Thioether and Ether Furofuran Lignans: Semisynthesis, Reaction Mechanism, and Inhibitory Effect against α-Glucosidase and Free Radicals

The transformation of sesame lignans is interesting because the derived products possess enhanced bioactivity and a wide range of potential applications. In this study, the semisynthesis of 28 furofuran lignans using samin (5) as the starting material is described. Our methodology involved the protonation of samin (5) to generate an oxocarbenium ion followed by the attack from two different nucleophiles, namely, thiols (RSH) and alcohols (ROH). The highly diastereoselective thioether and ether furofuran lignans were obtained, and their configurations were confirmed by 2D NMR and X-ray crystallography. The mechanism underlying the reaction was studied by monitoring 1H NMR and computational calculations, that is, the diastereomeric α- and β-products were equally formed through the SN1-like mechanism, while the β-product was gradually transformed via an SN2-like mechanism to the α-congener in the late step. Upon evaluation of the inhibitory effect of the synthesized lignans against α-glucosidases and free radicals, the lignans 7f and 7o of the phenolic hydroxyl group were the most potent inhibitors. Additionally, the mechanisms underlying the α-glucosidase inhibition of 7f and 7o were verified to be of a mixed manner and noncompetitive inhibition, respectively. The results indicated that both 7f and 7o possessed promising antidiabetic activity, while simultaneously inhibiting α-glucosidases and free radicals.


Introduction
Furofuran lignan is an important group characterized by the presence of a 2,6-diaryl-3,7-dioxabicyclo [3.3.0]octane skeleton. Numerous furofuran lignans have been isolated from various plants, and their biological activities have been evaluated; for example, antitumoral activity of syringaresinol (1) [1], cholesterol lowering effect of sesamin (2) in rat serum and liver [2], and antioxidant of sesamolin (3) [3] (Figure 1). Owing to the broad range of activities and structural diversity of furofuran lignans, the development of synthetic methodologies has been reported to improve their bioactivities [4][5][6]. However, no investigation into the structure activity relationship (SAR) has been reported, due to the lack of a practical synthetic method to produce diverse furofuran lignans.
We report a new synthetic approach to easily produce a wide series of furofuran lignans. Our approach is based on the semi-synthesis of sesaminol (4) using naturally available sesamolin (3) as a starting material in the presence of acidic resin as a catalyst [7]. The success of transforming sesamolin (3) into samin (5), a versatile lignan building block, enabled us to synthesize a variety of furofuran lignans. We initially applied phenolics and flavonoids as carbon nucleophiles to couple with samin (5), thereby producing various pairs of epimeric products (α-and β-forms) in fair to good yields [8,9]. Moreover, we also synthesized a series of furofuran lignans with catechol moieties via oxidative the cleavage of methylenedioxy moiety [10]. As a consequence, up to 50 furofuran lignans could be synthesized and evaluated for their inhibitory effect against α-glucosidase and We report a new synthetic approach to easily produce a wide series of furofuran lignans. Our approach is based on the semi-synthesis of sesaminol (4) using naturally available sesamolin (3) as a starting material in the presence of acidic resin as a catalyst [7]. The success of transforming sesamolin (3) into samin (5), a versatile lignan building block, enabled us to synthesize a variety of furofuran lignans. We initially applied phenolics and flavonoids as carbon nucleophiles to couple with samin (5), thereby producing various pairs of epimeric products (α-and β-forms) in fair to good yields [8,9]. Moreover, we also synthesized a series of furofuran lignans with catechol moieties via oxidative the cleavage of methylenedioxy moiety [10]. As a consequence, up to 50 furofuran lignans could be synthesized and evaluated for their inhibitory effect against αglucosidase and free radicals, thus elucidating the critical role of the catechol moiety in SAR. However, the carbon nucleophiles used in this reaction were in fact limited to certain phenolics and flavonoids containing more electron-donating groups.
