Synthesis, Crystal Structure, Inhibitory Activity and Molecular Docking of Coumarins/Sulfonamides Containing Triazolyl Pyridine Moiety as Potent Selective Carbonic Anhydrase IX and XII Inhibitors

: In this work, two classes of Carbonic Anhydrase (CA) inhibitors, sulfonamide and coumarin derivatives linked to pyta moiety ( 2a-b ) and their corresponding rhenium complexes ( 3a-b ), were designed. These compounds were synthesized and fully characterized by classical analytical methods and X-ray diffraction. All the synthesized compounds were evaluated for their inhibitory activity against the hCA isoforms I, II, IX and XII. They exhibited high inhibitory activities in the range of nanomolar for both hCA IX and hCA XII isoforms. The sulfonamide compound 2a showed the strongest inhibition against the tumour-associated hCA IX isoform with a Ki of 11.7 nM. The tumour-associated isoforms hCA IX and hCA XII were selectively inhibited by all the coumarin derivatives, with inhibition constants ranging from 12.7 nM ( 2b ) to 44.5 nM ( 3b ), while the hCA I and II isoforms were slightly inhibited (in the micromolar range), as expected. In terms of selectivity, compared to previously published rhenium complex-based CA inhibitors, complex 3b showed one of the highest selectivities against hCA IX and hCA XII compared to the off-target isoforms hCA I and hCA II, making it a potential anti-cancer drug candidate. Molecular docking calculations were performed to investigate the inhibition proﬁles of the investigated compounds at the tumour-associated hCA IX active site and to rationalize our results.


Introduction
Carbonic anhydrases (CAs, EC 4.2.1.1) are zinc-containing metalloenzymes that catalyze the reversible hydration of carbon dioxide into bicarbonate ions and protons, to maintain acid-base balance in tissues and blood. hCA IX, and XII are known to be highly rationalize our results, molecular docking calculations were also applied to investigate the inhibition profiles of the synthesized compounds at the tumour-associated hCA IX active site. Two different mechanisms of hCA inhibition are presented here: sulfonamide compounds (2a/3a) that bind directly to the zinc active site, and coumarin compounds (2b/3b) that block the entrance of the enzyme cavity in its closed (non-hydrolyzed) and hydrolyzed form.
Melting points (mp) were determined in open capillaries using a Mettler Toledo (VWR International SAS, Fontenay-sous-Bois, France) and are uncorrected. 1 H and 13 C nuclear magnetic resonance (NMR) spectra were recorded with a Bruker Avance 300 (75.5) MHz spectrometer (Bruker France SAS, Wissembourg, France) at room temperature using CDCl 3 or DMSO-d 6 as a solvent unless otherwise stated. Chemical shifts are reported in parts per million (ppm) relative to a residual solvent peak (CDCl 3 : 1 H: δ 7.26, 13 C: δ 77.2; DMSO-d 6 : 1 H: δ 2.50, 13 C: δ 39.5) and coupling constants (J) are given in Hertz (Hz). Multiplicities were recorded as s (singlet), br s (broad singlet), d (doublet), t (triplet), q (quadruplet) and m (multiplet). Infrared (IR) spectra were recorded with a Perkin-Elmer FTIR 1725 spectrophotometer (VWR International SAS, Fontenay-sous-Bois, France) in the range 4000-400 cm −1 . Electrospray (ESI) mass spectra were obtained on a Q TRAP Applied Biosystems spectrometer (Thermo-Fisher Scientific, Waltham, USA) and High-Resolution Mass Spectra (HRMS) on an LCT Premier Waters spectrometer (Waters SAS, Saint-Quentin en Yvelines, France). Microanalysis was performed by the microanalytical department of the Laboratoire de Chimie de Coordination de Toulouse (LCC, Toulouse, France). Electronic spectrum was measured on a Hewlett Packard 8453 temperature-controlled spectrometer in the range 1000-200 nm in methanol solution.
Literature methods were used to prepare intermediates 1a [33] and 1b [34] (it is known that azido compounds are potentially explosive intermediate products, however, we never observed suspicious reactions in this experimental procedure).

Ligands Synthesis
General Procedure: A mixture of the azido compound (1a or 1b, 1.1 eq.) and 2ethynylpyridine (1eq.) was suspended in acetonitrile (0.2-0.5 M final concentration). Copper(II) acetate monohydrate (0.2 eq.) and sodium ascorbate (0.4 eq.) were added and stirred in the dark at 45 • C overnight. The reaction solvent was removed under reduced pressure and the remaining residues were purified by column chromatography on silica gel using CH 2 Cl 2 /MeOH (95:5 for 2a and 98:2 for 2b) as eluent.  2

Re(I)-Complexes Synthesis
General Procedure: A solution of 2a/2b (1 eq.) and commercial [Re(CO) 5 Cl] (1.1 eq.) in methanol was stirred for 12 h at 65 • C. After cooling to room temperature, the solution was concentrated until 3 mL and then a precipitate was obtained. 5-8 mL of methanol was added to the precipitate. The mixture was heated then cooled and stored at 4 • C for 2 h before the supernatant was carefully removed. This process was repeated three times before the precipitate was dried under vacuum.

