Protein surface-assisted enhancement in the binding affinity of an inhibitor for recombinant human carbonic anhydrase-II

Article abstract of DOI:10.1021/ja047557p

We elaborate on a novel strategy for enhancing the binding affinity of an active-site directed inhibitor by attaching a tether group, designed to interact with the surface-exposed histidine residue(s) of enzymes. In this approach, we have utilized the rec

Full text of DOI:10.1021/ja047557p

Published on Web 08/07/2004
Protein Surface-Assisted Enhancement in the Binding Affinity
of an Inhibitor for Recombinant Human Carbonic Anhydrase-II
Abir L. Banerjee, Michael Swanson, Bidhan C. Roy, Xiao Jia, Manas K. Haldar,
Sanku Mallik,* and D. K. Srivastava*
Contribution from the Department of Chemistry and Molecular Biology,
North Dakota State UniVersity, Fargo, North Dakota 58105
Received April 27, 2004; E-mail: dk.srivastava@ndsu.nodak.edu; sanku.mallik@ndsu.nodak.edu
Abstract: We elaborate on a novel strategy for enhancing the binding affinity of an active-site directed
inhibitor by attaching a tether group, designed to interact with the surface-exposed histidine residue(s) of
enzymes. In this approach, we have utilized the recombinant form of human carbonic anhydrase-II (hCA-
II) as the enzyme source and benzenesulfonamide and its derivatives as inhibitors. The steady-state kinetic
and the ligand binding data revealed that the attachment of iminodiacetate (IDA)-Cu2+ to benzenesulfonamide
(via a triethylene glycol spacer) enhanced its binding affinity for hCA-II by about 40-fold. No energetic
contribution of either IDA or triethylene glycol spacer was found (at least in the ground state of the enzyme-
inhibitor complex) when Cu2+ was stripped off from the tether group-conjugated sulfonamide derivative.
Arguments are presented that the overall strategy of enhancing the binding affinities of known inhibitors by
attaching the IDA-Cu2+ groups to interact with the surface-exposed histidine residues will find a general
application in designing the isozyme-specific inhibitors as potential drugs.
Introduction
Following the completion of the human genome project, there
has been a growing awareness of the potential drug targets for
treating a variety of human diseases.1 The overall strategy in
the area of drug discovery has been greatly facilitated by the
availability of modern combinatorial and high-throughput
facilities.2 The lead drug compounds, identified by these
approaches,2d-e are often subject to optimizations by the aids
of the structural data of the target macromolecules (mostly
enzymes and proteins). However, there is an intrinsic limitation
in the structure-based drug design, primarily due to inherent
flexibility in the protein structure, as well as the limited
geometry of the ligand binding pockets (active sites in the case
of enzymes) to accommodate extensive variations in the inhibitor
structures.4 Besides, on the basis of the structural coordinates
of the selected enzyme-ligand complexes, it is difficult to
predict, a priori, whether the desired changes in the inhibitor
Figure 1. Surface-assisted enhancement in the binding affinity of an
inhibitor. The inhibitor is shown to bind at both the active site pocket as
well as a surface-exposed residue of an enzyme.
structure would be better accommodated within the active site
pockets.3 However, such limitations could be overcome, at least
in principle, by designing enzyme inhibitors, which would not
only bind to their cognate enzyme’s active site pockets, but also
bind (by looping around) to the surface-exposed amino acid
residues (Figure 1).
(1) (a) Lindsay M, A. Nat. ReV. Drug DiscoVery 2003, 2, 831-838. (b)
Hemmila, I. A.; Hurskainen, P. Drug DiscoVery Today 2002, 7, 150-156.
(C) Stumm, G.; Russ, A.; Nehls, M. Am. J. PharmacoGenomics 2002, 2,
263-271. (d) Guzey, C.; Spigset, O. Drug Saf. 2002, 25, 553-560.
(2) (a) Dolle, R. E. J. Comb. Chem. 2003 5, 693-753. (b) Lee, A.;
Breitenbucher, J. G. Curr. Opin. Drug DiscoVery DeV. 2003 6, 494-508.
(c) Jurinke, C.; van den Boom, D.; Cantor, C. R.; Koster, H. AdV. Biochem.
Eng. Biotechnol. 2002, 77, 57-74. (d) Geysen, H. M.; Schoenen, F.;
Wagner, D.; Wagner, R. Nat. ReV. Drug DiscoVery 2003, 2, 222-230. (e)
Jager, S.; Brand, L.; Eggeling, C. Curr. Pharm. Biotechnol. 2003, 4, 463-
476. (f) Bleicher, K. H.; Bohm, H. J.; Muller, K.; Alanine, A. I. Nat. ReV.
Drug DiscoVery 2003, 2, 369-378.
(3) (a) Wong, C. F.; McCammon, A. J. AdV. Protein Chem. 2003, 66, 87-
121. (b) Davis, A. M.; Teague, S. J.; Kleywegt, G. J. Angew. Chem., Int.
Ed. 2003, 42, 2718-2736. (c) Neamati, N.; Barchi, J. J., Jr. Curr. Top.
Med. Chem. 2002, 2, 211-227.
(4) (a) Carlson, H. A. Curr. Opin. Chem. Biol. 2002, 6, 447-452. (b) Wong,
C. F.; McCammon, J. A. Annu. ReV. Pharmacol. Toxicol. 2003, 43, 31-
45. (c) Noble, M. E.; Endicott, J. A.; Johnson, L. N. Science 2004, 303,
1800-1805.
In pursuit of identifying targetable surface-exposed amino
acid residues in the vicinity of the enzyme’s active site pockets,
it was realized that various transition-metal ions (Cu2+, Zn2+
,
Ni2+) bind to the imidazole group of the histidine residues.5
The ability of iminodiacetic acid-conjugated Cu2+ (IDA-Cu2+
)
to preferentially interact with the histidine residues of proteins
has been utilized to perform the two-dimensional crystallization
of proteins at lipid bilayers,6 purification of peptides and
proteins,7 probing the accessibility of histidine residues,8 map-
(5) (a) Bal, W.; Christodolou, J.; Sadler, P. J.; Tucker, A. J. Inorg. Biochem.
1998, 70, 33-39. (b) Sadler, P. J.; Viles. J. H. Inorg. Biochem. 1996, 35,
4490-4496.
