fisher GM, Andrews KT, Poulsen S-A, Capasso
C, Supuran CT (2014) Discovery of a new family of carbonic
anhydrases in the malaria pathogen Plasmodium falciparum—the
η-carbonic anhydrases. Bioorg Med Chem Lett 24:4389–4396
Dudutiene V, Matuliene J, Smirnov A, Timm DD, Zubriene A,
Baranauskiene L, Morkunaite V, Smirnoviene J, Michailoviene
V, Juozapaitiene V et al (2014) Discovery and characterization of
novel selective inhibitors of carbonic anhydrase IX. J Med Chem
57:9435–9446
Lane TW, Saito MA, George GN, Pickering IJ, Prince RC, Morel FMM
(2005) Biochemistry: a cadmium enzyme from a marine diatom.
Nature 435:42
Leslie AGW (2006) The integration of macromolecular difraction
data. Acta Crystallogr D 62:48–57
Linkuviene V, Matuliene J, Juozapaitiene V, Michailoviene V, Jachno
J, Matulis D (2016) Intrinsic thermodynamics of inhibitor bind-
ing to human carbonic anhydrase IX. Biochim Biophys Acta
1860:708–718
Dudutienė V, Zubrienė A, Smirnov A, Gylytė J, Timm D, Manakova
E, Gražulis S, Matulis D (2013) 4-Substituted-2,3,5,6-tetrafuor-
obenzenesulfonamides as inhibitors of carbonic anhydrases I, II,
VII, XII, and XIII. Bioorg Med Chem 21:2093–2106
Liu R, Bu W, Xi J, Mortazavi SR, Cheung-Lau JC, Dmochowski IJ,
Loll PJ (2012) Beyond the detergent efect: a binding site for
sodium dodecyl sulfate (SDS) in mammalian apoferritin. Acta
Crystallogr D 68:497–504
Dudutienė V, Zubrienė A, Smirnov A, Timm DD, Smirnovienė J,
Kazokaitė J, Michailovienė V, Zakšauskas A, Manakova E,
Gražulis S, Matulis D (2015) Functionalization of fuorinated ben-
zenesulfonamides and their inhibitory properties toward carbonic
anhydrases. ChemMedChem 10:662–687
Maryanoff BE, Costanzo MJ, Nortey SO, Greco MN, Shank RP,
Schupsky JJ, Ortegon MP, Vaught JL (1998) Structure-activity
studies on anticonvulsant sugar sulfamates related to topiramate.
Enhanced potency with cyclic sulfate derivatives. J Med Chem
41:1315–1343
Elleuche S, Pöggeler S (2010) Carbonic anhydrases in fungi. Micro-
biology 156:23–29
Matulis D, Lovrien R (1998) 1-Anilino-8-naphthalene sulfonate anion-
protein binding depends primarily on ion pair formation. Biophys
J 74:422–429
Emsley P, Cowtan K (2004) Coot: model-building tools for molecular
graphics. Acta Crystallogr D 60:2126–2132
Ferrari S, Di Iorio E, Barbaro V, Ponzin D, Sorrentino FS, Parmeggiani
F (2011) Retinitis pigmentosa: genes and disease mechanisms.
Curr Genomics 12:238–249
Matulis D, Todd MJ (2004) Thermodynamics-structure correlations
of sulfonamide inhibitor binding to carbonic anhydrase. In: Lad-
bury JE, Doyle ML (eds) Biocalorimetry 2. Wiley, Chichester,
pp 107–132
Ferry JG (2010) The gamma class of carbonic anhydrases. Biochim
Biophys Acta 1804:374–381
Matulis D, Baumann CG, Bloomfeld VA, Lovrien RE (1999) 1-ani-
lino-8-naphthalene sulfonate as a protein conformational tighten-
ing agent. Biopolymers 49:451–458
Fujikura T, Miigata K, Hashimoto S, Imai K, Takenaka T (1982) Stud-
ies on benzenesulfonamide derivatives with alpha- and beta-adr-
