Novel 2-substituted-benzimidazole-6-sulfonamides as carbonic anhydrase inhibitors: synthesis, biological evaluation against isoforms I, II, IX and XII and molecular docking studies

Abstract Inhibition of Carbonic Anhydrases (CAs) has been clinically exploited for many decades for a variety of therapeutic applications. Within a research project aimed at developing novel classes of CA inhibitors (CAIs) with a proper selectivity for certain isoforms, a series of derivatives featuring the 2-substituted-benzimidazole-6-sulfonamide scaffold, conceived as frozen analogs of Schiff bases and secondary amines previously reported in the literature as CAIs, were investigated. Enzyme inhibition assays on physiologically relevant human CA I, II, IX and XII isoforms revealed a number of potent CAIs, showing promising selectivity profiles towards the transmembrane tumor-associated CA IX and XII enzymes. Computational studies were attained to clarify the structural determinants behind the activities and selectivity profiles of the novel inhibitors.


Introduction
Carbonic anhydrases (CA) are a family of ubiquitary zinc metalloenzymes that catalyze the reversible reaction of hydration of CO 2 to HCO 3 À 1 . This simple transformation plays a physiological regulatory role in a number of processes associated with pH control, ion transport, fluid secretion and several biosynthetic pathways 1 . Fifteen CA isoenzymes are encoded in humans and other primates, that differ for their subcellular localization, catalytic activity, and susceptibility to different classes of inhibitors 1 . Specifically, cytosolic (CA I, CA II, CA III, CA VII, and CA XIII), membrane-bound (CA IV, CA IX, CA XII and CA XIV), mitochondrial (CA VA and CA VB), and secreted in saliva (CA VI) enzymes were characterized 1 . Most of these CA isoforms represent interesting therapeutic targets, and their inhibition has been exploited clinically for many decades for a variety of applications in treating a multitude of diseases such as glaucoma, edema, epilepsy, obesity, neuropathic pain and other neurological disorders [2][3][4] . More recently, hCA IX and XII have been implicated in tumor progression/metastasis, and their selective inhibition could represent an additional opportunity for drug intervention against hypoxic cancers 5 .
Because of the ubiquity of CAs, the selectivity of the inhibitors for certain isoforms is a crucial issue to be reached in a drug development campaign in order to target a disease without relevant side effects 7 . In this respect, expanding the chemical space by the exploration of novel scaffolds may aid the development of novel classes of CAIs featuring improved pharmacological properties in terms of inhibition potency and isoform-selectivity.
Among the several sulfonamide CAIs described, Schiff bases and secondary amines incorporating aromatic/heterocyclic sulfonamide moieties in their structure (compounds of type I in Figure 1) have been extensively investigated in recent years [8][9][10][11] . Comparing the activity of imines and their secondary amine counterparts, it is evident that the molecular flexibility markedly affects, positively or negatively, both activity and selectivity [8][9][10][11] .
Prompted by these outcomes and exploiting our experience in the application of the frozen analog approach 12,13 , we decided to further reduce the flexibility of the Schiff bases constraining the N¼C imine bond into a ring (II, Figure 1). As a rigidifying building block, we decided to introduce the benzimidazole (III, Figure 1), a privileged structure extensively used in medicinal chemistry, which has been only scarcely explored for its potential in the development of CAIs 14 . As substituent in the position 2 of benzimidazole core, we selected phenols and benzoic acid derivatives. In fact, different studies showed that these scaffolds are effective CAIs, giving interactions with the enzyme that may not involve a direct interaction with the active site zing ion [15][16][17] . Herein, we report the synthesis of 2-substituted-benzimidazole-6-sulfonamides 6-18 and the related carboxamide 19 and their inhibitory activity toward four physiologically relevant enzymes, the cytosolic isoforms hCA I and II as well as the transmembrane tumor-associated ones hCA IX and XII.

