Ординатура / Офтальмология / Английские материалы / Carbonic Anhydrase Its Inhibitors and Activators_Supuran, Scozzafava, Conway_2004

coordinate with a deprotonated anchoring group to the zinc ion (Liljas et al. 1994). In addition, both compounds make hydrogen bonds with the side chain oxygen atom of Thr-199 and the backbone nitrogen atom of the same residue. The hydroxy group of Thr-199 forms an additional hydrogen bond with Glu-106, such that the Thr-199 hydroxyl acts as a hydrogen-bond acceptor for inhibitor binding.

Despite the similarity in anchoring to the zinc according to a well-known interaction pattern (Liljas et al. 1994), a surprising difference is observed in the binding mode of the two inhibitors with respect to the attached ring system. Topiramate forms hydrogen bonds with amino acid side chains in a hydrophilic binding pocket (Asn-67, Gln-92) and to a water molecule that donates a hydrogen bond to Thr-200 (Figure 3.2A). In addition, this water interacts with the oxygen atom of the ligand’s six-membered ring. RWJ-37947, instead, shows a different binding mode in which the ring system is rotated by ca. 180˚ (mirrored), although an orientation as observed for topiramate would appear feasible (Figure 3.3). Therefore, surprisingly, the cyclic sulfate group points to a more hydrophobic pocket (Leu-198, Pro-202, Phe-131), and except for the sulfamate anchoring group, no further hydrogen bonds are observed (Figure 3.2B).

Given the apparent topological similarity of the two compounds and the surprising difference in their binding modes, the question arises as to whether such a behavior is predictable, and, in particular, whether computational tools can highlight this difference.

Here, the grid-based automated docking program AutoDock (Goodsell and Olson 1990; Morris et al. 1996) was used to address this question and to predict the binding mode of topiramate, RWJ-37947 and three additional commonly used CA inhibitors (see Figure 3.1: acetazolamide 3, brinzolamide 4 and dorzolamide 5). For four compounds, x-ray crystal structures were extracted from the Protein Data Bank (PDB codes 1a42, 1azm, 1cil, 1eou). The crystal structure of CA II in complex with topiramate was kindly provided by Dr. C.T. Supuran, University of Florence. All five protein structures were superimposed with Sybyl® 6.8 (Tripos, Inc., St. Louis, MO), based on all atoms of 20 conserved amino acids in the binding pocket (Ser-29, Pro-30, Asn-61, Gly-63, Gln-92, His-94, His-96, Trp-97, Glu-106, Glu-117, His-119, His-122, Ala-142, Val-143, Ser-197, Thr-199, Pro-201, Glu-205, Trp-209, Asn-244). Comparison of the amino acids in the binding pocket of the superimposed structures indicates slight conformational flexibility for His-64 only (Figure 3.4). This is further corroborated by superpositioning 55 complex structures of hCA II with Relibase (Gunther et al. 2003; Hendlich et al. 2003), which reveals that CA II has a fairly invariant binding site, except for the amino acid at position 64 (Klebe 2003). The biochemical reason for this observation is most likely related to the functional role of His-64 in catalysis, wherein it acts as a proton shuttle in regenerating the zincbound hydroxide (Duda et al. 2001, 2003). In general, two main conformations of His-64 are observed: the in conformation, in which His-64 points to the active site, and the out conformation, in which His-64 points to the solvent. In the five complexes mentioned, His-64 adopts both the in (1azm, 1eou) and the out (1a42, 1cil, crystal structure of topiramate) conformations.

Copyright © 2004 CRC Press, LLC

A

B

FIGURE 3.3 (See color insert following page 148.) The binding mode of RWJ-37947 observed in the crystal structure (A) and manually rotated by 220∞ around the single bond connecting the ring system with the sulfamate anchor (B). (B) reveals that a binding mode as observed in the topiramate complex is sterically possible. The solvent accessible surface of the binding pocket of CAII hosting topiramate is shown (A, B).

To allow as much space as possible for ligand placement, 1cil with His-64 in the out conformation was used to dock all five ligands. The structure was constructed for docking by removing all water and ligand atoms, adding polar hydrogens and assigning Amber atomic charges and solvation parameters as required by the

Copyright © 2004 CRC Press, LLC

FIGURE 3.4 (See color insert.) Binding mode of topiramate (A) and RWJ-37947 (B) in CA II. All amino acids in the binding site are highly conserved, except His 64, which adopts the out conformation in the topiramate complex (A) and the in conformation in complex with RWJ-37947 (B). All molecule representations are drawn with PyMOL. (DeLano, 2002.)

