Synthesis of 6-Membered-Ring Fused Thiazine-Dicarboxylates and Thiazole-Pyrimidines via One-Pot Three-Component Reactions

A facile and efficient one-pot three-component reaction method for the synthesis of thiazine-dicarboxylates is reported. Reaction of an isocyanide and dialkyl acetylenedicarboxylate with 2-amino-4H-1,3-thiazin-4-one derivatives containing both an acidic proton and an internal nucleophile gave the products in good yields of 76–85%. The reactivity of dialkyl acetylenedicarboxylates was further tested in the synthesis of thiazole-pyrimidines where a two-component reaction of 2-aminothiazole with dialkyl acetylenedicarboxylates was successfully converted to a more efficient three-component reaction of a thiourea, α-haloketone and dialkyl acetylenedicarboxylate (DMAD/DEtAD) to give thiazole-pyrimidines in good yields of 70–91%.


Introduction
Multicomponent reactions (MCRs) are one-pot reactions that comprise at least three reactants [1][2][3]. Some of the well-known MCRs include the classic Ugi four-component reaction [4] (U-4CR), the Passerini three-component reaction [5] (P-3CR), the Biginelli threecomponent reaction [6] (B-3CR) and the Hantzsch three-component reaction [7] (H-3CR). The history and discovery of MCRs can be traced back to Strecker's [8] multicomponent reaction (S-3CR), which was reported in 1850, and three decades on, the Hantzsch reaction for preparing dihydropyridines (DHP) was reported [7]. The Biginelli 3-CR was reported in 1891, whereas Mannich [9] reported his eponymous MCR in 1912. The first isocyanidebased MCRs were reported in 1921 and 1959 by Passerini [5] and Ugi [10], respectively. MCRs represent one of the most convenient synthetic approaches in chemistry, with the main advantages being a reduction in the number of sequential reaction steps required and often better yields [11,12]. The Passerini and Ugi reactions have gained a great deal of attention since the early 1990s [13,14]. The versatility of these reactions relies on the dual reactivity of the isocyanide carbon atom, leading to the formation of diverse and complex products [15,16].
The first reaction of isocyanides 1 with electron deficient alkynes such as DMAD 2 to give zwitterions 3 (Scheme 1) was reported by Winterfeldt and co-workers [17]. This zwitterion adduct 3 can readily undergo cycloaddition reactions with various third components, generating a range of heterocyclic scaffolds of interest [18]. Previously, our group reported the reaction of zwitterion adduct 3 and different five-membered rings, for example, 4, containing both an acidic proton and internal nucleophile, to give products 5 incorporating all three reaction components (Scheme 1). Depending on the nature of the five-membered ple, 4, containing both an acidic proton and internal nucleophile, to give products 5 incorporating all three reaction components (Scheme 1). Depending on the nature of the fivemembered ring, in certain cases such as for thiol 6, we observed that there was competition between the three-component reaction (to give 7) and the two-component reaction (to give 8). For the latter, DMAD 2 reacts directly with the five-membered ring, as shown in Scheme 1 for 1H-1,2,4-triazole-5-thiol 6, to give methyl 7-oxo-7H-[1, 2,4]triazolo [5,1b] [1,3]thiazine-5-carboxylate 8 [19]. We found that the most successful three-component reactions involved five-membered rings containing more acidic protons. In this work, we set out to investigate alternative rings to act as the third component substrate in these reactions and selected the acidic six-membered ring 2-amino-4H-1,3thiazin-4-one derivatives 9, with predicted pKa values in the range of −1.80 to −2.21. Thiazines are bioactive scaffolds with a wide range of pharmacological activities, such as antimicrobial [20,21], anti-inflammatory [22], anticancer [23], antitubercular [24] and antiviral [25] activities. Another aim of this work was the further investigation of the two-component reactions between various five-membered rings and DMAD to prepare compounds such as 8 and their analogues. Of particular interest to us were the thiazole-pyrimidine derivatives, which display interesting biological activity. Compounds derived from thiazolopyrimidines are known to exhibit antioxidant and antitumor activities [26]. Derivatives containing a thiazole-pyrimidine moiety as part of a more complex structure were shown to possess activity against Mycobacterium tuberculosis [27], while thiazolopyrimidine containing compounds ritanserin [28] and setoperone [29] have been used in the study of psychiatric disorders.

