A One-Pot Approach to Novel Pyridazine C-Nucleosides

The synthesis of glycosides and modified nucleosides represents a wide research field in organic chemistry. The classical methodology is based on coupling reactions between a glycosyl donor and an acceptor. An alternative strategy for new C-nucleosides is used in this approach, which consists of modifying a pre-existent furyl aglycone. This approach is applied to obtain novel pyridazine C-nucleosides starting with 2- and 3-(ribofuranosyl)furans. It is based on singlet oxygen [4+2] cycloaddition followed by reduction and hydrazine cyclization under neutral conditions. The mild three-step one-pot procedure leads stereoselectively to novel pyridazine C-nucleosides of pharmacological interest. The use of acetyls as protecting groups provides an elegant direct route to a deprotected new pyridazine C-nucleoside.


Introduction
The synthesis of nucleoside analogues has a prominent role in the field of organic chemistry and biology [1]. Among modified nucleosides, C-nucleosides represent a special moiety of this compound class due to their higher stability towards enzymatic and chemical hydrolysis than that of natural N-nucleosides, as well as owing to the interesting biological and pharmacological properties of some of their derivatives [1,2]. The first natural C-nucleoside isolated was pseudouridine [3], followed by showdomycin [4], and oxazinomycin [5]-all characterized by important pharmacological activities, from antitumor to antibacterial. C-nucleosides comprise a sugar moiety and a non-natural heterocyclic base connected by a carbon-carbon bond. The stability of this bond and the broad structural variation in the heterocyclic base account for the high interest in this compound class [6][7][8][9]. The main synthetic approach for C-glycosides is based on coupling reactions between a glycosyl donor and an acceptor in the presence of a promoter [10]. Alternatively, structural elaborations on a pre-existing aglycone are performed to afford the final C-glycoside [10]. In this context, the use of the furan as starting aglycone is of particular interest due to the easy preparation and high versatility of this heterocycle [11]. We elaborated a green procedure based on the use of a furan aglycone as a building block and the [4+2] cycloaddition of singlet oxygen [12] by dye-sensitized photooxygenation as a strategy for providing the final desired structures. Indeed, singlet oxygen addition to furan systems offers various elegant routes to differently functionalized derivatives through structural elaborations of the initial endoperoxides, the 2,3,7-trioxabicyclo[2.2.1]hept-5-enes [13]. Although the furan endoperoxides are thermally unstable, their reactivity can be controlled and opportunely addressed by working at low temperatures. With this strategy, it is possible to synthesize novel glycosyl derivatives, such as glycals [14] and spiroketals of monosaccharides, which are structural motifs of many products characterized by important and assorted biological properties [15]. Pyridazine [16], pyrazoline [14], bis-epoxide [17], and spirocyclic C-nucleosides [17] were also obtained. Key intermediates are often unsaturated Z-1,4-enediones, easily obtained by photooxygenation and followed by in situ reduction of the corresponding endoperoxides [14][15][16][17] at low temperatures. In many cases, the photochemical approach represents an alternative to other methods. As an example, Maeba described the synthesis of 3-(2 ,3 ,5 -tri-O-benzoyl-β-ribofuranosyl)pyridazine through a series of reactions involving the oxidation of the furan ring by bromine and methanol [18]. We obtained an analogue 4-(2 ,3 ,5 -tri-O-acetyl-β-ribofuranosyl)-3,6-dimetylpyridazine in 70% yield (based on starting furan) by oxygenation, followed by diethyl sulfide reduction and hydrazine hydrochloride cyclization [16]. The efficient formation of this compound prompted us to explore the possibility of extending this approach to novel furans in order to synthesize 3-and 4-(ribofuranosyl)pyridazines. The pyridazine ring is recognized as a versatile pharmacophore. In the last years, particular attention has been devoted to developing novel synthetic approaches to this system, starting with new precursors or by utilizing green methodologies [19][20][21][22].

