Aryl-4,5-dihydro-1H-pyrazole-1-carboxamide Derivatives Bearing a Sulfonamide Moiety Show Single-digit Nanomolar-to-Subnanomolar Inhibition Constants against the Tumor-associated Human Carbonic Anhydrases IX and XII.

A series of new 3-phenyl-5-aryl-N-(4-sulfamoylphenyl)-4,5-dihydro-1H-pyrazole-1-carboxamide derivatives was designed here, synthesized, and studied for carbonic anhydrase (CAs, EC 4.2.1.1) inhibitory activity against the human (h) isozymes I, II, and VII (cytosolic, off-target isoforms), and IX and XII (anticancer drug targets). Generally, CA I was not effectively inhibited, whereas effective inhibitors were identified against both CAs II (KIs in the range of 5.2–233 nM) and VII (KIs in the range of 2.3–350 nM). Nonetheless, CAs IX and XII were the most susceptible isoforms to this class of inhibitors. In particular, compounds bearing an unsubstituted phenyl ring at the pyrazoline 3 position showed 1.3–1.5 nM KIs against CA IX. In contrast, a subset of derivatives having a 4-halo-phenyl at the same position of the aromatic scaffold even reached subnanomolar KIs against CA XII (0.62–0.99 nM). Docking studies with CA IX and XII were used to shed light on the derivative binding mode driving the preferential inhibition of the tumor-associated CAs. The identified potent and selective CA IX/XII inhibitors are of interest as leads for the development of new anticancer strategies.


Introduction
Carbon dioxide (CO 2 ) plays a vital role in all life processes. It exists in equilibrium with bicarbonate, but the interconversion is a slow process [1]. Carbonic anhydrases (CAs, EC 4.2.1.1) are metalloenzymes that catalyze this reversible hydration reaction of carbon dioxide to bicarbonate and protons [1,2]. As a result, CAs play an important role in respiration, pH and CO 2 homeostasis, electrolyte secretion in a variety of tissues and organs, biosynthetic reactions (such as gluconeogenesis and lipid and urea synthesis), bone resorption, tumorigenicity, and many other physiological or pathological processes [3]. Fifteen different α-carbonic anhydrase isoforms have been isolated and characterized in humans and many of them are well-established therapeutic targets to treat a wide range of disorders [4]. Based on cellular and sub-cellular location, CAs are classified into four different groups: Cytosolic (CA I, II, III, VII, XIII), mitochondrial (CA VA, VB), secretory (CA VI), and membrane-associated (CA IV, IX, XII, XIV, XV) [5]. The active site of CA is a deep cleft, on one side of which only hydrophobic residues are present, whereas hydrophilic residues line the opposite side. The Zn 2+ ion is coordinated tetrahedrally by three imidazoles of conserved histidine residues (His94, His96 and His119), and a water molecule/hydroxide ion is a fourth ligand [2]. The membrane-bound CA IX and XII isoforms are known as the CAs associated with cancers, being expressed in a limited number of normal tissues [5][6][7][8][9]. Solid tumors usually have a dynamic microenvironment characterized by acidic pH and low levels of oxygen (i.e., hypoxia). For survival, cancer cells acquire different adaptive features, which are most probably responsible for the development of invasive and metastatic phenotypes. As a response to regional hypoxia, cancer cells obtain their need for high levels of ATP synthesis by switching their metabolism from aerobic respiration to fermentative (anaerobic) glycolysis [10]. CA IX and XII regulate tumor cell pH by maintaining an intracellular slightly alkaline pH (of 7.2-7.4) while acidifying the extracellular pH (which arrives at 6.0-6.5), and this is exploited by the cancer cells for survival and proliferation [6]. The hypoxic conditions in solid tumors induce overexpression of CA IX and CA XII, which maintain the acidic tumor environment by catalyzing the reversible hydration of tumor cell-generated CO 2 into a bicarbonate anion (HCO 3 − ) and a proton (H + ) and then trap H + extracellularly to lower the pH [5,6]. Hence, the overexpression of these isoforms contributes to the increased acidification of the extracellular hypoxic environment in contrast to normal tissues. Expression of both membrane-associated CAs is induced by hypoxia in breast tumors and in several cancer cell lines [7,8]. CA XII expression in breast tumors is indicative of lower grade disease, lower relapse rates, and better overall patient survival [9]. In contrast, CA expression in brain tumors is associated with poor prognosis [9,10]. Although inhibition of hCAs by aromatic and heterocyclic sulfonamides has been largely studied for 50 years with applications in the treatment of various disorders [2,4], these classes of CA inhibitors (CAIs) remain an attractive chemical family in the design of new inhibitors, as shown by the ongoing research in this field [11][12][13]. The primary sulfonamide group binds to the Zn 2+ ion in the CA enzyme active site as an anion (SO 2 NH − ) blocking the catalysis [11]. A wealth of recent studies has shown that the presence of urea as a linker in sulfonamide-containing compounds is beneficial to generating effective CA inhibitory activity [14,15]. In the present study, novel 3-phenyl-5-aryl-N-(4-sulfamoylphenyl)-4,5-dihydro-1H-pyrazole-1-carboxamide derivatives were designed and synthesized on the basis of sulfonyl semicarbazides and iminoureido CAIs previously synthesized in our laboratory [16,17]. In vitro inhibition studies were carried out for all newly synthesized compounds against CA I, II, VII, IX, and XII. The synthesized derivatives were thoroughly characterized by 1 H-NMR, 13 C-NMR, IR, and MS, and their purity was checked by HPLC.

