Chlorocarbonylsulfenyl Chloride Cyclizations Towards Piperidin-3-yl-oxathiazol-2-ones as Potential Covalent Inhibitors of Threonine Proteases

Using rescaffolding approach, we designed piperidine compounds decorated with an electrophilic oxathiazol-2-one moiety that is known to confer selectivity towards threonine proteases. Our efforts to prepare products according to the published procedures were not successful. Furthermore we identified major side products containing nitrile functional group, resulting from carboxamide dehydration. We systematically optimized reaction conditions towards our desired products to identify heating of carboxamides with chlorocarbonylsulfenyl chloride and sodium carbonate as base in dioxane at 100 °C. Our efforts culminated in the preparation of a small series of piperidin-3-yl-oxathiazol-2-ones that are suitable for further biological evaluation.


Introduction
Proteases play key roles in complex biological systems and in multiple structural and signalling pathways.They constitute a historically important field in medicinal chemistry and continue to represent a source of potential drug targets.They are involved in the pathology of hypertension, autoimmune and inflammatory diseases, reperfusion injury, blood clotting disorders, HIV and other viral infections, parasitic and bacterial infections, and last but not least, cancer. 1 Protease inhibitors are not valuable only as potential drugs but also as experimental tools for structural biology, 2 as they can be used as molecular probes in the elucidation of protease structures and protease pathway mechanisms. 3Recently, databases of proteases (sometimes also termed peptidases, proteinases or proteolytic enzymes) have been established as a resource in this immense research field; namely the Merops database with over 4000 individual entries. 4ur research efforts are mainly focused on the N-terminal threonine proteases that form stable covalent acyl-enzyme complexes and are subsequently hydrolyzed to afford product peptides.Threonine proteases constitute 99 entries in the Merops database, where we specifically study the threonine-type endopeptidases, such as the proteasomes. 5The proteasomes consist of a central proteolytic unit, known as the 20S proteasome, and the 19S regulators, which together make up a 26S structure (Figure 1).The constitutive isoform of the proteasome is expressed in all eukaryotic cells while its immunomodulatory isoform, the immunoproteasome, is mainly expressed in cells associated with the immune system, such as lymphocytes and Jukič et al.: Chlorocarbonylsulfenyl Chloride Cyclizations ... monocytes. 5,6The constitutive proteasome contains three enzymatically active subunits, namely the b1c (caspaselike), the b2c (trypsin-like), and the β5c (chymotrypsin-like) that are embedded into a barrel-shaped structure consisting of four rings of β-subunits and α-subunits in an abbα order.The immunoproteasome has essentially the same overall structure, only the catalytically active subunits of cCP are replaced by their counterparts b1i, b2i, and β5i (Figure 1).The 20S proteasome core particle of both isoforms is a protease of 720 kDa and 28 individual subunits and is responsible for essential proteolytic degradation during cellular inflammatory and oxidative stress. 7[10] There is an amounting body of research on the small-molecule inhibitors of proteasomes. 5,11Both marketed medicines, bortezomib and carfilzomib, equally inhibit the catalytically active β5 subunits of the constitutive proteasome and the immunoproteasome.The combined inhibition of both isoforms leads to cytotoxicity that limits the clinical application of these broad spectrum proteasome inhibitors. 6In addition, many of the investigational compounds are peptide-like compounds and this represents a serious limitation to their metabolic stability and bioavailability. 5To overcome these problems, multiple approaches can be found in literature: design of reversible proteasome inhibitors, 12 use of structural differences in the binding sites of both proteasomes in structure-based drug design, 13,14 design of highly selective and hydrolytically more stable peptidic compounds, 15 design of highly selective non-peptidic compounds, 16 use of non-catalytic residues or allosteric sites in inhibitor design, 17 and the design of selective electrophilic warheads. 18The majority of these compounds are covalent irreversible inhibitors bearing an electrophilic warhead that is capable of reacting with the N-terminal threonine residue in the catalytic active site of the examined protease. 5,11Electrophilic warheads belong to structural classes of aldehydes, α',β'-epoxyketones, α-keto aldehydes, β-lactones, vinyl sulfones, Michael-acceptor systems, and boronates. 19The active interest in this field is clearly represented by a very recent publication, 19 where a new mechanism for an existing warhead was reported, i.e. the formation of 1,4-oxazepane upon reaction of an α',β'-epoxyketone warhead with the N-terminal threonine rather than the previously reported morpholine ring. 14,19Such new developments provide invaluable data for the design of novel and selective irreversible inhibitors of threonine proteases.
In order to design targeted covalent inhibitors of threonine protease, we sought to examine the available electrophilic warheads. 20We were in particular interested in compounds that could provide a suitable reactivity and selectivity towards threonine proteases.Recently, oxathiazol-2-one moiety was identified in a high-throughput screening campaign as a promising candidate. 21The proposed mechanism of the covalent modification of N-terminal threonine induced by this electrophilic fragment is depicted in Figure 2 and proceeds through cyclocarbonylation. 18,21In current paper we describe an optimized synthetic approach towards oxathiazol-2-one electrophilic war- head in compounds with basic nitrogen atom and the preparation of a focused library of piperidin-3-yl-oxathiazol-2-ones that are suitable for further biological evaluation.

