Inhibition studies on a panel of human carbonic anhydrases with N1-substituted secondary sulfonamides incorporating thiazolinone or imidazolone-indole tails

Abstract Being the primary sulfonamide among the most efficient zinc binding group (ZBG) to design inhibitors for the metallo-enzymes carbonic anhydrases (CA, EC 4.2.1.1), herein, we propose an investigation on four physiologically important human (h) CAs (hCA I, II, IV, and IX) with N1-substituted secondary sulfonamides incorporating thiazolinone or imidazolone-indole tails. The effect of the functionalisation of the sulfonamide group with five different substitution patterns, namely acetyl, pyridine, thiazole, pyrimidine, and carbamimidoyl, was evaluated in relation to the inhibition profile of the corresponding primary sulfonamide analogues. With most of these latter being nanomolar inhibitors of all four considered isoforms, a totally counterproductive effect on the inhibition potency can be ascribed to N1-functionalisations of the ZBG primary sulfonamide structure with pyridine, thiazole, and pyrimidine moieties. On the other hand, incorporation of less hindered groups, such as sulfonylacetamides and sulfonylguanidines, maintained a certain degree of activity dependent on the tailing moiety, with KIs spanning in the low micromolar range.


Introduction
Primary sulfonamide is the most efficient zinc binding group (ZBG) to design inhibitors for the metallo-enzymes carbonic anhydrases (CA, EC 4.2.1.1), being its structural features ideal for the binding to the Zn 2þ ion present at the bottom of the active site cavity and the residues nearby 1-3 . The negatively charged nitrogen of SO 2 NH À coordinates the positively charged metal ion replacing the physiological zinc-bound nucleophile, with the proton on the coordinated nitrogen atom being at H-bond distance to Thr199 OG1 atom, which act as acceptor (Figure 1) 1 .
Benzenesulfonamides constitute the most common and best characterised class of CA inhibitors (CAIs) 3 . The presence of an aromatic/heteroaromatic scaffold bearing the sulfonamide group further stabilise the enzyme-ligand complex, by several Van der Waals interactions taking place with residues Gln92, Val121, Phe131, Leu198, and Thr200 (hCA II binding site (Figure 1)) 1,3 .
All CAs found in humans belong to the a-class (a-CA) and are characterised as sixteen isoforms, which differ by molecular features, oligomeric arrangement, cellular localisation, distribution in organs and tissues, expression levels, and kinetic properties 1,2 . Abnormal levels or activities of most sixteen hCA isoforms have been often associated with different human diseases, making these isozymes of great interest for the design of inhibitors which selectively target specific isoforms, in order to reduce the overall side effects exhibited by most non-isoform selective CAIs clinically used up to now 1-4 . Most efforts to design isoform selective CAIs have been pursued by modulating the ring directly linked to the ZBG (ring approach) 3,5 or appending different tails to the aromatic ring just mentioned (tail approach) 3,5-7 .
The previously accepted idea 8 that only primary sulfonamides may act as effective inhibitors has been shown to be incorrect, but only few modifications were carried out on the sulfonamide moiety 1 a, in comparison to the high number of tails appended at the aromatic ring incorporating this ZGB 6,3, 9,10 . The rational of this choice could be that of not losing the ideal features of primary sulfonamides, i.e. the negative charge on nitrogen and the presence of a proton on it, which are both important for the inhibition mechanism 1-3 . Conversely, it should be taken into account that N 1 -substitution with proper functional moieties could also elicit novel inhibition mechanisms 1 .
Secondary sulfonamides maintain the possibility to coordinate the Zn ion in the deprotonated form, as confirmed by X-ray crystallography. Di Fiore et al. demonstrated that N-hydroxy and Nmethoxy sulfonamides are able to bind to the Zn(II) ion in the deprotonated form 11 . An analogous inhibition mechanism has been identified for the cyclic secondary sulfonamide saccharin and its derivatives 12 . Likewise, we previously reported that the incorporation of a nitro-group at the N 1 -atom of aromatic sulfonamide moieties does not deprive them of CA inhibitory efficacy, although the precise inhibitory mode has not been clarified yet 13 .