To expand the scope of the synthetic methodology, the synthesis of a new series of furofuran lignans using thiols and alcohols as sulfur and oxygen nucleophiles, respectively, was established (Scheme 1). Thiols and alcohols as stronger nucleophiles would promote product yield, and offer diverse furofuran lignans. In this project, the reaction mechanism and inhibitory effect against α-glucosidases and free radicals, and a kinetic study are also described. This investigation provides insights into furofuran lignans in the scope of the synthetic methodology and the reaction mechanism together with the structure activity relationship with regard to α-glucosidase and free radicals.  To expand the scope of the synthetic methodology, the synthesis of a new series of furofuran lignans using thiols and alcohols as sulfur and oxygen nucleophiles, respectively, was established (Scheme 1). Thiols and alcohols as stronger nucleophiles would promote product yield, and offer diverse furofuran lignans. In this project, the reaction mechanism and inhibitory effect against α-glucosidases and free radicals, and a kinetic study are also described. This investigation provides insights into furofuran lignans in the scope of the synthetic methodology and the reaction mechanism together with the structure activity relationship with regard to α-glucosidase and free radicals. We report a new synthetic approach to easily produce a wide series of furofuran lignans. Our approach is based on the semi-synthesis of sesaminol (4) using naturally available sesamolin (3) as a starting material in the presence of acidic resin as a catalyst [7]. The success of transforming sesamolin (3) into samin (5), a versatile lignan building block, enabled us to synthesize a variety of furofuran lignans. We initially applied phenolics and flavonoids as carbon nucleophiles to couple with samin (5), thereby producing various pairs of epimeric products (α-and β-forms) in fair to good yields [8,9]. Moreover, we also synthesized a series of furofuran lignans with catechol moieties via oxidative the cleavage of methylenedioxy moiety [10]. As a consequence, up to 50 furofuran lignans could be synthesized and evaluated for their inhibitory effect against αglucosidase and free radicals, thus elucidating the critical role of the catechol moiety in SAR. However, the carbon nucleophiles used in this reaction were in fact limited to certain phenolics and flavonoids containing more electron-donating groups.
To expand the scope of the synthetic methodology, the synthesis of a new series of furofuran lignans using thiols and alcohols as sulfur and oxygen nucleophiles, respectively, was established (Scheme 1). Thiols and alcohols as stronger nucleophiles would promote product yield, and offer diverse furofuran lignans. In this project, the reaction mechanism and inhibitory effect against α-glucosidases and free radicals, and a kinetic study are also described. This investigation provides insights into furofuran lignans in the scope of the synthetic methodology and the reaction mechanism together with the structure activity relationship with regard to α-glucosidase and free radicals.

Chemistry
Prior to studying the scope of thioethers and ethers as nucleophiles, we screened a variety of acids using methanol as a model nucleophile in an acetonitrile solution of samin (5) (Table S1). Lewis acid (AlCl 3 and BF 3 .Et 2 O) and Brønsted acid (HCl, CF 3 CO 2 H, and Amberlyst-15) efficiently promoted the nucleophilic substitution of samin (5) with methanol, thereby providing a single diastereomer (2α-9a) in a high yield (95-98%), while acetic acid (weak Brønsted acid) failed to trigger the reaction. Although several Lewis and Brønsted acids can be applied in this reaction, the heterogeneous Amberlyst-15 was used for further investigation because no workup step is required when using this.
Generally, the reactions between samin (5) and thiols produced furofuran lignans in higher yield (78%-quant), and α-products were exclusively obtained as an isolated yield. Theoretically, the reaction proceeds through an oxocarbenium ion and generates a mixture of αand β-products with a diastereomeric ratio of 1:1 [11]. In our previous studies, we obtained isolated αand β-products in the ratios of 1:1 to 3:1 from the reactions between samin (5) and a series of phenolics and flavonoids as carbon nucleophiles. Therefore, the unexpected results prompted us to determine a diastereomeric ratio of αand β-thioether furofuran lignans using an 1 H NMR analysis of the crude reaction mixture (Table 1). Apparently, the reactions between the samin (5) and thiols produced αand β-products in the ratio of 5:1 to 8:1, while the steric hindrance resulted in a higher diastereomeric ratio of 12:1. This led to the inference that the alpha-product was obtained as an isolated yield.
