Accessible Published by De Gruyter June 22, 2016

Chemistry of aza- and diazafulvenium methides in heterocyclic synthesis

Teresa M. V. D. Pinho e Melo

Abstract

Aza- and diazafulvenium methide systems are versatile building blocks for the synthesis of pyrroles and pyrazoles. These extended dipoles participate in sigmatropic [1,8]H shifts and 1,7-electrocyclizations giving vinyl pyrroles and pyrazoles. Under flash vacuum pyrolysis conditions these heterocycles undergo interesting rearrangements. Aza- and diazafulvenium methides can be intercepted by dipolarophiles. Derivatives with carboxylate groups at C-4 and/or C-5 act exclusively as 1,7-dipoles affording products resulting from the addition across the 1,7-positions. These 1,7-cycloadducts include chlorin and bacteriochlorin type macrocycles as well as steroidal analogues, compounds with relevance in medicinal chemistry. In contrast with this chemical behavior, 5-trifluoromethylazafulvenium methides can participate in both 1,7- and 1,3-dipolar cycloadditions. The generation and reactivity of benzodiazafulvenium methides is also discussed.

Introduction

Pericyclic reactions of azafulvenium methides (1) and diazafulvenium methides (2) are an interesting and versatile approach to pyrroles and pyrazoles, including heterocyclic-annulated derivatives. These extended dipoles can be considered “higher-order” azomethine ylides or azomethine imides and, in principle, can act as 4π 1,3-dipoles or as 8π 1,7-dipoles. After the pioneer work of Storr et al. [1, 2] significant advances have been made, allowing the gathering of new data on the chemistry of these transient dipoles. A reactivity pattern, which depends on the nature of the substituents, has been established. Furthermore, the study has been extended to benzodiazafulveniummethides which are intermediates in the synthesis of indazole and imidazole derivatives.

Herein, the more relevant aspects of the chemistry of aza- and diazafulvenium methides are described, including the use of diazafulvenium methides as precursors of molecules with relevance in Medicinal Chemistry. The generation and reactivity of non-classical heterocyclic-fused-[c]thiazoles is also discussed.

Chemistry of 4,5-dimethoxycarbonyl aza- and diazafulvenium methides

Storr and co-workers found that the generation of azafulvenium methides 6 by the thermal extrusion of sulfur dioxide from the corresponding 2,2-dioxo-1H,3H-pyrrolo[1,2-c]thiazoles 5 could be achieved under flash vacuum pyrolysis (FVP) reaction conditions. They described the first evidence for trapping of transient azafulvenium methides in pericyclic reactions. Azafulvenium methides 6a and 6b undergo sigmatropic [1,8]H shifts giving vinylpyrroles 7 whereas the acyl derivatives 6c electrocyclize to give pyrrolo[1,2-c]-[1,3]oxazines 9. It was observed that the presence of the acyl group lowers the temperature required for the elimination of SO2 allowing the reaction to be carried out in solution at 200°C using 1,2,4-trichlorobenzene (1,2,4-TCB) as solvent (Scheme 1) [1, 2].

Scheme 1: Sigmatropic [1,8]H shifts and electrocyclization of azafulvenium methides giving vinylpyrroles and pyrrolo[1,2-c]-[1,3]oxazines, respectively.

Scheme 1:

Sigmatropic [1,8]H shifts and electrocyclization of azafulvenium methides giving vinylpyrroles and pyrrolo[1,2-c]-[1,3]oxazines, respectively.

In relation with our research on the synthesis of chiral 1H,3H-pyrrolo[1,2-c]thiazoles [35], we became interested in exploring the generation of azafulvenium methides in order to get further knowledge on the reactivity of these transient 8π 1,7-dipoles. We found that the vinylpyrroles 7 could be obtained from the corresponding 2,2-dioxo-1H,3H-pyrrolo[1,2-c]thiazoles under thermolysis in a sealed tube, without the need for FVP conditions. Interestingly, our FVP conditions led to a different outcome, the synthesis of methyl 5-oxo-5H-pyrrolizine-2-carboxylates 10. The pyrolysis of vinylpyrroles 7 also affords heterocycles 10, indicating that pyrroles 7 are intermediates in the formation of 5-oxo-5H-pyrrolizine-2-carboxylates 10 from sulfones 5 (Scheme 2) [6, 7].

Scheme 2: Chemical behavior of 1H,3H-pyrrolo[1,2-c]thiazoles 5a and 5c under thermolysis.

Scheme 2:

Chemical behavior of 1H,3H-pyrrolo[1,2-c]thiazoles 5a and 5c under thermolysis.

It is known that 2-substituted 3-(pyrrol-2-yl)propionate methyl esters undergo concerted elimination of methanol on FVP to give pyrrol-2-ylideneketenes which electrocyclize to give pyrrolizinones [8]. This led to postulate pyrrol-2-ylpropionates 11 as intermediates of the synthesis of 5-oxo-5H-pyrrolizines, formed from the N-vinylpyrroles 7 through a sequence of sigmatropic shifts (Scheme 3).

Scheme 3: Mechanism proposal for the synthesis of 5-oxo-5H-pyrrolizines 10 from N-vinylpyrroles 7.

Scheme 3:

Mechanism proposal for the synthesis of 5-oxo-5H-pyrrolizines 10 from N-vinylpyrroles 7.

Cheletropic extrusion of SO2 from 3-phenyl-2,2-dioxo-1H,3H-pyrrolo[1,2-c]thiazoles 13 under sealed tube solution thermolysis leads to 2-styryl-1H-pyrroles 16. In this case, azafulvenium methides are generated and participate in 1,7-electrocyclizations to afford bicyclic heterocycles, which undergo ring-opening reactions to give the final products (Scheme 4) [6, 7].

Scheme 4: Synthesis of 2-styryl-1H-pyrroles 16 via electrocyclization of azafulvenium methides 14.

Scheme 4:

Synthesis of 2-styryl-1H-pyrroles 16 via electrocyclization of azafulvenium methides 14.