Chemistry
All chemicals were purchased from Sigma Aldrich Srl (Milan, Italy) or from Fluorochem Ltd. (Hadfield, UK) and were of the highest purity. All solvents were reagent grade and, when necessary, were purified and dried by standard methods. All reactions requiring anhydrous conditions were conducted under a positive atmosphere of nitrogen in oven-dried glassware. Standard syringe techniques were used for anhydrous addition of liquids. Reactions were routinely monitored by TLC performed on aluminum-backed silica gel plates (Merck DC, Alufolien Kieselgel 60 F 254 ) with spots visualized by UV light (k ¼ 254, 365 nm) or using a KMnO 4 alkaline solution. Solvents were removed using a rotary evaporator operating at a reduced pressure of $10 Torr. Organic solutions were dried over anhydrous Na 2 SO 4 . Chromatographic purification was done on an automated flash-chromatography system (IsoleraTM Dalton 2000, Biotage) using cartridges packed with KP-SIL, 60 Å  (40-63 mm particle size). All microwave assisted reactions were conducted in a CEM Discover V R SP microwave synthesizer equipped with a vertically focused IR temperature sensor. Analytical high performance liquid chromatography (HPLC) was performed on a Shimadzu SPD 20 A UV/VIS detector (k ¼ 220 and 254 nm) using C-18 column Phenomenex Synergi Fusion -RP 80 A (75 Â 4.60 mm; 4 mm) at 25 C using a mobile phase A (water þ 0.1% TFA) and B (ACN þ 0.1% TFA) at a flow rate of 1 ml/min. 1 H spectra were recorded at 400 MHz on a Bruker Ascend 400 spectrometer while 13 C NMR spectra were obtained by distortionless enhancement by polarization transfer quaternary (DEPTQ) spectroscopy on the same spectrometer. Chemical shifts are reported in d (ppm) relative to the internal reference tetramethylsilane (TMS). Due to the existence of tautomers, some 1 H and 13 C NMR signals could not be detected for some of the prepared benzimidazoles so only the distinct signals are reported. Low resolution mass spectra were recorded on a Finnigan LCQ DECA TermoQuest mass spectrometer in electrospray positive and negative ionization modes (ESI-MS). High resolution mass spectra were recorded on a Bruker solariX MRMS in electrospray positive ionization modes (ESI-FTMS). All tested compounds possessed a purity of at least 95% established by HPLC unless otherwise noted. Acids 27 and 28a were commercially available, acid 28b was obtained by previously reported procedure (see Supplementary Data).

2-(4-Hydroxyphenyl
imidazole-6-sulfonamide 26a (174 mg, 0.504 mmol) was dissolved in 2 ml of a solution DCM/TFA (1:1) and the mixture was stirred for 18 h. The solvent was evaporated, and the resulting solid was crystallized with ethanol to give the title compound as a brown solid (110 mg, 75%). 1 Compound 35 (150 mg, 0.408 mmol) was dissolved in 2 ml of a solution DCM/TFA (1:1) and the mixture was stirred for 24 h. The solvent was evaporated, and the resulting solid was taken up with 2 ml of THF. To the resulting mixture, a water solution (2 ml) of LiOH (62 mg, 2.58 mmol) was added, and the reaction mixture was stirred at room temperature for 4 h. The reaction was concentrated under vacuum, the aqueous phase was washed with CHCl 3 then acidified with 3N HCl until a white precipitate formed that was recovered by filtration. Compound 19 was obtained as white solid (110 mg, 91%) after recrystallization from ethanol 1

3,4-Diamino-N-(tert-butyl)benzenesulfonamide (20a)