To fully characterize the structures of the thioether furofuran lignans and assign the configuration of the αand β-products, we scaled up the reaction between samin (5) and thiol 6f to afford an adequate amount of diastereomeric products in a pure form. The structures of 2α-7f and 2β-7f were verified by MS and 2D NMR (see supporting information). The configuration of the αand β-products was assigned by coupling constant ( Figure 2) and NOESY ( Figure 3) data analysis. A singlet signal of H-2 appearing at δ H 5.37 ppm ( Figure 2) was characteristic of the α-product, which was contributed by the dihedral angle (θ) of 90 • between H-1 and H-2. On the other hand, the β-product showed a doublet signal of H-2 at δ H 5.10 ppm together with a typical coupling constant of 6.1 Hz ( Figure 2). The characteristic splitting pattern of H-2 would be useful to address the configuration of the αand β-products synthesized from samin (5) and thiols. mation). The configuration of the α-and β-products was assigned by coupling constant ( Figure 2) and NOESY ( Figure 3) data analysis. A singlet signal of H-2 appearing at δH 5.37 ppm ( Figure 2) was characteristic of the α-product, which was contributed by the dihedral angle (θ) of 90° between H-1 and H-2. On the other hand, the β-product showed a doublet signal of H-2 at δH 5.10 ppm together with a typical coupling constant of 6.1 Hz ( Figure 2). The characteristic splitting pattern of H-2 would be useful to address the configuration of the α-and β-products synthesized from samin (5) and thiols.    Furthermore, the structures of the thioether furofuran lignans were unequivocally verified by X-ray crystallography using a pair of diastereomers 7f. Prior to performing the X-ray analysis, modification of 2β-7f was required to convert it from a greasy appearance to a crystalline solid by reacting it with 4-bromobenzyl bromide (see Experimental section for details). The X-ray crystal structures of 2α-7f and 2β-7fBr ( Figure 3) clearly showed the cis fusion of the two furan moieties. The five-membered rings adopted an envelope conformation with the oxygen O-3 and O-7 atoms as a flip. The substituents at C-2 and C-6 of furofuran moiety occupied a syn orientation for 2α-7f (Figure 3a) while an anti orientation was noticed for 2β-7fBr ( Figure 3b). Notably, the dihedral angle between H-1 and H-2 of 2α-7f and 2β-7fBr showed 96 • and 28 • , respectively. These observations were identical to those of 1 H NMR results.

Investigation of Reaction Mechanism of Ether Furofuran Lignan Formation
To investigate how an α-product exclusively formed during the nucleophilic substitution of samin (5) with alcohols (8b-8m), 1 H NMR was used to follow the reaction progress. The model reaction comprising of samin (5), methanol-d4 as a nucleophile and solvent, along with 10 mol% of TFA-d1 as an acid catalyst, was carried out in an NMR tube. Figure 5 shows an overlaid 1 H NMR spectra of the model reaction monitored at eight different times (t = 0, 5, 15, 30, 60, 120, 240, and 1440 min). The occurrence of 2α-10 (α-product) and 2β-10 (β-product) was observed by the presence of H-2 signals at TM H 4.85 (s) and 4.93 (d, J = 5.5 Hz), respectively, while the depletion of samin (5) was noted by the reduction in signal intensity at TM H 5.27 (s). This observation preliminarily suggested that the substitution of samin (5) by alcohol potentially proceeded through the formation of an oxocarbenium ion followed by a nucleophilic attack. To probe the reaction progress after α-and β-products were generated, the time-course plot of the evolution of the starting material and products was constructed ( Figure 6).

Investigation of Reaction Mechanism of Ether Furofuran Lignan Formation
To investigate how an α-product exclusively formed during the nucleophilic substitution of samin (5) with alcohols (8b-8m), 1 H NMR was used to follow the reaction progress. The model reaction comprising of samin (5), methanol-d 4 as a nucleophile and solvent, along with 10 mol% of TFA-d 1 as an acid catalyst, was carried out in an NMR tube. Figure 5 shows an overlaid 1 H NMR spectra of the model reaction monitored at eight different times (t = 0, 5, 15, 30, 60, 120, 240, and 1440 min). The occurrence of 2α-10 (α-product) and 2β-10 (β-product) was observed by the presence of H-2 signals at TM H 4.85 (s) and 4.93 (d, J = 5.5 Hz), respectively, while the depletion of samin (5) was noted by the reduction in signal intensity at TM H 5.27 (s). This observation preliminarily suggested that the substitution of samin (5) by alcohol potentially proceeded through the formation of an oxocarbenium ion followed by a nucleophilic attack. To probe the reaction progress after αand β-products were generated, the time-course plot of the evolution of the starting material and products was constructed ( Figure 6).