Under FVP 3-phenyl-2,2-dioxo-1H,3H-pyrrolo[1,2-c]thiazole 13 are converted into 4-oxo-1,4-dihydro-1-azabenzo[f]azulene-3-carboxylates (e.g. 17). 2-Styryl-1H-pyrroles 16 are also converted into these heterocycles upon FVP via a mechanism with pyrrol-3-ylketenes (e.g. 18) as key intermediates which undergo electrocyclization followed by sigmatropic H-shifts affording 4-oxo-1,4-dihydro-1-azabenzo[f]azulenes (Scheme 5) [6, 7].

Scheme 5: Synthesis of 4-oxo-1,4-dihydro-1-azabenzo[f]azulene-3-carboxylate 17 from 1H,3H-pyrrolo[1,2-c]thiazole 13 and 2-styryl-1H-pyrrole 16 (R = Me).

Scheme 5:

Synthesis of 4-oxo-1,4-dihydro-1-azabenzo[f]azulene-3-carboxylate 17 from 1H,3H-pyrrolo[1,2-c]thiazole 13 and 2-styryl-1H-pyrrole 16 (R = Me).

The reactivity of 1,1-dimethyl- and 1-methyl-2,2-dioxo-1H,3H-pyrrolo[1,2-c]thiazole derivatives has also been reported [9, 10]. Azafulvenium methides generated from these sulfones undergo [1,8]H sigmatropic shifts to give C-vinylpyrroles. Interestingly, under FVP some of these C-vinylpyrroles are converted into 5-oxo-5H-pyrrolizines or functionalized C-allyl-1H-pyrroles. The rearrangement to allylpyrroles can be regarded as a formal insertion of the N-R2CH group into the C–C sigma bond involving the pyrrole C-2 carbon and the vinylic group via a sequence of sigmatropic shifts. Selected examples of this type of transformation, using optimized FVP conditions, are shown in Scheme 6.

Scheme 6: Synthesis of C-vinylpyrroles and C-allylpyrroles from 1-methyl-1H,3H-pyrrolo[1,2-c]thiazole derivatives.

Scheme 6:

Synthesis of C-vinylpyrroles and C-allylpyrroles from 1-methyl-1H,3H-pyrrolo[1,2-c]thiazole derivatives.

2,2-Dioxo-1H,3H-pyrazolo[1,5-c]thiazoles are masked diazafulvenium methides which can also be trapped in pericyclic reactions [1, 2, 11, 12]. In fact, 7,7-dimethyl- and 1-methyl-diazafulvenium methides 26, generated under FVP, are converted efficiently into C-vinylpyrazoles and N-vinylpyrazoles, respectively (Scheme 7) [11, 12].

Scheme 7: Sigmatropic [1,8]H shifts of diazafulvenium methides 26.

Scheme 7:

Sigmatropic [1,8]H shifts of diazafulvenium methides 26.

4,5-Dimethoxycarbonyl azafulvenium methides 6, generated under microwave irradiation, can be intercepted in [8π+2π] cycloadditions. In fact, only under microwave irradiation was it possible to observe for the first time azafulvenium methides acting as 1,7-dipole on reacting with dipolarophiles [13, 14]. 4,5-Dimethoxycarbonyl diazafulvenium methides 26 also behave as 8π 1,7-dipoles either under microwave irradiation or conventional heating, affording the corresponding 1,7-cycloadducts (Scheme 8) [1114].

Scheme 8: [8π + 2π] Cycloaddition of aza- and diazafulvenium methides.

Scheme 8:

[8π + 2π] Cycloaddition of aza- and diazafulvenium methides.

Storr et al. reported that diazafulvenium methide 26a did not react with N-phenylmaleimide (NPM) or dimethyl acetylenedicarboxylate but could be intercepted in [8π+2π] cycloadditions with silylated acetylenes giving adducts resulting from the addition across the 1,7-positions (Scheme 9) [2]. We could later observe that diazafulvenium methide 26a, unsubstituted at C-1 and C-7, react with both electron-rich and electron-deficient dipolarophiles. However, other derivatives react preferentially with electron-deficient dipolarophiles.

Scheme 9: [8π + 2π] Cycloaddition of diazafulvenium methide 26a with 1,2-bis(trimethylsilyl)ethyne.

Scheme 9:

[8π + 2π] Cycloaddition of diazafulvenium methide 26a with 1,2-bis(trimethylsilyl)ethyne.

For instance, 1,2-diazafulvenium methide 26b, derived from sulfone 25b under conventional heating, participates in [8π+2π] cycloadditions with a range of electron-deficient dipolarophiles including hererodipolarophiles to give functionalized pyrazolo-annulated systems in good yield (Scheme 10). These cycloadditions can also be carried out under microwave irradiation which results in a significant shortening of the reaction time (MW at 230°C in 1,2,4-TCB for 10 min) [1114].

Scheme 10: [8π + 2π] Cycloaddition of diazafulvenium methide 26b with electron-deficient dipolarophiles.

Scheme 10:

[8π + 2π] Cycloaddition of diazafulvenium methide 26b with electron-deficient dipolarophiles.

The microwave irradiation of 2,2-dioxo-1H,3H-pyrrolo[1,2-c]thiazoles in presence of dipolarophiles affords the corresponding 1,7-cycloadducts. As previously mentioned, only the microwave-induced reaction conditions allowed the trapping of 4,5-dimethoxycarbonyl azafulvenium methides (e.g. 6d) via [8π+2π] cycloadditions (Scheme 11) [13, 14].

Scheme 11: Microwave-induced [8π + 2π] cycloaddition of azafulvenium methide 6d with electron-deficient dipolarophiles.

Scheme 11:

Microwave-induced [8π + 2π] cycloaddition of azafulvenium methide 6d with electron-deficient dipolarophiles.

Chemistry of 5-trifluoromethyl-azafulvenium methides

The high electronegativity of the CF3 group results in very different electron density distribution and may lead to significant changes in the reactivity of molecules. In this context, the chemistry of 5-(trifluoromethyl)azafulvenium methides was also explored. In fact, in contrast with the reactivity observed for methoxycarbonyl aza- and diazafulvenium methides, 5-trifluoromethylazafulvenium methides can participate in both 1,7- and 1,3-dipolar cycloadditions leading to trifluoromethylpyrrole-annulated systems Scheme 12 [1517].

Scheme 12: Cycloaddition of 5-trifluoromethyl-azafulvenium methides.