Results and Discussion
For this purpose, the suitable furans 2a,b were initially prepared by the β-stereoselective reduction [23] of the corresponding furyl ketoses 1a,b (Scheme 1). The latter were obtained by coupling reactions between 2,3,5-tri-O-benzyl-D-ribono-1,4-lactone as the glycosyl donor, and 2-or 3-furyllithium, according to a previously reported procedure [15]. The reduction times were particularly long (overnight), probably owing to the openchain structures of the furyl derivatives 1a,b in equilibrium with very small amounts of the corresponding cyclic structures, as evidenced by their proton spectra. Although the 1 H-NMR spectra of both crude reaction mixtures showed only the presence of products 2a,b, a considerable loss of material occurred during purification by silica gel chromatography, even if performed under N2, as reported in the literature for 2a [24]. Firstly, 2a was photo-oxygenated at −20 °C in dichloromethane with methylene blue as the sensitizer (Scheme 2). When the reaction was complete (90 min, TLC or 1 H-NMR), Et2S (1.2 equiv.) was added to the crude mixture, which was kept at −20 °C to avoid thermal rearrangement of the intermediate endoperoxide [12][13][14][15][16][17]. The reduction times were particularly long (overnight), probably owing to the openchain structures of the furyl derivatives 1a,b in equilibrium with very small amounts of the corresponding cyclic structures, as evidenced by their proton spectra. Although the 1 H-NMR spectra of both crude reaction mixtures showed only the presence of products 2a,b, a considerable loss of material occurred during purification by silica gel chromatography, even if performed under N 2 , as reported in the literature for 2a [24]. Firstly, 2a was photooxygenated at −20 • C in dichloromethane with methylene blue as the sensitizer (Scheme 2). When the reaction was complete (90 min, TLC or 1 H-NMR), Et 2 S (1.2 equiv.) was added to the crude mixture, which was kept at −20 • C to avoid thermal rearrangement of the intermediate endoperoxide [12][13][14][15][16][17].
After 2 h, the solvent and unreacted Et2S were removed under reduced pressure, and the crude 4a was treated with hydrazine hydrochloride, as reported in [16]. Under these conditions, compound 4a rapidly isomerized into the more stable E-isomer 4a', owing to the configurational instability of the Z-enedione 4a under acidic conditions, as observed in similar cases [12,16] (Scheme 2). However, we found that when a solution of hydrazine in THF (2.0 M) was added to the crude glycosyl enedione 4a, the expected pyridazine 5a was promptly obtained (Scheme 2). The synthesis of 5a was then realized through a onepot three-step procedure in good total yield, as shown in Scheme 3. The β-configuration of pyridazine 5a was confirmed by the 2D-NOESY experiment, which evidenced correlation between the H-1′ and the H-4′ of the sugar ring ( Figure 1). With the exception of the protecting groups, the 3-(ribofuranosyl)pyridazine 5a was the same as that reported by Maeba et al. [18].  The novel conditions were applied to furan 2b and, in good total yield, led to the new 4-(ribofuranosyl)pyridazine 5b (Scheme 4). The β-configuration was also confirmed for 5b by the 2D-NOESY experiment. After 2 h, the solvent and unreacted Et 2 S were removed under reduced pressure, and the crude 4a was treated with hydrazine hydrochloride, as reported in [16]. Under these conditions, compound 4a rapidly isomerized into the more stable E-isomer 4a', owing to the configurational instability of the Z-enedione 4a under acidic conditions, as observed in similar cases [12,16] (Scheme 2). However, we found that when a solution of hydrazine in THF (2.0 M) was added to the crude glycosyl enedione 4a, the expected pyridazine 5a was promptly obtained (Scheme 2). The synthesis of 5a was then realized through a one-pot three-step procedure in good total yield, as shown in Scheme 3. The β-configuration of pyridazine 5a was confirmed by the 2D-NOESY experiment, which evidenced correlation between the H-1 and the H-4 of the sugar ring ( Figure 1). With the exception of the protecting groups, the 3-(ribofuranosyl)pyridazine 5a was the