Chemistry
In the present work, a new series of 3-phenyl-5-aryl-N-(4-sulfamoylphenyl)-4,5-dihydro-1H-pyrazole-1-carboxamide (5-25) is reported. Initially, 4-ureidobenzenesulfonamide 2 was obtained from reaction of sulfanilamide with sodium cyanate in ethanol and water, which was further refluxed with hydrazine hydrate in ethanol to obtain 4-aminosulphonylphenyl semicarbazide 3. Various chalcones of the general formula 4 were prepared by Claisen-Schmidt condensation of the appropriate 4-chloro/bromophenyl acetophenone with substituted aromatic/heteroaromatic aldehydes in the presence of NaOH. Cyclization of semicarbazide 3 with chalcones in the presence of a base afforded the corresponding 3-phenyl-5-aryl-N-(4-sulfamoylphenyl)-4,5-dihydro-1H-pyrazole-1-carboxamide (5-25) (Scheme 1), accomplishing the chemical hybridization of sulfanilamide with pyrazoline. The chemical structures of titled compounds 5-25 were established on the basis of spectroscopic analysis. In detail, compounds 5-25 were characterized by the presence of the three prominent doublets of doublets, each having the integration of one proton. Formation of the pyrazoline was confirmed by the presence of a the pyrazoline was confirmed by the presence of a doublet of doublets for CH2 at C4 giving a signal at δ 3.26-3.13 and δ 3.93-3.83 ppm, and CH at C5 giving a signal at δ 5.54-5.48 ppm. The general procedures for the synthesis and spectral data are described in the Experimental Section. SO

Carbonic Anhydrases Inhibition
Sulfonamides 5-25 were evaluated for their inhibition against the cytosolic CA I, II, and VII and the membrane-bound CA IX and XII by using a stopped-flow CO2 hydrase assay method [18]. The clinically used acetazolamide (AAZ) was used as standard drug in the kinetic evaluation. The following SAR can be worked out from the data reported in Table 1.

Carbonic Anhydrases Inhibition
Sulfonamides 5-25 were evaluated for their inhibition against the cytosolic CA I, II, and VII and the membrane-bound CA IX and XII by using a stopped-flow CO 2 hydrase assay method [18]. The clinically used acetazolamide (AAZ) was used as standard drug in the kinetic evaluation. The following SAR can be worked out from the data reported in Table 1.