Scheme 12:

Cycloaddition of 5-trifluoromethyl-azafulvenium methides.

5-(Trifluoromethyl)azafulvenium methide (36a, R=H) showed site selectivity in the reaction with strong electron-deficient dipolarophiles, such as DMAD and NPM, leading exclusively to 1,3-cycloadducts. However, in the cycloaddition with less activated dipolarophiles, 1,7-cycloadducts resulting from [8π+2π] cycloaddition are also formed Scheme 13 [15].

Scheme 13: Cycloaddition of azafulvenium methide 36a with electron-deficient dipolarophiles.

Scheme 13:

Cycloaddition of azafulvenium methide 36a with electron-deficient dipolarophiles.

The microwave-induced reaction of 1-benzyl-5-trifluoromethyl-azafulvenium methide (36b, R=Bn) with DMAD affords a mixture of 1,3- and 1,7-cycloadducts 39 and 40 in 89:11 ratio, respectively. Two cycloadducts are also obtained in a diastereoselective manner from the reaction with N-substituted maleimides. However, the major products resulted from the addition across the 1,7-position, with selectivity opposite to the one observed with DMAD (Scheme 14).

Scheme 14: Reactivity of 5-trifluoromethyl-azafulvenium methide 36b.

Scheme 14:

Reactivity of 5-trifluoromethyl-azafulvenium methide 36b.

Calculations carried out at the DFT level, considering cycloadditions of 1-benzyl-5-trifluoromethyl-azafulvenium methide with N-substituted maleimides as model reactions, revealed that the exo-cycloaddition is the main reaction path for the 1,7-cycloaddition, while the endo approach is the main mode of reaction leading to 1,3-cycloadducts. The higher selectivity of 5-trifluoromethyl-azafulvenium methide for the formation of 1,3-cycloadducts when compared with 1-benzyl-5-trifluoromethyl-azafulvenium methide, bearing an additional benzyl group at C-1, indicates that a combination of electronic and steric factors determines the reaction outcome [16]. 1-Methyl-5-(trifluoromethyl)azafulvenium methide showed similar chemoselectivity towards N-substituted maleimides [17].

Under FVP or conventional thermolysis 1-methyl- and 1-benzyl-5-trifluoromethyl-azafulvenium methide undergoes allowed suprafacial sigmatropic [1,8]H shifts leading to the efficient formation of 2-methyl-3-(trifluoromethyl)-1-vinyl-1H-pyrrole and 2-methyl-1-styryl-3-trifluoromethyl-1H-pyrrole (43), respectively [16, 17].

Reactivity of azafulvenium methides with internal dipolarophiles

The reactivity of azafulvenium methides with internal dipolarophiles was described. It was observed that under microwave irradiation 4,5-dimethoxycarbonylazafulvenium methide, with a prop-2-ynyloxyphenyl substituent, are converted into 2-[2-(2H-chromen-8-yl)vinyl]-1H-pyrrole 45. The process involves the formation of styryl-1H-pyrrole 46, derived from the electrocyclization of the in situ generated 1,7-dipole, followed by the rearrangement of the aryl propargyl ether moiety to form the chromene unit by an initial Claisen rearrangement (Scheme 15) [18].

Scheme 15: Synthesis of 2-[2-(2H-chromen-8-yl)vinyl]-5-methyl-1Hpyrrole 45.

Scheme 15:

Synthesis of 2-[2-(2H-chromen-8-yl)vinyl]-5-methyl-1Hpyrrole 45.

4,5-Dimethoxycarbonylazafulvenium methide with a more-activated dipolarophile, as well as the more-activated 5-(trifluoromethyl)azafulvenium methide with an unactivated alkyne, could be trapped by [8π+2π] cycloadditions to afford 8,12a-dihydro-6H-chromeno[4,3-e]indolizines 50 and 53. From these reactions, styryl-1H-pyrrole derivatives were also isolated (Scheme 16) [18].

Scheme 16: Microwave-induced thermolysis of 1H,3H-pyrrolo[1,2-c]-thiazole 2,2-dioxides 49 and 52.

Scheme 16:

Microwave-induced thermolysis of 1H,3H-pyrrolo[1,2-c]-thiazole 2,2-dioxides 49 and 52.

Interestingly, 5-(trifluoromethyl)azafulvenium methide bearing a 2-[(3-methoxycarbonylprop-2-enyl)oxy]phenyl substituent at C-1 participated in both intramolecular [8π+2π] and [4π+2π] cycloadditions to form the corresponding tetrahydro-6H-chromeno[4,3-e]indolizine 56 and 10-tetrahydrochromeno[3,4-b]pyrrolizine 57, respectively (Scheme 17) [18].

Scheme 17: Reactivity of 7-(trifluoromethyl)-1H,3H-pyrrolo[1,2-c]-thiazole-2,2-dioxide (55).

Scheme 17:

Reactivity of 7-(trifluoromethyl)-1H,3H-pyrrolo[1,2-c]-thiazole-2,2-dioxide (55).

The importance of the chromene derivatives led us to explore the reactivity of 1H,3H-pyrrolo[1,2-c]thiazoles 58 under microwave irradiation as an approach to novel chromene derivatives (Scheme 18) [18]. In fact, the chromene (benzopyran) scaffold is frequently found in several natural products and bioactive compounds and has attracted great attention [1922]. In addition, due to their spectroscopic properties, chromene derivatives can also be used as photochromes and dyes [23, 24]. Starting from chiral 1H,3H-pyrrolo[1,2-c]thiazoles 58, instead of the corresponding sulfones, the competitive sulfur dioxide extrusion was avoided. The microwave-induced rearrangement of 1H,3H-pyrrolo[1,2-c]thiazoles 58 allowed the efficient synthesis of chiral 1H,3H-pyrrolo[1,2-c]thiazoles 59 and 60 bearing a chromene substituent at C-3. It is worth noting that the rearrangement of the prop-2-ynyloxy derivative was more efficient than the rearrangement of the compound bearing a (3-ethoxycarbonylprop-2-ynyl)oxy substituent.

Scheme 18: Synthesis of 3-(2H-chromen-8-yl)-1H,3H-pyrrolo[1,2-c]-thiazoles 59 and 60.