To a stirred suspension of 24a (1.65 g, 6.04 mmol) in 250 ml of MeOH, ammonium formate (7.61 g, 120.74 mmol) and palladium on carbon 10% wt. (160 mg) were added. The resulting mixture was heated at reflux for 4 h. After cooling, the mixture was filtered, and the solvent evaporated under reduced pressure. The crude material was taken up with 100 ml of water and extracted with EtOAc (3 Â 60 ml). The combined organic phases were washed with brine, dried over Na 2 SO 4 , filtered and evaporated. The product, obtained as a light brown solid (1.30 g, 88%), was used for the next step without further purification. 1 To a solution of 2-nitroaniline (2.00 g, 14.48 mmol) in Et 2 O (100 ml), ethyl chlorooxoacetate (2.17 g, 1.78 ml, 15.93 mmol) was added portionwise with continuous stirring. Once the addition was complete, the resulting yellow suspension was stirred for 18 h at room temperature and then concentrated under vacuo. The crude residue was dissolved in EtOAc (100 ml), washed with saturated NaHCO 3 (3 Â 30 ml) and with brine (30 ml). The organic phase was dried over anhydrous Na 2 SO 4 , and evaporated to dryness, giving the desired product as a yellow solid (3.38 g, 98%). 1

4-Amino-3-nitrobenzenesulfonyl chloride (23)
A solution of ethyl 2-(2-nitrophenylamino)-2-oxoacetate 22 (2.00 g, 8.40 mmol) in 4.5 ml of chlorosulfonic acid was heated at 80 C for 3 h. The red mixture was poured slowly into ice À water (150 ml) and stirred for 30 min. The product was extracted from the aqueous solution using Et 2 O (3 Â 30 ml). The combined organic phases were washed with brine (10 ml), dried (Na 2 SO 4 ), filtered, and concentrated in vacuo to give the title compound as a brown solid which was immediately used for the next reaction without purification. 1 To a stirred solution at 0 C of crude 23 (1.04 g, 4.39 mmol) in dry THF (25 ml) was added dropwise, under nitrogen atmosphere, tertbutylamine (1.85 ml, 17.56 mmol). The reaction was allowed to reach room temperature and was stirred for 18 h. The solvent was removed at reduced pressure and the residue was taken up with 50 ml of water and extracted with EtOAc (3 Â 20 ml). The combined organic phases were washed with brine, dried over Na 2 SO 4 , filtered and evaporated under reduced pressure. Purification by silica gel chromatography (DCM/MeOH) yield pure 24a (0.96 g, 80%) as a light yellow solid. 1 To a stirred solution of 20a (150 mg, 0.62 mmol) in dry DMF (7 ml), 4-hydroxybenzaldehyde (75 mg, 0.61 mmol) and Na 2 S 2 O 5 (0.165 g, 0.793 mmol) were added. The resulting mixture was stirred at 80 C for 18 h. After cooling at room temperature, water was added. The brown precipitate formed was recovered by filtration and was washed several times with water and 1N HCl. After recrystallization from EtOH, compound 26a was obtained as light brown solid (160 mg, 72%). 1

N-(tert-butyl)-3,4-dinitrobenzamide (33)
A solution of 3,4-dinitrobenzoic acid 32 (1.00 g, 4.71 mmol) in thionyl chloride (15 ml) was refluxed for 2 h under a nitrogen atmosphere. After the solution was cooled at room temperature, the excess thionyl chloride was removed at reduced pressure and the crude material dried under vacuum. To the residue, dissolved in dry THF (20 ml) and cooled to 0 C, was added dropwise a mixture of tert-butylamine (525 mL, 5.00 mmol) and triethylamine (695 mL, 5.00 mmol) in dry THF (5 ml). The mixture was stirred at room temperature for 18 h, filtered, and evaporated. The crude residue was dissolved in DCM (20 ml), washed with 1N HCl, saturated NaHCO 3 , and water, dried, and concentrated in vacuum. Recrystallization from EtOH yielded compound 33 (1.05 g, 84%) as a light yellow solid. 1

Enzyme activity assays