Carbonic Anhydrases Inhibition
Sulfonamides 5-25 were evaluated for their inhibition against the cytosolic CA I, II, and VII and the membrane-bound CA IX and XII by using a stopped-flow CO2 hydrase assay method [18]. The clinically used acetazolamide (AAZ) was used as standard drug in the kinetic evaluation. The following SAR can be worked out from the data reported in Table 1. CA I was the least inhibited isoform by compounds 5-25 with inhibition constants (K I s) ranging between 36.2 and 769.6 nM. The 4-Br substitution as R (23-25) stood out as the most efficient for providing <100 nM K I s (i.e., 38.2-45.5 nM), though a comparison can be made on a limited set of Ar substituents. By contrast, when analog Ar groups are considered, a H atom (5-11) produced better CA I inhibition than a chlorine atom (12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22).
The most physiologically relevant isoform, CA II, was efficiently inhibited by most derivates with K I s nearby or even below 10 nM (5.2-13.5 nM), except for 5, 7, 12, and 15. The phenyl derivative 12 induced significantly worse CA II inhibition than most of its relative compounds (K I of 233 nM).
The last tested cytosolic CA, isoform VII, was even more affected than CA II, with the K I s compacting in a very narrow range spanning between 2.3 and 7.7 nM, comparably to the standard AAZ (K I of 2.5 nM). Compounds 5 and 17 (K I s of 33.2 and 27.6 nM) reported a 10-fold worse CA VII inhibition, whereas a drop in efficacy was measured for the 4-Cl derivative 12 (K I of 350 nM), which also exhibited a reduced action against CA II when compared to the other derivatives.
Undoubtedly, the activity of the tumor-associated CAs IX and XII was the most susceptible to inhibition by the sulfonamides 5-25 reported here. Most derivatives showed an extremely narrow range of inhibition of CA IX as K I s spanned in the range 1.3-2.9 nM. Again, compound 12 together with 14 and 24 exhibited slightly increased K I s (11.4-16.5 nM), whereas the 2-hydroxy derivative 20 turned out as the less active CA IX inhibitor in the study. It should be noted that compounds bearing an unsubstituted phenyl ring at the pyrazoline 3 position showed the best K I s against CA IX (1.3-1.5 nM).
Even subnanomolar inhibition was achieved against CA XII by some screened compounds (K I s of 0.62-0.99 nM), where another number oddly showed K I s settling around 1 nM. It is interesting to note that subnanomolar inhibitions are achieved uniquely by derivatives having a 4-halo-phenyl at the 3 position of the aromatic scaffold. The less active inhibitors 12 and 13 exhibited a K I of 7.7 nM. Further discussion on the SAR cannot be produced, because of the extremely narrow trend of inhibition measured for 5-25 against all CAs.
Derivatives 5-25 act as selective inhibitors against the tumor-associated CAs IX and XII over CA II and mostly CA I, with a selectivity index even reaching four-digit values (e.g., CA I/XII SI of compound 18).

Molecular Docking
The binding mode of sulfonamides 5-25 to CA IX (pdb 5FL4) [19] and XII (pdb 1JD0) [8], the most susceptible isoforms to this class of CAIs, was predicted by docking. The effectiveness of the adopted computational procedure was testified by a root-mean-square deviation (RMSD) of only 0.49 and 0.35 Å between docked and crystallographic poses of the ligands present in pdb 5FL4 and pdb 1JD0, respectively ( Figure 1). CA I was the least inhibited isoform by compounds 5-25 with inhibition constants (KIs) ranging between 36.2 and 769.6 nM. The 4-Br substitution as R (23-25) stood out as the most efficient for providing <100 nM KIs (i.e., 38.2-45.5 nM), though a comparison can be made on a limited set of Ar substituents. By contrast, when analog Ar groups are considered, a H atom (5-11) produced better CA I inhibition than a chlorine atom (12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22).
The most physiologically relevant isoform, CA II, was efficiently inhibited by most derivates with KIs nearby or even below 10 nM (5.2-13.5 nM), except for 5, 7, 12, and 15. The phenyl derivative 12 induced significantly worse CA II inhibition than most of its relative compounds (KI of 233 nM).
The last tested cytosolic CA, isoform VII, was even more affected than CA II, with the KIs compacting in a very narrow range spanning between 2.3 and 7.7 nM, comparably to the standard AAZ (KI of 2.5 nM). Compounds 5 and 17 (KIs of 33.2 and 27.6 nM) reported a 10-fold worse CA VII inhibition, whereas a drop in efficacy was measured for the 4-Cl derivative 12 (KI of 350 nM), which also exhibited a reduced action against CA II when compared to the other derivatives.
Undoubtedly, the activity of the tumor-associated CAs IX and XII was the most susceptible to inhibition by the sulfonamides 5-25 reported here. Most derivatives showed an extremely narrow range of inhibition of CA IX as KIs spanned in the range 1.3-2.9 nM. Again, compound 12 together with 14 and 24 exhibited slightly increased KIs (11.4-16.5 nM), whereas the 2-hydroxy derivative 20 turned out as the less active CA IX inhibitor in the study. It should be noted that compounds bearing an unsubstituted phenyl ring at the pyrazoline 3 position showed the best KIs against CA IX (1.3-1.5 nM).
Even subnanomolar inhibition was achieved against CA XII by some screened compounds (KIs of 0.62-0.99 nM), where another number oddly showed KIs settling around 1 nM. It is interesting to note that subnanomolar inhibitions are achieved uniquely by derivatives having a 4-halo-phenyl at the 3 position of the aromatic scaffold. The less active inhibitors 12 and 13 exhibited a KI of 7.7 nM. Further discussion on the SAR cannot be produced, because of the extremely narrow trend of inhibition measured for 5-25 against all CAs.
Derivatives 5-25 act as selective inhibitors against the tumor-associated CAs IX and XII over CA II and mostly CA I, with a selectivity index even reaching four-digit values (e.g., CA I/XII SI of compound 18).

