Aryl derivatives of 3H-1,2-benzoxathiepine 2,2-dioxide as carbonic anhydrase inhibitors

Abstract A new series of homosulfocoumarins (3H-1,2-benzoxathiepine 2,2-dioxides) possessing various substitution patterns and moieties in the 7, 8 or 9 position of the heterocylic ring were prepared by original procedures and investigated for the inhibition of four physiologically relevant carbonic anhydrase (CA, EC 4.2.1.1) isoforms, the human (h) hCA I, II, IX and XII. The 8-substituted homosulfocoumarins were the most effective hCA IX/XII inhibitors followed by the 7-substituted derivatives, whereas the substitution pattern in position 9 led to less effective binders for the transmembrane, tumour-associated isoforms IX/XII. The cytosolic isoforms hCA I and II were not inhibited by these compounds, similar to the sulfocoumarins/coumarins investigated earlier. As hCA IX and XII are validated anti-tumour targets, with one sulphonamide (SLC-0111) in Phase Ib/II clinical trials, finding derivatives with better selectivity for inhibiting the tumour-associated isoforms over the cytosolic ones, as the homosulfocoumarins reported here, is of crucial importance.


Introduction
Carbonic anhydrases (CAs, EC 4.2.1.1) are metalloenzymes widespread in nature, being encoded by at least eight different genetic families, which have been identified in organisms all over the phylogenetic tree 1-3 . By catalysing a crucial physiologic reaction, by which CO 2 is hydrated with the formation of a weak base (bicarbonate) and a strong acid (hydronium ions), these enzymes are involved in a multitude of physiologic processes, starting with pH regulation and ending with metabolism 1,3-6 . As thus, CAs are drug targets for decades, with their inhibitors having pharmacological applications in a multitude of fields 1,3-5 . The primary sulphonamides were discovered as CA inhibitors (CAIs) in the 40 s, and most of the drugs that were launched in the next decades as diuretics, antiepileptics, or antiglaucoma agents targeting CAs belonged to this class of compounds 1,3-5 . Although highly effective as CAIs 1 , the sulphonamides generally indiscriminately inhibit most a-CA isoforms present in mammals (at least 15 in humans, and 16 in other vertebrates 1 ) as well as CAs belonging to the other genetic families (b-, c-, d-, f-, g-, hand i-CAs) 2-5 and for this reason alternative CAI classes were searched for. In fact, in the last 10 years, a multitude of new chemotypes as well as novel CA inhibition mechanisms were reported 1,4,7-9 , which highly enriched our understanding of these enzymes and also allowed for obtaining isoform-selective CAIs targeting all the mammalian isoforms 4, [7][8][9] . Among the new such chemotypes, which also showed the highest levels of isoform selectivity, were the coumarins 9 , the sulfocoumarins 7,8 and their congeners, homosulfocoumarins (3H-1,2-benzoxathiepine 2,2-dioxides) 10 . Considering the fact that this last chemotype was only recently reported and rather poorly investigated 10 , we report here a series of new aryl-3H-1,2-benzoxathiepine 2,2-dioxides substituted in various positions of the heterocyclic ring, which have been designed in order to explore the chemical space around this new CA inhibitory chemotype and to see whether the presence of various moieties in position 7, 8 or 9 of the heterocyclic system maintains the desired enzyme inhibitory activity and selectivity for the target isoforms.