Scheme 18:

Synthesis of 3-(2H-chromen-8-yl)-1H,3H-pyrrolo[1,2-c]-thiazoles 59 and 60.

Chemistry of benzodiazafulvenium methides

The generation and reactivity of a novel class of diazafulvenium methides generated from thiazolo[3,4-b]indazole-2,2-dioxides, the benzo-2,3-diazafulvenium methides, has been recently described. The aim of this study was to explore new extended dipolar systems capable of participating in pericyclic reactions, as a way of developing new synthetic routes to indazoles [25].

Thiazolo[3,4-b]indazoles 62 were obtained by cycloaddition of sydnones 61 with benzyne and converted into the corresponding sulfones 63, the precursors of the target benzo-2,3-diazafulvenium methides (Scheme 19).

Scheme 19: Synthesis of 1,3-dihydrothiazolo[3,4-b]indazoles 62 and 1,3-dihydrothiazolo[3,4-b]indazole-2,2-dioxides 63.

Scheme 19:

Synthesis of 1,3-dihydrothiazolo[3,4-b]indazoles 62 and 1,3-dihydrothiazolo[3,4-b]indazole-2,2-dioxides 63.

It was observed that benzo-2,3-diazafulvenium methides behave as 1,3-dipoles on reacting with N-substituted maleimides affording 1H-indazoles 67 in moderate yields (Scheme 20) [25]. This process can be rationalized considering that the initially formed benzodiazafulvenium methides 64 react with maleimides affording cycloadducts 65, resulting from the addition across the 1,3-position. Cycloadducts 65 undergo pyrazolidine ring-opening, followed by a sigmatropic H-shift giving the more stable 1H-indazoles 67. This chemical behavior is in contrast with the previously observed reactivity for 4,5-methoxycarbonyl-diazafulvenium methides, which participate exclusively in [8π+2π] cycloadditions to give 1,7-cycloadducts. Quantum chemical calculations, carried out at the DFT level of theory, were in agreement with the observed reactivity.

Scheme 20: Reaction of benzo-2,3-diazafulvenium methides 64 with N-substituted maleimides.

Scheme 20:

Reaction of benzo-2,3-diazafulvenium methides 64 with N-substituted maleimides.

Under FVP or under microwave irradiation 1-methyl- and 7,7-dimethyl-benzo-2,3-diazafulvenium methides undergo sigmatropic [1,8]H shifts allowing the efficient synthesis of N-vinyl- and C-vinyl-2H-indazoles (Scheme 21) [25].

Scheme 21: Chemical behavior of 1,3-dihydrothiazolo[3,4-b]indazole-2,2-dioxides 68 under FVP.

Scheme 21:

Chemical behavior of 1,3-dihydrothiazolo[3,4-b]indazole-2,2-dioxides 68 under FVP.

Aiming to broaden this chemistry, the generation and reactivity of novel benzo-2,5-diazafulvenium methides was explored. The synthetic strategy selected to achieve this goal involved the use of thiazolo[3,4-a]benzimidazole-2,2-dioxides as the dipole precursor [26].

The first evidence for the generation of benzo-2,5-diazafulvenium methides came from the chelotropic extrusion of SO2 from 3-methyl-1H,3H-thiazolo[3,4-a]benzimidazole-2,2-dioxide (71) in the presence of N-substituted maleimides under conventional solution thermolysis and under microwave-induced conditions leading to benzo[d]imidazole annulated heterocycles 73 (Scheme 22) [26].

Scheme 22: [8π + 2π] Cycloaddition of benzo-2,5-diazafulvenium methide 72 with N-substituted maleimides.

Scheme 22:

[8π + 2π] Cycloaddition of benzo-2,5-diazafulvenium methide 72 with N-substituted maleimides.

Benzo-2,5-diazafulvenium methides, having one methyl group at C-7 or alternatively one methyl or benzyl group a C-1, have the structural requirements to participate in a sigmatropic [1,8]H shift. In fact, FVP of sulfones 74 carried out at 650°C afforded vinylbenzo[d]imidazoles 76 in good yield (Scheme 23) [26].

Scheme 23: Sigmatropic [1,8]H shifts of benzo-2,5-diazafulvenium methides derived from thiazolo[3,4-a]benzimidazole-2,2-dioxides 74 leading to N-vinylbenzimidazoles 75.

Scheme 23:

Sigmatropic [1,8]H shifts of benzo-2,5-diazafulvenium methides derived from thiazolo[3,4-a]benzimidazole-2,2-dioxides 74 leading to N-vinylbenzimidazoles 75.

Non-classical heterocyclic-fused-[c]thiazoles reacting as thiocarbonyl ylides and azomethine ylides

Pummerer-type dehydration of 2-oxo-1H,3H-pyrrolo[1,2-c]thiazole 76a and 2-oxo-1H,3H-pyrazolo[1,5-c]thiazole 76b generates the non-classical heterocyclic-fused-[c]thiazoles 77a and 77b, respectively. These intermediates, containing a tetravalent sulfur atom, participate in [4π+2π] cycloadditions to give nitrogen-bridged heterocyclic compounds namely, pyrazolo[1,5-a]pyridines, thiazolo[2,3,4-cd]pyrrolizines and indolizines. Heteropentalene 77a reacts as thiocarbonyl ylide with N-substituted maleimides, whereas with dimethyl acetylene dicarboxylate it acts as an azomethine ylide. Non-classical 1H,3H-pyrazolo[1,5-c]thiazole 77b reacts with both types of dipolarophiles, giving products resulting from the addition across the thiocarbonyl ylide moiety (Scheme 24) [2729].

Scheme 24: Generation and [4π + 2π] cycloaddition of the heteropentalene 77 with N-substituted maleimides and dimethyl acetylene dicarboxylate.

Scheme 24:

Generation and [4π + 2π] cycloaddition of the heteropentalene 77 with N-substituted maleimides and dimethyl acetylene dicarboxylate.

The generation and reactivity of heteropentalene 77c–77e was also reported leading to a range of nitrogen-heterocycles. For the first time, one non-classical pyrrolo[1,2-c]thiazole 77c was isolated and its structure determined by X-ray crystallography [29].