Crystal Structure Determination
X-ray intensity data of ligands 2a, 2b and corresponding rhenium complexes 3a, 3b were collected on a Bruker D8 VENTURE diffractometer (Karlsruhe, Germany), and using graphite mono-chromated Mo Kα radiation (λ = 0.71073 Å) for 2b, 3a and 3b, and Cu Kα radiation (λ = 1.54178 Å) for 2a at 193 K. The semi-empirical absorption corrections were employed [SADABS, Program for data correction, Bruker-AXS]. Crystallographic data and refinement details are given in Table 1. The structures were solved using an intrinsic phasing method [35], and refined by full matrix least squares procedures on F2 [36]. All non-H atoms were refined with anisotropic displacement parameters. Hydrogen atoms were refined isotropically at calculated positions, using a riding model with their isotropic displacement parameters constrained to equal 1.5 times the equivalent isotropic displacement parameters of their pivot atoms for terminal sp3 carbon, and 1.2 times for all other carbon atoms. H atoms on N (2a and 3a) were located by difference Fourier maps and freely refined. In complexes 3a and 3b some parts of the molecules were disordered. Several restraints (SAME, SIMU, DELU) were applied to refine some moieties of the molecules and to avoid the collapse of the structures during the least-squares refinement by the large anisotropic displacement parameters.  The supplementary crystallographic data for this paper for compounds 2a (CCDC 2098097), 2b (CCDC 2098098), 3a (CCDC 2098099 ) and 3b (CCDC 2098100 ) can be obtained free of charge from CCDC.

Carbonic Anhydrase Inhibition Assays
An Applied Photophysics stopped-flow instrument was used for assaying the CA catalyzed CO 2 hydration activity. Phenol red (at a concentration of 0.2 mM) was 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 the constant ionic strength), following the initial rates of the CA-catalyzed CO 2 hydration reaction for a period of 10-100 s. The CO 2 concentrations ranged from 1.7-17 mM for the determination of the kinetic parameters and inhibition constants. In particular, CO 2 was bubbled in distilled deionized water for 30 min so that the water was saturated (the concentration at a specific temperature is known from literature). In addition, a CO 2 assay kit (from Sigma) was used to measure the concentration in variously diluted solutions obtained from the saturated solution (which was kept at the same temperature and a constant bubbling during the experiments). For each inhibitor, at least six traces of the initial 5-10% of the reaction were used for determining the initial velocity [37]. The 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 distilled-deionized water and dilutions up to 0.01 nM were completed thereafter with distilled-deionised water. Inhibitor and enzyme solutions were pre-incubated together for 15 min-2 h (or longer, i.e., 4-6 h) at room temperature (at 4 • C for the incubation periods longer than 15 min) prior to assay, to allow for the formation of the E-I complex. The inhibition constants were obtained by non-linear least-squares methods using PRISM 3 and the Cheng-Prusoff equation [38,39], as reported earlier, and represent the mean from at least three different determinations. hCA I was purchased by Sigma-Aldrich and used without further purification, whereas all the other hCA isoforms were recombinant ones obtained in-house as reported earlier [40,41].