Synthesis of 4-Oxo-4,6-Dihydropyrimido[2,1-b][1,3]Thiazine-6,7-Dicarboxylates by Three-Component Reaction
The six-membered rings 9a-e, which were selected for use in these three-component reactions, are shown in Figure 1. To the best of our knowledge, none of these compounds In this work, we set out to investigate alternative rings to act as the third component substrate in these reactions and selected the acidic six-membered ring 2-amino-4H-1,3thiazin-4-one derivatives 9, with predicted pK a values in the range of −1.80 to −2.21. Thiazines are bioactive scaffolds with a wide range of pharmacological activities, such as antimicrobial [20,21], anti-inflammatory [22], anticancer [23], antitubercular [24] and antiviral [25] activities. Another aim of this work was the further investigation of the two-component reactions between various five-membered rings and DMAD to prepare compounds such as 8 and their analogues. Of particular interest to us were the thiazolepyrimidine derivatives, which display interesting biological activity. Compounds derived from thiazolopyrimidines are known to exhibit antioxidant and antitumor activities [26]. Derivatives containing a thiazole-pyrimidine moiety as part of a more complex structure were shown to possess activity against Mycobacterium tuberculosis [27], while thiazolopyrimidine containing compounds ritanserin [28] and setoperone [29] have been used in the study of psychiatric disorders.
We initially investigated the direct reactivity of dialkyl acetylenedicarboxylate towards 2-aminothiazoles for the preparation of thiazole-pyrimidine derivatives and subsequently converted this two-component reaction into a one-pot, three-component reaction of dialkyl acetylenedicarboxylate with thiourea and an α-haloketone for the formation of thiazolepyrimidines. Here we present a facile and efficient one-pot, three component reaction method for the synthesis of novel dihydropyrimido-thiazine-6,7-dicarboxylates from an isocyanide, dialkyl acetylenedicarboxylate and 2-amino-4H-1,3-thiazin-4-one derivatives; and the synthesis of 5H-thiazolo[3,2-a]pyrimidine-7-carboxylates from thiourea, an αhaloketone and dialkyl acetylenedicarboxylates. The six-membered rings 9a-e, which were selected for use in these three-component reactions, are shown in Figure 1. To the best of our knowledge, none of these compounds has previously been tested in reaction with acetylenedicarboxylates 2 and isocyanides 1. In addition, compounds 9c and 9d have not been previously reported. The initial reaction of 2-amino-4H-1,3-thiazin-4-one 9a [30] with isocyanide 1a and acetylene dicarboxylates 2a-b resulted in products 10a-b (Table 1). This success encouraged us to further test the reaction of 2-amino-6-methyl-4H-1,3-thiazin-4-one 9b [31], 2-amino-6-ethyl-4H-1,3-thiazin-4-one 9c, 2-amino-6-propyl-4H-1,3-thiazin-4-one 9d and 2-amino-6-phenyl-4H-1,3-thiazin-4-one 9e [31] with the zwitterion adduct formed by the reaction of aliphatic isocyanides 1a-c and acetylenedicarboxylates 2a-b. Generally, the reactions were carried out by the addition of isocyanide 1 to acetylenedicarboxylate 2 at 0 • C under inert conditions, as shown in Table 1, followed by slow introduction of compound 9a-e as a solution in dry dichloromethane to the reaction mixture at room temperature. After purification, novel products 10a-u were obtained in good yields (Table 1). When selected reactions were tested using the green solvents ethanol and isopropanol instead of dichloromethane, low yields of thiazine compounds 10 were obtained. When acetone was employed as a solvent for the synthesis of thiazine compounds 10a-u, using starting materials 9a-e, average yields were obtained. Nonetheless, acetone can be considered as an alternative greener solvent to afford thiazine compounds 10a-u using a three-component reaction. Compound 10o gave the highest yield in acetone of 58%, while the lowest yield obtained was for 10g, at 41%. has previously been tested in reaction with acetylenedicarboxylates 2 and isocyanides 1.