Interestingly, the heteropentalene derived from 2-oxo-1H,3H-pyrrolo[1,2-c]thiazole 78, bearing a trifluoromethyl group, participates in [4π+2π] cycloadditions with N-substituted maleimides, but only as an azomethine ylide. The reaction with NPM led to a mixture of the exo- and endo-cycloadducts 79 (R=Ph). Sigmatropic H shift of these cycloadducts afforded the final products 80a in moderate yield. The dehydration of sulfoxide 78 in the presence of N-methylmeleimide afforded the endo-cycloadduct 80b exclusively (Scheme 25) [29].

Scheme 25: Cycloaddition of the heteropentalene generated from 2-oxo-1H,3H-pyrrolo[1,2-c]thiazole 78 with N-substituted maleimides.

Scheme 25:

Cycloaddition of the heteropentalene generated from 2-oxo-1H,3H-pyrrolo[1,2-c]thiazole 78 with N-substituted maleimides.

Synthesis of molecules with relevance in medicinal chemistry

4,5,6,7-Tetrahydropyrazolo[1,5-a]pyridine fused chlorin derivatives as theranostics agents for cancer

Porphyrins, have a rich pattern of absorption bands, present excellent interaction properties with human fluids and cells, being accumulated in tumors. On the other hand, hydroporphyrins such as chlorins and bacteriochlorins present strong absorbance within the phototherapeutic window (600–800 nm), with good yields for singlet oxygen generation and therefore are preferable photosensitizers to be used in photodynamic therapy (PDT) of malignant diseases. In this context, we explored the reactivity of porphyrins towards diazafulvenium methides, which led to the synthesis of a new type of stable 4,5,6,7-tetrahydropyrazolo[1,5-a]pyridine-fused chlorins and bacteriochlorins, via unprecedented [8π+2π] cycloaddition, as potential photodynamic agents. Tetraarylporphyrins 81 reacted with diazafulvenium methide 26a affording chlorins 82 as single products. The best reaction conditions were achieved by carrying out the reaction with excess of porphyrin (2 equiv) under microwave irradiation for 20 min at 250°C. Chlorins 82 participated in [8π+2π] cycloaddition with diazafulvenium methide 26a giving bacteriochlorins 83 in a selective fashion. In this case, conventional heating proved to be more efficient than the microwave methodology (Scheme 26) [30, 31].

Scheme 26: [8π + 2π] Cycloaddition of diazafulvenium methide 26a with porphyrins and chlorins.

Scheme 26:

[8π + 2π] Cycloaddition of diazafulvenium methide 26a with porphyrins and chlorins.

PDT has been successfully used in the treatment of skin cancers, however, treatment of melanoma with this method can be compromised due to the natural resistance mechanism of some melanoma cancer cells [32, 33]. In particular, high melanin levels in such pigmented tumors can lead to optical interference via competition with the photosensitizer for light absorption. This, together with the antioxidant effect of melanin, can affect the efficiency of PDT. Melanin is the dominant absorber in the 500–600 nm spectral window and therefore photosensitizers that absorb at longer wavelengths are interesting targets. On the other hand, UV-Vis absorption spectra of these chlorins 82 and bacteriochlorins revealed intense absorption and emission bands at 650 nm and 730 nm, respectively. These favorable photophysical characteristics, combined with high stability, led us to evaluate the phototoxicity of some of these tetrahydropyrazolo[1,5-a]pyridine-fused chlorins, the sensitizer 82b, and the more hydrophilic derivative 84 in melanoma cell lines (Scheme 27) [34].

Scheme 27: Synthesis of meso-tetraarylchlorin 84 and IC50 values for chlorins 82b and 85 in human melanoma cell lines (A375 and C32).

Scheme 27:

Synthesis of meso-tetraarylchlorin 84 and IC50 values for chlorins 82b and 85 in human melanoma cell lines (A375 and C32).

Preliminary studies on the photodynamic activity of 4,5,6,7-tetrahydropyrazolo[1,5-a]pyridine-fused chlorins 82b and 84 proved this class of compounds to be very active as photodynamic agents against melanotic (A375) and amelanotic (C32) cancer cells [34]. Interestingly, di(hydroxymethyl)-chlorin 84 was particularly active against human melanocytic melanoma cells (IC50=31 nM). Comparison between 82b and 84 demonstrated that the increased hydrophilicity of the latter is crucial to ensure nanomolar activity against melanoma cells. This difference in photocytotoxic activity can be related with the difference observed in cellular uptake, the latter showing cellular uptake values 10-times higher than chlorin 82b. The metabolic activity studies of A375 and C32 cells after photodynamic treatment with the chlorins, both in the presence and absence of the singlet oxygen quencher sodium azide and in the presence of the superoxide scavenger D-mannitol, were carried out. Addition of these inhibitors resulted in a decrease in the growth inhibition rate. Thus, both singlet oxygen and superoxide must be involved in the observed photodynamic activity. Experiments with A375 and C32 cells in the absence of light demonstrated that the cytotoxicity is light-dependent. PDT studies with different irradiances demonstrated that the cytotoxicity is also light-dose dependent.

NIR emitters are particularly important as their light output is in a region where organisms are nearly transparent. The incorporation of high atomic number metals such as platinum, can enhance the triplet state properties of tetrapyrrolic macrocycles, and in many cases leads to long wavelength, room temperature phosphorescence. In this context, NIR luminescent compounds based on platinum(II) derivatives of 4,5,6,7-tetrahydropyrazolo[1,5-a]pyridine-fused chlorins 86 and 87 have been prepared as outlined in Scheme 28 [35].

Scheme 28: Synthesis of Pt(II) 4,5,6,7-tetrahydropyrazolo[1,5-a]pyridine–fused chlorins 86 and 87.

Scheme 28:

Synthesis of Pt(II) 4,5,6,7-tetrahydropyrazolo[1,5-a]pyridine–fused chlorins 86 and 87.

It was observed that the intensity of the phosphorescence of chlorins 86 and 87 is strongly quenched by the presence of oxygen. In contrast, the fluorescence is relatively unaffected. This can provide a quantitative measure of oxygen concentration inside cells, allowing distinction between normal and malignant cells. Finally, the phosphorescence lifetime is also reduced in the presence of oxygen, which makes this class of compounds suitable for applications in fluorescence lifetime imaging microscopy. Singlet oxygen formation quantum yield for chlorin 87 (0.58) indicated the potential of this compound for applications in PDT. Therefore, these novel and stable near infrared luminescent compounds based on platinum(II) derivatives of 4,5,6,7-tetrahydropyrazolo[1,5-a]pyridine-fused chlorins are likely to be useful for theranostics.