AutoDock program (Morris et al. 1998). AutoDock was then allowed to dock ligands flexibly into the rigid binding pocket of CA II. For this purpose, the protein was first embedded in a three-dimensional grid, a probe atom was placed at each grid point and the interaction energy of the probe atom with the protein was assigned to the grid point. Hence, an affinity grid was calculated for each atom type in the ligand and the grids were then used as protein representation to speed up the docking process. For ligand preparation, Gasteiger–Marsili charges were assigned to the ligand atoms and rotatable bonds were explicitly defined. The largest ring system of the inhibitor was chosen as root fragment, and all rotatable bonds were kept

Copyright © 2004 CRC Press, LLC

TABLE 3.1

Docking Results for CA Inhibitors

Note: Ligands were docked into CA II (PDB code 1cil).

a The total number of clusters obtained in each case is 5 (dorzolamide), 14 (brinzolamide), 13 (acetazolamide), 11 (topiramate) and 9 (RWJ-37947).

flexible. Docking was then carried out by using an empirical binding free energy function and the Lamarckian genetic algorithm (Morris et al. 1998). One hundred independent docking runs were performed for each ligand, applying a standard AutoDock protocol, with a grid spacing of 1 Å, an initial population of 50 randomly placed individuals, a maximum number of 1.5 ∞ 106 energy evaluations, a mutation rate of 0.02, a crossover rate of 0.80 and an elitism value of 1. Results differing by less than 1 Å root-mean-square deviation (rmsd) were merged together in one cluster.

To evaluate the docking results, the rmsd values of the docked solutions with respect to the crystal structure and the corresponding energy scores were taken into account. Table 3.1 and Figure 3.5 show the docking results.

In Table 3.1, the docking solutions corresponding to the first rank of the top cluster and the first rank of the largest cluster are presented along with the corresponding cluster size, the scoring value and the rmsd with respect to the crystal structure. Except for topiramate, the rmsd of the top cluster is generally >3.5 Å and

thus unsatisfactorily high; the corresponding cluster size, however, is rather small (≤11). Considering the first rank in the largest cluster, all rmsd values are below 1.4 Å,

implying that the resulting binding modes are close to the crystal structure (Figure 3.5).

Copyright © 2004 CRC Press, LLC

A B

C D

E

FIGURE 3.5 (See color insert.) Dorzolamide (A), brinzolamide (B), topiramate (C), RWJ37947 (D) and acetazolamide (E) docked into the binding site of CA II (PDB code 1cil). The first rank of the largest cluster is represented. X-ray crystal structures of the corresponding ligands are shown in green.

Copyright © 2004 CRC Press, LLC

TABLE 3.2

Observed Docking Conformations of Topiramate

and RWJ-37947

Note: The position of the ring systems as observed in the x-ray structure and the coordination of the sulfamate anchoring group to zinc and Thr 199 is regarded as correct.

The overall best results in terms of rmsd are observed for dorzolamide and RWJ37947 (in both cases, 0.54 Å for the first rank of the largest cluster; Figure 3.4A

and D). Comparing the scoring values between the first rank of the top cluster and the first rank of the largest cluster, the differences are small (≤0.45 kcal/mol) and

all within the degeneracy range identified previously for the same docking methodology (Sotriffer et al. 2002). The small differences in the scoring values indicate that, in the present case, the scoring function implemented in AutoDock does not significantly differentiate between these binding geometries.

Table 3.2 compares the docked solutions of topiramate and RWJ-37947 with respect to the conformation of the ring system and the coordination of the sulfamate anchoring group to the zinc atom. Assuming that the conformation of the ring system observed in the crystal structure represents the correct binding mode, 74 correct binding modes for topiramate and 90 correct binding modes for RWJ-37947 were generated by the 100 docking runs. As regards the zinc-coordination of the sulfamate group, an approximately equal presentation of correctly docked solutions was obtained for both cases (74 runs for topiramate and 73 runs for RWJ). Docked solutions were considered as wrong if the sulfamate group was placed far away from the zinc atom (>4 Å) or if the anchoring group interacted with Thr-200 instead of Thr-199. Regarding the wrong binding modes of the ring system of topiramate, a correct sulfamate coordination was obtained in none of the cases. In this context, the situation is different for RWJ-37947: among the 10 wrong binding modes of the ring system, 7 show correct coordination of the sulfamate group. This finding can be interpreted as the rotated binding mode of the RWJ ring system being, in principle, possible and compatible with a correct coordination of the sulfamate anchor. Nevertheless, the binding mode observed in the crystal structure is clearly preferred (73 correctly docked solutions with respect to the ring system and zinc coordination).