Results and Discussion
In addition, compounds 9c and 9d have not been previously reported. The initial reaction of 2-amino-4H-1,3-thiazin-4-one 9a [30] with isocyanide 1a and acetylene dicarboxylates 2a-b resulted in products 10a-b (Table 1). This success encouraged us to further test the reaction of 2-amino-6-methyl-4H-1,3-thiazin-4-one 9b [31], 2-amino-6-ethyl-4H-1,3-thiazin-4-one 9c, 2-amino-6-propyl-4H-1,3-thiazin-4-one 9d and 2-amino-6-phenyl-4H-1,3-thiazin-4-one 9e [31] with the zwitterion adduct formed by the reaction of aliphatic isocyanides 1a-c and acetylenedicarboxylates 2a-b. Generally, the reactions were carried out by the addition of isocyanide 1 to acetylenedicarboxylate 2 at 0 °C under inert conditions, as shown in Table 1, followed by slow introduction of compound 9a-e as a solution in dry dichloromethane to the reaction mixture at room temperature. After purification, novel products 10a-u were obtained in good yields (Table 1). When selected reactions were tested using the green solvents ethanol and isopropanol instead of dichloromethane, low yields of thiazine compounds 10 were obtained. When acetone was employed as a solvent for the synthesis of thiazine compounds 10a-u, using starting materials 9a-e, average yields were obtained. Nonetheless, acetone can be considered as an alternative greener solvent to afford thiazine compounds 10a-u using a three-component reaction. Compound 10o gave the highest yield in acetone of 58%, while the lowest yield obtained was for 10g, at 41%.    has previously been tested in reaction with acetylenedicarboxylates 2 and isocyanides 1 In addition, compounds 9c and 9d have not been previously reported. The initial reactio of 2-amino-4H-1,3-thiazin-4-one 9a [30] with isocyanide 1a and acetylene dicarboxylate 2a-b resulted in products 10a-b (Table 1). This success encouraged us to further test th reaction of 2-amino-6-methyl-4H-1,3-thiazin-4-one 9b [31], 2-amino-6-ethyl-4H-1,3-thia zin-4-one 9c, 2-amino-6-propyl-4H-1,3-thiazin-4-one 9d and 2-amino-6-phenyl-4H-1,3-th azin-4-one 9e [31] with the zwitterion adduct formed by the reaction of aliphatic isocya nides 1a-c and acetylenedicarboxylates 2a-b. Generally, the reactions were carried out b the addition of isocyanide 1 to acetylenedicarboxylate 2 at 0 °C under inert conditions, a shown in Table 1, followed by slow introduction of compound 9a-e as a solution in dr dichloromethane to the reaction mixture at room temperature. After purification, nove products 10a-u were obtained in good yields (Table 1). When selected reactions wer tested using the green solvents ethanol and isopropanol instead of dichloromethane, low yields of thiazine compounds 10 were obtained. When acetone was employed as a solven for the synthesis of thiazine compounds 10a-u, using starting materials 9a-e, averag yields were obtained. Nonetheless, acetone can be considered as an alternative greene solvent to afford thiazine compounds 10a-u using a three-component reaction. Com pound 10o gave the highest yield in acetone of 58%, while the lowest yield obtained wa for 10g, at 41%.  Based on our previous results, we believe the very good yields obtained for 10a-u were due to the presence of an acidic proton on the 2-amino-4H-1,3-thiazin-4-one derivatives 9ae which enables the reaction with the zwitterion adduct to occur readily in dichloromethane. The highest yields obtained were 85% for 10b, 10h and 10n and the lowest yield was 76% for 10h. Based on the isolated yields it is clear that the 2-amino-4H-1,3-thiazin-4-one derivatives 9 were completely inert towards direct reaction with either dialkyl acetylenedicarboxylates 2 or alkyl isocyanides 1, while they demonstrated very high reactivity towards the zwitterion adduct. Although we have not definitively established the mechanism for the formation of 10, a plausible mechanism for this reaction is proposed (after Esmaeili and co-workers [32]) and shown in Scheme 2. Zwitterion adduct 3 deprotonates 9, followed by 1,4-nucleophilic attack on the nitrilium ion A. The resulting ketenimine intermediate B cyclizes to afford compound 10.