Synthesis of steroid fused heterocycles

Steroids are a widely important class of both naturally occurring and synthetic compounds with a great diversity of applications in human physiology and medicine. On the other hand, synthetic steroids fused to heterocyclic compounds at positions 16 and 17 of the D-ring have unique biological properties. In this context, [8π+2π] cycloaddition of steroidal derivatives with diazafulvenium methides was explored as a route to chiral hexacyclic steroids. 16-Dehydropregnenolone acetate (16-DPA, 91) reacted with diazafulvenium methides 90, generated in situ from the corresponding 2,2-dioxo-1H,3H-pyrazolo[1,5-c][1,3]thiazole, under microwave irradiation or in refluxing 1,2,4-trichlorobenzene to afford steroids 92 in a regio- and stereoselective fashion (Scheme 29) [36].

Scheme 29: Synthesis of hexacyclic steroids from 16-dehydropregnenolone acetate via [8π + 2π] cycloaddition of diazafulvenium methides.

Scheme 29:

Synthesis of hexacyclic steroids from 16-dehydropregnenolone acetate via [8π + 2π] cycloaddition of diazafulvenium methides.

The thermal reactivity of 3-benzyl-2,2-dioxo-1H,3H-pyrazolo[1,5-c][1,3]thiazole 93 in the presence of 16-DPA was then explored. In this case, along with the desired 1,7-cycloadduct 94 the competitive synthesis of N-styryl-1H-pyrazole 95, resulting from an allowed suprafacial sigmatropic [1,8]H shift of the in situ generated diazafulvenium methide, was observed. Hexacyclic steroid 94, having an additional chiral center (C-24), was obtained regio- and stereoselectively. The diazafulvenium methide derived from sulfone 25b, reacted with 16-DPA via [8π+2π] cycloadditon leading to stereoisomers 96 and 97. From these reactions N-vinylpyrazoles were also obtained (Scheme 30).

Scheme 30: Synthesis of hexacyclic steroids from 16-dehydropregnenolone acetate via [8π + 2π] cycloaddition of diazafulvenium methides.

Scheme 30:

Synthesis of hexacyclic steroids from 16-dehydropregnenolone acetate via [8π + 2π] cycloaddition of diazafulvenium methides.

The cycloadducts, obtained exclusively or as the major product, could result from an endo cycloaddition of conformers II or alternatively could be formed via an exo cycloaddition of conformers I, which have the 1-substituent group pointing inward, considering the approach of the dipole by the less hindered α-face of the steroid (Scheme 31).

Scheme 31: Exo and endo [8π + 2π] cycloaddition of 16-DPA with 1-substituted diazafulvenium methides, considering the approach of the dipole by the α-face.

Scheme 31:

Exo and endo [8π + 2π] cycloaddition of 16-DPA with 1-substituted diazafulvenium methides, considering the approach of the dipole by the α-face.

In order to be able to rationalize the stereoselectivity observed in the [8π+2π] cycloadditions of 16-DPA towards 1-methyl- and 1-benzyldiazafulvenium methides, quantum chemical calculations were carried out at the DFT level of theory, considering the cycloadditions of these diazafulvenium methides with NPM as the model reactions. Starting from sulfone 25b two racemic diastereoisomeric products were obtained, compound 98 isolated as the major product and compound 99, whereas the microwave-induced reaction of 2,2-dioxo-1H,3H-pyrazolo[1,5-c]thiazole 93 with NPM affords 1,7-cycloadduct 100 as single product (Scheme 32). The molecular structure of this compound was unambiguously established by X-ray crystallography.

Scheme 32: Cycloaddition of diazafulvenium methides generated from 1H,3H-pyrazolo[1,5-c]thiazole-2,2-dioxides 25b and 93 with N-phenylmaleimide.

Scheme 32:

Cycloaddition of diazafulvenium methides generated from 1H,3H-pyrazolo[1,5-c]thiazole-2,2-dioxides 25b and 93 with N-phenylmaleimide.

DFT calculations show that conformers II of the studied diazafulvenium methides are more stable than conformers I. Energy barriers of the transition states associated with these involving the exo approach, calculated at the B3LYP/6-31G(d) level of theory and considering the lower energy conformer of each compound, are unfavorable by more than 20 kJ/mol relative to the alternative endo approach. Therefore, we could conclude that 16-dehydropregnenolone acetate reacted with diazafulvenium methides to afford chiral 4,5,6,7-tetrahydropyrazolo[1,5-a]pyridine-fused steroids via endo selective [8π+2π] cycloadditions, with the approach of the dipole for the less hindered α-face of the steroid.

Chiral 6,7-bis(hydroxymethyl)-1H,3H-pyrrolo[1,2-c]thiazoles with anti-breast cancer properties

Fused hetero-bicyclic systems represent a class of compounds of major relevance in Medicinal Chemistry due to their broad spectrum of biological activities. Among these, pyrrolo-thiazoles and bis(hydroxymethyl)pyrrole derivatives have been proved to possess interesting anti-tumor properties (e.g. anti-leukemia, anti-breast cancer) [37, 38]. On the other hand, among cytotoxic agents, bifunctional alkylating drugs (DNA cross-linking agents) are widely used as chemotherapeutic agents in clinic [39].

Having access to a wide range of 1H,3H-pyrrolo[1,2-c]thiazoles, precursors of azafulvenium methides, chiral hydroxymethyl-1H,3H-pyrrolo[1,2-c]thiazole derivatives were explored as promising anti-breast cancer agents [4042].