same as that reported by Maeba et al. [18]. After 2 h, the solvent and unreacted Et2S were removed under reduced pressure, and the crude 4a was treated with hydrazine hydrochloride, as reported in [16]. Under these conditions, compound 4a rapidly isomerized into the more stable E-isomer 4a', owing to the configurational instability of the Z-enedione 4a under acidic conditions, as observed in similar cases [12,16] (Scheme 2). However, we found that when a solution of hydrazine in THF (2.0 M) was added to the crude glycosyl enedione 4a, the expected pyridazine 5a was promptly obtained (Scheme 2). The synthesis of 5a was then realized through a onepot three-step procedure in good total yield, as shown in Scheme 3. The β-configuration of pyridazine 5a was confirmed by the 2D-NOESY experiment, which evidenced correlation between the H-1′ and the H-4′ of the sugar ring ( Figure 1). With the exception of the protecting groups, the 3-(ribofuranosyl)pyridazine 5a was the same as that reported by Maeba et al. [18].  The novel conditions were applied to furan 2b and, in good total yield, led to the new 4-(ribofuranosyl)pyridazine 5b (Scheme 4). The β-configuration was also confirmed for 5b by the 2D-NOESY experiment.  After 2 h, the solvent and unreacted Et2S were removed under reduced pressure, and the crude 4a was treated with hydrazine hydrochloride, as reported in [16]. Under these conditions, compound 4a rapidly isomerized into the more stable E-isomer 4a', owing to the configurational instability of the Z-enedione 4a under acidic conditions, as observed in similar cases [12,16] (Scheme 2). However, we found that when a solution of hydrazine in THF (2.0 M) was added to the crude glycosyl enedione 4a, the expected pyridazine 5a was promptly obtained (Scheme 2). The synthesis of 5a was then realized through a onepot three-step procedure in good total yield, as shown in Scheme 3. The β-configuration of pyridazine 5a was confirmed by the 2D-NOESY experiment, which evidenced correlation between the H-1′ and the H-4′ of the sugar ring ( Figure 1). With the exception of the protecting groups, the 3-(ribofuranosyl)pyridazine 5a was the same as that reported by Maeba et al. [18].  The novel conditions were applied to furan 2b and, in good total yield, led to the new 4-(ribofuranosyl)pyridazine 5b (Scheme 4). The β-configuration was also confirmed for 5b by the 2D-NOESY experiment.  The novel conditions were applied to furan 2b and, in good total yield, led to the new 4-(ribofuranosyl)pyridazine 5b (Scheme 4). The β-configuration was also confirmed for 5b by the 2D-NOESY experiment. The relevance of the protecting groups in the glycoside chemistry, as well as the possibility of using easily removable groups, motivated us to extend the one-pot route to a sugarfuran protected with acetyls. Hence, we carried out a coupling reaction with the 2,3,5-tetra-1-O-acetyl-β-D-ribofuranose as the glycosyl donor and methyl furan-2-carboxylate as the acceptor-both commercially available-in the presence of SnCl4 as the promoter, according to a reported procedure [17]. This synthetic approach was described to lead mainly to β-anomers due to the participation of the neighboring acetyl group at the C-2′ of the sugar ring during the departure of the leaving group at C-1′ [25,26].
Surprisingly, the reaction afforded only the ribofuranosyl furan 2c, whose α-configuration was assigned by the 2D-NOESY experiment (Scheme 5). Indeed, a NOE effect between the H-1′ and the H-3′ of the sugar ring in C6D6 was evidenced ( Figure 2). In this anisotropic solvent, the H-3′ and H-4′ signals do not overlap as it occurs when the 1 H-NMR spectrum is recorded in CDCl3 (aromatic-solvent-induced shift (ASIS) effects) [27]. The previously incorrect β-configuration for 2c was assigned on the basis of the constant value of the coupling, in comparison with those of the two anomers of showdomycin, and the synthetic procedure used [17].  Attempts to synthesize the β-anomer of 2c by carrying out the coupling reaction in acetonitrile and/or by using different promoters (BF3 or TMSOTf) failed, and only the α-1c was obtained, as evidenced spectroscopically and chromatographically.