Chemistry
Reagents, starting materials and solvents were obtained from commercial sources and used as received. Thin-layer chromatography was performed on silica gel, spots were visualised with UV light (254 and 365 nm). Melting points were determined on an OptiMelt automated melting point system. IR spectra were recorded on Shimadzu FTIR IR Prestige-21 spectrometer. NMR spectra were recorded on Bruker Advance Neo (400 MHz) spectrometer with chemical shifts values (d) in ppm relative to TMS using the residual DMSO-d 6 signal ( 1 H 2.50; 13 C 39.52) or CDCl 3 signal ( 1 H 7.26; 13 C 77.16) as an internal standard. High-resolution mass spectra (HRMS) were recorded on a mass spectrometer with a Q-TOF micro mass analyser using the ESI technique. Elemental analyses were measured using Carlo Erba (EA1108) apparatus (Milan, Italy).
General procedure for the synthesis of ethenylphenoles (3,14,18,27) To a stirred solution of methyltriphenylphosphonium bromide (2.60 eq) in dry THF (5 mL/1 mmol of methyltriphenylphosphonium bromide), was added tBuOK (3.2 eq) in several portions over 20 min. Reaction mixture was stirred for 1 h at r.t. Corresponding benzaldehyde (1 eq) was added and stirring continued at room temperature for 24 h. Reaction mixture was diluted with CH 2 Cl 2 (4 mL/1 mmol of methyltriphenylphosphonium bromide). Organic layer was washed with water (2 Â 20 mL) and brine (2 Â 20 mL), and dried over Na 2 SO 4 , filtered and concentrated. The crude product was purified by column chromatography on silica gel (PE/EtOAc 4:1).

CA inhibitory assay
An Applied Photophysics stopped-flow instrument has been used for assaying the CA catalysed CO 2 hydration activity 15 . 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 constant the ionic strength), following the initial rates of the CA-catalysed CO 2 hydration reaction for a period of 10 À 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 À 10% of the reaction have been used for determining the initial velocity. The uncatalysed 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 À deionised water, and dilutions up to 0.01 nM were done thereafter with the assay buffer. Inhibitor and enzyme solutions were preincubated together for 6 h at room temperature prior to assay in order to allow for the formation of the E À I complex. The inhibition constants were obtained by nonlinear least-squares methods using PRISM 3 and the Cheng À Prusoff equation, as reported earlier [16][17][18][19] , and represent the mean from at least three different determinations. All CA isoforms were recombinant ones obtained in-house as reported earlier 19,20 .
In an attempt to prepare 6-aryl 3H-1,2-benzoxathiepine 2,2-dioxides, the commercially available bromo salicylaldehyde 13 was first converted to olefin 14 under Wittig reaction conditions, followed by treatment with sulphonyl chloride 4, thus providing bis-olefin 15 for olefin metathesis ring closure reaction (Scheme 2). Utilisation of the Ru-catalyst 6 as described above did not provide the formation of the desired key intermediate 6-bromo 3H-1,2-benzoxathiepine 2,2-dioxide (16) even at prolonged reaction times. By doubling catalyst 6 amount (10 mol%) only traces of compound 16 were observed after 40 h. No product formation was observed also when using Schrock and Schrock-Hoveyda Mo-catalysts. Probably olefin metathesis ring closure reaction did not take place due to sterical constraints due to the bulky Br atom at 3-positon of bis-olefin 15.
The synthesis of 8-bromo intermediate 20 was started from commercially available aldehyde 17, when under Wittig reaction conditions olefin 18 was obtained, which was thereafter treated with sulphonyl chloride 4 and provided the bis-olefin 19 in good yield (Scheme 3). Ru-catalysed olefin metathesis afforded the key intermediate 20 which in turn, by reaction with a series of aryl boronic acids under Suzuki reaction condition, provided the desired compounds 21-25.
The same strategy was successfully utilised for the synthesis of a series of 9-aryl 3H-1,2-benzoxathiepine 2,2-dioxides starting by the treatment of aldehyde 26 with methyltriphenylphosphonium bromide under Wittig reaction conditions (Scheme 4). The obtained phenol 27 was reacted with sulphonyl chloride 4 and ring closure of isolated 28 was successfully performed in Ru-catalysed olefin metathesis conditions, providing bromide 29. Further reaction of compound 9 with aryl boronic acids provided the desired derivatives 30-34 in moderate yields.