Tertiary sulfonamides lose their acid character, paving the way to a totally new inhibition mechanism. Carradori et al. 14,15 recently synthesised different series of tertiary sulfonamides, namely probenecid and saccharin derivatives, obtaining interesting and selective inhibition profile. However, the inhibition mechanism with such compounds was not yet been elucidated. A plethora of series of benzenesulfonamide bearing indole or thiazolidinone moieties have also been reported so far and demonstrated to possess an effective inhibitory profile against a wide panel of hCAs [16][17][18][19][20] .
In an earlier investigation of some of us, a series of benzenesulfonamides linked to a chromone nucleus through thiazolinone and imidazolone spacers demonstrated excellent inhibitory activity against a panel of hCAs in a nanomolar range 21 . In continuation of the previous work, and with the aim to pursue the identification of new potent and selective CAIs, herein we explore the effect of the functionalisation of the sulfonamide group on the inhibitory properties by appending several moieties, namely acetyl, pyridine, thiazole, pyrimidine, and carbamimidoyl, at the N 1 -atom. A consistent series of imine-, thiazolinone-, and imidazolone benzenesulfonamides appending an indole ring instead of the chromone nucleus (from our earlier work) was thus obtained. All the synthesised derivatives were tested in vitro in order to evaluate their inhibition profiles against four physiologically important hCAs, i.e. hCA I, II, IV, and IX. Comparison of the inhibitory efficacy of primary sulfonamide derivatives with the N 1 -substituted analogues further strengthened the role of the sulfonamide in the pharmaceutical field related to CAs.

Chemistry
Melting points were recorded on a Stuart SMP10 digital melting point apparatus and were uncorrected. Infrared (IR) spectra were recorded as KBr disks using a Shimadzu FT-IR 8400S infrared spectrophotometer. Mass spectral data are given as m/z (intensity %). 1 H NMR spectra were recorded on either on a Varian Mercury VX-300 MHz spectrophotometer or Bruker AVANCE III Nano Bay 400 MHz FT-NMR spectrophotometer. 13 C NMR spectra were run at 100 MHz in deuterated dimethylsulfoxide (DMSO-d 6 ) on Bruker AVANCE III Nano Bay 400 MHz FT-NMR spectrophotometer. Chemical shifts are expressed in d values (ppm) using the solvent peak as internal standard. All coupling constant (J) values are given in Hz. The abbreviations used are as follows: s, singlet; d, doublet; t, triplet; m, multiplet. Elemental microanalyses were carried out at the Regional Center for Mycology and Biotechnology, Al-Azhar University, Egypt. Reaction were routinely monitored by Thin Layer Chromatography (TLC) on silica gel precoated F254 Merck plates. Unless otherwise noted, all solvents and reagents were commercially available and used without further purification. Compounds 2 22 , 4a-f 23,24 , 6a-f 23,24 , 7a 25 , 8a,b 26,27 and 9a 28 were prepared according to the reported procedures.
General procedure for the preparation of compounds 4a-f A mixture of indole-3-carboxaldehyde 2 (1.45 g, 10 mmol) and the appropriate sulfonamide 3a-f (10 mmol) were grinded and mixed together using a mortar and a pestle. The mixture was fused on a watch glass using drops of glacial acetic acid. The resulted paste was washed with ethanol/water mixture then crystallised from ethanol to afford compounds 4a-f.      Ar-H, J ¼ 9.6), 6.86 (t, 1H, Ar-H, J ¼ 6.2), 7.04 (d, 2H, Ar-H, J ¼ 8.

General procedure for the preparation of compounds 7a-f
A mixture of indole-3-carboxaldehyde 2 (1.45 g, 10 mmol), the appropriate thiazolinone derivative 6a-f (10 mmol) and fused sodium acetate (0.82 g, 10 mmol) in glacial acetic acid (50 ml) was refluxed for 12 h. The separated solid was filtered off while hot, washed with hot water and crystallised from DMF/H 2 O.