9
10.1021/ja047557p CCC: $27.50 © 2004 American Chemical Society
J. AM. CHEM. SOC. 2004, 126, 10875-10883
10875

A R T I C L E S
Banerjee et al.
ping the distribution of the histidine residues on the protein
surface,9 and designing inhibitors.10 Upon a cursory examination
of the X-ray crystallographic structures of the enzymes, which
are potential targets of drug designing, we found several
enzymes containing the surface-exposed histidine residues within
10-15 Å of the active sites.11 Of these, carbonic anhydrase was
selected as a model system for testing our strategy because of
the following reasons: (1) the enzyme has been the target of
drug discovery over the last 40 years, for the treatment of a
variety of diseases, such as glaucoma, eplipetic seizure, high
altititude sickness, and recently for cancer,10a,12,14 (2) there are
14 different types of isozymes of human carbonic anhydrase,
and most of these isozymes have different distribution of the
surface-exposed histidine residues,11e,13 (3) several carbonic
anhydrase isozymes have been cloned, expressed, and purified
in bulk quantities, and their structural-functional properties have
been investigated in considerable detail,15 and (4) a variety of
sulfonamide and hydroxamate derivatives as well as other
inhibitors of carbonic anhydrases have been recognized, and
their binding affinities have been determined by the kinetic
method.10,14 A preliminary account of our strategy involving
the commercially available preparation of bovine carbonic
anhydrase (which contained a mixture of carbonic anhydrase
isozymes) and a few IDA-Cu2+ ligand sulfonamide inhibitors
has been recently presented.10b To derive quantitative conclu-
sions on the influence of attaching a tether group to benzene-
sulfonamide on the enzyme-inhibitor interaction, we designed
compound-1 (Figure 2) in the light of the structural coordinates
of hCA-II-4-fluorobenzenesulfonamide (FBS) complex (see
results) and performed the detailed kinetic and thermodynamic
studies.
Figure 2. Structure of compound-1. The active site-directed group,
benzenesulfonamide, and the histidine binding residue, IDA-Cu2+, are shown
to be connected via a triethylene glycol spacer.
(6) (a) Frey, W.; Schief, W. R., Jr.; Pack, D. W.; Chen, C.-T.; Chikotti, A.;
Stayton, P.; Vogel, V.; Arnold, F. H. Proc. Natl. Acad. Sci. U.S.A. 1996,
93, 4937-4941. (b) Pack, D. W.; Chen, G.; Maloney, K. M.; Chen, C. T.;
Arnold, F. H. J. Am. Chem. Soc. 1997, 119, 2479-2487.
Figure 3. Surface topology of human carbonic anhydrase-II with bound
FBS. The surface-exposed histidine residues, the active site Zn2+ (sur-
rounded by the histidine residues), and FBS are shown. Generated by the
software GRASP.17
(7) (a) Nakagawa, Y.; Yip, T. T.; Belew, M.; Porath, J. Anal. Biochem. 1988,
168, 75-81. (b) Lanmango, N. S.; Zhu, X.; Lindberg, I. Arch. Biochem.
Biophys. 1996, 330, 238-250. (c) Mrabet, N. T. Biochemistry 1992, 31,
2690-2702.
As will be shown in the subsequent sections, the binding
affinity of benzenesulfonamide is enhanced by about 40-fold
upon attachment of IDA-Cu2+ via the triethylene spacer
(compound-1). This quantitative conclusion has been derived
both by the steady-state kinetics of the enzyme inhibition data
and the direct measurement of enzyme-inhibitor binding
affinities. Our overall strategy toward the structure-based design
of isozyme-specific inhibitors and the role of the tether group
in stabilizing the ground states of the enzyme-inhibitor
complexes are discussed in the following sections.
(8) Berna, P. P.; Mrabet, N. T.; Van Beeumen, J.; Devreese, B.; Porath, J.;
Vijayalakshmi, M. A. Biochemistry 1997, 36, 6896-905.
(9) Fazal, M. A.; Roy, B. C.; Sun, S.; Mallik, S.; Rodgers, K. R. J. Am. Chem.
Soc. 2001, 123, 6283-6290.
(10) (a) Supuran, C. T.; Scozzafava, A.; Casini, A. Med. Res. ReV. 2003, 23,
146-189. (b) Roy, B. C.; Hegge, R.; Rosendahl, T.; Jia, X.; Lareau, R.;
Mallik, S.; Srivastava, D. K. J. Chem. Soc., Chem. Commun. 2003, 2328-
2329.
(11) (a) Oka, M.; Matsumoto, Y.; Sugiyama, S.; Tsuruta, N.; Matsuhima, M. J.
Med. Chem. 2000, 43, 2479-2483. (b) Phillips, R. S.; Demidkina, T. V.;
Faleev, N. G. Biochim. Biophys. Acta 2003, 1647, 167-172. (c) Pang, S.
S.; Guddat, L. W.; Duggleby, R. G. J. Biol. Chem. 2003, 278, 7639-7644.
(d) Qiu, W.; Campbell, R. L.; Gangloff, A.; Dupuis, P.; Boivin, R. P.;
Tremblay, M. R.; Poirier, D.; Lin, S. X. FASEB J. 2002, 16, 1829-1831.
(e) Stams, T.; Christianson, D. W. In The Carbonic Anhydrases: New
Horizons; Chegwidden, W. R., Carter, N. D., Edwards, Y. H., Eds.;
Birkhauser Verlag: Basel, 2000; pp 159-174. (f) Vihinen, P. P.; Pyrhonen,
S. O.; Kahari, V. M. Ann Med. 2003, 35, 66-78. (g) Meltzer, H. Y. J.
Clin. Psychiatry 1991, 52, 58-62.
(12) (a) Supuran, C. T. Expert Opin. InVest. Drugs 2003, 12, 283-287. (b)
Rafajova, M.; Zatovicova, M.; Kettmann, R.; Pastorek, J.; Pastorekova, S.
Int. J. Oncol. 2004, 24, 995-1004.
(13) Tureci, O.; Sahin, U.; Vollmar, E.; Siemer, S.; Gottert, E.; Seitz, G.;
Parkkila, A.-K.; Shah, G. N.; Grubb, J. H.; Pfreundschuh, M.; Sly, W. S.
Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 7608-7613.
(14) (a) Scozzafava, A.; Owa, T.; Mastrolorenzo, A.; Supuran, C. T. Curr. Med.
Chem. 2003, 10, 925-953. (b) Herker, U.; Pfeiffer, N. Curr. Opin.
Opthalmol. 2001, 12, 88-93. (c) Potter, C. P. S.; Harris, A. L. Br. J. Cancer
2003, 89, 2-7. (d) Mansoor, U. F.; Zhang, X.-R.; Blackburn, G. M In The
Carbonic Anhydrases: New Horizons; Chegwidden, W. R., Carter, N. D.,
Edwards, Y. H., Eds.; Birkhauser Verlag: Basel, 2000; pp 437-459.
(15) (a) Maren, T. H. Physiol. ReV. 1967, 47, 595-781. (b) Tashian, R. E. AdV.
Genet. 1992, 30, 321-356. (c) Sly, W. S.; Hu, P. Y. Annu. ReV. Biochem.
1995, 64, 375-401. (d) Tripp, B. C.; Smith, K.; Ferry, J. G. J. Biol. Chem.
2001, 276, 48615-48618. (e) Schwartz, G. J. J. Nephrol. 2002, 15, S61-
S64.
Results
The X-ray crystallographic structure of recombinant human
carbonic anhydrase-II (hCA-II) in the presence of FBS (pdb
file: 1IF4.pdb) has been known to atomic resolution.16 The
surface topology of the enzyme, its active site binding pocket
(containing the histidine-fenced Zn2+ and FBS), and the surface-
exposed histidine residues are shown in Figure 3.