Carbonic anhydrase inhibition
The obtained homosulfocoumarins 7-34 were investigated for their CA inhibitory properties by using a stopped-flow CO 2 hydrase assay 15 and four human CA isoforms (hCA I, II, IX and XII) known to be drug targets 1 ( Table 1).
The following structure-activity relationship (SAR) can be observed from the inhibition data of Table 1.
(i) as the previously reported homosulfocoumarins 10 and similar to sulfocoumarins 7-9 , also the derivatives reported here did not significantly inhibit the cytosolic isoforms hCA I and II, unlike the sulphonamide acetazolamide (used as standard CAI), which has a very good affinity (in the nanomolar range) for hCA II and a micromolar one for hCA I (Table 1).
(ii) the transmembrane, tumour-associated isoforms hCA IX and XII were effectively inhibited by derivatives 7-29 reported here (in the lowmedium nanomolar arneg) and were poorly inhibited, in the micromolar range by the 9-substituted-homosulfocoumarins 30-34 (K I s in the range of 16.4-60.9 mM against hCA IX and >100 mM Scheme 2. Reagents and conditions: (i) KOtBu, CH 3 P(C 6 H 5 ) 3 Br, THF, RT, 18 h, 82%; (ii) 4, NEt 3 , CH 2 Cl 2 , 0 C to RT, 4 h, 66%; (iii) a) 6 (5 mol% and 10 mol%), toluene, against hCA XII). Thus, although weak inhibitors, these sulfocoumarins are anyhow highly selective for the inhibition of hCA IX, whereas their activity against hCA I, II and XII is absent (Table 1). As already anticipated above, the most important factors associated with CA IX/ XII inhibitory activity are the position and the nature of the moieties present on the six-membered ring of the homosulfocoumarin. Indeed, for 9-substituted derivatives, the presence of bulky, substituted aryls as in 30-34 leads to low activity, as mentioned above. Only the 9-bromo-derivative 29 had a medium potency inhibitory action against the two isoforms, with K I s in the range of 754.8 -3824 nM. On the contrary, the 8-substituted derivatives 20-25 showed a much better inhibitory power against both isoforms, being generally more potent than the corresponding 7-substituted derivatives 7-12. Indeed, for the 7-substituted homosulfocoumarins the K I s were in the range of 66.2 -620.8 nM against hCA IX and of 455.5 -2934 nM against hCA XII. On the contrary, for the 8-substituted homosulfocoumarins, the K I s were in the range of 44.0 -104.8 nM against hCA IX and in the range of 77.9 -473.2 nM for hCA XII ( Table 1). The 8-(4-trifluoromethyl)phenyl-substituted homosulfocoumarin 24 was the most effective hCA IX inhibitor (potency in the same range as AAZ), whereas the corresponding 4-fluorophenyl derivative 23 was the best hCA XII inhibitor in the new series of compounds investigated here but it was an order of magnitude less effective compared to acetazolamide.

Conclusions
A new series of homosulfocoumarins (3H-1,2-benzoxathiepine 2,2dioxides) possessing various moieties in the 7, 8 or 9 position of the heterocylic ring were prepared by original procedures and investigated for the inhibition of four physiologically relevant CA isoforms, hCA I, II, IX and XII. The 8-substituted homosulfocoumarins were the most effective hCA IX/XII inhibitors followed by the 7-substituted derivatives, whereas the substitution pattern in position 9 led to less effective inhibitors for these transmembrane, tumourassociated isoforms. The cytosolic isoforms hCA I and II were not inhibited by these compounds, similar to the sulfocoumarins/coumarins investigated earlier. As hCA IX and XII are validated antitumour targets 5 , with one sulphonamide (SLC-0111) in Phase Ib/II clinical trials, finding derivatives with a better selectivity for inhibiting the tumour-associated isoforms over the cytosolic ones, as the homosulfocoumarins reported here, is of crucial importance.