Structure–activity relationship (SAR) studies of several 1H,3H-pyrrolo[1,2-c]thiazole derivatives (e.g. 101–106) against breast adenocarcinoma MCF7 cell line allowed the establishment of structural features for antitumor activity, Fig. 1. It was demonstrated that only 1H,3H-pyrrolo[1,2-c]thiazoles bearing hydroxymethyl substituents show in vitro anticancer activity. On the other hand, the position of the hydroxymethyl substituent is crucial. In fact, derivative 101 with this group at C-6 is active whereas the C-7 substituted derivative 102 showed low activity. Nevertheless, the presence of two hydroxymethyl groups leads to a slight improvement in activity against MCF7 breast cancer (BC) cell lines as illustrated by 6,7-bis(hydroxymethyl)-1H,3H-pyrrolo[1,2-c]thiazole 103 with IC50 value of 1.1 μM. Furthermore, the combined presence of a phenyl group at C-3 and a methyl group at C-5 in the 1H,3H-pyrrolo[1,2-c]thiazole ring system is essential to ensure high cytotoxicity. Interestingly, pyrrolo[1,2-c]thiazole 106 showed even better performance against BC cell lines than the corresponding enantiomer 103 [40, 41].

Fig. 1: Cytoxicity against breast cancer human cell line MCF7 (72 h incubation time) [40, 41].

Fig. 1:

Cytoxicity against breast cancer human cell line MCF7 (72 h incubation time) [40, 41].

In an effort to produce new structures with better anticancer activity based on lead structure 103, we synthesized and evaluated the effect of replacing the phenyl substituent at C-3 by a 4-methoxyphenyl group 107a and by the more hydrophilic 4-hydroxyphenyl group 107b (Fig. 2) [42]. The in vitro cytotoxicity on several human breast cancer cell lines (MCF7, HCC1954 and HCC1806 cell lines) was studied. Particularly interesting were the results obtained for compound 109b, which proved to be the most promising compound regarding HCC1806 cell line, a triple-negative breast cancer (TNBC). TNBC is an extremely aggressive form of BC, occurs more often in younger women and is difficult to treat successfully as it lacks the receptors used in the currently available therapies. TNBC patients present a higher incidence of metastases and a shorter time of recurrence relatively to other BC types. The lack of adequate therapeutic response for TNBC makes the search new anti-cancer agents against this type BC particularly relevant. Additional studies in order to understand the mechanism of action and the impact on cancer biology, namely the effects of these compounds on cell survival, viability, cell cycle, DNA damage and expression of proteins related to cell death pathways were also performed indicating that these compounds may induce DNA damage.

Fig. 2: Cytoxicity against breast cancer human cell line MCF7 and HCC1808 (72 h incubation time) [42].

Fig. 2:

Cytoxicity against breast cancer human cell line MCF7 and HCC1808 (72 h incubation time) [42].

Conclusion

The chemistry of aza-, diazafulvenium and benzodiazafulvenium methides discussed in this review has demonstrated the potential of pericyclic reactions of these reactive intermediates as an approach to a wide range of heterocyclic compounds namely pyrroles, pyrazoles, indazoles and imidazoles. Chlorin and bacteriochlorin type macrocycles, as well as steroidal analogues, are among the synthesized heterocycles, compounds with relevance in medicinal chemistry.

Acknowledgments

Thanks are due to Coimbra Chemistry Centre (CQC), supported by the Portuguese Agency for Scientific Research, “Fundação para a Ciência e a Tecnologia” (FCT), through Project N° 007630 UID/QUI/00313/2013, co-funded by COMPETE2020-UE.

References

[1] O. B. Sutcliffe, R. C. Storr, T. L. Gilchrist, P. Rafferty, A. P. A. Crew. Chem.Comm.2000, 675 (2000). Search in Google Scholar

[2] O. B. Sutcliffe, R. C. Storr, T. L. Gilchrist, P. Rafferty. J. Chem. Soc. Perkin Trans. 12001, 1795 (2001). Search in Google Scholar

[3] T. M. V. D. Pinho e Melo, D. M. Barbosa, P. J. R. S. Ramos, A. M. d’A. Rocha Gonsalves, T. L. Gilchrist, A. M. Beja, J. A. Paixão, M. R. Silva, L. Alte da Veiga. J. Chem. Soc. Perkin Trans. I1999, 1219 (1999). Search in Google Scholar

[4] T. M. V. D. Pinho e Melo, M. I. L. Soares, A. M. d’A. Rocha Gonsalves, J. A. Paixão, A. M. Beja, M. Ramos Silva, L. Alte da Veiga, J. Costa Pessoa. J. Org. Chem. 67, 4045 (2002). Search in Google Scholar

[5] T. M. V. D. Pinho e Melo, C. S. B. Gomes, A. M. d’A. Rocha Gonsalves, J. A. Paixão, A. M. Beja, M. Ramos Silva, L. Alte da Veiga. Tetrahedron58, 5093 (2002). Search in Google Scholar

[6] T. M. V. D. Pinho e Melo, M. I. L. Soares, A. M. d’A. Rocha Gonsalves, H. McNab. Tetrahedron Lett.45, 3889 (2004). Search in Google Scholar

[7] T. M. V. D. Pinho e Melo, M. I. L. Soares, A. M. d’A. Rocha Gonsalves, J. A. Paixão, A. Matos Beja, M. Ramos Silva. J. Org. Chem.70, 6629 (2005). Search in Google Scholar

[8] H. McNab, S. Parson, E. Stevenson. J. Chem. Soc., Perkin Trans.11999, 2047 (1999). Search in Google Scholar

[9] T. M. V. D. Pinho e Melo, M. I. L. Soares, C. M. Nunes. Tetrahedron63, 1833 (2007). Search in Google Scholar

[10] M. I. L. Soares, S. M. M. Lopes, P. F. Cruz, R. M. M. Brito, T. M. V. D. Pinho e Melo. Tetrahedron64, 9745 (2008). Search in Google Scholar

[11] T. M. V. D. Pinho e Melo, M. I. L. Soares, A. M. d’A. Rocha Gonsalves. Tetrahedron Lett. 47, 791 (2006). Search in Google Scholar

[12] T. M. V. D. Pinho e Melo, C. M. Nunes, M. I. L. Soares, J. A. Paixão, A. Matos Beja, M. Ramos Silva. J. Org. Chem. 72, 4406 (2007). Search in Google Scholar

[13] M. I. L. Soares, T. M. V. D. Pinho e Melo. Tetrahedron Lett.49, 4889 (2008). Search in Google Scholar

[14] M. I. L. Soares, T. M. V. D. Pinho e Melo. Curr. Microw. Chem.1, 22 (2014). Search in Google Scholar