An Applied Photophysics stopped-flow instrument has been used for assaying the CA-catalysed CO 2 hydration activity 18 . Phenol red (at a concentration of 0.2 mM) has been used as indicator, working at the absorbance maximum of 557 nm, with 20 mM Hepes (pH 7.5) as buffer, and 20 mM Na 2 SO 4 (for maintaining constant the ionic strength), following the initial rates of the CA-catalysed CO 2 hydration reaction for a period of 10-100 s. CO 2 concentrations ranged from 1.7 to 17 mM for the determining inhibition constants. For each inhibitor, at least six traces of the initial 5-10% of the reaction have been used for measuring the initial velocity. Uncatalyzed rates were determined in the same manner and subtracted from the total observed rates. Stock solutions of inhibitor (0.1 mM) were prepared in buffer with a maximum 3% DMSO, and dilutions up to 0.01 nM were done with the assay buffer. Inhibitor and enzyme solutions were preincubated for 15 min at room temperature prior to assay, in order to allow for the formation of the E-I complex. The inhibition constants were obtained by nonlinear least-squares methods using PRISM 3 and the Cheng-Prusoff equation, as reported earlier, and represent the mean from at least three different determinations [19][20][21][22][23][24][25][26] . All CA isoforms were recombinant ones obtained in-house as reported earlier [27][28][29] .

Molecular modeling methods
The latest version of the AD4 docking software (version 4.2) 30 together with its GUI AutoDockTools (ADT) and the AutoDock4(Zn) force field 31 , were employed. The hCA IX X-ray structure used for the experiment had the PDB code 5FL4 32 . The protein structure was prepared for the docking using the Protein Preparation Wizard of the Maestro suite 33 that adds bond orders, adds hydrogen atoms, deletes water molecules and produces the appropriate protonation states. The co-crystal ligand of 5FL4 was separated from the cognate protein. The 2 D Sketcher tool of Maestro was used to build compounds 13, 14 and 17. For the three ligands, the protonation and tautomeric state, as well as their geometry, were optimized through LigPrep, part of the same suite. Through Maestro, the X-ray structures of hCA I (PDB 6F3B) 34 , hCA II (PDB 3K34) 35 , and hCA XII (PDB 5MSA) 36 , were downloaded and superimposed on the structure of hCA IX. The ligands were translated in the AD4 specific file format (PDBQT) using the python scripts prepare_ligand4.py and prepare_receptor4.py, part of ADT, applying the standard settings. Following the AutoDock4(Zn) force field protocol 37 , to add the tetrahedral zinc pseudo atoms to the receptor PDBQT the script zinc_pseudo.py, part of the material provided with the force field, was employed. The docking area was centered on the active site. The zinc-specific non bonded pairwise potentials were included in the creation of the grid parameter file. A set of grids of 60 Å Â 40 Å Â 50 Å with 0.375 Å spacing was calculated considering the docking area for all the ligands atom types employing AutoGrid4. For every ligand, 200 independent docking simulations were achieved. Each docking calculation comprised 20 million energy evaluations employing the Lamarckian genetic algorithm local search (GALS) method. This latter assesses a population of viable docking solutions and propagates the best individuals from each generation into the following generation of feasible solutions. A low-frequency local search according to the method of Solis and Wets was applied to every docking attempt to guarantee that the final solution represented a local minimum. All dockings were performed with a population size of 250, and 300 iterations of Solis and Wets local search were applied with a probability of 0.06. A rate of mutation of 0.02 and a crossover rate of 0.8 were used to produce new docking attempts for following generations, and the best individual from each generation was propagated over the following generation. The docking results from every (200) independent docking calculation were clustered based on the of root-meansquare deviation (rmsd) (solutions differing by less than 2.0 Å) between the Cartesian coordinates of the atoms and were ranked on the basis of free energy of binding (DGAD4).