A possible rationalization of this unexpected result is that the coupling proceeds here through anchimeric assistance by the acetyl group at the C-5′. A similar stereochemical trend is described in the iodination of some acetylated oxathiolanes [28]. Otherwise, the use of the only β-anomer of 1,2,3,5-tetra-O-acetyl-D-ribofuranose as a glycosyl donor together with the observed full stereoselectivity suggest that the reaction to methyl furan-2carboxylate could occur through a concerted pathway leading to the only α-anomer of 2c. This hypothesis could be confirmed by the use of the only α-anomer of 1,2,3,5-tetra-Oacetyl-D-ribofuranose as a glycosyl donor, which should lead to the β-anomer of 1c. Control experiments showed that the peracetylation reaction in pyridine or catalyzed by the The relevance of the protecting groups in the glycoside chemistry, as well as the possibility of using easily removable groups, motivated us to extend the one-pot route to a sugarfuran protected with acetyls. Hence, we carried out a coupling reaction with the 2,3,5tetra-1-O-acetyl-β-D-ribofuranose as the glycosyl donor and methyl furan-2-carboxylate as the acceptor-both commercially available-in the presence of SnCl 4 as the promoter, according to a reported procedure [17]. This synthetic approach was described to lead mainly to β-anomers due to the participation of the neighboring acetyl group at the C-2 of the sugar ring during the departure of the leaving group at C-1 [25,26].
Surprisingly, the reaction afforded only the ribofuranosyl furan 2c, whose α-configuration was assigned by the 2D-NOESY experiment (Scheme 5). Indeed, a NOE effect between the H-1 and the H-3 of the sugar ring in C 6 D 6 was evidenced ( Figure 2). In this anisotropic solvent, the H-3 and H-4 signals do not overlap as it occurs when the 1 H-NMR spectrum is recorded in CDCl 3 (aromatic-solvent-induced shift (ASIS) effects) [27]. The previously incorrect β-configuration for 2c was assigned on the basis of the constant value of the coupling, in comparison with those of the two anomers of showdomycin, and the synthetic procedure used [17]. The relevance of the protecting groups in the glycoside chemistry, as well as the possibility of using easily removable groups, motivated us to extend the one-pot route to a sugarfuran protected with acetyls. Hence, we carried out a coupling reaction with the 2,3,5-tetra-1-O-acetyl-β-D-ribofuranose as the glycosyl donor and methyl furan-2-carboxylate as the acceptor-both commercially available-in the presence of SnCl4 as the promoter, according to a reported procedure [17]. This synthetic approach was described to lead mainly to β-anomers due to the participation of the neighboring acetyl group at the C-2′ of the sugar ring during the departure of the leaving group at C-1′ [25,26].
Surprisingly, the reaction afforded only the ribofuranosyl furan 2c, whose α-configuration was assigned by the 2D-NOESY experiment (Scheme 5). Indeed, a NOE effect between the H-1′ and the H-3′ of the sugar ring in C6D6 was evidenced ( Figure 2). In this anisotropic solvent, the H-3′ and H-4′ signals do not overlap as it occurs when the 1 H-NMR spectrum is recorded in CDCl3 (aromatic-solvent-induced shift (ASIS) effects) [27]. The previously incorrect β-configuration for 2c was assigned on the basis of the constant value of the coupling, in comparison with those of the two anomers of showdomycin, and the synthetic procedure used [17].  Attempts to synthesize the β-anomer of 2c by carrying out the coupling reaction in acetonitrile and/or by using different promoters (BF3 or TMSOTf) failed, and only the α-1c was obtained, as evidenced spectroscopically and chromatographically.