energic antagonistic and antihypertensive activities. Chem Pharm
Bull 30:4092–4101
Matulis D, Kranz JK, Salemme FR, Todd MJ (2005) Thermody-
namic stability of carbonic anhydrase: measurements of bind-
ing afnity and stoichiometry using ThermoFluor. Biochemistry
44:5258–5266
Goldberg RMRN, Lennen Kishore N (2002) Thermodynamic quanti-
ties for the ionization reactions of bufers. J Phys Chem Ref Data
31:231–370
McMurtrie HL, Cleary HJ, Alvarez BV, Loiselle FB, Sterling D, Mor-
gan PE, Johnson DE, Casey JR (2004) The bicarbonate transport
metabolon. J Enzyme Inhib Med Chem 19:231–236
Morris AL, MacArthur MW, Hutchinson EG, Thornton JM (1992)
Stereochemical quality of protein structure coordinates. Proteins
12:345–364
Hanwell MD, Curtis DE, Lonie DC, Vandermeersch T, Zurek E, Hutch-
ison GR (2012) Avogadro: an advanced semantic chemical editor,
visualization, and analysis platform. J Cheminform 4:17
Innocenti A, Villar R, Martinez-Merino V, Gil MJ, Scozzafava A, Vullo
D, Supuran CT (2005) Carbonic anhydrase inhibitors: inhibition
of cytosolic/tumor-associated carbonic anhydrase isozymes I, II,
and IX with benzo[b]thiophene 1,1-dioxide sulfonamides. Bioorg
Med Chem Lett 15:4872–4876
Murshudov GN, Vagin AA, Dodson EJ (1997) Refnement of macro-
molecular structures by the maximum-likelihood method. Acta
Crystallogr D 53:240–255
Jogaitė V, Zubrienė A, Michailovienė V, Gylytė J, Morkūnaitė V, Matu-
lis D (2013) Characterization of human carbonic anhydrase XII
stability and inhibitor binding. Bioorg Med Chem 21:1431–1436
Kikutani S, Nakajima K, Nagasato C, Tsuji Y, Miyatake A, Matsuda Y
(2016) Thylakoid luminal θ-carbonic anhydrase critical for growth
and photosynthesis in the marine diatom Phaeodactylum tricor-
nutum. Proc Natl Acad Sci USA 113:9828–9833
Nishimori I, Vullo D, Innocenti A, Scozzafava A, Mastrolorenzo A,
Supuran CT (2005) Carbonic anhydrase inhibitors: inhibition of
the transmembrane isozyme XIV with sulfonamides. Bioorg Med
Chem Lett 15:3828–3833
Oelschlager H (1961) 3-Alkyl-6-Halogen-Aniline Aus p-Halogenierten
Fettaromatischen Ketonen. Justus Liebigs Ann Chem 641:81–94
Otwinowski Z, Minor W (1997) Processing of x-ray difraction data
collected in oscillation mode. Methods Enzymol 276:307–326
Pastorekova S, Parkkila S, Pastorek J, Supuran CT (2004) Carbonic
anhydrases: current state of the art, therapeutic applications and
future prospects. J Enzyme Inhib Med Chem 19:199–229
Purkerson JM, Schwartz GJ (2007) The role of carbonic anhydrases in
renal physiology. Kidney Int 71:103–115
Kišonaitė M, Zubrienė A, Čapkauskaitė E, Alexey S, Smirnovienė
Joana V, Kairys V, Michailovienė E, Manakova S Gražulis,
Matulis D (2014) Intrinsic thermodynamics and structure cor-
relation of benzenesulfonamides with a pyrimidine moiety bind-
ing to carbonic anhydrases I, II, VII, XII, and XIII. PLoS ONE
9:e114106
Kivelä A-J, Kivelä J, Saarnio J, Parkkila S (2005) Carbonic anhydrases
in normal gastrointestinal tract and gastrointestinal tumours.