Copyright © 2004 CRC Press, LLC

Also, for topiramate, the binding mode observed in the crystal structure is preferentially obtained (74 correctly docked solutions). Here, in addition, a wrongly placed ring system seems to impair the correct coordination of the sulfamate group. This finding is compatible with the tight binding mode of topiramate embedded in a well defined hydrogen bond network (Figure 3.2A).

3.3 CONCLUSIONS

It is surprising that current docking algorithms such as AutoDock yielded, in principle, the observed binding modes for topiramate as well as for RWJ-37947 in their respective crystallographically determined enzyme–inhibitor complexes. However, these binding modes were not necessarily top-ranked based on energy evaluation only, but were deemed top ranked when energy and cluster size were jointly considered. The latter criterion indirectly maps the shape of the energy surface used for docking. The most frequently detected local minimum also supposedly corresponds to the most likely geometry. More CA–inhibitor complexes have to be investigated, first to confirm the presented results on the prediction of the binding mode and second to correlate the derived scoring value with binding affinity. The implementation of a frequency factor (Sotriffer et al. 1996) that takes into account the cluster size should be beneficial to rank docked solutions and better separate cluster ranks. Other docking studies have also revealed (Sotriffer et al. 2002) that the application of knowledge-based pair potentials (Gohlke et al. 2000a, 2000b) instead of the implemented AutoDock energy function can further improve the results. Nevertheless, by incorporating the size of the cluster, the docking results presented here demonstrate that it is possible to accurately predict the binding mode of five CAIs. This observation can lead to the following conclusions: if considerably different binding modes show up as highly ranked within virtual screening or docking studies, more than one should be considered as a meaningful structural template for the design of either a single compound or of even better targeted libraries. Integration of combinatorial chemistry and structure-based drug design might be one of the most promising ways to adopt, although not the only one. Considerations of multiple binding modes will be essential for the conceptual design of the core for a chemical library aiming at high hit rates (Figure 3.6). Lead optimization will remain largely driven by intuition, because of its multidimensional character, but the quality of supporting experimental and computational tools is rapidly improving and will lead to significantly faster identification and optimization of new lead compounds.

An optimal fit in a target site neither guarantees that the desired activity of the drug will be enhanced or that undesired side effects will be diminished, nor considers the pharmacokinetics of the drug. Screening of well-designed targeted libraries with a view to pharmacodynamic and pharmacokinetic profiles will finally facilitate the discovery of optimal candidates for further development.

Where observation is concerned, chance favors only the prepared mind.

Louis Pasteur, Speech (1854)

Copyright © 2004 CRC Press, LLC

X X

FIGURE 3.6 Consideration of multiple binding modes should have an impact on the conceptual design of compounds and compounds libraries. Although moieties responsible for specific interactions and accounting for a considerable percentage of overall intrinsic affinity can be kept constant (indicated as X), the substituent list for R1 and R2 might be exchangable to a certain extent.

REFERENCES

Abbate, F., Supuran, C.T., Scozzafava, A., Orioli, P., Stubbs, M.T. and Klebe, G. (2002) Nonaromatic sulfonamide group as an ideal anchor for potent human carbonic anhydrase inhibitors: Role of hydrogen-bonding networks in ligand binding and drug design. Journal of Medicinal Chemistry 45, 3583–3587.

Antel, J. (1999) Integration of combinatorial chemistry and structure-based drug design.

Current Opinion in Drug Discovery and Development 2, 224–233.

Aribi, A.M., and Stringer, J.L. (2002) Effects of antiepileptic drugs on extracellular pH regulation in the hippocampal CA1 region in vivo. Epilepsy Research 49, 143–151.

Babine, R.E., and Bender, S.L. (1997) Molecular recognition of protein–ligand complexes: Applications to drug design. Chemical Reviews 97, 1359–1472.

Baldwin, J.J., Ponticello, G.S., Anderson, P.S., Christy, M.E., Murcko, M.A., Randall, W.C., Schwam, H., Sugrue, M.F., Springer, J.P., Gautheron, P. et al. (1989) Thienothiopyran- 2-sulfonamides: novel topically active carbonic anhydrase inhibitors for the treatment of glaucoma. Journal of Medicinal Chemistry 32, 2510–2513.