A mixture of indole-3-carboxaldehyde 2 (1.45 g, 10 mmol), N-(4methoxybenzoyl)glycine 8 b (2.09 gm, 10 mmol) and fused sodium acetate (0.5 g, 6 mmol) in acetic anhydride (20 ml) was heated in a boiling water bath for 5 h. The mixture was cooled and ethanol (20 ml) was added slowly and allowed to stand overnight in the refrigerator. The crystalline product was filtered off, washed with hot water and recrystallised from benzene to give compound 9 b.
Buff General procedure for the preparation of compounds 10a-l An equimolar amount of compound 9a,b (10 mmol) and the appropriate sulfonamide 3a-f (10 mmol) in glacial acetic acid (20 ml) containing freshly fused sodium acetate (0.03 g, 0.36 mmol) was heated in a boiling water bath with constant stirring for 10-14 h. The separated solid was filtered off, washed with water, and crystallised from DMF/water to give compounds 10a-l.

Carbonic anhydrase inhibition assay
An applied photophysics stopped-flow instrument has been used for assaying the CA catalysed CO 2 hydration activity 29 . 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. The CO 2 concentrations ranged from 1.7 to 17 mM for the determination of the kinetic parameters and inhibition constants. For each inhibitor at least six traces of the initial 5-10% of the reaction have been used for determining the initial rate. The uncatalysed rates were determine d in the same manner and subtracted from the total observed rates. Stock solutions of inhibitor (0.1 mM) were prepared in distilled-deionised water and dilutions up to 0.01 nM were done thereafter with the assay buffer. Inhibitor and enzyme solutions were preincubated together 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 non-linear least-squares methods using PRISM 3 and the Cheng-Prusoff equation, as reported earlier [30][31][32][33] , and represent the mean from at least three different determinations. All CA isoforms were recombinant ones obtained in-house as reported earlier [34][35][36] .

Chemistry
The target compounds 4a-f, 7a-f, and 10a-l were synthesised as illustrated in Schemes 1-3. The key stating material, indole-3-carboxaldehyde 2, was prepared via Vilsemier-Hack formylation reaction, as reported in literature 22 . IR spectrum of compound 2 revealed the presence of (NH) stretching band at 3363 cm À1 along with a (C ¼ O) stretching band at 1633 cm À1 . In addition, 1 H NMR showed a singlet signal at 9.94 ppm attributed to the proton of the (CH ¼ O) group and a D 2 O exchangeable signal of the (NH) proton at 12.15 ppm. Schiff's bases 4a-f were obtained via the condensation of the aldehyde compound 2 with different sulfonamide derivatives 3a-f. The reported method for the preparation of these compounds failed to give the expected compounds, and in some derivatives a very poor yield was obtained 37 . Searching for an alternative method, Naqvi et al. 38 reported an energy efficient greener methodology for the preparation of Schiff's bases involving a mechano-chemical solvent-free procedure. Applying this new methodology afforded compounds 4a-f in good yields and high purity in a short reaction time. IR spectra of compounds 4a-f were characterised by the disappearance of the (C ¼ O) stretching band of the aldehyde 2 and broadening of the (NH) stretching band at 3350-3200 cm À1 due to the additional (NH) groups. Also, two characteristic stretching bands corresponding to the (SO 2 ) group were identified at 1334-1311 and 1149-1126 cm À1 . 1 H NMR of compounds 4a-f displayed the presence of sharp singlet signal at 9.86-9.93 ppm attributed to the azomethine proton (CH ¼ N) in addition to the (NH) singlet signals, all disappeared by D 2 O exchange. As for compound 4 b; a singlet signal at 2.06 ppm integrated for the three protons of the (CH 3 ) group in 1 H NMR spectrum and a signal at 27.1 ppm in 13 C NMR spectrum confirmed its structure. In addition, 13 C NMR of compounds 4c-f adopted the expected pattern of their carbon content. Mass spectra of compounds 4a-c gave m/z at 299.0, 341.42 and 377.42 corresponding to their molecular weights, respectively.