World J Gastroenterol 11:155–163
Slavik J, Horak J, Rihova L, Kotyk A (1982) Anilinonaphthalene sul-
fonate fuorescence and amino acid transport in yeast. J Membr
Biol 64:175–179
Köhn L, Burstedt MSI, Jonsson F, Kadzhaev K, Haamer E, Sandgren
O, Golovleva I (2008) Carrier of R14 W in carbonic anhydrase
IV presents Bothnia dystrophy phenotype caused by two allelic
mutations in RLBP1. Invest Ophthalmol Vis Sci 49:3172–3177
Krishnamurthy VM, Kaufman GK, Urbach AR, Gitlin I, Gudiksen
KL, Weibel DB, Whitesides GM (2008) Carbonic anhydrase as a
model for biophysical and physical-organic studies of proteins and
protein-ligand binding. Chem Rev 108:946–1051
Stams T, Christianson DW (2000) X-ray crystallographic studies of
mammalian carbonic anhydrase isozymes. Birkhäuser, Basel, pp
159–174
Stams T, Nair SK, Okuyama T, Waheed A, Sly WS, Christianson DW
(1996) Crystal structure of the secretory form of membrane-asso-
ciated human carbonic anhydrase IV at 2.8-Å resolution. Proc Natl
Acad Sci USA 93:13589–13594
1 3

Eur Biophys J
Sterling D, Alvarez BV, Casey JR (2002) The extracellular component
of a transport metabolon. Extracellular loop 4 of the human AE1
Cl-/HCO3- exchanger binds carbonic anhydrase IV. J Biol Chem
277:25239–25246
similarity to the monomeric restriction enzyme MspI. Nucleic
Acids Res 33:1892–1901
Zhang J, Tsoi H, Li X, Wang H, Gao J, Wang K, Go MY, Ng SC, Chan
FK, Sung JJ, Yu J (2016) Carbonic anhydrase IV inhibits colon
cancer development by inhibiting the Wnt signalling pathway
through targeting the WTAP-WT1-TBL1 axis. Gut 65:1482–1493
Zolfaghari Emameh R, Barker HR, Syrjänen L, Urbański L, Supuran
CT, Parkkila S (2016a) Identification and inhibition of car-
bonic anhydrases from nematodes. J Enzyme Inhib Med Chem
31:176–184
Stryer L (1965) The interaction of a naphthalene dye with apomyoglo-
bin and apohemoglobin. A fuorescent probe of non-polar binding
sites. J Mol Biol 13:482–495
Supuran CT (2008) Carbonic anhydrases—an overview. Curr Pharm
Des 14:603–614
Supuran CT (2004) Carbonic anhydrases: catalytic and inhibition
mechanisms, distribution and physiological roles. In: Supuran CT,
Scozzafawa A, Conway J (eds) Carbonic anhydrase–its inhibitors
and activators. CRC Press, Boca Raton, pp 1–24
Zolfaghari Emameh R, Barker HR, Tolvanen MEE, Parkkila S,
Hytönen VP (2016b) Horizontal transfer of β-carbonic anhydrase
genes from prokaryotes to protozoans, insects, and nematodes.
Parasit Vectors 9:152
Svichar N, Waheed A, Sly WS, Hennings JC, Hübner CA, Chesler M
(2009) Carbonic anhydrases CA4 and CA14 both enhance AE3-
mediated Cl–HCO3- exchange in hippocampal neurons. J Neuro-
sci 29:3252–3258
Zubrienė A, Čapkauskaitė E, Gylytė J, Kišonaitė M, Tumkevičius S,
Matulis D (2014) Benzenesulfonamides with benzimidazole moie-
ties as inhibitors of carbonic anhydrases I, II, VII, XII and XIII. J
Enzyme Inhib Med Chem 29:124–131
Tamai S, Waheed A, Cody LB, Sly WS (1996) Gly-63–>Gln sub-
stitution adjacent to His-64 in rodent carbonic anhydrase IVs
largely explains their reduced activity. Proc Natl Acad Sci USA
93:13647–13652
Zubrienė A, Smirnovienė J, Smirnov A, Morkūnaitė V, Michailovienė
V, Jachno J, Juozapaitienė V, Norvaišas P, Manakova E, Gražulis
S, Matulis D (2015) Intrinsic thermodynamics of 4-substituted-
2,3,5,6-tetrafluorobenzenesulfonamide binding to carbonic
anhydrases by isothermal titration calorimetry. Biophys Chem
205:51–65
Vagin A, Teplyakov A (1997) MOLREP: an automated program for
molecular replacement. J Appl Cryst 30:1022–1025
Vernier W, Chong W, Rewolinski D, Greasley S, Pauly T, Shaw
M, Dinh D, Ferre RA, Nukui S, Ornelas M, Reyner E (2010)
Thioether benzenesulfonamide inhibitors of carbonic anhydrases
II and IV: structure-based drug design, synthesis, and biological
evaluation. Bioorg Med Chem 18:3307–3319
Zubrienė A, Smirnov A, Dudutienė V, Timm DD, Matulienė J,
Michailovienė V, Zakšauskas A, Manakova E, Gražulis S, Matu-
lis D (2017) Intrinsic thermodynamics and structures of 2,4- and
3,4-substituted fuorinated benzenesulfonamides binding to car-
bonic anhydrases. ChemMedChem 12:161–176
Waheed A, Okuyama T, Heyduk T, Sly WS (1996) Carbonic anhydrase
IV: purifcation of a secretory form of the recombinant human
enzyme and identifcation of the positions and importance of its
disulfde bonds. Arch Biochem Biophys 333:432–438
Yang Z, Horton JR, Maunus R, Wilson GG, Roberts RJ, Cheng X
(2005) Structure of HinP1I endonuclease reveals a striking
1 3