Note that the N-terminal segment of the enzyme contains six
surface-exposed histidine residues, namely His-3, His-4, His-
10, His-15, His-17, and His-64. Of these, His-64 has been
demonstrated to mediate the proton transfer between exterior
solvent molecule and the zinc-bound water at the active site of
(16) Kim, C.-Y.; Christianson, D. W. PDB file name: 1IF4.pdb.
9
10876 J. AM. CHEM. SOC. VOL. 126, NO. 35, 2004

Carbonic Anhydrase Inhibitor
A R T I C L E S
not parametrized in our modeling software, no further attempt
was made to energy-minimize the resultant hCA-II-compound-1
structure. However, on the basis of these preliminary modeling
data, we have tentatively assigned His-4 as the most likely target
for the interaction of IDA-Cu2+ of compound-1 (see Discussion
section). While we are currently testing the validity of this model
by creating the site-specific mutation in His-4 as well as other
neighboring residues, we proceeded to ascertain whether the
incorporation of IDA-Cu2+ to benzenesulfonamide (via the
spacer) increased the binding affinity of the parent inhibitor.
Using benzenesulfonamide, compound-1, and Cu2+-depleted
compound-1 as the inhibitors, and recombinant human carbonic
anhydrase as the enzyme, we performed the steady-state kinetic
and ligand binding experiments in 25 mM HEPES buffer, pH
7.0, containing 10% acetonitrile (referred to as the standard
HEPES buffer). As detailed in the Experimental Section, the
plasmid encoding the full length of human carbonic anhydrase-
II was expressed in the Escherichia coli BL21 (codon plus DE3
RIL) cells and purified by affinity chromatography on a
p-aminomethyl benzenesulfonamide-agarose column.30 The
resultant enzyme preparation was judged to be homogeneous
by SDS-PAGE. Compound-1 was synthesized by following the
protocol outlined in Scheme 1.
Figure 4. Structural model of hCAII with modeled compound-1. The bond,
which was rotated (from -179° to -9°) to position the IDA-conjugated
Cu2+ in the vicinity of His-4, is highlighted.
Steady-State Kinetic Studies. The time course of the hCA-
II-catalyzed reaction was determined by monitoring the hy-
drolysis of p-nitrophenyl acetate at 348 nm.20 This esterase
reaction has been frequently employed for determining the
inhibition constants (Ki) of a variety of inhibitors of different
carbonic anhydrase isozymes.21 As noted with other sulfonamide
derivatives, compound-1 was also found to serve as a competi-
tive inhibitor of the enzyme against p-nitrophenyl acetate as
substrate. Figure 5 shows the inhibitor concentration dependence
of the initial rates of the hCA-II catalysis in the standard HEPES
buffer.
the enzyme during catalysis.18 The linear distance between the
active site zinc atom and the NE2 nitrogens of the surface-
exposed histidine residues was determined to be 12, 15, 24,
22, 21, and 15 Å for His-64, His-4, His-3, His-17, His-15, and
His-10, respectively. In light of these, as well as the location
of FBS residue within the enzyme’s active site pocket, we
decided to attach a triethylene glycol spacer between benzene-
sulfonamide and the IDA-Cu2+ moieties and synthesize com-
pound-1 (Figure 2). Upon performing the energy minimization
of compound-1, it was noted that the distance between the NH2
nitrogen of benzenesulfonamide and the Cu2+ conjugated to the
IDA moiety was 19 Å. Hence, it appeared likely that benzene-
sulfonamide and IDA-Cu2+ moieties of compound-1 would be
able to bridge (by small adjustments in the protein and ligand
structures) between the active site pocket and one of the surface-
exposed histidine residues of the enzyme.
To probe such a possibility, we superimposed the benzene-
sulfonamide residue of compound-1 on the known structural
coordinate of FBS (bound to hCA-II) and adjusted a few
torsional angles while monitoring the steric hindrance (“bump
check”). It appeared that just by rotation of the bond between
the two methylene groups (next to the amide nitrogen of
compound-1) from -179° (original position) to -9.0°, the IDA-
conjugated Cu2+ approached within 2.2 Å distance of the NE2
nitrogen of His-4 (Figure 4). No other surface-exposed histidine
residue showed closer proximity (to IDA-Cu2+) than His-4 upon
manually adjusting (while maintaining the original position of
benzenesulfonamide within the enzyme’s active site pocket)
other bonds of compound-1. Since the His-Cu2+ interaction is
For an easy comparison, the inhibition data are presented as
% activity of the enzyme as a function of the total concentration
of the inhibitor. Note that as the concentration of the inhibitors
increases, the rate of enzyme catalysis decreases. However, such
(19) Roy, B. C.; Mallik, S. J. Org. Chem. 1999, 64, 2969-2974.
(20) Pocker, Y.; Stone, J. T. Biochemistry 1967, 6, 668-678.
(21) (a) Forster, R. E. In The Carbonic Anhydrases: Cellular Physiology and
Molecular Genetics; Dodgson, S. J., Tashian, R. E., Gros, G., Carter, N.
D., Eds.; Plenum Press: New York, 1991; pp 79-98. (b) Winum, J.-Y.;
Vullo, D.; Casini, A.; Montero, J.-L.; Scozzafava, A.; Supuran, C. T. J.
Med. Chem. 2003, 46, 5471-5477. (c) Vullo, D.; Franchi, M.; Gallori, E.;
Antel, J.; Scozzafava, A.; Supuran, C. T. J. Med. Chem. 2004, 47, 1272-
1279.
(22) (a) Chen, R. F.; Kernohan, J. C. J. Biol. Chem. 1967, 242, 5813-5823.
(b) Day, Y. S. N.; Baird, C. L.; Rich, R. L.; Myszka, D. G. Protein Sci.
2002, 11, 1017-1025. (c) Fierke, C. A.; Calderone, T. L.; Krebs, J. F.
Biochemistry 1991, 30, 11054-11063.
(23) Qin, L.; Srivastava, D. K. Biochemistry 1998, 37, 3499-3508.
(24) (a) Holm, R. H.; Kennepohl, P.; Solomon, E. I. Chem. ReV. 1996, 96, 2239-
2314. (b) Glusker, J. P.; Katz, A. K.; Bock, C. W. Rigaku J. 1999, 16,
8-16.
(25) (a) Gray, W. D.; Rauh, C. E. J. Pharmacol. Exp. Ther. 1967, 156, 383-
396. (b) Gray, W. D.; Maren, T. H.; Sisson, G. M.; Smith, F. H. J.
Pharmacol. Exp. Ther. 1957, 121, 160-168. (c) Barnish, I. T.; Cross, P.
E.; Dickinson, R. P.; Gadsby, B.; Parry, M. J.; Randall, M. J.; Sinclair, I.
W. J. Med. Chem. 1980, 23, 117-121. (d) Maren, T. H.; Sanyal, G. Annu.
ReV. Pharmacol. Toxicol. 1983, 23, 439-59.
(26) Sharma, S.; Agarwal, G. P. Anal. Biochem. 2001, 288, 126-140.
(27) Kumar, N. R.; Srivastava, D. K. Biochemistry 1995, 34, 9434-9443.
(28) Shnek, D. R.; Pack, D. W.; Sasaki, D. Y.; Arnold, F. H. Langmuir 1994,
10, 2382-2388.