The amino group of 2-nitroaniline 21 was protected by acylation with ethyl chlorooxoacetate in diethyl ether. Reaction of the resulting ethyl 2-(2-nitrophenylamino)-2-oxoacetate 22 with chlorosulfonic acid at 80 C, followed by aqueous workup, yielded the unprotected sulfonyl chloride derivative 23, which reacted with tert-butylor ethyl-amine to afford the N-substituted-4-amino-3nitrobenzenesulfonamides 24a and 24b, respectively. Catalytic hydrogenation with ammonium formate and palladium catalyst converted the nitro derivatives into the corresponding amino derivatives 20a and 20b.
Coupling reaction of benzenesulfonamide 20a with 2-(4hydroxyphenyl)acetic acid 27 or 3-arylpropanoic acids 28a and 28b, in the presence of the peptide coupling reagents hydroxybenzotriazole (HOBt) and 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC) and N-methylmorpholine in dry DMF, furnished the amides 29, 30a and 30b as regioisomeric mixtures, which were purified without separation of regioisomers. Benzimidazoles derivatives 8 and 17 were straightforwardly obtained by p-toluenesulfonic acid-mediated cyclization and deprotection of the corresponding 2-amido anilines 30a and 30b in refluxing toluene. The carboxylic acid 18 was obtained by deprotecting the methyl ester 17 with lithium hydroxide. Our attempts to cyclize compound 29 using the same reaction conditions were not successful. On the other hand, benzimidazole 31 was obtained in good yield (74%) by using acetic acid under microwave irradiation. Finally, deprotection with trifluoracetic acid at room temperature furnished the primary sulfonamide 7.
CA inhibition assays and structure-activity relationship (SAR) considerations Table 1 lists the enzyme inhibitory activities of the newly synthesized compounds 6-19 against the human (h) CA I, II, IX and XII isoforms, assessed by a stopped-flow CO 2 hydrase assay [19]. AAZ 1 was used as the standard drug in the assay 18 . Selectivity ratios (SR) for inhibiting the tumor-associated transmembrane isoforms (hCA IX and XII) over the physiologically dominant cytosolic one (hCA II) are also reported for the most active compounds.
First, as putative leads for the development of selective CAIs, we synthesized compound 6 and its homologs with a methyl or ethyl linker between the benzimidazole and the phenol moieties (compounds 7 and 8 in Table 1). They all resulted in mediumpotency inhibition of the slow cytosolic isoform hCA I, with K i values ranging from 213.6 nM to 442.1 nM. The 4 0 -hydroxybenzyl derivative 7 showed good inhibition activity for hCA II, IX, and XII but was not selective (K i 91.9 nM for hCA II, 73.9 nM for hCA IX, 63.8 nM for hCA XII). On the other hand, the 4 0 -hydroxyphenyl (6) and 4 0 -hydroxyphenylethyl (8) analogs were effective inhibitors of hCA IX and XII, respectively, with K i s in the low nanomolar range (6, K i 17.4 nM for hCA IX, 44.1 nM for hCA XII; 8, K i 14.4 nM for hCA IX, 9.8 nM for hCA XII) basically comparable to those of the reference 1 (K i 25 nM for hCA IX, 5.7 nM for hCA XII), and interesting selectivity ratio (SR) vs hCA II. Based on these data, compound 7 was not further investigated, and a structure-activity relationship (SAR) study was undertaken on 6. In particular, different substitution patterns on the pendant 2-phenyl ring at 5-position of benzimidazole were investigated (compounds 9-17). Deletion of the 4 0 -hydroxy substituent (9), as well as its replacement with methoxycarbonyl (10) or carboxy (11) groups, produces good but unselective inhibitors; in fact, a modest gain (9, 11) or a subsistence (10) in activity toward the isoform I and a general decrease in inhibitory potency for hCA II, IX, and XII with respect to the lead 6 were observed, suggesting a precise role played by the hydroxy group in the interaction with the enzyme. This is confirmed by the inhibitory activities showed by compounds 12-14, where the introduction of an o-substituent to the p-hydroxy group on the 2-phenyl ring resulted in a moderate increase in activity toward the two cytosolic CA evaluated, and a more considerable improvement for the isoform XII, with compound 14 being the most potent hCA XII inhibitor (K i for hCA XII 3.8 nM), showing also a good SR with respect to hCA II (SR 18.1).