Results and Discussion
We designed our compounds on the basis of their synthetic accessibility and their potential to be modified accordingly during further optimizations.Therefore, we selected a piperidine central core derivatized with an electrophilic oxathiazol-2-one warhead that could confer the selectivity towards threonine proteases as reported beforehand (Figure 3). 18,21e started the synthesis with the alkylation of nipecotamide employing a set of alkyl bromides in DMF as a solvent and Na 2 CO 3 as a base to obtain compounds 2a, 2b and 2c-e.In the case of compound 2f, alkylation with p-nitrobenzylbromide was followed by hydrogenation in MeOH with final acylation using benzyl chloride.The key step in the synthesis was the cyclization of suitably substituted nipecotamides 2a-f into piperidin-3-yl-oxathiazol-2-ones 7-3e using chlorocarbonylsulfenyl chloride as a reagent (Figure 4).This synthetic approach was reported by Gryder et al. when they described the synthesis of the oxathiazol-2-one analogue of bortezomib.The penultimate carboxamide dipeptide was successfully transformed into the oxathiazol-2-one-bortezomib in high yield by using chlorocarbonylsulfenyl chloride in refluxing THF. 22espite our numerous attempts to obtain the final oxathiazol-2-ones 3a-f by following the original procedure no product could be isolated.Initial experiments in refluxing THF resulted in a complex mixture of products. 23If the experiments were performed at lower temperature (0 °C, room temperature), no apparent conversion was observed.Our first modification of the original procedure was to use relatively nonpolar and system-inert toluene as a solvent that could provide an alternative reactant/intermediate stabilization pattern and would enable a broader temperature sweep.This system was also described by Gurjar et al.where they heated the mixture of amide and chlorocarbonylsulfenyl chloride in toluene from 60 to 90 °C until the settlement of HCl evolution, followed by 1 h of reflux; this yielded > 50% of isolated oxathiazol-2-one. 23No conversion was observed in our case at lower temperatures (0 °C, room temperature) with a formation of complex mixture of products at 60 °C and reflux conditions.Further experiments using pyridine as solvent afforded similar results.Nevertheless, a difference in reaction scope can be observed as besides previously mentioned report by Gryder et al., 22 literature only describes a relatively simple case of benzamide cyclization towards final 5-phenyl-1,3,4-oxathiazol-2-one.In our case, the reaction incorporated a piperidin-3-yl central scaffold (compounds 2a-e) containing an additional basic centre.We also conducted a thor- ough separation of complex product mixtures in the case of cyclization of compound 2a and identified a dominant side product (> 30% yield) flanked by a myriad of other chemical species that could not be obtained at a significant quantity.The dominant side product was identified when examining its 13 C NMR spectrum.Namely, the carbon atom of the carboxamide 2a can be found as expected at 178.3 ppm (400 MHz, DMSO-d 6 ), whereas the carbon of the dominant side product species was found upfield at 121.8 ppm.When recording IR spectrum, a marked peak at 2240 cm -1 was found indicating the presence of a nitrile functionality; the formation of the side product 1-benzylpiperidine-3-carbonitrile 4a (Figure 5) was then further confirmed by HRMS.The nature of this reaction outcome can be rationalized as presented in Figure 5.
In our reaction system, the dehydration process is facilitated by the primary amide 2a (Figure 5) that readily couples with the chlorocarbonylsulfenyl chloride to form an active intermediate (Figure 5).The coupling is followed by rapid elimination that is catalyzed either with the starting substituted piperidine as a base or is assisted by other bases in the reaction system (such as pyridine) to form the corresponding nitrile 4a (Figure 5).Indeed, similar dehydrations of primary carboxamides using an acidic reagent such as POCl 3 , SOCl 2 are well documented in literature. 24,25ore recent, chemoselective and milder methods were  also reported, where ethyl dichlorophosphate/DBU system or methyl (carboxysulfamoyl)triethylammonium hydroxide (Burgess reagent) were used as the dehydrating reagents. 26,272][33] The myriad of reaction side products that was observed is a consequence of multitude of side reactions that can occur during dehydration reactions, such as thermal decomposition of the formed oxathiazol-2-one and hydrolysis reactions (Figure 6).The formed oxathiazol-2-one can also take part in the 1,3-dipolar nitrile sulphide cycloaddition reaction with available nitrile to obtain thiadiazoles as side products. 34The nitrile sulphide is  a Yield after purification using column chromatography (SiO 2 support with n-hexane:EtOAc solvent system as an eluent).
Jukič et al.: Chlorocarbonylsulfenyl Chloride Cyclizations ... generated in situ by thermal decomposition of oxathiazol-2-one. 35][38] After initial unsuccessful attempts to prepare the desired compounds 3a-f, we turned our attention to microwave-assisted report on flow-chemistry synthesis of oxathiazol-2-one in dioxane at 200 °C and residence time of 1 min in a flow reactor reported by Öhrngren et al. 39 On this basis, we modified the reaction procedure and dissolved the carboxamides 2a-f (Figure 4) in dry dioxane (27 mL/1 mmol carboxamide), used an excess of solid Na-2 CO 3 (5 eq) and chlorocarbonylsulfenyl chloride (2 eq), and stirred the reaction mixture at 100 °C for 16 h under argon to obtain the desired oxathiazol-2-ones 3a-f (Figure 4) in 16 to 68% yields (Table 1).