(17) Nicholls, A.; Sharp, K.; Honig, B. Proteins: Struct., Funct., Genet. 1991,
11, 281-296.
(18) (a) Tu, C. K.; Silverman, D. N.; Forsman, C.; Jonsson, B. H.; Lindskog, S.
Biochemistry 1989, 28, 7913-7918. (b) Duda, D.; Tu, C.; Qian, M.; Laipis,
P.; McKenna, M. A.; Silverman, D. N.; McKena, R. Biochemistry 2001,
40, 1741-1748.
(29) (a) Studier, F.; Rosenberg, A.; Dunn, J.; Dubendorff, J. Methods Enzymol.
1990, 185, 60-89. (b) Sambrook, J.; Fritsch, E.; Maniatis, T. Molecular
Cloning: A Laboratory Manual, 2nd ed.; Cold Spring Harbor Laboratory:
Cold Spring Harbor, NY, 1989.
(30) Khalifah, R. G. J. Biol. Chem. 1971 246, 2561-2573.
9
J. AM. CHEM. SOC. VOL. 126, NO. 35, 2004 10877

A R T I C L E S
Banerjee et al.
Scheme 1
a decrease is considerably steeper with compound-1 (9) than
that obtained with the parent compound, benzenesulfonamide
(b). These data qualitatively implied that the Ki value of
compound-1 was much lower than that of benzenesulfonamide.
For quantitative comparison, the data of Figure 5 were analyzed
by eq 3. The solid lines are the best fit of the experimental data
for the Ki values of sulfonamide and compound-1 being 1.50
µM and 35 nM, respectively. Hence, the incorporation of the
tether residue (IDA-Cu2+ via the triethylene glycol spacer) to
benzenesulfonamide increased the inhibitory potency of com-
pound-1 by about 40-fold.
(b). In fact, the fitted line for the benzenesulfonamide-dependent
inhibition of the enzyme passes through these data points. Hence,
the Ki value for the inhibition of the enzyme by Cu2+-depleted
compound-1 has also been taken to be 1.5 µM. These compara-
tive experimental results clearly demonstrate that the enhance-
ment of the inhibitory potency of compound-1 is solely dictated
by the presence of Cu2+ at the IDA site.
However, on the basis of the above data, we could not rule
out the possibility of compensatory effect between different
regions of compound-1 (in the absence of Cu2+), yielding the
same binding affinity of the Cu2+-depleted compound-1 for
hCA-II vis a vis that of benzenesulfonamide. For example, it is
possible that triethylene glycol (spacer) and iminodiacetate
(chelator) exhibit stabilizing and destabilizing (or vice versa)
effects, respectively, and these opposite effects cancel each other.
To probe such a possibility, we performed the steady-state
kinetic studies for the inhibition of hCA-II, using triethylene
glycol-conjugated benzenesulfonamide as an inhibitor. Note that
the latter compound is devoid of the iminodiacetate (chelator)
moiety. The analysis of the inhibition data (not shown to avoid
crowding in Figure 5) by eq 3 produced a Ki value of 0.75 µM
for the triethyleneglycol-benzenesuflonamide conjugate. Note
that the latter value is only 2-fold lower than that obtained either
with benzenesulfonamide or the Cu2+-depleted compound-1.
This 2-fold decrease in the Ki value is presumably due to the
interaction of triethylene glycol moiety with some group at the
enzyme surface. Since iminodiacetic acid does not inhibit hCA-
II, we could not perform the above studies. However, if the
iminodiacetate moiety of compound-1 exhibits the (opposite)
destabilizing effect, its magnitude would, at best, be by 2-fold.
Clearly, these effects are not significant as compared to the 40-
fold reduction in the Ki value of benzenesulfonamide upon
attachment of IDA-Cu2+ via the triethylene glycol spacer.
To further ascertain the contribution of Cu2+ vis a vis the
remainder of the tether moiety (including IDA and the spacer),
we performed the above inhibition studies utilizing Cu2+
-
depleted compound-1. The data are shown as open triangles in
Figure 5. Note a close correspondence between these data and
the data for the inhibition of the enzyme by benzenesulfonamide
Figure 5. Steady-state kinetic data for the inhibition of hCA-II. The initial
rates of the enzyme-catalyzed hydrolysis of p-nitrophenyl acetate substrate
were measured as a function of the inhibitor concentrations. [Enzyme] )
1 µM, [p-nitrophenyl acetate] ) 0.5 mM. The solid smooth lines are the
best fits of the data according to eq 3 for the Ki values of benzenesulfona-
mide (b), compound-1 (9), and Cu2+-depleted compound-1 (4) of 1.5,
35, 1 nM, and 1.5 µM, respectively.
Binding Studies for the Interactions of the Inhibitors to
hCA-II. To ensure that the Ki values determined in the previous
section were the true measures of the binding affinities of the
inhibitors to hCA-II and not due to some unforeseen kinetic
9
10878 J. AM. CHEM. SOC. VOL. 126, NO. 35, 2004

Carbonic Anhydrase Inhibitor
A R T I C L E S
Figure 6. Fluorescence spectra of dansylamide in the absence and presence
of hCA-II. [Dansylamide] ) 10 µM, [hCA-II] ) 1 µM. λex ) 330 nm.
Figure 7. Binding isotherm of hCA-II-dansylamide complex. The increase
in the fluorescence emission intensity at 448 nm (λex ) 330 nm) for the
titration of a fixed concentration of the enzyme (1 µM) with increasing
concentrations of dansylamide has been plotted. The solid smooth line is
the best fit of the data23 for the Kd of 3.2 ( 0.24 µM.
complexity, we determined the dissociation constants (Kd) of
the enzyme-inhibitor complexes by the dansylamide displace-
ment method.22 Dansylamide has been known to serve as a
fluorogenic ligand for carbonic anhydrases, and it exhibits
changes in the fluorescence emission profile upon binding to
the enzyme sites. However, since the magnitude of the above
spectral changes of dansylamide upon binding to the enzymes
(as well as its binding constant) is dependent on the biological
origin of the enzyme (including isozymes), buffer composition
and its pH, and the stability of dansylamide in the reaction
mixture, we acquired the emission spectra of dansylamide in
the absence and presence of hCA-II in the standard HEPES
buffer. The presence of 10% acetonitrile in our buffer media
was found to stabilize the fluorophore during the course of our
experiments. Figure 6 shows the fluorescence emission spectra
of 10 µM dansylamide (λex ) 330 nm) in the absence and
presence of 1 µM hCA-II. Note that whereas the emission
spectrum of free dansylamide shows an emission maximum at
536 nm, the presence of 1 µM enzyme creates a shoulder peak
at 448 nm. The change in the fluorescence emission intensity
of dansylamide upon binding to hCA-II (at 448 nm) was utilized
to determine the dissociation constant of the enzyme-dansyl-
amide complex. Figure 7 shows the titration of a limiting
concentration of enzyme by increasing concentrations of dan-
sylamide. Note the increase in fluorescence emission intensity
at 448 nm (λex ) 330 nm) upon increasing the dansylamide
concentration.