Specifically, the introduction of a second polar group (OH or COOH) at the ortho position to the phenol ring positively affected the interaction with isoform XII (compounds 12 and 14), while the presence of the methoxycarbonyl group (compound 13) resulted in a minor increase in hCA XII inhibition with respect to parent compound 6.
Concerning the IX isoform, an exactly opposite trend can be observed for compounds [12][13][14]. Decoration of the 2-phenyl moiety with a 4 0 -hydroxy and 3 0 -methoxycarbonyl groups gives a highly effective hCA IX inhibitor (13), which shows low nanomolar K i (2.2 nM), with a relevant gain in activity compared to the reference 1 (K i 25 nM), and good SR vs hCA II. Differently, compounds 12 and 14, featuring 3 0 ,4 0 -dihydroxy and 3 0 -carboxy-4 0 -hydroxy     substituents, although being rather effective hCA IX inhibitors (K i for hCA IX 41.6 nM and 34.2 nM for 12 and 14, respectively), show a slight decrease in activity with respect to reference compound 6 (K i for hCA IX 17.4 nM). These data suggested the ortho-carboxy phenol ring of compound 14 and the ortho-carboxymethyl phenol ring of compound 13 as proper scaffolds for activity and selectivity for hCA XII and hCA IX, respectively. For these derivatives, further SARs were investigated.
First, we explored structural modifications of the 5-sulfonamide moiety, including the insertion of a small alkyl group at the nitrogen atom to produce secondary sulfonamides (compounds 15 and 16), and replacement of the sulfonamide with a carboxamide (19). The obtained compounds proved to be scarcely active or completely inactive inhibitors of all the CA isoforms tested (K i s varying from 390.5 nM to micromolar values), strongly supporting the crucial role played by the primary sulfonamide group in the interaction with the enzyme.
Finally, we explored the effect of the combination of the ethyl linker between the benzimidazole and the 2-substituted phenol moieties (compounds 17 and 18 in Table 1). The potency for the isoform hCA I decreased, while hCA II was slightly more (18) or equally inhibited (17) with respect to related compounds 14 and 13, respectively. Both 17 and 18 derivatives resulted to be potent hCA IX and hCA XII inhibitors, showing low nanomolar K i values (17, K i for hCA IX 5.9 nM, K i for hCA XII 7.9 nM; 18, K i for hCA IX 7.6 nM, K i for hCA XII 4.2 nM). However, the presence of the ethylene linker between the benzimidazole scaffold and the side phenyl ring abolished the selectivity for hCA IX and XII isoforms, that was observed for compounds 13 and 14, respectively.
Noteworthy, compounds 17 and 18 are better inhibitors than the phenol derivative 8, proving the role of the ortho-carboxymethyl phenol and the ortho-carboxy phenol rings as proper scaffolds in the development of potent and selective hCA IX and hCA XII inhibitors.

Molecular docking studies
To clarify the reasons for the activities displayed by the newly designed compounds, molecular docking studies were attained. Docking calculations were performed using a protocol already successfully applied in our previous work on CA inhibitors 39 . Namely, AutoDock4.2 (AD4) 30,37 was employed together with the AutoDock4(Zn) forcefield 31 , which was specifically designed to accurately predict the binding interactions of ligands docking to zinc metalloproteins.