Molecular Docking Studies
There are five known mechanisms of carbonic anhydrase inhibition [42]. Two of these are considered to be inhibitors that (i) coordinate directly with the zinc ion in the CA active site and (ii) block the entrance to the active site [42]. The former classes are the classical inhibitors, such as sulfonamides and their isosteres, and some of them are co-crystallized with hCA IX. The latter (CAIs), such as coumarin derivatives, are known to bind to the CA cavity in a different region than the sulfonamide class [30][31][32]. Therefore, the 3D crystal structures of the proteins used in molecular docking were extracted from the RCSB Protein Data Bank web server. The X-ray crystal structure of hCA-IX complexed with 5-(1-naphthalen-1-yl-1,2,3-triazol-4-yl)thiophene-2-sulfonamide (PDB: 5FL4, Ligand ID: 9FK, 1.82Å) [43] was used to predict the binding mode interactions of the synthesized sulfonamide compounds. Similarly, two other PDB crystal structures were downloaded and used to predict the binding mode of our synthesized coumarin-containing compounds (3F8E [30] and 5DVX [44] for hCA II and IX, respectively).
The downloaded crystal structures were prepared using the Discovery Studio visualization code [45]. A single chain was retained if there were multiple protein chains, and all other small non-protein molecules were removed from the crystal structures. To correctly estimate the interaction mode of the native coumarin, co-crystallized in hCA II, and thus the coumarin compounds synthesized against isoform IX, the prepared structure of hCA IX (pdb: 5DVX) was overlaid on the prepared structure of hCA II-coumarin (pdb: 3F8E). Next, the co-crystallized ligand (2-amino-2-hydroxymethyl-propene-1,3-diol) inside the hCA IX isoform was removed, and the hydrolyzed coumarin of the hCA II isoform, together with its Zn 2+ -coordinated water molecule, was inserted inside the prepared structure of the hCA IX isoform. The alignment of the protein structures was performed by the Chimera code [46] and is visualized in Figure S1 in the Supplementary Information. These enzyme-inhibitor complexes were taken as initial structures for docking studies using AutoDock4.2 software.
The AutoDock program has the ability to take into account the flexibility of docked inhibitors during the docking process. To explore the binding modes of the protein with the inhibitors, the Lamarckian genetic algorithm (LGA) [47] was run. Using the AutoDock tools [48], the partial charges of the Kollman unit atom and the essential H atoms were added. Gasteiger partial charges were added to the ligand atoms. The non-polar hydrogen atoms were merged. Interestingly, a charge of +2 was assigned to the Zn atom which is attached by three histidine residues. All proteins were kept rigid and all rotational torsional bonds of the ligands were left free to allow flexible docking. The docking area was selected by constructing a grid box of size 40 × 40 × 40 Å points centered on the x, y and z coordinates of 14.09, −27.414 and 59.605 for sulfonamide-based compounds, and −5.095, 6.414 and 9.427 for coumarin-based compounds, respectively, with a grid spacing of 0.375 Å, using the AutoGrid program [48]. In addition, to predict atom-specific affinity maps, electrostatic, and desolvation potential maps, a distance-dependent dielectric function was assigned. Docking calculations were performed with the Lamarckian genetic algorithm LGA. 500 docking trials were defined as conformational searches and the root mean square (RMS) cluster tolerance was set at 2.0 Å. All other parameters were kept at their default settings [48].
To validate the docking protocol detailed above, we docked native inhibitors [5-(1naphthalen-1-yl-1,2,3-triazol-4-yl)thiophene-2-sulfonamide and 2-amino-2-hydroxymethylpropene-1,3-diol] of hCA isoforms. The positions and orientations occupied by the native ligands were compared to the conformations found in the X-ray structures available in the Protein Data Bank. The root mean square deviation (RMSD) between the experimental structure and the docked structure, and the docked free energy values returned by AutoDock were computed to assess the accuracy of the calculated poses.
In the case of our compounds (the organic ligands 2a/2b and their corresponding rhenium(I) complexes 3a/3b), the validated method was used to predict their binding conformation in the same regions of the hCA IX cavity occupied by the two co-crystallized ligands. The initial geometries of the synthesized compounds were taken from the Xray structures and were optimized by the DFT method to minimize energy using the GAUSSIAN-09 program [49]. In the case of the two rhenium compounds, DFT methods were used to give correct results for transition metal-containing systems [50]. Among the different density function methods, the hybrid B3LYP [51,52] proves to be a suitable choice for fac-Re(CO) 3 complexes specifically. Furthermore, the visualization program GaussView [53], included in GAUSSIAN-09, considers the coordination of the rhenium ring to two nitrogen atoms of the pyta chelating unit as rigid bonds. The Re and Cl atoms were assigned a charge of +1 and −1, respectively, the sulfonamide group was constructed in its negative charge (-SO 2 -NH ) and the coumarin unit was constructed in the hydrolyzed and native forms.
The synthesis of the targeted compounds 3a/3b is depicted in Scheme 1. The synthetic strategy adopted is similar to that developed recently by our group [24]. It consists of a three-step procedure. Firstly, two azido compounds 1a and 1b were synthesized from the commercial compounds 4-(2-aminoethyl)-benzenesulfonamide and 7-hydroxy-4methylumbelliferone, respectively, according to literature procedures [33,34]. These azido intermediates 1a/1b were combined with commercial 2-ethynylpyridine via Copper(I)catalyzed Alkyne-Azide Cycloaddition reaction to get click derivatives containing sulfonamide/coumarin moieties (2a and 2b). After purification, the pyta ligands 2a/2b were obtained in good yield (average yield of 81%). Both organic pyta compounds were then reacted with a slight excess of rhenium pentacarbonyl chloride in methanol at 65 • C over 12h to give the two tricarbonyl rhenium complexes 3a/3b in good yields (c.a. 83%).
derivatives have been easily prepared by a Copper-catalyzed Alkyne-Azide Cycloaddition reaction (CuAAC). The mild reaction conditions associated with the click approach allowed the preparation of a large number of mono-functionalized pyta derivatives in which 1,4-disubstituted 1,2,3-triazoles form an integral part of the metal chelating system (so called "click-to-chelate" concept) [64,65]. We were the first group to use this "click-tochelate" concept to develop tricarbonyl rhenium(I) complexes as carbonic anhydrase IX inhibitors [24]. It should be noted that while numerous 1,2,3-triazole-based CA inhibitors have been reported, the triazole ring (formed by the CuAAC reaction) only acts as a connector between the pharmacophore (coumarin/sulfonamide derivatives) and a metal complex or an organic moiety [66][67][68][69][70][71][72][73].