The binding isotherm was analyzed by a complete solution
of the quadratic equation describing the enzyme-ligand interac-
tion, as elaborated by Qin and Srivastava.23 The solid smooth
line is the best fit of the data for the Kd value of the enzyme-
dansylamide complex as being equal to 3.2 µM. It should be
pointed out the above value is higher than that reported in the
literature (0.25-1.2 µM).22 At this time, we do not understand
the origin of this discrepancy.
For determining the dissociation constants of the enzyme-
inhibitor complexes, we titrated a fixed concentration of the
mixture of the enzyme and dansylamide with increasing
concentrations of the inhibitors. The titration data of Figure 8
show that as the concentration of the inhibitor increases, the
fluorescence emission intensity (448 nm) of the enzyme-
dansylamide complex decreases.
Figure 8. Determination of the dissociation constants of the hCA-II-
inhibitor complexes by the dansylamide displacement method. (b), (9),
and (4) represent the data for benzenesulfonamide, compound-1, and Cu2+
-
depleted compound-1, respectively. [hCA-II] ) 1 to 1.1 µM. [Dansylamide]
) 10.8 µM. The Kd values of the enzyme-benzenesulfonamide (as well as
the Cu2+-depleted compound-1) and compound-1 were determined (by the
best fit of the data by eq 9) to be 0.66 ( 0.05 µM, and 23.5 ( 3.4 nM,
respectively.
Since the concentrations of some of the inhibitors (e.g.,
compound-1) were comparable to the concentration of the
enzyme, the experimental data were analyzed (solid lines) by a
complete solution of the quadratic equation (eqs 8 and 9) for
the competitive displacement of the ligands, as described in the
Experimental Section. The solid smooth lines are the best fit of
the experimental data according to eq 9 for the dissociation
constants of the enzyme-benzenesulfonamide (b) and enzyme-
compound-1 (9) complexes as being 0.66 µM and 23.5 nM,
respectively. Note that these values are similar to the corre-
sponding Ki values (determined by the steady-state kinetic
method) of the enzyme-inhibitor complexes. Such a similarity
strengthens our deduction that the incorporation of IDA-Cu2+
(in conjunction with the tether residue) to benzenesulfonamide
indeed enhances the binding affinity of compound-1. To further
ascertain the role of Cu2+ of compound-1 vis a vis the remainder
of the tether chain in enhancing the binding affinity, we
performed the above experiment using the Cu2+-depleted
9
J. AM. CHEM. SOC. VOL. 126, NO. 35, 2004 10879

A R T I C L E S
Banerjee et al.
that would promote and/or preclude protein-protein, protein-
lipid, protein-carbohydrate, and protein-nucleic acid interac-
tions, crystallization of membrane-bound proteins, etc. However,
to successfully employ this methodology, the following criteria
must be met: (1) The targetable enzyme structures must have
their surface-exposed residues (potential of interacting with
Cu2+) within 15-25 Å from the active site-directed ligands.
(2) The active site pockets of the enzymes must not be buried
deep into the protein structures. (3) The inhibitory features must
be desirable at or near the neutral pH. The latter is particularly
so since at higher pH, Cu2+ interacts with different side chain
groups of proteins.6
Various transition-metal ions are known to interact with
different types of organic compounds, including the amino acid
side chain groups of proteins.24 However, the molecular
mechanism of protein-metal interaction is very complex and
is contributed by a combination of electrostatic, hydrophobic,
and/or donor-acceptor (coordination) factors.26 The overall
energetics of one type of interaction versus the other is dictated
by the nature of chelating ligand, metal ion, surface amino acid
composition, and the surrounding chemical environment (e.g.,
pH, ionic strength, etc.). However, on the basis of a variety of
the affinity chromatography data, it has been argued that at
neutral pH, the binding of proteins to IDA-Cu2+ columns is
generally limited to metal coordination by surface-accessible
histidine residues.6 At higher pH values, other basic groups of
amino acid side chains (or N-terminus group) can interact with
IDA-Cu2+. Since the imidazole group of histidine has a pKa of
6.8, it serves as an ideal ligand for interaction with IDA-Cu2+
at neutral pH. By using the diethylpyrocarbonate modified
proteins, Arnold and her collaborators have shown (via the ESR
spectroscopy) that the native proteins are targeted to IDA-Cu2+
lipid assemblies through coordination by surface histidines.28
However, there are reports of involvement of other amino acid
side chains (e.g., Trp, Phe, Tyr, Cys) of proteins in binding to
metal affinity columns, but the mechanism of such interaction
is not yet clear.26
Although the above discussion suggests that IDA-Cu2+ moiety
of compound-1 has potential to interact with the surface-exposed
histidine (presumably His-4 as discerned via the modeling
studies) of hCA-II, we cannot rule out the possibility of
involvement of the thiol group of cysteine and the indole
nitrogen of tryptophan in such interactions. Although there is
no surface-exposed cysteine residue in hCA-II, there are two
surface-exposed tryptophans (Trp-5 and Trp-16) in the vicinity
of the histidine clusters near the N-terminal end of hCA-II. We
are currently in the process of performing the site-specific
mutations of different N-terminal histidine (as well as tryp-
tophan) residues in hCA-II to ascertain their contributions in
binding to IDA-Cu2+, and we will report these findings
subsequently.
Figure 9. Effect of EDTA on hCA-II-catalyzed reaction in the presence
of compound-1. The enzyme-catalyzed hydrolysis of p-nitrophenyl acetate
in the presence of compound-1, used as such (trace 1) or preincubated with
5 mM EDTA (trace 2). While the reaction was in progress, 5 mM EDTA
was added (as shown by arrow) in trace 1. [hCA-II] ) 1 µM. [p-nitrophenyl
acetate] ) 0.5 mM. [Compound-1] ) 1 µM. The slopes of the trace 1 before
and after addition of EDTA were calculated (∆A348/s) to be 0.0006 and
0.0016, respectively. The slope of trace 2 was calculated to be 0.0028.
compound-1 (Figure 8, 4). As observed during the Ki measure-
ment (Figure 5), the displacement of the enzyme-bound dans-
ylamide either by benzenesulfonamide or by the Cu2+-depleted
compound-1 could be described by a common fitted line for a
dissociation constant of 0.65 µM. Hence, in the absence of Cu2+
,
compound-1 behaves just like benzenesulfonamide, suggesting
that the IDA and triethylene glycol spacer do not play any role
in stabilizing the enzyme-compound-1 complex.
To ensure that compound-1 served as inhibitor and did not
irreversibly inactivate the enzyme, we initiated a steady-state
enzyme-catalyzed reaction using p-nitrophenyl acetate as a
substrate in the presence of compound-1 (Figure 9, trace 1),
and a few minutes after the start of the reaction, a small aliquot
of EDTA (5 mM) was added. Since EDTA has much higher
affinity for Cu2+ than IDA,32 it could easily remove the IDA-
bound Cu2+. As shown in Figure 9 (trace 1), as soon as EDTA
is added to the reaction mixture, the time course of the reaction
progress increases sharply. The rate of the latter process was
found to be about 3-fold faster than that observed prior to the
addition of EDTA. When we performed the above steady-state
experiment utilizing EDTA-incubated compound-1, the initial
rate of the enzyme catalysis was found to be faster than that
observed without EDTA. These qualitative experiments clearly
rule out the possibility of inactivation of hCA-II in the presence
of compound-1.