[15] C. M. Nunes, M. Ramos Silva, A. Matos Beja, R. Fausto, T. M. V. D. Pinho e Melo. Tetrahedron Lett.51, 411 (2010). Search in Google Scholar

[16] W. J. Peláez, T. M. V. D. Pinho e Melo. Tetrahedron69, 3646 (2013). Search in Google Scholar

[17] W. J. Peláez, A. J. Pepino, G. A. Argüello, T. M. V. D. Pinho e Melo. Eur. J. Org. Chem.2014, 2933 (2014). Search in Google Scholar

[18] F. M. R. Laia, M. I. L. Soares, C. S. B. Gomes, T. M. V. D. Pinho e Melo. Eur. J. Org. Chem.2015, 1341 (2015). Search in Google Scholar

[19] T. Lee, Y. D. Gong. Molecules17, 5467 (2012). Search in Google Scholar

[20] S. A. Patil, R. Patil, L. M. Pfeffer, D. D. Miller. Future Med. Chem.5, 1647 (2013). Search in Google Scholar

[21] N. Thomas, S. M. Zachariah. Asian J. Pharm. Clin. Res.6, Suppl. 2, 11 (2013). Search in Google Scholar

[22] K. C. Nicolaou, J. A. Pfefferkorn, A. J. Roecker, G. Q. Cao, S. Barluenga, H. J. Mitchell. J. Am. Chem. Soc.122, 9939 (2000). Search in Google Scholar

[23] S. V. Paramonov, V. Lokshin, O. A. Fedorova. J. Photochem. Photobiol. C12, 209 (2011). Search in Google Scholar

[24] N. Azizi, M. Mariami, M. Edrisi. Dyes Pigment.100, 215 (2014). Search in Google Scholar

[25] M. I. L. Soares, C. M. Nunes, C. S. B. Gomes, T. M. V. D. Pinho e Melo. J. Org. Chem.78, 628 (2013). Search in Google Scholar

[26] M. I. L. Soares, T.M. V. D. Pinho e Melo. Tetrahedron71, 4227 (2015). Search in Google Scholar

[27] J. M. Kane. J. Org. Chem.45, 5396 (1980). Search in Google Scholar

[28] O. B. Sutcliffe, R. C. Storr, T. L. Gilchrist, P. Rafferty. Tetrahedron56, 10011 (2000). Search in Google Scholar

[29] M. I. L. Soares, C. S. B. Gomes, T. M. V. D. Pinho e Melo. Tetrahedron67, 8392 (2011). Search in Google Scholar

[30] N. A. M. Pereira, A. C. Serra, T. M. V. D. Pinho e Melo. Eur. J. Org. Chem.2010, 6539 (2010). Search in Google Scholar

[31] N. A. M. Pereira, S. M. Fonseca, A. C. Serra, T. M. V. D. Pinho e Melo, H. D. Burrows. Eur. J. Org. Chem.2011, 3970 (2011). Search in Google Scholar

[32] Y.-Y. Huang, D. Vecchio, P. Avci, R. Yin, M. Garcia-Diaz, M. R. Hamblin. Biolog. Chem.394, 239 (2013). Search in Google Scholar

[33] A. Kawczyk-Krupka, A. M. Bugaj, W. Latos, K. Zaremba, A. Sieroń. Photodiagnosis and Photodyn. Ther.10, 503 (2013). Search in Google Scholar

[34] N. A. M. Pereira, M. Laranjo, M. Pineiro, A. C. Serra, K. Santos, R. Teixo, A. M. Abrantes, A. C. Gonçalves, A. B. Sarmento Ribeiro, M. F. Botelho, T. M. V. D. Pinho e Melo. Eur. J. Med. Chem.103, 374 (2015). Search in Google Scholar

[35] Stable Metal Organic Near Infrared Light Emitting Platinum(II) Derivatives of 4,5,6,7-Tetrahydropyrazolo[1,5-a]pyridine-fused Chlorins for Imaging, Sensing and Optoelectronics, Provisional Patent Application n° 20161000013344 (2016/02/24, Portuguese). Inventors: Hugh D. Burrows, Teresa M. V. D. Pinho e Melo, Arménio C. Serra, Nelson A. M. Pereira, Liliana M. Martelo, Marta Pineiro. Applicant: University of Coimbra. Search in Google Scholar

[36] S. M. M. Lopes, C. F. O. Correia, S. C. C. Nunes, N. A. M. Pereira, A. R. F. Ferreira, E. P. Sousa, C. S. B. Gomes, J. A. R. Salvador, A. A. C. C. Pais, T. M. V. D. Pinho e Melo. Org. Biomol. Chem.13, 9127 (2015). Search in Google Scholar

[37] M. L. Fascio, M. I. Errea, N. B. D’Accorso. Eur. J. Med. Chem.90, 666 (2015). Search in Google Scholar

[38] O. D. Scharer. ChemBioChem6, 27 (2005). Search in Google Scholar

[39] T. L. Su, T. C. Lee, R. Kakadiya. Eur. J. Med. Chem.69, 609 (2013). Search in Google Scholar

[40] M. I. L. Soares, A. F. Brito, M. Laranjo, A. M. Abrantes, M. F. Botelho, J. A. Paixão, A. M. Beja, M. R. Silva, T. M. V. D. Pinho e Melo. Eur. J. Med. Chem.45, 4676 (2010). Search in Google Scholar

[41] M. I. L. Soares, A. F. Brito, M. Laranjo, J. A. Paixao, M. F. Botelho, T. M. V. D. Pinho e Melo. Eur. J. Med. Chem.60, 254 (2013). Search in Google Scholar

[42] K. Santos, M. Laranjo, A. M. Abrantes, A. F. Brito, C. Goncalves, A. B. Sarmento Ribeiro, M. Filomena Botelho, M. I. L. Soares, A. S. R. Oliveira, T. M. V. D. Pinho e Melo. Eur. J. Med. Chem.79, 273 (2014). Search in Google Scholar

Published Online: 2016-6-22
Published in Print: 2016-5-1

©2016 IUPAC & De Gruyter. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. For more information, please visit: http://creativecommons.org/licenses/by-nc-nd/4.0/