Berman, H.M., Westbrook, J., Feng, Z., Gilliland, G., Bhat, T.N., Weissig, H., Shindyalov, I.N., and Bourne, P.E. (2000) The protein data bank. Nucleic Acids Research 28, 235–242.

Böhm, H.J., and Klebe, G. (1996) What can we learn from molecular recognition in pro- tein–ligand complexes for the design of new drugs? Angewandte Chemie International Edition in English 35, 2588–2614.

Böhm, H.J., Klebe, G., and Kubinyi, H. (1996) Wirkstoffdesign, Spektrum Akademischer Verlag, Heidelberg, Berlin, Oxford.

Copyright © 2004 CRC Press, LLC

Bohm, H.J., and Stahl, M. (2000) Structure-based library design: molecular modelling merges with combinatorial chemistry. Current Opinion in Chemical Biology 4, 283–286.

Breton, S. (2001) The cellular physiology of carbonic anhydrases. Journal of Pancreas (Online) 2, 159–164.

Busenbark, K., Pahwa, R., Hubble, J., and Koller, W. (1992) The effect of acetazolamide on essential tremor: an open-label trial. Neurology 42, 1394–1395.

Carrion, E., Hertzog, J.H., Medlock, M.D., Hauser, G.J., and Dalton, H.J. (2001) Use of acetazolamide to decrease cerebrospinal fluid production in chronically ventilated patients with ventriculopleural shunts. Archives of Disease of the Child 84, 68–71.

Casini, A., Antel, J., Abbate, F., Scozzafava, A., David, S., Waldeck, H., Schafer, S., and Supuran, C.T. (2003) Carbonic anhydrase inhibitors: SAR and x-ray crystallographic study for the interaction of sugar sulfamates/sulfamides with isozymes I, II and IV.

Bioorganic and Medicinal Chemistry Letters 13, 841–845.

Casini, A., Scozzafava, A., Mastrolorenzo, A., and Supuran, L.T. (2002) Sulfonamides and sulfonylated derivatives as anticancer agents. Current Cancer Drug Targets 2, 55–75.

Chazalette, C., Riviere-Baudet, M., Scozzafava, A., Abbate, F., Ben Maarouf, Z., and Supuran, C.T. (2001) Carbonic anhydrase inhibitors, interaction of boron derivatives with isozymes I and II: A new binding site for hydrophobic inhibitors at the entrance of the active site as shown by docking studies. Journal of Enzyme Inhibition 16, 125–133.

Christianson, D.W., and Cox, J.D. (1999) Catalysis by metal-activated hydroxide in zinc and manganese metalloenzymes. Annual Reviews in Biochemistry 68, 33–57.

Cowen, M.A., Green, M., Bertollo, D.N., and Abbott, K. (1997) A treatment for tardive dyskinesia and some other extrapyramidal symptoms. Journal of Clinical Psychopharmacology 17, 190–193.

Cox, J.D., Hunt, J.A., Compher, K.M., Fierke, C.A., and Christianson, D.W. (2000) Structural influence of hydrophobic core residues on metal binding and specificity in carbonic anhydrase II. Biochemistry 39, 13687–13694.

Cramer, R.D., III, Patterson, D.E., and Bunce, J.D. (1989) Recent advances in comparative molecular field analysis (CoMFA). Progress in Clinical and Biological Research 291, 161–165.

DeLano, W.L. (2002) The PyMol molecular graphics system, DeLano Scientific, San Carlos, CA, USA.

DesJarlais, R.L., Sheridan, R.P., Seibel, G.L., Dixon, J.S., Kuntz, I.D., and Venkataraghavan, R. (1988) Using shape complementarity as an initial screen in designing ligands for a receptor binding site of known three-dimensional structure. Journal of Medicinal Chemistry 31, 722–729.

Dodgson, S.J., Shank, R.P., and Maryanoff, B.E. (2000) Topiramate as an inhibitor of carbonic anhydrase isoenzymes. Epilepsia 41, S35–S39.

Duda, D., Govindasamy, L., Agbandje-McKenna, M., Tu, C., Silverman, D.N., and McKenna, R. (2003) The refined atomic structure of carbonic anhydrase II at 1.05 A resolution: implications of chemical rescue of proton transfer. Acta Crystallografica D Biological Crystallography 59, 93–104.

Duda, D., Tu, C., Qian, M., Laipis, P., Agbandje-McKenna, M., Silverman, D.N., and McKenna, R. (2001) Structural and kinetic analysis of the chemical rescue of the proton transfer function of carbonic anhydrase II. Biochemistry 40, 1741–1748.