The synthesis of the targeted compounds 3a/3b is depicted in Scheme 1. The synthetic strategy adopted is similar to that developed recently by our group [24]. It consists of a three-step procedure. Firstly, two azido compounds 1a and 1b were synthesized from the commercial compounds 4-(2-aminoethyl)-benzenesulfonamide and 7-hydroxy-4methylumbelliferone, respectively, according to literature procedures [33,34]. These azido intermediates 1a/1b were combined with commercial 2-ethynylpyridine via Copper(I)catalyzed Alkyne-Azide Cycloaddition reaction to get click derivatives containing sulfonamide/coumarin moieties (2a and 2b). After purification, the pyta ligands 2a/2b were obtained in good yield (average yield of 81%). Both organic pyta compounds were then reacted with a slight excess of rhenium pentacarbonyl chloride in methanol at 65 °C over 12h to give the two tricarbonyl rhenium complexes 3a/3b in good yields (c.a. 83%). The synthesized compounds (ligands and complexes) are stable in air, either as a solid or in solution, as expected. They were fully characterized by 1 H NMR, 13 C NMR, IR and MS as well as single-crystal X-ray diffraction. Selected parameters of the crystal structures are listed in Table 1, and selected bond lengths and angles values can be found in the supplementary data (supplementary information, Tables S1 and S2). Our four pytabased compounds showed classic spectral features (experimental section and/or supplementary information for more detail). Briefly, both mono-functionalized pyridyl-1,2,3-triazole derivatives 2a/2b were crystallized in the orthorhombic and monoclinic system groups and their crystals were solved in space groups Pbca and P2 1 /c, respectively ( Figure  S2 in supplementary information for the molecular views of 2a and 2b). Structural features are very similar and compare well to those observed for related complexes, i.e., an anti arrangement adopted by the N1 and N2 atoms (2a), or the N3 and N4 atoms (2b) of the pyridine and triazole rings, respectively, and an azo character of the 1,2,3-triazole ring [24,[54][55][56][57][58][59][60][61][62][63]. Additionally, a significant contribution to crystal cohesion is provided by a Scheme 1. General synthetic scheme of compounds 2a, 2b and 3a, 3b.
The synthesized compounds (ligands and complexes) are stable in air, either as a solid or in solution, as expected. They were fully characterized by 1 H NMR, 13 C NMR, IR and MS as well as single-crystal X-ray diffraction. Selected parameters of the crystal structures are listed in Table 1, and selected bond lengths and angles values can be found in the Supplementary Data (Supplementary Information, Tables S1 and S2). Our four pyta-based compounds showed classic spectral features (experimental section and/or Supplementary  Information for more detail). Briefly, both mono-functionalized pyridyl-1,2,3-triazole derivatives 2a/2b were crystallized in the orthorhombic and monoclinic system groups and their crystals were solved in space groups Pbca and P2 1 /c, respectively ( Figure S2 in Supplementary Information for the molecular views of 2a and 2b). Structural features are very similar and compare well to those observed for related complexes, i.e., an anti arrangement adopted by the N1 and N2 atoms (2a), or the N3 and N4 atoms (2b) of the pyridine and triazole rings, respectively, and an azo character of the 1,2,3-triazole ring [24,[54][55][56][57][58][59][60][61][62][63]. Additionally, a significant contribution to crystal cohesion is provided by a network of intra-layer hydrogen bonds and π-π intermolecular hydrogen interactions for 2a (Table S3 in Supplementary Information).
According to the Infra-Red spectra of the two rhenium complexes 3a and 3b, the facial arrangement of the CO ligands in the complexes is confirmed by the pattern of the CO stretching frequencies, i.e., one sharp, intense absorption (2024 cm −1 for 3a and 2021 cm −1 for 3b) and a second broad, intense band in the 1894-1925 cm −1 region [74][75][76]. The average value of the CO stretching frequencies of 1948 cm −1 is consistent with those reported for other pyta ligands and indicates an overall donor strength similar to bipyridine ligands, as expected [74]. In the mass spectra, the prominent peak observed at m/z 656 for 3a and m/z 652 for 3b is assigned to [M + Na] + on the basis of the isotopic distribution of 185/187 Re. In the 1 H NMR spectra, the triazole proton signals were strongly downfield shifted compared to the spectra of the pyta ligands (0 .61 ppm for 3a vs. 2a and 0.67 ppm for 3b vs. 2b). Additionally, these spectroscopic assignments were confirmed by the solid-state structure of 3a/3b performed by X-ray crystallography (as mentioned previously, the selected parameters of the crystal structures are listed in Table 1, and selected bond lengths and angles values can be found in the Supplementary Information, Table S2). As is commonly found for tricarbonyl rhenium(I) complexes, both rhenium(I) centres present a distorted octahedral geometry, as highlighted in Figure 1 [54][55][56][57][58][59][60][61][62][63]74]. In the 'ReC 3 N 2 Cl' coordination sphere, the three carbon monoxide moieties are facially disposed to maximize Re-CO backbonding. The chelate ring is satisfactorily planar, with unexceptional bond lengths and bond angles close to those observed in similar complexes [Re(CO) 3 Cl(pyta derivative)] (the N1-C8-C9-N2 dihedral angles are 0.8(5) • for 3a and 2.0(2) • for 3b). The crystal-packing analysis proved that the molecular structures of both complexes 3a/3b are stabilized through a network of non-covalent interactions. A network of intermolecular hydrogen interactions was observed between the sulfonamide group and the rhenium core of the adjacent molecule for 3a, while more interactions were detected for the second complex 3b involving, in particular, the lattice DMSO molecules ( crystal-packing analysis proved that the molecular structures of both complexes 3a/3b are stabilized through a network of non-covalent interactions. A network of intermolecular hydrogen interactions was observed between the sulfonamide group and the rhenium core of the adjacent molecule for 3a, while more interactions were detected for the second complex 3b involving, in particular, the lattice DMSO molecules (Table S3 in