Discussion
The strategy of enhancing the binding affinities of the active
site-directed inhibitors by incorporating IDA-Cu2+ moieties via
suitable spacers is expected to find a variety of applications in
pharmaceutical, agricultural, and biotechnological industries.
Such applications may range from the de novo designing of
highly potent and isozyme-specific inhibitors such as potential
drugs, insecticide, and herbicide agents, designing compounds
However, one of the major advantages of the imidazole-
IDA-Cu2+ interaction is its prevalence in the aqueous medium.24
This is presumably because in aqueous solution, the IDA-
conjugated Cu2+ contains a bound water molecule, which is
exchanged by the interaction of the lone pair electrons of the
nitrogen of the imidazole residues (pKa ) 6.8), forming a stable
Lewis acid-base pair. Hence, unlike other polar amino acid
side chains of proteins, the histidine-Cu2+ interaction does not
involve high energetic costs for desolvating the interacting
(31) Dill, K. A. J. Biol. Chem. 1997 272, 701-704.
(32) Sillen, L. G.; Martell, A. E. In Stability Constants of Metal Ion Complexes,
2nd ed.; The Chemical Society: London, 1964.
9
10880 J. AM. CHEM. SOC. VOL. 126, NO. 35, 2004

Carbonic Anhydrase Inhibitor
A R T I C L E S
species. Evidently, targeting the surface-exposed histidines may
emerge to be an energetically facile strategy in the inhibitor
designing endeavors. A casual perusal of the X-ray crystal-
lographic structures in the Brookhaven Protein Data Bank
reveals that several enzymes, which are the potential targets of
drug/herbicide designs, contain the surface-exposed histidines
in the vicinity of their active site pockets. These enzymes include
17â-hydroxysterioid dehydrogenase, (target for breast cancer),
tyrosinase (target for cutaneous melanoma),11f adenylate kinase
(target for neurological disorder),11g aldose reductase (target for
diabetes),11a and acetolactate synthase (target for the herbicide
design).11c The binding affinities of the putative inhibitors of
glycol spacer and IDA groups in the absence of Cu2+. This
suggests that neither IDA nor triethylene glycol spacer has any
additive contribution in stabilizing the ground state of the
enzyme-inhibitor complex.
Experimental Section
Materials. Zinc sulfate, ampicillin, chloramphenicol, and IPTG were
purchased from Life Science Resources, Milwaukee, WI. Yeast extracts
and tryptone were purchased from Becton Dickinson, Sparks, MD.
Acetonitrile was purchased from Aldrich Chemical Co., Milwaukee,-
WI. HEPES, p-aminomethylbenzenesulfonamide-agarose, p-nitro-
phenyl acetate, and PMSF were obtained from Sigma, St. Louis, MO.
Dansylamide was purchased from Avocado Research Chemicals,
Heysham, Lancs, U.K. All the chemicals needed for synthesis of
compound-1 were purchased from Aldrich Chemical. The expression
cells BL21 codon plus DE3(RIL) were from Stratagene, La Jolla, CA.
All other chemicals were of reagent grade and were used without further
purification.
these enzymes can be significantly enhanced by attaching Cu2+
-
conjugated tether residues (to the active site-directed inhibitors),
which would interact with one of the surface-exposed histidine
residues of the enzymes.
It should be mentioned that the length of spacers between
the active site-directed and IDA-Cu2+ moieties need not be
precise in designing the inhibitors to bridge between the active
site and surface-exposed histidine residues in enzymes. This is
presumably because of the inherent flexibility in the enzyme
structures as well as the adjustments in inhibitor structures.
Although the above features have entropic disadvantage, the
latter is compensated by the enthalpic gain due to the IDA-
Cu2+-histidine interaction. For example, while designing
compound-1, we selected triethylene glycol as spacer just to
provide an approximate length so that benzenesulfonamide could
bind to the active site and the IDA-Cu2+ could bind to His-4.
Our recent isothermal titration microcalorimetric data reveal that
the enthalpic changes of compound-1 is much higher than those
accounted for the binding of individual moieties. On the other
hand, the free-energy changes of compound-1 is less than those
envisaged for the “additive” effect31 (our unpublished results).
Methods. Synthesis of Compound-1. To a suspension of 4-car-
boxybenzenesulfonamide (280 mg, 1.39 mmol) in dichloromethane (13
mL) and triethylamine (0.7 mL, 5 mmol), DMF (5 mL) was added to
make a solution. BOP reagent (616 mg, 1.39 mmol) was added to it,
followed by the addition of 219 (575 mg, 1.52 mmol) in dichloromethane
(2 mL). The reaction was continued for 12 h at room temperature with
stirring and then quenched with a saturated aqueous solution of NaCl.
The aqueous phase was extracted with ethyl acetate. The combined
organic layer was washed successively with 5% citric acid, water, 5%
NaHCO3, and water. The organic layer was dried over Na2SO4 and
evaporated under reduced pressure to provide the crude product as a
viscous liquid. The crude product was purified by column chromatog-
raphy (silica gel, initially with dichloromethane, and then 5% methanol
in dichloromethane) to give 3 as a viscous oil, 520 mg (61%). 1H NMR
(CDCl3, 300 MHz): δ 1.45 (s, 18H), 2.64 (d, 2H, J ) 9 Hz), 2.88 (s,
2H), 2.96 (s, 2H), 3.39 (s, 2H), 3.48-3.72 (m, 8H), 6.14 (br, 1H),
7.71-8.01 (m, 4H).
The other advantage of using the surface-exposed histidine
residues as the second anchor site is the possibility of designing
isozyme-specific inhibitors. For example, because of the marked
similarity in the active site geometry of different carbonic
anhydrase isozymes,11e benzenesulfonamide cannot serve as an
ideal drug for inhibiting one isozyme (e.g., hCA-II) without
affecting the others. This is why several sulfonamide inhibitors,
initially approved as the anticonvulsants, cerebral vasodilators,
diuretics, and antiglaucoma agents, were subsequently with-
drawn from the market because of serious side effects.25 Since
the surface distribution of the histidine residues are different
among different carbonic anhydrase isozymes, it is highly
plausible to achieve the isozyme selectivity by employing our
“two-prong” approach in inhibitor designing.
Compound 3 (305 mg, 0.54 mmol) was taken in a solution of 4 N
HCl in 1,4 dioxane (6 mL) and was stirred for 5 h at room temperature.
The solvents were removed by rotary evaporator, and the residue was
dried under vacuum. The desired acid was obtained as the HCl salt
(263 mg, 99%) and used as such for the next step without further
1
purification. H NMR (DMSO-d6, 500 MHz): δ 3.41-3.52 (m, 9H),
3.59-3.75 (m, 3H), 4.17 (s, 4H), 7.49 (br, 2H), 7.87 (d, 2H, J ) 10
Hz), 8.0 (d, 2H, J ) 10 Hz). 13C NMR (DMSO-d6, 125 MHz): δ 55.41,
65.97, 69.41, 69.86, 70.18, 126.28, 128.56, 137.86, 146.88, 165.95,
168.21.