Experimental
Chemicals from commercial sources were used without further purification.Anhydrous THF, DCM and Et 3 N were dried and purified by distillation over Na, K 2 CO 3 and KOH, respectively.Analytical thin-layer chromatography (TLC) was performed on Merck silica gel (60F 254 ) plates (0.25 mm).Column chromatography was performed on silica gel 60 (Merck, particle size 0.040-0.063mm).Melting points were determined on a Reichert hot stage microscope and are uncorrected. 1H-, COSY-, HMQC-and 13 C-NMR spectra were recorded on a Bruker AVANCE DPX 400 spectrometer in CDCl 3 or DMSO-d 6 solution with TMS as internal standard.Chemical shifts are reported in ppm (δ) downfield from TMS.All the coupling constants (J) are in hertz.IR spectra were recorded on a PerkinElmer Spectrum BX System FT-IR spectrometer.Mass spectra were obtained with a VG-Analytical Autospec Q mass spectrometer with ESI ionization (MS Centre, Jožef Stefan Institute, Ljubljana).All reported yields are those of purified products.

Conclusion
Based on the previously reported oxathiazol-2one-bearing and nonpeptidic inhibitors of the chymotrypsin-like (β5i) subunit of the immunoproteasome, we designed a novel series of piperidin-3-yl-oxathiazol-2-ones as potential covalent inhibitors of threonine proteases.Compounds were designed with a synthetically accessible piperidine central core derivatized with an oxathiazol-2-one electrophilic moiety.In lieu of previously reported synthetic approaches, we identified a synthetic protocol that enables the cyclization of carboxamides incorporating a basic centre into oxathiazol-2-ones.This straightforward protocol using chlorocarbonylsulfenyl chloride as a reagent in dioxane afforded the desired products in moderate to good yields.Thus, a vast chemical space of 5-substituted oxathiazol-2-ones can be explored and various chemical libraries of inhibitors of threonine proteases can be compiled.

Figure 2 .Figure 3 .
Figure 2. Oxathiazol-2-one electrophilic warhead and its interaction mechanism with the N-terminal threonine in the active site

Figure 5 .
Figure 5.The proposed mechanism of dexydration of primary amides to nitriles using chlorocarbonylsulfenyl chloride.