CA Inhibition Assays
Five compounds, azido intermediate 1b, both ligands 2a and 2b, and rhenium complexes 3a and 3b were tested in vitro for their inhibitory activity against the physiologically relevant hCA isoforms I, II, IX and XII by means of the stopped-flow carbon dioxide

CA Inhibition Assays
Five compounds, azido intermediate 1b, both ligands 2a and 2b, and rhenium complexes 3a and 3b were tested in vitro for their inhibitory activity against the physiologically relevant hCA isoforms I, II, IX and XII by means of the stopped-flow carbon dioxide hydration assay. Their activities were compared to the standard CAI acetazolamide (AAZ) ( Table 2). All compounds reported here were very efficient as hCA IX and XII inhibitors in the nanomolar range (except for compound 1b with a K i of 227 nM for hCA IX) with Ki values close to those of the reference AAZ. Among the sulfonamides derivatives, 2a presented the highest hCA IX and XII inhibitory activities, with K i of 11.7 and 9.8 nM, for hCA IX and hCA XII, respectively. Additionally, 2a strongly inhibited the cytosolic isoform hCA II with K i of 6.8 nM, while the inhibition of the cytosolic isoform hCA I was in the sub micromolar range (K i of 103.5 nM). The corresponding rhenium complex 3a showed a similar behaviour to 2a. By comparing these results with those obtained previously [24], several structure-activity properties can be established for compounds bearing an arylsulfonamide moiety. Firstly, a better inhibition profile is observed with ligands compared to their corresponding tricarbonyl rhenium analogues. In this study, 2a showed a 2.5-and 4.6-fold higher affinity against hCA IX and hCA XII, respectively, than its rhenium complex 3a, (K i (2a) = 11.7 and 9.8 nM vs. K i (3a) = 29.7 and 45.5 nM). In contrast, the presence of a metal centre increased selectivity against hCA IX and hCA XII compared to the off-target isoforms hCA I and hCA II. From Table 2, 3a exhibited hCA I/hCA IX and hCA I/hCA XII ratios of 29 and 19, respectively, compared to 9 and 10 for 2a. On the other hand, there is no clear linear relationship between the linker length (between the chelate part and the sulfonamide unit) and the inhibitory activity. While the introduction of a methylene group between the metal part and the sulfonamide unit had a slight negative effect on the inhibition values [24], a linker with an ethylene group increased the inhibitory activity. Thus, for hCA II, K i values are 4.6 nM (no linker, see ref. [24]), 56.1 nM (one -CH 2 -group, see ref. [24]) and 6.8 nM (-CH 2 CH 2 -group, compound 2a, Table 2). Similar behaviour has been reported recently for compounds with linkers with longer carbon chains, and has been explained in particular by a slight intra-molecular movement ability [75].
The tumour-associated isoforms hCA IX and hCA XII were selectively inhibited by all the coumarin derivatives, with inhibition constants ranging from 12.7 nM to 44.5 nM, while the hCA I and II isoforms were slightly inhibited (in the micromolar range), as expected. Additionally, the presence of the chelate unit had a positive effect on hCA IX and hCA XII inhibitions. 2b showed a 5-to 7-fold higher affinity against hCA IX and hCA XII, respectively, than the azido intermediate 1b, (K i (2b) = 44.5 nM and 12.7 nM vs. K i (1b) = 227.6 nM and 94.9 nM). The same trend was observed for rhenium complex 3b compared to 1b.
All of the non-metalated free ligands analyzed here showed better affinity for hCA isoforms I, II, IX and XII than their corresponding rhenium complexes, with the exception of complex 3b, whose inhibitory activity against hCA IX was very slightly superior to that of ligand 2b (K i (3b) = 34.9 nM vs. K i (2b) = 44.5 nM). Despite this slight decrease in affinity, rhenium derivatives seem more interesting because of their potential use as imaging agents (with technetium-99m analogues) for the in vivo detection of hCA IX-and hCA XII-expressing tumours in patients. Moreover, among the rhenium (or technetium-99m) complex-based CA inhibitors [23,76], coumarin derivative 3b showed one of the highest selectivities against hCA IX and hCA XII compared to the off-target isoforms hCA I and hCA II, making it a promising anti-cancer drug candidate (hCA I or hCA II/hCA IX and hCA I or hCA II/hCA XII ratios of 287 and 315 respectively).

Molecular Docking Studies
Molecular docking calculations were performed to predict and understand the binding modes of the synthesized inhibitors 2/3 to the hCA IX active site cavity. Since all synthesized compounds do not contain a sulfonamide moiety, we assumed that in addition to the direct interactions between the ligand and the zinc ion, the studied ligands could also interact with the pocket entrance of the enzyme away from the zinc ion. As previously reported by Maresca et al. [30][31][32], coumarin derivatives could hydrolyze to form 2-hydroxy-cinnamic acid derivatives, i.e., two possibilities could be found for each molecule, the closed or open coumarin form. In our calculations, we considered this possibility by distinguishing 2b and 3b in their non-hydrolyzed forms (Figure 2, compounds 1P-2b and 1P-3b) from their hydrolyzed forms (Figure 2, compounds 2P-2b and 2P-3b). To this end, the 5FL4 [43] and 5DVX [44] pdb codes were proposed as hCA IX adduct models for docking studies of synthesized sulfonamide (2a/3a) and coumarin (2b/3b) compounds, respectively. Using molecular docking processes with AutoDock software, 500 conformations were clustered for each case, while the preferred conformations were selected by examining the cluster with the lowest mean binding energy, the RMSD value between the X-ray and the docked structure, coordination, hydrogen bond interactions and hydrophobic contacts.

Molecular Docking Studies
Molecular docking calculations were performed to predict and understand the binding modes of the synthesized inhibitors 2/3 to the hCA IX active site cavity. Since all synthesized compounds do not contain a sulfonamide moiety, we assumed that in addition to the direct interactions between the ligand and the zinc ion, the studied ligands could also interact with the pocket entrance of the enzyme away from the zinc ion. As previously reported by Maresca et al. [30][31][32], coumarin derivatives could hydrolyze to form 2-hydroxy-cinnamic acid derivatives, i.e., two possibilities could be found for each molecule, the closed or open coumarin form. In our calculations, we considered this possibility by distinguishing 2b and 3b in their non-hydrolyzed forms (Figure 2, compounds 1P-2b and 1P-3b) from their hydrolyzed forms (Figure 2, compounds 2P-2b and 2P-3b). To this end, the 5FL4 [43] and 5DVX [44] pdb codes were proposed as hCA IX adduct models for docking studies of synthesized sulfonamide (2a/3a) and coumarin (2b/3b) compounds, respectively. Using molecular docking processes with AutoDock software, 500 conformations were clustered for each case, while the preferred conformations were selected by examining the cluster with the lowest mean binding energy, the RMSD value between the X-ray and the docked structure, coordination, hydrogen bond interactions and hydrophobic contacts.  (1A,1B,1P-2b and 1P-3b) and their hydrolysis products catalyzed by the CA enzyme (2A,2B,2P-2b and 2P-3b) (1A,1B,1P-2b and 1P-3b) and their hydrolysis products catalyzed by the CA enzyme (2A,2B,2P-2b and 2P-3b).