Erdei, A., Gyori, I., Gedeon, A., and Szabo, I. (1990) Successful treatment of intractable gastric ulcers with acetazolamide. Acta Medica Hungarica 47, 171–178.

Esposito, E.X., Baran, K., Kelly, K., and Madura, J.D. (2000) Docking of sulfonamides to carbonic anhydrase II and IV. Journal of Molecular Graph Model 18, 283–289, 307–288.

Copyright © 2004 CRC Press, LLC

Fakhoury, T., Murray, L., Seger, D., McLean, M., and Abou-Khalil, B. (2002) Topiramate overdose: Clinical and laboratory features. Epilepsy Behavior 3, 185–189.

Gibbs, J.W., III, Sombati, S., DeLorenzo, R.J., and Coulter, D.A. (2000) Cellular actions of topiramate: Blockade of kainate-evoked inward currents in cultured hippocampal neurons. Epilepsia 41, S10–S16.

Gohlke, H., Hendlich, M., and Klebe, G. (2000a) Knowledge-based scoring function to predict protein-ligand interactions. Journal of Molecular Biology 295, 337–356.

Gohlke, H., Hendlich, M., and Klebe, G. (2000b) Predicting binding modes, binding affinities and ‘hot spots’ for protein-ligand complexes using a knowledge-based scoring function. Perspectives in Drug Discovery and Design 20, 115–144.

Goodsell, D.S., and Olson, A. J. (1990) Automated docking of substrates to proteins by simulated annealing. Proteins 8, 195–202.

Greer, J., Erickson, J.W., Baldwin, J.J., and Varney, M.D. (1994) Application of the threedimensional structures of protein target molecules in structure-based drug design.

Journal of Medicinal Chemistry 37, 1035–1054.

Gruneberg, S., Stubbs, M.T., and Klebe, G. (2002) Successful virtual screening for novel inhibitors of human carbonic anhydrase: strategy and experimental confirmation.

Journal of Medicinal Chemistry 45, 3588–3602.

Grzybowski, B.A., Ishchenko, A.V., Kim, C.Y., Topalov, G., Chapman, R., Christianson, D.W., Whitesides, G.M., and Shakhnovich, E.I. (2002a) Combinatorial computational method gives new picomolar ligands for a known enzyme. Proceedings of the National Academy of Sciences of the United States of America 99, 1270–1273.

Grzybowski, B.A., Ishchenko, A.V., Shimada, J., and Shakhnovich, E.I. (2002b) From knowl- edge-based potentials to combinatorial lead design in silico. Accounts of Chemical Research 35, 261–269.

Gunther, J., Bergner, A., Hendlich, M., and Klebe, G. (2003) Utilising structural knowledge in drug design strategies: applications using Relibase. Journal of Molecular Biology 326, 621–636.

Hansch, C., and Fujita, T. (1964) r-s-p analysis: A method for the correlation of biological activity and chemical structure. Journal of the American Chemical Society 86, 1616–1626.

Hansch, C., McClarin, J., Klein, T., and Langridge, R. (1985) A quantitative structure–activity relationship and molecular graphics study of carbonic anhydrase inhibitors. Molecular Pharmacology 27, 493–498.

Hebebrand, J., Antel, J., Preuschoff, U., David, S., Sann, H., and Weske, M., (2002) Drug screening method for the treatment and prophylaxis of obesity. WO Patent 02/07821.

Hendlich, M., Bergner, A., Gunther, J., and Klebe, G. (2003) Relibase: Design and development of a database for comprehensive analysis of protein-ligand interactions. Journal of Molecular Biology 326, 607–620.

Herrero, A. I., Del Olmo, N., Gonzalez-Escalada, J. R., and Solis, J. M. (2002) Two new actions of topiramate: inhibition of depolarizing GABA(A)-mediated responses and activation of a potassium conductance. Neuropharmacology 42, 210–220.

Hindle, S.A., Rarey, M., Buning, C., and Lengauer, T. (2002) Flexible docking under pharmacophore type constraints. Journal of Computer Aided Molecular Design 16, 129–149.

Hunt, C.A., Mallorga, P.J., Michelson, S. R., Schwam, H., Sondey, J. M., Smith, R.L., Sugrue, M.F., and Shepard, K.L. (1994) 3-Substituted thieno[2,3-b][1,4]thiazine-6-sulfona- mides. A novel class of topically active carbonic anhydrase inhibitors. Journal of Medicinal Chemistry, 37, 240–247.

Copyright © 2004 CRC Press, LLC