The thiazolinone derivatives 7a-f were synthesised according to Scheme 2. The appropriate sulfonamide derivatives 3a-f were reacted with chloroacetyl chloride at r.t. in DMF to give the intermediate chloroacetyl derivatives 5a-f which were then intramolecularly cyclised by reflux with ammonium thiocyanate in ethanol, to obtain the thiazolinone compounds 6a-f 23,24 . Knoevenagel condensation of the aldehyde 2 with the active methylene group in compounds 6a-f afforded the newly synthesised compounds 7a-f. IR spectra of compounds 7a-f were consistent with their proposed structures where broad stretching bands appeared at 3433-3199 cm À1 corresponding to the (NH) groups. The (C ¼ O) band of the thiazolinone ring appeared at 1697-1627 cm À1 with an additional (C ¼ O) band at 1701 cm À1 in compound 7 b. The two (SO 2 ) stretching bands of all derivatives of the series were displayed at 1334-1311 and 1151-1126 cm À1 . Furthermore, compounds 7a-f showed characteristic 1 H NMR signal in the region 7.78-8.87 ppm corresponding to the alkene (CH¼) proton accompanied with the absence of the singlet signal corresponding to the aliphatic (CH 2 ) protons with the usual pattern of the protons of both indole and benzene sulfonamide rings. The three protons of the acetamide group of compound 7 b appeared as singlet signal at 1.94 ppm. 13 C NMR spectra of compounds 7a and 7f were consistent with their carbon skeleton.
Scheme 3 illustrated the synthesis of the imidazolinone derivatives 10a-l. The oxazolone intermediates 9a 28 and 9 b were synthesised by Erlenmeyer reaction of hippuric acid or 4-methoxy hippuric acid with indole-3-carboxaldehyde 2 in acetic anhydride and fused sodium acetate. Interestingly, the IR spectra of compounds 9a,b explored two (C ¼ O) stretching bands for each at 1634, 1689 and 1788, 1743 cm À1 , respectively, assigned for the Nacetyl and oxazolone carbonyl groups, respectively. This confirmed that the (NH) group of the indole ring was acetylated under the

Carbonic anhydrase inhibitory activity
The CA inhibitory activities of compounds 4a-f, 7a-f, and 10a-l in addition to acetazolamide (AAZ) as standard inhibitor, were measured against hCAI, hCA II, hCA IV and hCA IX by a stopped flow CO 2 hydrase assay 29 . The rational for the choice of these four isoforms is that hCA II and IV are targets for antiglaucoma drugs 1-3 , whereas hCA IX have been validated as targets for the treatment and prognosis of hypoxic cancers 39,40 . Otherwise hCA I is one of the main off-target isoforms both for the antiglaucoma or anticancer CAIs therapeutic application 1,3 .
The following structure-activity-relationships (SARs) were obtained from the inhibition data reported in Table 1.
Scheme 3. Synthesis of compounds 10a-l. Reagents and reaction conditions: (i) Glycine, 10% sodium hydroxide, ice bath, (0 C); (ii) Indole-3-carboxaldehyde 2, acetic anhydride, fused sodium acetate, (100 C); (iii) The appropriate sulfonamide 3a-f, glacial acetic acid, fused sodium acetate, (100 C). (i) As expected the inhibitory profile of the tested compounds against the four CA isoforms was strictly dependent on the functionalisation mode of the sulfonamide group, with the primary SO 2 NH 2 bearing derivatives (4a, 7a, 10a, and 10 g) arisen as the best hCA inhibitors herein reported. Sulfonylacetamido (4 b, 7 b, 10 b, and 10 h) and sulfonylguanidino (4c, 7c, 10c, and 10i) compounds maintained a certain degree of activity, dependent on the tailing moiety and the considered isozymes. Conversely, a totally counterproductive effect on the inhibition potency can be ascribed to all other N 1 -functionalisations, with all following considerations regarding uniquely the active ZBG bearing-derivatives.
(ii) The cytosolic isoform hCA I was significantly inhibited by all primary sulfonamide compounds with K I s in the range of 88.5-96.6 nM. It is pertinent to mention that the corresponding sulfonylguanidine (10f) and sulfonylacetamide (10 h) derivatives lose to a great extent their efficacy, showing a K I of 5764.1 and of 5719.1 nM, respectively.