Molecular Docking
The binding mode of sulfonamides 5-25 to CA IX (pdb 5FL4) [19] and XII (pdb 1JD0) [8], the most susceptible isoforms to this class of CAIs, was predicted by docking. The effectiveness of the adopted computational procedure was testified by a root-mean-square deviation (RMSD) of only 0.49 and 0.35 Å between docked and crystallographic poses of the ligands present in pdb 5FL4 and pdb 1JD0, respectively (Figure 1).  The in silico simulation was extended to both enantiomers (R) and (S) for each ligand. Coordination of the Zn ion from both isozymes occurs by the SO 2 NH − group of all ligands, which is further involved in two H-bonds with the OH group and backbone NH of T199 ( Figure 2). The aromatic core bearing the SO 2 NH 2 forms π-alkyl contacts with L198, H94, and V121. The in silico simulation was extended to both enantiomers (R) and (S) for each ligand. Coordination of the Zn ion from both isozymes occurs by the SO2NH -group of all ligands, which is further involved in two H-bonds with the OH group and backbone NH of T199 (Figure 2). The aromatic core bearing the SO2NH2 forms π-alkyl contacts with L198, H94, and V121.  Similarly, an almost opposite binding orientation was found for the molecular tail of the (R)and (S)-enantiomers of derivative 18 when bound to CA XII ( Figure 4B). By contrast, by a 120° rotation of the pyrazoline orientation, the 3-phenyl groups from the two enantiomers of compounds 6 and 21, as well as their 5-aryl groups, occupy the same area within the active site of CA XII ( Figure  4A,C). It can be speculated that the markedly less lipophilic character of the "hydrophobic region" of the CA XII active site with respect to CA IX (Figure 2) attracts the most hydrophilic pendant between those at the pyrazoline position 3 and 5, in the case where a significant difference among them exists (e.g., compounds 6 and 21). The in silico simulation was extended to both enantiomers (R) and (S) for each ligand. Coordination of the Zn ion from both isozymes occurs by the SO2NH -group of all ligands, which is further involved in two H-bonds with the OH group and backbone NH of T199 (Figure 2). The aromatic core bearing the SO2NH2 forms π-alkyl contacts with L198, H94, and V121.  Similarly, an almost opposite binding orientation was found for the molecular tail of the (R)and (S)-enantiomers of derivative 18 when bound to CA XII ( Figure 4B). By contrast, by a 120° rotation of the pyrazoline orientation, the 3-phenyl groups from the two enantiomers of compounds 6 and 21, as well as their 5-aryl groups, occupy the same area within the active site of CA XII ( Figure  4A,C). It can be speculated that the markedly less lipophilic character of the "hydrophobic region" of the CA XII active site with respect to CA IX (Figure 2) attracts the most hydrophilic pendant between those at the pyrazoline position 3 and 5, in the case where a significant difference among them exists (e.g., compounds 6 and 21). Similarly, an almost opposite binding orientation was found for the molecular tail of the (R)-and (S)-enantiomers of derivative 18 when bound to CA XII ( Figure 4B). By contrast, by a 120 • rotation of the pyrazoline orientation, the 3-phenyl groups from the two enantiomers of compounds 6 and 21, as well as their 5-aryl groups, occupy the same area within the active site of CA XII ( Figure 4A,C). It can be speculated that the markedly less lipophilic character of the "hydrophobic region" of the CA XII active site with respect to CA IX (Figure 2) attracts the most hydrophilic pendant between those at the pyrazoline position 3 and 5, in the case where a significant difference among them exists (e.g., compounds 6 and 21).
It can also be noted that the ureido carbonyl group of the (R)-enantiomers can accept a H-bond from Q92 side chain CONH 2 within both CAs active sites (Figures 3 and 4). It can also be noted that the ureido carbonyl group of the (R)-enantiomers can accept a H-bond from Q92 side chain CONH2 within both CAs active sites (Figures 3 and 4).
The binding modes depicted in Figures 2-4 are expected to be hindered within the CA II active site where the mutation of V/A131 to F131 significantly narrows the volume of the cavity. As a result, the CA II inhibition efficiency of 5-25 drops when compared to the tumor-related CAs, and so that against CA I does so because of the tight active site of such an isoform. According to the docking scores computed for both optical isomers of compounds 5-25 reported in Table 2, the role of eutomers might be ascribed to (S)-enantiomers in the binding to CA IX (GlideScore -5.86-7.01 kcal/mol) and to (R)-enantiomers in the binding to CA XII (GlideScore -6.48-7.92 kcal/mol). No further comparison can be made between Glide-predicted scores and KIs determined in vitro. In fact, the GlideScore function is assumed to distinguish compounds that bind strongly from those that do not, and compounds 5-25 all act as low nanomolar CA IX and XII inhibitors (and even nanomolar CAIs against isoforms I and II). The docking scores computed for AAZ when bound within CA IX and XII active sites are indeed comparable to those of the eutomers of 5-25. Table 2. Docking score (Glide) of the predicted poses of (R)-and (S)-enantiomers of 5-25 and AAZ bound to CA IX (pdb 5FL4) and CA XII (pdb 1JD0).