A possible rationalization of this unexpected result is that the coupling proceeds here through anchimeric assistance by the acetyl group at the C-5′. A similar stereochemical trend is described in the iodination of some acetylated oxathiolanes [28]. Otherwise, the use of the only β-anomer of 1,2,3,5-tetra-O-acetyl-D-ribofuranose as a glycosyl donor together with the observed full stereoselectivity suggest that the reaction to methyl furan-2carboxylate could occur through a concerted pathway leading to the only α-anomer of 2c. This hypothesis could be confirmed by the use of the only α-anomer of 1,2,3,5-tetra-Oacetyl-D-ribofuranose as a glycosyl donor, which should lead to the β-anomer of 1c. Control experiments showed that the peracetylation reaction in pyridine or catalyzed by the  The relevance of the protecting groups in the glycoside chemistry, as well as the possibility of using easily removable groups, motivated us to extend the one-pot route to a sugarfuran protected with acetyls. Hence, we carried out a coupling reaction with the 2,3,5-tetra-1-O-acetyl-β-D-ribofuranose as the glycosyl donor and methyl furan-2-carboxylate as the acceptor-both commercially available-in the presence of SnCl4 as the promoter, according to a reported procedure [17]. This synthetic approach was described to lead mainly to β-anomers due to the participation of the neighboring acetyl group at the C-2′ of the sugar ring during the departure of the leaving group at C-1′ [25,26].
Surprisingly, the reaction afforded only the ribofuranosyl furan 2c, whose α-configuration was assigned by the 2D-NOESY experiment (Scheme 5). Indeed, a NOE effect between the H-1′ and the H-3′ of the sugar ring in C6D6 was evidenced ( Figure 2). In this anisotropic solvent, the H-3′ and H-4′ signals do not overlap as it occurs when the 1 H-NMR spectrum is recorded in CDCl3 (aromatic-solvent-induced shift (ASIS) effects) [27]. The previously incorrect β-configuration for 2c was assigned on the basis of the constant value of the coupling, in comparison with those of the two anomers of showdomycin, and the synthetic procedure used [17].  Attempts to synthesize the β-anomer of 2c by carrying out the coupling reaction in acetonitrile and/or by using different promoters (BF3 or TMSOTf) failed, and only the α-1c was obtained, as evidenced spectroscopically and chromatographically.
A possible rationalization of this unexpected result is that the coupling proceeds here through anchimeric assistance by the acetyl group at the C-5′. A similar stereochemical trend is described in the iodination of some acetylated oxathiolanes [28]. Otherwise, the use of the only β-anomer of 1,2,3,5-tetra-O-acetyl-D-ribofuranose as a glycosyl donor together with the observed full stereoselectivity suggest that the reaction to methyl furan-2carboxylate could occur through a concerted pathway leading to the only α-anomer of 2c. This hypothesis could be confirmed by the use of the only α-anomer of 1,2,3,5-tetra-Oacetyl-D-ribofuranose as a glycosyl donor, which should lead to the β-anomer of 1c. Control experiments showed that the peracetylation reaction in pyridine or catalyzed by the Attempts to synthesize the β-anomer of 2c by carrying out the coupling reaction in acetonitrile and/or by using different promoters (BF 3 or TMSOTf) failed, and only the α-1c was obtained, as evidenced spectroscopically and chromatographically.
A possible rationalization of this unexpected result is that the coupling proceeds here through anchimeric assistance by the acetyl group at the C-5 . A similar stereochemical trend is described in the iodination of some acetylated oxathiolanes [28]. Otherwise, the use of the only β-anomer of 1,2,3,5-tetra-O-acetyl-D-ribofuranose as a glycosyl donor together with the observed full stereoselectivity suggest that the reaction to methyl furan-2-carboxylate could occur through a concerted pathway leading to the only α-anomer of 2c. This hypothesis could be confirmed by the use of the only α-anomer of 1,2,3,5-tetra-O-acetyl-D-ribofuranose as a glycosyl donor, which should lead to the β-anomer of 1c. Control experiments showed that the peracetylation reaction in pyridine or catalyzed by the TMSOTf afforded the peracetylated pyranosic form of the sugar as the main product, as well as the β-anomer of 1,2,3,5-tetra-O-acetyl-D-ribofuranose, which is the only commercially available form.