Binding Mode Interactions of Sulfonamide Compounds 2a and 3a with hCA IX
Both sulfonamide compounds, organic ligand 2a and its corresponding rhenium complex 3a, were docked into the same active hCA IX-pocket occupied by the native inhibitor (ligand ID: 9FK), and all rotational bonds were considered flexible. The top conformations generated for the docked ligands 2a and 3a in the hCA IX pocket were aligned and compared with the native ligand 9FK as illustrated in Figure 3. Generally, the binding mode obtained by docking for the two docked inhibitors 2a/3a shows that the benzene-sulfonamide moiety resembles the binding mode of ligands found in the crystal structure of 9FK-hCA IX complex (PDB ID: 5FL4). The deprotonated nitrogen of the sulfonamide tail in both cases was located close to the zinc atom, and was similar to that of the native ligand. Similarly, the Zn-N distance in the binding region of the CA-inhibitor complexes ranged between 2.18 and 2.25 Å (Table S4 in Supplementary Information), values which are close to those reported for other CA-inhibitor complexes containing similar inhibitors (1.95-2.57 Å) [77]. hibitor (ligand ID: 9FK), and all rotational bonds were considered flexible. The top conformations generated for the docked ligands 2a and 3a in the hCA IX pocket were aligned and compared with the native ligand 9FK as illustrated in Figure 3. Generally, the binding mode obtained by docking for the two docked inhibitors 2a/3a shows that the benzenesulfonamide moiety resembles the binding mode of ligands found in the crystal structure of 9FK-hCA IX complex (PDB ID: 5FL4). The deprotonated nitrogen of the sulfonamide tail in both cases was located close to the zinc atom, and was similar to that of the native ligand. Similarly, the Zn-N distance in the binding region of the CA-inhibitor complexes ranged between 2.18 and 2.25 Å (  yellow (for 3a) and aquamarine (for side amino acid residues); oxygen atoms in red; nitrogen atoms in blue; sulfur atoms in yellow; chlorine and rhenium atoms in green. Figure 3 reveals that inhibitors 2a and 3a were stabilized not only by H-bonding interactions, but also through several hydrophobic/hydrophilic interactions with side chain amino acid residues of the protein, depending on the chelate part of the inhibitors. For the metallic compound 3a, the nitrogen and oxygen atoms of the sulfonamide moiety formed two hydrogen bonds with Thr200 residue, while the aromatic ring formed hydrophobic interactions with the side chain Val121 and Leu199. Thus, the -CH2CH2-linker, which made a weak hydrophobic interaction with the end of Leu199, seems to give more flexibility to the Re-compound, allowing the accommodation of the fac-Re(CO)3 core near residues Thr201, Pro202 and Trp9. On the other hand, the best conformation of the organic compound 2a orients the arylsulfonamide moiety deep into the hCA IX binding cavity, in a similar mode to its corresponding metallic compound, forming three hydrogen bonds, two with the Thr200 residue and another one between the N-triazolyl atom and Gln92. This preferred conformation is chosen because of its shorter Zn-N distance of 2.18 Å , the lowest mean binding energy of −8.24 kcal mol −1 , and favourable van der Waals interactions with the most important side residues (e.g., His119, Trp210, His94, Leu199, Val121, Gln92).
From the obtained results, we can suggest that the shorter Zn-N distance and the additional H-bonding interaction, provided by the free triazolyl nitrogen of the pyta scaffold, reflect the nanomolar inhibition data found for the organic compound 2a against hCA-IX compared to its corresponding Re-based derivative 3a (Ki(2a) = 11.7 nM vs. Ki(3a) = 29.7 nM).  yellow (for 3a) and aquamarine (for side amino acid residues); oxygen atoms in red; nitrogen atoms in blue; sulfur atoms in yellow; chlorine and rhenium atoms in green. Figure 3 reveals that inhibitors 2a and 3a were stabilized not only by H-bonding interactions, but also through several hydrophobic/hydrophilic interactions with side chain amino acid residues of the protein, depending on the chelate part of the inhibitors. For the metallic compound 3a, the nitrogen and oxygen atoms of the sulfonamide moiety formed two hydrogen bonds with Thr200 residue, while the aromatic ring formed hydrophobic interactions with the side chain Val121 and Leu199. Thus, the -CH 2 CH 2 -linker, which made a weak hydrophobic interaction with the end of Leu199, seems to give more flexibility to the Re-compound, allowing the accommodation of the fac-Re(CO) 3 core near residues Thr201, Pro202 and Trp9. On the other hand, the best conformation of the organic compound 2a orients the arylsulfonamide moiety deep into the hCA IX binding cavity, in a similar mode to its corresponding metallic compound, forming three hydrogen bonds, two with the Thr200 residue and another one between the N-triazolyl atom and Gln92. This preferred conformation is chosen because of its shorter Zn-N distance of 2.18 Å, the lowest mean binding energy of −8.24 kcal mol −1 , and favourable van der Waals interactions with the most important side residues (e.g., His119, Trp210, His94, Leu199, Val121, Gln92).
From the obtained results, we can suggest that the shorter Zn-N distance and the additional H-bonding interaction, provided by the free triazolyl nitrogen of the pyta scaffold, reflect the nanomolar inhibition data found for the organic compound 2a against hCA-IX compared to its corresponding Re-based derivative 3a (K i (2a) = 11.7 nM vs. K i (3a) = 29.7 nM).