Ligands 13, 14 and 17 were selected for the in silico experiments as representative of the whole set. These compounds were first docked in the active site of hCA IX. For the latter, a high-resolution X-ray crystal structure bound to a small molecule inhibitor (PDB code 5FL4) 32 was chosen. According to our theoretical model (Figures 2(A), 3(A) and 4(A)), in the three inspected ligands, the negative nitrogen of the sulfonamide group chelates the zinc ion of the active site. The sulfonamide also engages an H-bond with the backbone of T200. Furthermore, the benzimidazole nitrogen is in a potential H-bond accepting position with Q92 side chain. The Notably, it would appear that the pendant 2-phenyl ring, in the three ligands, points towards what has been defined as a "selectivity hot spot" in CAs 40 (Figures 2(A), 3(A) and 4(A)): a highvariability region in CAs binding site, that can be exploited for the rational design of selective compounds among different CAs. Here, ligand 13 phenyl ring and its 3 0 -methoxycarbonyl moiety are able to establish favorable contacts with the lipophilic sidechains of L91 and V130 (Figure 2(A)). On the other hand, the 4 0hydroxy group is pointing outside of the binding site, probably establishing a network of stabilizing H-bonding interactions with the solvent water molecules (Figure 2(A)). It can be argued that this accounts for the higher potency displayed by compounds featuring the 4 0 -hydroxy group. The same holds true for compound 14. As for ligand 17, while the benzimidazole core binding mode is conserved, the ethyl linker engenders greater flexibility which allows the side phenyl ring to expand further into the hotspot gorge. This allows enhancing the positive contacts with L91 and V130. Moreover, an additional H-bond between the 3 0 -methoxycarbonyl group and Q71 is formed (Figure 4(A)).
With the aim of rationalizing the selectivity of the compounds, the crystal structures of hCA I (PDB 6F3B) 34 , hCA II (PDB 3K34) 35 , and hCA XII (PDB 5MSA) 36 were downloaded and their binding sites analyzed. To ascertain how the predicted docked poses of ligands 13, 14 and 17 in the hCA IX active site would fit in the other CAs, the 3 D structures of the enzymes were superimposed. This analysis revealed that the recognition pattern achieved for hCA IX, conducive of potent enzyme inhibition, is unlikely to be confirmed for hCA I and hCA II due to major steric clashes (Figures 2(C,D), 3(C,D), and 4(C,D)). Indeed, their binding sites in the hot spot region feature bulky substituents that would unfavorably affect the binding mode of the compounds, especially in the case of ligand 17 (Figure 4(C,D)).
Conversely, hCA XII and hCA IX binding sites share a higher degree of homology. As such, the binding poses found for 13, 14 and 17 in hCA IX also fit in hCA XII (Figures 2(B) and 3(B)). Still, few key differences in the hot spot region can be found. Specifically, some hydrophobic residues in hCA IX are replaced by polar ones (L91, L123, V130 in CA IX become T88, Y121, S130 in hCA XII, respectively). Importantly, in hCA XII the positive K69 takes the place of Q71 in hCA IX. It is possible to infer that the more hydrophilic and positively charged binding site of hCA XII provides a better fit for compounds with a 3 0 -carboxyl group on the pendant 2-phenyl ring (see compound 14, Figure 3(B)). Instead, compounds bearing the 3 0 -methoxycarbonyl moiety can interact more favorably with the lipophilic and neutral hot spot region of hCA IX. Purportedly, the presence of the ethyl linker grants the possibility for the ligand to maximize the favorable interactions in hCA IX and hCA XII, with both substitution patterns on the 2-phenyl ring, the 3 0 -carboxyl (18) or the 3 0 -methoxycarbonyl (17) moiety. On the other hand, the same ethyl linker, by enhancing the ligand flexibility, should also allow for a better fit into the hCA I and hCA II isoform structures, thereby negatively impacting on the ligand selectivity profile (Figures 2, 3 and 4). X-ray structure. The protein is shown as green ribbons and its molecular surface as transparent gray. The ligand is shown as yellow sticks. (d) 17 hCA IX docked binding pose within the hCA II (PDB 3K34) structure. The protein is shown as pink ribbons and its molecular surface in transparent gray. The ligand is depicted as yellow sticks. The images were rendered using the UCSF Chimera software 30 .