Based on our previous results, we believe the very good yields obtained for 10a-u were due to the presence of an acidic proton on the 2-amino-4H-1,3-thiazin-4-one derivatives 9a-e which enables the reaction with the zwitterion adduct to occur readily in dichloromethane. The highest yields obtained were 85% for 10b, 10h and 10n and the lowest yield was 76% for 10h. Based on the isolated yields it is clear that the 2-amino-4H-1,3thiazin-4-one derivatives 9 were completely inert towards direct reaction with either dialkyl acetylenedicarboxylates 2 or alkyl isocyanides 1, while they demonstrated very high reactivity towards the zwitterion adduct. Although we have not definitively established the mechanism for the formation of 10, a plausible mechanism for this reaction is proposed (after Esmaeili and co-workers [32]) and shown in Scheme 2. Zwitterion adduct 3 deprotonates 9, followed by 1,4-nucleophilic attack on the nitrilium ion A. The resulting ketenimine intermediate B cyclizes to afford compound 10. Scheme 2. Proposed reaction mechanism for the formation of 10.

Synthesis of 5H-Thiazolo[3,2-a]Pyrimidine-7-Carboxylates Using Two-Component and Three-Component Reactions
We went on to investigate the direct reaction of various five-membered ring substrates with electron deficient alkynes (DMAD/DEtAD). This study was started when we noticed the presence of co-products resulting from direct reaction of DMAD with some of the five-membered rings when we were attempting three-component reactions including an isocyanide [19]. Crank and co-workers [33] previously reported the reaction of 4-methylthiazol-2-amine 11 with DMAD 2 to afford Diels-Alder adduct 12, methyl 3-methyl-7-oxo-7H-thiazolo[3,2-a]pyrimidine-5-carboxylate 13 and product 14, resulting from reac-Scheme 2. Proposed reaction mechanism for the formation of 10.
DMAD 2a and pre-prepared five-membered ring substrates 11 directly to obtain fused 5membered ring thiazole-pyrimidines 13. However, the moderate yields achieved using this approach led us to optimize the preparation as a one-pot, three-component reaction from thiourea 15, α-haloketones 16 and DMAD or DEtAD. To the best of our knowledge, this is the first report of such a one-pot preparation. Initially, in order to achieve thiazole-pyrimidines 13, we prepared 5-membered ring substrates 11a-g [39][40][41][42] (Figure 2) to react with DMAD 2a. The reactions were carried out by treating 11c and 11e-g with 2a under ethanol reflux conditions to obtain 13a-d (Method A). Repetition of these reactions using DEtAD 2b afforded 13e-i, whilst di-tert-butyl acetylenedicarboxylate (DTAD) 2c and substrate 11a gave rise to 13j (Table 2). Despite the disappointing yields, the success of this reaction led us to extend it to the use of alternative five-membered ring reactants 17 [43], 18 [44] and 5 [45], with these substrates giving rise to compounds 19a-b, 20 and 8, respectively, in good yields.  Initially, in order to achieve thiazole-pyrimidines 13, we prepared 5-membered ring substrates 11a-g [39][40][41][42] (Figure 2) to react with DMAD 2a. The reactions were carried out by treating 11c and 11e-g with 2a under ethanol reflux conditions to obtain 13a-d (Method A). Repetition of these reactions using DEtAD 2b afforded 13e-i, whilst di-tert-butyl acetylenedicarboxylate (DTAD) 2c and substrate 11a gave rise to 13j (Table 2). Despite the disappointing yields, the success of this reaction led us to extend it to the use of alternative five-membered ring reactants 17 [43], 18 [44] and 5 [45], with these substrates giving rise to compounds 19a-b, 20 and 8, respectively, in good yields. from thiourea 15, α-haloketones 16 and DMAD or DEtAD. To the best of our knowledge, this is the first report of such a one-pot preparation. Initially, in order to achieve thiazole-pyrimidines 13, we prepared 5-membered ring substrates 11a-g [39][40][41][42] (Figure 2) to react with DMAD 2a. The reactions were carried out by treating 11c and 11e-g with 2a under ethanol reflux conditions to obtain 13a-d (Method A). Repetition of these reactions using DEtAD 2b afforded 13e-i, whilst di-tert-butyl acetylenedicarboxylate (DTAD) 2c and substrate 11a gave rise to 13j (Table 2). Despite the disappointing yields, the success of this reaction led us to extend it to the use of alternative five-membered ring reactants 17 [43], 18 [44] and 5 [45], with these substrates giving rise to compounds 19a-b, 20 and 8, respectively, in good yields.   