TMSOTf afforded the peracetylated pyranosic form of the sugar as the main product, as well as the β-anomer of 1,2,3,5-tetra-O-acetyl-D-ribofuranose, which is the only commercially available form.
With the aim of preserving the methyl ester function, as well as the protecting acetyl groups on the sugar ring, the reaction was carried out under the same conditions except for the use of 1.2 equiv. of hydrazine. In this case, the one-pot procedure afforded the protected pyridazine C-nucleoside 5d good yield, showing that the cyclization proceeds more rapidly than the attack to the ester functions (Scheme 7). The structure of compound 5d, assigned by NMR data, was confirmed by X-ray crystallographic analysis (Figure 3). Pyridazine 5d was quite stable; however, it slowly aromatized after several days at room temperature, leading to the corresponding furan 6d (Scheme 8). Similar elimination was previously observed in a benzoylated pyridazine C-nucleoside [18].  With the aim of preserving the methyl ester function, as well as the protecting acetyl groups on the sugar ring, the reaction was carried out under the same conditions except for the use of 1.2 equiv. of hydrazine. In this case, the one-pot procedure afforded the protected pyridazine C-nucleoside 5d good yield, showing that the cyclization proceeds more rapidly than the attack to the ester functions (Scheme 7). TMSOTf afforded the peracetylated pyranosic form of the sugar as the main product, as well as the β-anomer of 1,2,3,5-tetra-O-acetyl-D-ribofuranose, which is the only commercially available form. The mild procedure for pyridazine C-nucleoside was later applied to the 2-(2′,3′,5′tri-O-acetyl-α-D-ribofuranosyl)furan (2c). Unexpectedly, the sequence of reactions on 2c led in high yield to the deprotected pyridazine C-nucleoside 5c (Scheme 6). Scheme 6. One-pot synthesis of the 6-(α-D-ribofuranosyl)pyridazine-3-carbohydrazide (5c).
With the aim of preserving the methyl ester function, as well as the protecting acetyl groups on the sugar ring, the reaction was carried out under the same conditions except for the use of 1.2 equiv. of hydrazine. In this case, the one-pot procedure afforded the protected pyridazine C-nucleoside 5d good yield, showing that the cyclization proceeds more rapidly than the attack to the ester functions (Scheme 7). The structure of compound 5d, assigned by NMR data, was confirmed by X-ray crystallographic analysis (Figure 3). Pyridazine 5d was quite stable; however, it slowly aromatized after several days at room temperature, leading to the corresponding furan 6d (Scheme 8). Similar elimination was previously observed in a benzoylated pyridazine C-nucleoside [18]. The structure of compound 5d, assigned by NMR data, was confirmed by X-ray crystallographic analysis (Figure 3). TMSOTf afforded the peracetylated pyranosic form of the sugar as the main product, as well as the β-anomer of 1,2,3,5-tetra-O-acetyl-D-ribofuranose, which is the only commercially available form. The mild procedure for pyridazine C-nucleoside was later applied to the 2-(2′,3′,5′tri-O-acetyl-α-D-ribofuranosyl)furan (2c). Unexpectedly, the sequence of reactions on 2c led in high yield to the deprotected pyridazine C-nucleoside 5c (Scheme 6). Scheme 6. One-pot synthesis of the 6-(α-D-ribofuranosyl)pyridazine-3-carbohydrazide (5c).
With the aim of preserving the methyl ester function, as well as the protecting acetyl groups on the sugar ring, the reaction was carried out under the same conditions except for the use of 1.2 equiv. of hydrazine. In this case, the one-pot procedure afforded the protected pyridazine C-nucleoside 5d good yield, showing that the cyclization proceeds more rapidly than the attack to the ester functions (Scheme 7). The structure of compound 5d, assigned by NMR data, was confirmed by X-ray crystallographic analysis (Figure 3). Pyridazine 5d was quite stable; however, it slowly aromatized after several days at room temperature, leading to the corresponding furan 6d (Scheme 8). Similar elimination was previously observed in a benzoylated pyridazine C-nucleoside [18]. Pyridazine 5d was quite stable; however, it slowly aromatized after several days at room temperature, leading to the corresponding furan 6d (Scheme 8). Similar elimination was previously observed in a benzoylated pyridazine C-nucleoside [18].