Cmpd
GlideScore (  The binding modes depicted in Figures 2-4 are expected to be hindered within the CA II active site where the mutation of V/A131 to F131 significantly narrows the volume of the cavity. As a result, the CA II inhibition efficiency of 5-25 drops when compared to the tumor-related CAs, and so that against CA I does so because of the tight active site of such an isoform. According to the docking scores computed for both optical isomers of compounds 5-25 reported in Table 2, the role of eutomers might be ascribed to (S)-enantiomers in the binding to CA IX (GlideScore -5.86-7.01 kcal/mol) and to (R)-enantiomers in the binding to CA XII (GlideScore -6.48-7.92 kcal/mol). No further comparison can be made between Glide-predicted scores and K I s determined in vitro. In fact, the GlideScore function is assumed to distinguish compounds that bind strongly from those that do not, and compounds 5-25 all act as low nanomolar CA IX and XII inhibitors (and even nanomolar CAIs against isoforms I and II). The docking scores computed for AAZ when bound within CA IX and XII active sites are indeed comparable to those of the eutomers of 5-25. Table 2. Docking score (Glide) of the predicted poses of (R)and (S)-enantiomers of 5-25 and AAZ bound to CA IX (pdb 5FL4) and CA XII (pdb 1JD0).

Conclusions
A series of new 3-phenyl-5-aryl-N-(4-sulfamoylphenyl)-4,5-dihydro-1H-pyrazole-1-carboxamide derivatives was designed here, synthesized, and studied for hCA inhibitory activity against the cytosolic, off-target isoforms I, II, and VII and the anticancer drug targets IX and XII. While CA I was not effectively inhibited, effective inhibitors were identified against both CAs II (K I s in the range of 5.2-233 nM) and VII (K I s in the range of 2.3-350 nM). Nonetheless, CAs IX and XII were the most susceptible isoforms to this class of inhibitors. In particular, compounds bearing an unsubstituted phenyl ring at the pyrazoline 3 position showed 1.3-1.5 nM K I s against CA IX. By contrast, a subset of derivatives having a 4-halo-phenyl at the same position of the aromatic scaffold even reached subnanomolar K I s against CA XII (0.62-0.99 nM). Docking studies with CA IX and XII, performed on both enantiomers of the ligands, showed the binding mode driving such a preferential inhibition of the tumor-associated CAs. The identified potent and selective CA IX/XII inhibitors could represent interesting leads for the development of new anticancer medications.

Chemistry
All the reagents and solvents were obtained from commercial suppliers and were used as received unless otherwise indicated. Solvents were dried, wherever necessary, according to standard procedures. All reactions were performed under N 2 atmosphere, unless otherwise indicated. Analytical silica gel 60 F 254 -coated TLC plates were purchased from Sigma-Aldrich and were visualized with UV light. IR spectra (ATR) were recorded on a Quest ATR Diamond Accessory (Black) P31482 and Shimadzu 8100 infrared spectrophotometer. 1 H-NMR was recorded at 300 MHz in DMSO-d 6 solvents using TMS as an internal reference standard at Sophisticated Analytical Instrument Facility (SAIF) IIT Powai, Mumbai. Molecular ion peaks of synthesized compounds were recorded using HRLCMS at Laxai-Avanti Life Sciences Pvt. Ltd. Hyderabad. Melting points were recorded using a Veego ® (VMP)-D capillary melting point apparatus (Veego Instruments Corp. Mumbai, India) and were uncorrected.