strate 11a gave a 70% yield of product 13j. When substrate 17 was treated with dialkyl acetylenedicarboxylates 2a-b the reaction gave 82% and 80% yield of 19a and 19b, respectively. Substrates 5 and 18 were also treated with DEtAD 2b and gave rise to compounds 8 and 20 in good yields of 88% and 82%, respectively. When comparing the reactivity of the substrates it is evident that substrates 5, 17 and 18 were more reactive towards 2a-b than 11 under these conditions, giving consistently better yields. Using Method A, low-to-average yields were obtained for most of the products 13. The highest yield from this set of compounds was for 13g with 79% yield and 13f with 71%. However, when using substrates 11d-g with phenyl groups at R 1 the yields dropped slightly due to the weakly activating phenyl ring and the deactivating groups on the phenyl ring (such as fluorine and chlorine) for both DMAD and DEtAD reactions. The presence of fluorine and chlorine at the para-position of substrates 11f and 11g resulted in a slight drop in yield for compounds 13c, 13d, 13h and 13i compared to 13b, with a methyl substituent at the para-position. The reaction of dialkyl acetylenedicarboxylate 2c and substrate 11a gave a 70% yield of product 13j. When substrate 17 was treated with dialkyl acetylenedicarboxylates 2a-b the reaction gave 82% and 80% yield of 19a and 19b, respectively. Substrates 5 and 18 were also treated with DEtAD 2b and gave rise to compounds 8 and 20 in good yields of 88% and 82%, respectively. When comparing the reactivity of the substrates it is evident that substrates 5, 17 and 18 were more reactive towards 2a-b than 11 under these conditions, giving consistently better yields.
The disappointing yields obtained from the direct reaction of substrates 11a-h with dialkyl acetylenedicarboxylates 2a-b prompted us to apply a second approach via a 3-CR to give product 13 (Method B). The reactions were carried out by initially mixing thiourea 15 with α-haloketone 16, followed by introduction of dialkyl acetylenedicarboxylate 2 in a 1:1 ratio of ethanol and THF at room temperature, after which the reactions were heated to ethanol reflux temperature for at least 12 h. This gave rise to products 13a-d in improved yields compared to the first approach ( Table 3). The reactions were repeated using DEtAD 2b and gave rise to products 13e-i as expected. Room temperature reactions resulted in incomplete reactions, with recovery of some of the starting materials, even after 48 h. The three-component reaction (3-CR) was thus found to be suitable for the synthesis of products 13a-i, which were isolated in good yields of 70-91%, based on the nature of the functional groups in the reactants 16a-e and 2a-b, as shown in Table 3. The highest yields for this set of compounds, using DMAD 2a, was 82% for 13a. The percentage yields of compounds 13b-d were found to be slightly lower when compared to 13a. Even starting materials substituted with fluorine or chlorine at the para-position (16c and 16d) gave very good yields of product. Yields increased when using DEtAD 2b rather than DMAD 2a. Table 3. Isolated yields of 13a-i from one-pot, three-component reaction.
48 h. The three-component reaction (3-CR) was thus found to be suitable for the synt of products 13a-i, which were isolated in good yields of 70-91%, based on the natu the functional groups in the reactants 16a-e and 2a-b, as shown in Table 3. The hig yields for this set of compounds, using DMAD 2a, was 82% for 13a. The percentage y of compounds 13b-d were found to be slightly lower when compared to 13a. Even ing materials substituted with fluorine or chlorine at the para-position (16c and 16d) very good yields of product. Yields increased when using DEtAD 2b rather than DM 2a.