Acid sulfonamide (130 mg, 0.27 mmol) was dissolved in MeOH (6
mL), and an aqueous solution of saturated NaHCO3 (1 mL) was added.
Solid CuCl2‚2H2O (45.8 mg, 0.27 mmol) was then added with stirring.
Immediate precipitation took place. The reaction mixture was stirred
for 30 min at room temperature. The solvents were evaporated to
dryness under reduced pressure. The residue was dissolved in a
minimum quantity of water, and MeOH was added to it, causing
precipitation. The precipitate was filtered, again dissolved in a minimum
quantity of water, and precipitated by absolute ethanol. Filtration of
the precipitate and drying under vacuum provided the pure compound-1
as a blue solid; 130 mg (91%). Anal. Calcd For C17H23CuN3O9S‚H2O:
C, 38.74; H, 4.78; N, 7.97. Found: C, 39.03; H, 4.64; N, 7.68.
From the mechanistic point of view, the experimental data
presented in the previous section clearly demonstrates that the
incorporation of IDA-Cu2+ to benzenesulfonamide via trieth-
ylene glycol enhances the binding affinity of the parent
compound by about 40-fold. This has been demonstrated by
both the comparative steady-state kinetic and the ligand
displacement method. The fact that the above enhancement in
the binding affinity is primarily contributed by the presence of
Cu2+ at the IDA site comes from the experiments involving
Cu2+-depleted compound-1 (see Figures 5 and 8). The binding
affinity of Cu2+-depleted compound-1 exhibits nearly the same
affinity for hCA-II as that of benzenesulfonamide, and there is
no major compensatory effect for the binding of triethylene
Expression and Purification of the Recombinant Human Car-
bonic Anhydrase-II. The expression vector pMA-5-8 containing human
carbonic anhydrase-II gene (at the Sma I and BamH I restriction
cleavage sites) was obtained as a gift from Dr. Carol Fierke’s laboratory.
The plasmid was transformed into E. coli BL21 codon plusDE3(RIL)
host cells by following the standard molecular biology protocol.29 The
transformed cells were grown in LB medium, supplemented with 100
9
J. AM. CHEM. SOC. VOL. 126, NO. 35, 2004 10881

A R T I C L E S
Banerjee et al.
µg of ampicillin/mL, 50 µg/mL of chloramphenicol, and 60 µg/mL of
ZnSO4 at 37 °C until A600 was 0.6. The expression of hCA-II was
induced by addition of 400 µM IPTG and 400 µM ZnSO4. The cells
were incubated further at 25 °C overnight. The cells were centrifuged
at 5000g for 15 min. The pellet was washed in 20 mM Tris-HCl, pH
8.7, and resuspended in the same buffer containing 0.5 mM EDTA. A
working concentration of 1 mM PMSF in 2-propanol was added prior
to sonication. The cells were sonicated for a total time of 10 min in a
Branson bath sonifier utilizing 40% duty cycle in an ice cold bath.
The sonicated extract was centrifuged at 15 000 rpm for 30 min, and
the supernatant (crude extract) was collected for further purification.
The enzyme activity of the recombinant human carbonic anhydrase
II was measured in 25 mM HEPES buffer, pH 7.0, containing 10%
acetonitrile (the standard buffer) at 25 °C, utilizing 0.4 mM p-
nitrophenyl acetate as substrate.20 The protein concentration was
determined according to the Bradford method, utilizing BSA as the
standard protein.
The recombinant form of human carbonic anhydrase II was purified
from the crude extract using p-aminomethyl benzenesulfonamide-
agarose column.30 The enzyme concentrations were determined spec-
trophotometrically using ꢀ280 ) 54 mM-1 cm-1 based on MW ) 30 000.
The purified enzyme was subjected to the SDS-PAGE analysis to
confirm the homogeneity of the enzyme.
Steady-State Kinetics for the Inhibition of hCA-II. The steady-
state kinetic experiments for the hCA-II catalyzed reaction were
performed on Perkin-Elmer Lambda 3B spectrophotometer. All the
steady-state experiments were performed in the standard HEPES buffer
containing 10% acetonitrile. The latter solvent ensured the solubility
of the substrate during the course of the reaction, and it did not exhibit
any influence on the catalytic activity of the enzyme. The initial rates
of the enzyme-catalyzed reactions were measured by following the
hydrolysis of the chromogenic substrate, p-nitrophenyl acetate, at 348
nm.20
for the hCA-II-catalyzed esterase reaction, since the p-nitrophenyl
acetate is not very soluble in the reaction buffer.
Vmax*[S]/Km
V )
(2)
1 + [I]f/Ki
In the absence of any inhibitor, the numerator term of eq 1 can be
taken as the measure of V0 (the initial velocity of the enzyme-catalyzed
reaction in the absence of inhibitor). Besides, [I]f can be represented
in the form of total concentration of the enzyme and inhibitor by a
complete solution of the quadratic equation describing their interaction.
With these solutions, the final form of the competitive steady-state
model can be given by eq 3.
V )
V0*Ki
K + ([I] - 0.5{([I] + [E] + K ) - ([I] + [E] + K )2 - 4*[I] *[E] })
x
i
t
t
t
i
t
t
i
t
t
(3)
The inhibition data of Figure 5 have been analyzed by eq 3 using the
nonlinear regression analysis software Grafit 4.0.
Spectrofluorometric Studies. All the spectrofluorometric studies
involving dansylamide were performed on Perkin-Elmer Lambda 50-B
spectrofluorometer equipped with a magnetic stirrer and thermostated
water bath. To ensure the stability of dansylamide during the course
of the titration as well as kinetic experiments, its stock solution (1 mM)
was prepared in 10 mM HCl, and it was diluted in the standard HEPES
buffer containing 10% acetonitrile. The emission spectra of dansylamide
in the absence and presence of hCA-II were acquired by fixing the
excitation wavelength at 330 nm, and both excitation and emission slits
were maintained at 5 mm with a cutoff filter at 390 nm.
The dissociation constant of the hCA-II-dansylamide complex was
determined by titrating a fixed concentration of the enzyme (1 µM)
with increasing concentrations of dansylamide in the standard HEPES
buffer, pH 7.0, containing 10% acetonitrile. The initial reaction volume
was 2.0 mL. The excitation and emission wavelengths were maintained
at 330 and 448 nm with a cutoff filter at 390 nm. The excitation and
emission slits were 5 mm each. The dissociation constant of the
enzyme-dansylamide complex was determined by analyzing the
binding isotherm as described by Qin and Srivastava.23
The steady-state kinetic experiments were performed in a total
reaction volume of 1.3 mL (in the standard buffer), containing 0.5 mM
p-nitrophenylacetate and varied concentrations of the inhibitors. The
initial rates of the enzyme catalysis were determined by taking the slopes
of the reaction traces. For easy comparison of the inhibition data, the
initial rates were translated into % activity as a function of the inhibitor
concentrations.