(iii) hCA II was the most inhibited isoform, among those considered here by the primary sulfonamide derivatives (4a, 7a, 10a, and 10 g) with K I s value 575.8, 76.4, 56.4, 602.3 nM, respectively. In analogy to the SAR reported for the previous isoform, it is interesting to note that the nature of the tail appended at the benzenesulfonamide scaffold greatly affected the inhibition of this isoform in the sulfonylacetamide subset, being the imidazolones (10 b) and (10 h) medium nanomolar inhibitors (K I 487.3 and 548.9 nM), and imine (4 b) and thiazolinone (7 b) devoid of any efficacy.
(iv) Uniquely the primary sulfonamides (4a, 7a, 10a, and 10 g) exhibited significant activity against the membrane-associated isoform IV (K I s spanning between 262.7 and 4429.3 nM), with exception of sulfonylacetamide (10 h) which showed a micromolar inhibition with a K I of 3048.1 nM.
(v) The general tendencies described above were also applicable to the tumour-associated isoform hCA IX, which was moderately inhibited by the primary sulfonamides (4a, 7a, 10a, and 10 g) with binding affinities ranging between 308.0 and 328.0 nM. Sulfonylacetamido (4 b, 7 b, 10 b, and 10 h) and sulfonylguanidino (4c, 7c, 10c, and 10i) compounds were shown to possess comparable effects on hCA IX, inhibiting it in the low micromolar range (K I s of 778.3-1615.9 nM and 1206.6-1636.4 nM).
(vi) The inhibition data reported in Table 1 unexpectedly highlighted a total loss of efficacy upon functionalisation of the ZBG primary sulfonamide structure by pyridine, thiazolinone and pyrimidine moieties, at least within the present set of imine-, thiazolidinoneand imidazolone-indole benzenesulfamides. On the other hand, incorporation of less hindered groups as in case of sulfonylacetamides and sulfonylguanidines maintained a certain degree of activity dependent on the tailing moiety and the considered isozymes.
(vii) Comparing the results of the current study with the earlier investigation in which a chromone nucleus was used instead of the indole ring 21 , revealed that the chromone derivatives showed excellent activity profile despite being secondary sulfonamides and bearing similar N 1 -substitutions. So, it can be claimed that activity is not solely linked to the absence or presence of the N 1substitution, but also affected by other distant fragments in the molecule (chromone vs. indole). In other words, the presence of secondary sulfonamide moiety is not to be exclusively accused of the dramatic decrease in activity of the indole derivatives compared to their chromone counterparts.

Conclusions
Since the primary sulfonamide is the most efficient zinc binding group (ZBG) to design inhibitors for the metallo-enzymes CAs, herein, we propose an investigation on four physiologically important human (h) CAs (I, II, IV, and IX) of N 1 -substituted secondary sulfonamides incorporating a thiazolinone or imidazoloneindole scaffold. The effect of functionalisation of the sulfonamide group with five different substitution patterns, namely acetyl, pyridine, thiazole, pyrimidine and carbamimidoyl, was evaluated in relation to the inhibition profile of the primary sulfonamide analogues. With most of these latter compounds acting as nanomolar inhibitors of all four considered isoforms, a total loss of inhibitory efficacy upon functionalisation of the ZBG primary sulfonamide structure arose in case of incorporation of pyridine, thiazole and pyrimidine moieties. On the other hand, incorporation of less hindered groups, such as sulfonylacetamide (4 b, 7 b, 10 b, and 10 h) and sulfonylguanidino (4c, 7c, 10c, and 10i) maintained a certain degree of activity dependent on the tailing moiety and the considered isozymes, with K I s spanning in the low micromolar range. Nevertheless, it is worth to be mentioned that the same functionalisation of the sulfonamide group but with chromone scaffolds, led to potent inhibitors, in the nanomolar range as revealed in a previous report 21 . This drew our attention to the role of the indole nucleus as a key fragment responsible for switching the activity of this type of compounds.