Scheme 4 shows the proposed mechanism for the formation of compound 13 Acheson and Wallis [34]). The reaction proceeds by initial nucleophilic attack of the nitrogen of substrate 11 on the α,β-unsaturated ester of DMAD forming an interme which undergoes internal cyclisation through nucleophilic attack of the side-chain n gen on the ester carbonyl carbon, followed by the loss of methanol, to afford produc  Scheme 4 shows the proposed mechanism for the formation of compound 13 (see Acheson and Wallis [34]). The reaction proceeds by initial nucleophilic attack of the ringnitrogen of substrate 11 on the α,β-unsaturated ester of DMAD forming an intermediate which undergoes internal cyclisation through nucleophilic attack of the side-chain nitrogen on the ester carbonyl carbon, followed by the loss of methanol, to afford product 13.
les 2021, 26, x FOR PEER REVIEW 7 The disappointing yields obtained from the direct reaction of substrates 11a-h dialkyl acetylenedicarboxylates 2a-b prompted us to apply a second approach via a to give product 13 (Method B). The reactions were carried out by initially mixing thio 15 with α-haloketone 16, followed by introduction of dialkyl acetylenedicarboxylate a 1:1 ratio of ethanol and THF at room temperature, after which the reactions were he to ethanol reflux temperature for at least 12 h. This gave rise to products 13a-d in proved yields compared to the first approach ( Table 3). The reactions were repeated u DEtAD 2b and gave rise to products 13e-i as expected. Room temperature reaction sulted in incomplete reactions, with recovery of some of the starting materials, even 48 h. The three-component reaction (3-CR) was thus found to be suitable for the synt of products 13a-i, which were isolated in good yields of 70-91%, based on the natu the functional groups in the reactants 16a-e and 2a-b, as shown in Table 3. The hig yields for this set of compounds, using DMAD 2a, was 82% for 13a. The percentage y of compounds 13b-d were found to be slightly lower when compared to 13a. Even ing materials substituted with fluorine or chlorine at the para-position (16c and 16d) very good yields of product. Yields increased when using DEtAD 2b rather than DM 2a.
Scheme 4 shows the proposed mechanism for the formation of compound 13 Acheson and Wallis [34]). The reaction proceeds by initial nucleophilic attack of the nitrogen of substrate 11 on the α,β-unsaturated ester of DMAD forming an interme which undergoes internal cyclisation through nucleophilic attack of the side-chain n gen on the ester carbonyl carbon, followed by the loss of methanol, to afford produc   The structures of the synthesized compounds were confirmed by mass spectrometry, NMR spectroscopy and FTIR spectroscopy (Supplementary Materials). Explicit confirmation for the structures of 13a, 13b, 13d and 8 was obtained from single-crystal X-ray analysis (Figure 3). The structures of the synthesized compounds were confirmed by mass spectrometry, NMR spectroscopy and FTIR spectroscopy (Supplementary Materials). Explicit confirmation for the structures of 13a, 13b, 13d and 8 was obtained from single-crystal X-ray analysis ( Figure 3). These compounds were prepared as part of our ongoing efforts towards identifying novel heterocycles with biological activity [46][47][48]. Compounds showing cytotoxicity values of >200 µM in MT-4 cells were evaluated for activity in an HIV assay (see Supplementary Materials). Unfortunately, these compounds were not found to be active as antiviral agents. However, due to the relatively high cytotoxicity values seen for some of the compounds, the possibility of their application as anticancer agents is being further investigated.

General Information
All commercially available reagents were supplied by Sigma Aldrich (Schnelldorf, Germany) and used without further purification. Dry solvents were used directly from an LC-Tech SP-1 Solvent Purification System (LC Technology Solutions Inc., Seabrook, TX, USA) stored under argon. All solvents used for chromatographic purposes were supplied by RadChem (Johannesburg, South Africa) and were used without further distillation.  These compounds were prepared as part of our ongoing efforts towards identifying novel heterocycles with biological activity [46][47][48]. Compounds showing cytotoxicity values of >200 µM in MT-4 cells were evaluated for activity in an HIV assay (see Supplementary Materials). Unfortunately, these compounds were not found to be active as antiviral agents. However, due to the relatively high cytotoxicity values seen for some of the compounds, the possibility of their application as anticancer agents is being further investigated.

General Information
All commercially available reagents were supplied by Sigma Aldrich (Schnelldorf, Germany) and used without further purification. Dry solvents were used directly from an LC-Tech SP-1 Solvent Purification System (LC Technology Solutions Inc., Seabrook, TX, USA) stored under argon. All solvents used for chromatographic purposes were supplied by RadChem (Johannesburg, South Africa) and were used without further distillation.