General Information
Melting points are uncorrected. The 1 H-and 13 C-NMR spectra were recorded at 500 and 125 MHz, respectively, on a Fourier Transform NMR Varian 500 Unity Inova spectrometer. The carbon multiplicity was evidenced by DEPT experiments. The proton couplings were evidenced by 1 H-1 H COSY experiments. The heteronuclear chemical shift correlations were determined by HMQC and HMBC pulse sequences. 1 H-1 H proximities through space within a molecule were determined by NOESY experiments. X-ray analysis was performed on a Bruker-Nonius Kappa CCD (Nonius BV, Delft, The Netherlands) diffractometer (graphite monochromated Mo Kα radiation, λ = 0.71073 Å, CCD rotation images, thick slices, φ and ω scans to fill asymmetric unit). Analytical TLC was performed on precoated silica gel plates (Macherey-Nagel, Düren, Germany) with 0.2 mm film thickness. Spots were visualized by UV light and by spraying with EtOH/H2SO4 (95:5 v/v), followed by heating for 5 min at 110 °C. Column chromatography was performed on silica gel (0.063-0.2 mm) (Macherey-Nagel). Reagent-grade commercially available reagents and solvents were used.

Synthesis of 2a,b
A solution of 243 mg (0.5 mmol) of 1a [15] in 5.2 mL of acetonitrile, cooled to −40 °C, was added to 240 µL (3 equiv.) of triethylsilane and 70 µL (1 equiv.) of BF3·Et2O. The solution was stirred at −40 °C for 4 h, which was allowed to rise to r.t. while the mixture was further stirred overnight. A saturated aqueous solution of K2CO3 was later added (10 mL), and the mixture was kept under stirring for 10 min. The organic layer was extracted with diethyl ether (3 × 30 mL), washed with brine, dried over anhydrous Na2SO4, and filtered. The solvent was removed under reduced pressure, and the residue was chromatographed on flash silica gel (n-hexane/ether 1:1 v/v), affording the C-nucleoside β-2a [24] with 35% yield.

General Information
Melting points are uncorrected. The 1 H-and 13 C-NMR spectra were recorded at 500 and 125 MHz, respectively, on a Fourier Transform NMR Varian 500 Unity Inova spectrometer. The carbon multiplicity was evidenced by DEPT experiments. The proton couplings were evidenced by 1 H-1 H COSY experiments. The heteronuclear chemical shift correlations were determined by HMQC and HMBC pulse sequences. 1 H-1 H proximities through space within a molecule were determined by NOESY experiments. X-ray analysis was performed on a Bruker-Nonius Kappa CCD (Nonius BV, Delft, The Netherlands) diffractometer (graphite monochromated Mo K α radiation, λ = 0.71073 Å, CCD rotation images, thick slices, ϕ and ω scans to fill asymmetric unit). Analytical TLC was performed on precoated silica gel plates (Macherey-Nagel, Düren, Germany) with 0.2 mm film thickness. Spots were visualized by UV light and by spraying with EtOH/H 2 SO 4 (95:5 v/v), followed by heating for 5 min at 110 • C. Column chromatography was performed on silica gel (0.063-0.2 mm) (Macherey-Nagel). Reagent-grade commercially available reagents and solvents were used.