Because of slow esterase activity of hCA-II, fairly high concentration
of the enzyme (∼1 µM) is utilized to reliably measure the initial rates
of the enzyme reaction at 348 nm. With such a high concentration of
the enzyme, part of the substrate and/or inhibitor can be envisaged to
be bound to the enzyme site. Since the Michaelis-Menten equation
relies on the free rather than the total concentrations of the substrates
and inhibitors, the fraction of the bound species must be subtracted
from their total concentrations for analyzing the kinetic data. For the
hCA-II catalyzed esterase reaction, the above constraint is not a major
problem since the Km for p-nitrophenyl acetate is about 4 mM, and as
long as the substrate concentration is maintained in the range of 1 mM
or higher, the total concentration of the substrate could be taken as the
measure of the free concentration. However, this would not be the case
with tight binding inhibitors (Ki in the range of nM). Hence, for
inhibitors, the free concentrations must be calculated from the total
concentration by complete solution of the quadratic equation describing
enzyme-inhibitor interaction. Equation 1 is a classic steady-state
equation for the competitive inhibition of the enzyme.
The dissociation constant of the hCA-II-inhibitor complex was
determined by competitive displacement of a fixed concentration of
dansylamide by increasing concentrations of the inhibitor. In a typical
experiment, an equilibration mixture containing 1 µM hCA-II and 10
µM dansylamide in the standard HEPES buffer (pH 7.0, containing
10% acetonitrile) was titrated with increasing concentrations of the
inhibitor. The initial reaction volume was 2.0 mL, and the reaction
content was continuously stirred during the course of the titration. The
excitation and emission wavelengths were maintained at 330 and 448
nm, respectively, with a cutoff filter at 390 nm. The excitation and
emission slits were 5 mm each. Since the binding of the inhibitor
competitively displaced dansylamide from the enzyme site, the
fluorescence intensity at 448 nm decreased, and the data were analyzed
by a modified form of the competitive binding model as elaborated by
Kumar and Srivastava.27
If two ligands (e.g., dansylamide and benzenesulfonamide, repre-
sented by D and B, respectively) compete for an enzyme site (E), it is
intuitively obvious that the dissociation constant for the EB complex
can be determined if the dissociation constant of ED complex is already
known. Such a determination would be relatively easy if [E]total , Kd
(D) and Kd (B), and [E]total , [D]total and [B]total. Under this situation,
the total concentrations of ligands can be taken as being equal to their
free concentrations. On the other hand, if this ideal situation is not
satisfied and the enzyme-ligand dissociation constants are comparable
to the initial concentrations of the enzyme and ligands (such as
encountered in the fluorescence titration data of Figure 8), the
Vmax
V )
(1)
1 + Km/[S] (1 + [I]f/Ki)
where [I]f is the free concentration of the inhibitor. If by experimental
design the substrate concentration can be maintained considerably lower
than the Km value ([S]total , Km), eq 1 can be further simplified to the
form of eq 2. In fact, the above condition is an obligatory requirement
9
10882 J. AM. CHEM. SOC. VOL. 126, NO. 35, 2004

Carbonic Anhydrase Inhibitor
A R T I C L E S
dissociation constant determination of one enzyme-ligand complex
by the competitive displacement method becomes cumbersome. How-
ever, recourse can be made if the dissociation constant of the one-
enzyme ligand complex is relatively higher than that of the other, as
observed for the binding of dansylamide versus compound-1 to the
hCA-II site. The competitive binding of dansylamide (D) versus
benzenesulfonamide (B) to the hCA-II (E) site can be formalistically
represented by eq 4.
cence (Fobsd) as a function of the inhibitor ([B]) concentration can be
given by eq 9
Fobsd ) ∆Fmax/[E]t*([E]t - [EB])
(9)
where ∆Fmax is the maximum change in fluorescence upon complete
displacement of dansylamide from the enzyme site. The titration data
of Figure 8 were fitted by the nonlinear regression analysis of eq 9
using the software package Grafit 4.
Molecular Modeling Studies. The molecular modeling studies were
performed on a Silicon Graphics-O2 molecular modeling workstation
with the aids of Accelrys softwares, InsightII(98), Discover, and
biopolymers. The coordinates for the X-ray crystallographic structures
of human carbonic anhydrase-II complexed with FBS (PDB file:
1IF4.pdb) were downloaded from the Brookhaven Protein Data Bank.
The structure of compound-1 was first subjected to energy minimization
by the aid of Discover. The sulfonamide ring of compound-1 was
manually superimposed on the enzyme-bound structural coordinate of
FBS. This was followed by manually adjusting the torsion angles of
different single bonds of the spacer moiety, while measuring the distance
between the Cu2+ (bound to IDA) and the surface-exposed histidine
residues. It was noted that just by rotating the bond between two
methylene group (as shown in Figure 4) from its original angle of
-179.7° to -9.0°, the IDA-bound Cu2+ reached closest (2.2 Å away)
to His-4.
In eq 4, Kd and Kb are the dissociation constants of the hCA-II-
dansylamide (ED) and hCA-II-benzenesulfonamide (EB) complexes,
respectively. If the free and total concentrations of E, D, and B are
represented by [E], [D], and [B] and [E]t, [D]t, and [B]t, respectively,
Kd and Kb can be represented as follows.
Kd ) [E][D]/[ED] ) [E]([D]t - [ED])/[ED]
Kb ) [E][B]/[EB] ) [E]([B]t - [EB])/[EB]
(5)
(6)
Conclusions
If [E]t , [D]t, then [D]t can be taken to be the measure of [D], and by
taking into account the mass balance ([E]t ) [E] + [ED] + [EB]), [E]t
can be represented by eq 7.
The present article provides a novel strategy for enhancing
the binding affinity of an active site-directed inhibitor by
attaching an iminodiacetic acid (IDA)-conjugated Cu2+ (via
triethylene glycol as spacer), which could interact with one of
the surface-exposed histidine (His-4) residues. Although the
overall approach has been developed for human carbonic
anhydrase-II, the overall strategy can be adopted for designing
highly potent inhibitors for other pathogenic enzymes as
potential drugs. Since the surface residues are unlikely to be
conserved during the course of the natural selection process,
our contemplated approach has the potential to generate
isozyme-specific inhibitors for selected enzymes.
Kb [EB]
Kb[EB][D]
[E]t ) [EB] +
+
(7)
[B]t - [EB]
([B]t - [EB])Kd
Equation 7 can be simplified to yield the dependence of [EB] on the
concentrations and the binding constants of other species (eq 8).
[EB] ) ([E]t + [Bt + Kb + (Kb/Kd)[D]t) -
([E] + [B + K + (K /K )[D] )2 - 4[E] [B]
x
t
t
b
b
d
t
t
t
(8)
2
Acknowledgment. This research was supported by the
National Institutes of Health Grants 1R15 DK56681-01A1 to
D.K.S. and 1R01 GM 63404-01A1 to S.M.
The dissociation constant of the enzyme-inhibitor complex is deter-
mined by monitoring the decrease in the fluorescence of the enzyme-
dansylmide complex (λex ) 330 nm, λem ) 448 nm) as a function of
increasing concentration of the inhibitor. Hence, the observed fluores-
JA047557P
9
J. AM. CHEM. SOC. VOL. 126, NO. 35, 2004 10883