Binding Mode Interactions of Coumarin Compounds 2b and 3b with hCA IX
For the docked coumarin compounds 2b/3b, only the open form (2P-2b and 2P-3b, as highlighted in Figure 2) was discussed for each type within the hCA IX active site. According to Table S4 (Supplementary Information), the predicted hydrolyzed coumarinproducts have lowest mean binding energies, compared to their mimic closed forms, which means more flexibility for the 2-hydroxycinnamic acid moiety, and more interactions with the CA IX side residues.
For the hCA IX isoform, in vitro results showed that 3b exhibited a slight best inhibition compared to its organic derivative 2b (K i (3b) = 34.9 nM) vs. K i (2b) = 44.5 nM). In the docking results, we also found that the 2P-3b (and/or its closed form 1P-3b) compound has a promising docking score ( Supplementary Information, Table S4).
In Figure 4, the top conformations generated for the docked hydrolyzed compounds 2P-2b and 2P-3b were superimposed in the same hCA IX binding pocket and then compared with the native inhibitor 2B (Ligand ID: TE1). Docking analysis shows that the open form of organic compound 2P-2b, which allows for more direct interactions with the side residues, is linked by five H-bonds to residues Arg196, The205, Gln203 and Gln224. The side chain residues that are in close contact with 2P-2b are Arg194, Leu223 and Vall262 (see Figure 4). In other words, the hydrolyzed form of the metallic inhibitor 2P-3b shares a binding mode somewhat similar to that occupied by its organic mimic. In contrast, the metallic inhibitor 2P-3b exhibits additional polar/hydrophobic interactions between the fac-Re(CO) 3 unit and some residues such as His200, Leu331, Pro334 and Pro335. We can also observe an additional interaction between residue Asn198 and 2-hydroxy-cinnamic acid moiety, which is absent in the corresponding organic compound 2P-2b. For the docked coumarin compounds 2b/3b, only the open form (2P-2b and 2P-3b, as highlighted in Figure 2) was discussed for each type within the hCA IX active site. According to table S4 (supplementary information), the predicted hydrolyzed coumarinproducts have lowest mean binding energies, compared to their mimic closed forms, which means more flexibility for the 2-hydroxycinnamic acid moiety, and more interactions with the CA IX side residues.
For the hCA IX isoform, in vitro results showed that 3b exhibited a slight best inhibition compared to its organic derivative 2b (Ki(3b) = 34.9 nM) vs. Ki(2b) = 44.5 nM). In the docking results, we also found that the 2P-3b (and/or its closed form 1P-3b) compound has a promising docking score (supplementary information, table S4).
In Figure 4, the top conformations generated for the docked hydrolyzed compounds 2P-2b and 2P-3b were superimposed in the same hCA IX binding pocket and then compared with the native inhibitor 2B (Ligand ID: TE1). Docking analysis shows that the open form of organic compound 2P-2b, which allows for more direct interactions with the side residues, is linked by five H-bonds to residues Arg196, The205, Gln203 and Gln224. The side chain residues that are in close contact with 2P-2b are Arg194, Leu223 and Vall262 (see Figure 4). In other words, the hydrolyzed form of the metallic inhibitor 2P-3b shares a binding mode somewhat similar to that occupied by its organic mimic. In contrast, the metallic inhibitor 2P-3b exhibits additional polar/hydrophobic interactions between the fac-Re(CO)3 unit and some residues such as His200, Leu331, Pro334 and Pro335. We can also observe an additional interaction between residue Asn198 and 2-hydroxy-cinnamic acid moiety, which is absent in the corresponding organic compound 2P-2b.  2P-2b), yellow (for 2P-3b) and aquamarine (for side amino acid residues); oxygen atoms in red; nitrogen atoms in blue; sulfur atoms in yellow; chlorine and rhenium atoms in green. Furthermore, Figure 5 demonstrates that the increased affinity of the metallic compound 2P-3b towards hCA IX is clearly related to the molecule overall shape, where the contribution of fac-Re(CO)3 head plays a very important role in filling the active site entrance. In contrast, the organic mimic 2P-2b provides an elongated form, resulting in a less bulky ligand which cannot interact efficiently with side chain residues of the active site pocket.  2P-2b), yellow (for 2P-3b) and aquamarine (for side amino acid residues); oxygen atoms in red; nitrogen atoms in blue; sulfur atoms in yellow; chlorine and rhenium atoms in green. Furthermore, Figure 5 demonstrates that the increased affinity of the metallic compound 2P-3b towards hCA IX is clearly related to the molecule overall shape, where the contribution of fac-Re(CO) 3 head plays a very important role in filling the active site entrance. In contrast, the organic mimic 2P-2b provides an elongated form, resulting in a less bulky ligand which cannot interact efficiently with side chain residues of the active site pocket. Crystals 2021, 11, x FOR PEER REVIEW 15 of 19 Figure 5. View of the shape complementarities between hCA-IX active site and 2P-2b (left) and 2P-3b (right). The inhibitor atoms are shown as spheres and the enzyme is represented as a surface-model. Color Code: Carbon atoms in green (for 2P-2b), yellow (for 2P-3b); oxygen atoms in red; nitrogen atoms in blue; sulfur atoms in yellow; chlorine and rhenium atoms in green.