Synthesis of 2a,b
A solution of 243 mg (0.5 mmol) of 1a [15] in 5.2 mL of acetonitrile, cooled to −40 • C, was added to 240 µL (3 equiv.) of triethylsilane and 70 µL (1 equiv.) of BF 3 ·Et 2 O. The solution was stirred at −40 • C for 4 h, which was allowed to rise to r.t. while the mixture was further stirred overnight. A saturated aqueous solution of K 2 CO 3 was later added (10 mL), and the mixture was kept under stirring for 10 min. The organic layer was extracted with diethyl ether (3 × 30 mL), washed with brine, dried over anhydrous Na 2 SO 4 , and filtered. The solvent was removed under reduced pressure, and the residue was chromatographed on flash silica gel (n-hexane/ether 1:1 v/v), affording the C-nucleoside β-2a [24] with 35% yield. The synthesis of 2b was performed as reported above for 2a, starting from 1b [15]. The flash silica gel (n-hexane/ether 1:1 v/v) afforded the novel C-nucleoside β-2b with 30% yield.

X-ray Crystallography of 5d
X-ray analysis was performed on single crystals of 5d obtained as colorless blocks by slow evaporation of a DCM/hexane solution at room temperature. One selected crystal was mounted at ambient temperature on a Bruker-Nonius KappaCCD diffractometer (graphite monochromated Mo K α radiation, λ = 0.71073 Å, CCD rotation images, thick slices, ϕ and ω scans to fill asymmetric unit). Semiempirical absorption corrections (SADABS [29]) were applied. The structure was solved by direct methods (SIR97 program [30]) and anisotropically refined by the full matrix least-squares method on F 2 against all independently measured reflections using SHELXL-2018/3 To SHELXL (version 2018/3) [31] and WinGX software (version 2014.1) [32]. All of the hydrogen atoms were introduced in calculated positions and refined according to a riding model with C-H distances in the range of 0.93-0.98 Å and with Uiso (H) equal to 1.2 Ueq or 1.5 Ueq (C methyl ) of the carrier atom. Compound 5d crystallizes in P 2 1 2 1 2 1 space group with two independent molecules in the asymmetric unit (see Figure S31 of Supporting Information). The two molecules are very similar to each to the other; in molecule A, one acetyl group is split into two positions with refined occupancy factors of 0.61 and 0.39. Some restraints were introduced in the last stage of refinement to regularize the geometry. In the absence of strong anomalous scatterer atoms, the Flack parameter is meaningless, so it was not possible to assign the absolute configuration by anomalous dispersion effects with diffraction measurements on the crystal. The absolute configuration has been assigned by reference to unchanging chiral centers in the synthetic procedure. A rather high residual electronic density (0.472 eA −3 ) is explained by residual thermal disorder of some acetyl groups at ambient temperature. Unfortunately, it was not possible to re-collect data at low temperatures. Crystal data and structure refinement details are reported in Table S1 of SI. Figures were generated using program Ortep-3 [32]. CCDC-1493362.

Conclusions
In summary, we highlight that novel pyridazine C-nucleosides can be easily obtained starting with protected ribofuranosyl furans. Appropriate mild conditions to preserve the Zconfiguration of the intermediate 1,4-dicarbonyl compound, necessary for the final step that involves the reaction with hydrazine. It is noteworthy that the use of 2a and 2b provides the corresponding 3-and 4-(ribofuranosyl)pyridazines 5a and 5b, respectively, which would be hard to synthesize through direct coupling reactions with complete regioselectivity. Moreover, the use of an acetylated sugar provides a direct route to deprotected pyridazine-C-nucleosides. The three-step one-pot procedure is completely stereoselective and gives the product 5 the same anomeric configuration as the starting sugar furans.
The ease of methodology, together with the good yields of pyridazine-C-nucleosides 5, foresee novel applications in the field of C-nucleosides synthesis since pyridazinesare useful intermediates for constructing heterocycle derivatives [33]. These systems are considered by GlaxoSmithKline as one of the "most developable" heteroaromatic rings [34] and are proposed as privileged structures for drug design [35,36]. Moreover, a broad array of significant biological activities has been evidenced in several compounds with pyridazine rings [37][38][39].
Supplementary Materials: The following are available online. 1 H-and 13 C-NMR, COSY and NOESY spectra for all new compounds. X-ray crystallographic data for 5d. Supplementary data associated with this article can be found in the online version, at CCDC-1493362, which contains the supplementary crystallographic data for this paper. These data can be obtained free of charge from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.