Conclusions
In brief, we report here the synthesis of two new pyridyl-triazole derivatives containing sulfonamide/coumarin moieties and their corresponding rhenium complexes as potent carbonic anhydrase inhibitors. All the newly synthetized compounds using a click chemistry strategy were fully characterized in solution (IR, NMR, MS) and solid state (X-ray). All bond length and angle values are in line with those reported for other pyta ligands and their corresponding rhenium complexes. In addition, the ethylene bridge keeps the sulfonamide or coumarin group at a large distance from the chelating part of the molecule, and prevents the folding of the pendant arm towards the metal centre, as expected.
The four compounds showed selective inhibitory activity in the high nanomolar range against both tumour-associated isoforms hCA IX and hCA XII (Ki values range from 9.8 nM to 44.5 nM). In terms of selectivity, coumarin derivative 2b and 3b showed high selectivities against hCA IX and hCA XII compared to the off-target isoforms hCA I and hCA II (ratio hCA I or hCA II/hCA IX range of 225 (2b) and 287 (3b) and ratio hCA I or hCA II/hCA XII of 787 (2b) and 315 (3b)). Despite this slight decrease in affinity against hCA XII, 3b seems more promising than 2b because of its potential use as an imaging agent (with its technetium-99m analogue) for the in vivo detection of cancer.
The biological results obtained here are well rationalized by the molecular docking studies (for hCA IX). For the sulfonamide derivatives, the additional H-bonding and van der Waals interactions provided by the two (free) nitrogens of the pyta unit explain the stronger inhibition of 2a against hCA IX compared to its corresponding metallic complex 3a. In contrast, the docking simulation revealed that coumarin-based rhenium complex 3b (in its hydrolyzed form 2P-3b) exhibited greater inhibition against hCA IX than its ligand 2b, due to the 3-dimensional architecture of the molecule, where the bulky of fac-Re(CO)3 head plays a very important role in filling the active site entrance. This finding is very encouraging for further developments in the field of rhenium(I) complexes of coumarin with CA inhibitory action. However, while the data (in vitro study and predictive docking experiments) indicated that 3b may be promising drug candidate, further biochemical research is required to establish the efficacy of this compound prior to further steps. In vivo studies will be conducted in the near future to obtain more information about its inhibitory power and cytotoxicity. Figure 5. View of the shape complementarities between hCA-IX active site and 2P-2b (left) and 2P-3b (right). The inhibitor atoms are shown as spheres and the enzyme is represented as a surface-model. Color Code: Carbon atoms in green (for 2P-2b), yellow (for 2P-3b); oxygen atoms in red; nitrogen atoms in blue; sulfur atoms in yellow; chlorine and rhenium atoms in green.

Conclusions
In brief, we report here the synthesis of two new pyridyl-triazole derivatives containing sulfonamide/coumarin moieties and their corresponding rhenium complexes as potent carbonic anhydrase inhibitors. All the newly synthetized compounds using a click chemistry strategy were fully characterized in solution (IR, NMR, MS) and solid state (X-ray). All bond length and angle values are in line with those reported for other pyta ligands and their corresponding rhenium complexes. In addition, the ethylene bridge keeps the sulfonamide or coumarin group at a large distance from the chelating part of the molecule, and prevents the folding of the pendant arm towards the metal centre, as expected.
The four compounds showed selective inhibitory activity in the high nanomolar range against both tumour-associated isoforms hCA IX and hCA XII (K i values range from 9.8 nM to 44.5 nM). In terms of selectivity, coumarin derivative 2b and 3b showed high selectivities against hCA IX and hCA XII compared to the off-target isoforms hCA I and hCA II (ratio hCA I or hCA II/hCA IX range of 225 (2b) and 287 (3b) and ratio hCA I or hCA II/hCA XII of 787 (2b) and 315 (3b)). Despite this slight decrease in affinity against hCA XII, 3b seems more promising than 2b because of its potential use as an imaging agent (with its technetium-99m analogue) for the in vivo detection of cancer.
The biological results obtained here are well rationalized by the molecular docking studies (for hCA IX). For the sulfonamide derivatives, the additional H-bonding and van der Waals interactions provided by the two (free) nitrogens of the pyta unit explain the stronger inhibition of 2a against hCA IX compared to its corresponding metallic complex 3a. In contrast, the docking simulation revealed that coumarin-based rhenium complex 3b (in its hydrolyzed form 2P-3b) exhibited greater inhibition against hCA IX than its ligand 2b, due to the 3-dimensional architecture of the molecule, where the bulky of fac-Re(CO) 3 head plays a very important role in filling the active site entrance. This finding is very encouraging for further developments in the field of rhenium(I) complexes of coumarin with CA inhibitory action. However, while the data (in vitro study and predictive docking experiments) indicated that 3b may be promising drug candidate, further biochemical research is required to establish the efficacy of this compound prior to further steps. In vivo studies will be conducted in the near future to obtain more information about its inhibitory power and cytotoxicity.  Table S4: Estimated free energy of binding (Kcal mol −1 ), Hbond interactions number and Zn-N distances (Å) in the interaction region of the HCA IX-inhibitor complexes, Figure S1: Alignment of isoforms IX of hCA (pdb codes: 5fl4 and 5dvx) within the active site of hCA II (pdb code: 3f8e), Figure S2: Molecular views of 2a (top) and 2b (bottom) with 50% ellipsoidal probability; hydrogen atoms have been omitted for clarity, except H on N5 atom (2a).