To probe how the β-product transformed to an α-product after samin (5) was completely consumed, we set two independent reactions of 2β-10 with and without methanold 4 as the nucleophile in the presence of TFA-d 1 as the acid catalyst and CDCl 3 as the solvent (Scheme 2). Previously, the S N 1-like mechanism (Route i) had been proposed for the epimerization of certain furofuran lignans [4,12], which involved the protonation of O-3 to generate intermediate A followed by ring closure to yield an α-product. Alternatively, the S N 2-like mechanism (Route ii) possibly took place via the displacement of -OCD 3 by the introduction of CD 3 OD through the transition state B.
The results of this experiment are presented in Figure 7. The 1 H NMR spectra clearly showed that 10 min after the addition of methanol-d 4 into the reaction, 2α-10 readily formed (Figure 7a). Additionally, we also performed the reaction of 2β-10 with methanol-d 4 as a nucleophile in CDCl 3 and the results displayed that it had no product 2α-10 produced in this condition (data not shown). Hence, we hypothesized that the transformation of a β-product to an α-product using alcohol nucleophile could occur through an S N 2like mechanism.   To probe how the β-product transformed to an α-product after samin (5) was com pletely consumed, we set two independent reactions of 2β-10 with and without metha nol-d4 as the nucleophile in the presence of TFA-d1 as the acid catalyst and CDCl3 as the solvent (Scheme 2). Previously, the SN1-like mechanism (Route i) had been proposed for the epimerization of certain furofuran lignans [4,12], which involved the protonation o O-3 to generate intermediate A followed by ring closure to yield an α-product. Alterna tively, the SN2-like mechanism (Route ii) possibly took place via the displacement of -OCD3 by the introduction of CD3OD through the transition state B.  The results of this experiment are presented in Figure 7. The 1 H NMR spectra clearly showed that 10 min after the addition of methanol-d4 into the reaction, 2α-10 readily formed (Figure 7a). Additionally, we also performed the reaction of 2β-10 with methanol-d4 as a nucleophile in CDCl3 and the results displayed that it had no product 2α-10 produced in this condition (data not shown). Hence, we hypothesized that the transformation of a β-product to an α-product using alcohol nucleophile could occur through an SN2-like mechanism. Our finding is strikingly different from that of Johansson [12], who proposed that the epimerization of furanosides proceeded through an SN1-like mechanism without experimental data to support this claim. We also confirmed a transformational possibility of 2β-10 by bond distance observation of the structure at the transition state, using the PCM/B3LYP/6-31+G(d,p) method to optimize the transition state structure ( Figure 8). As the nucleophile and leaving group were totally the same alcohols. The C-O (breaking bond) and C-O b (forming bond) bond distances at the transition state were not significantly different. However, the slightly shorter new C-O b (2.51 Å) bond had a higher preference than the C-O bond. The short H a -O bond (0.97 Å) also supported the fact that the -OCD3 of 2β-10 tends to leave after deuteration by TFA-d1. All of the illustrated evidence reasonably demonstrates the transformation of 2β-10 to 2α-10 occurring through a SN2-like transition state [13].  The results of this experiment are presented in Figure 7. The 1 H NMR spectra clearly showed that 10 min after the addition of methanol-d4 into the reaction, 2α-10 readily formed ( Figure 7a). Additionally, we also performed the reaction of 2β-10 with methanol-d4 as a nucleophile in CDCl3 and the results displayed that it had no product 2α-10 produced in this condition (data not shown). Hence, we hypothesized that the transformation of a β-product to an α-product using alcohol nucleophile could occur through an SN2-like mechanism. Our finding is strikingly different from that of Johansson [12], who proposed that the epimerization of furanosides proceeded through an SN1-like mechanism without experimental data to support this claim. We also confirmed a transformational possibility of 2β-10 by bond distance observation of the structure at the transition state, using the PCM/B3LYP/6-31+G(d,p) method to optimize the transition state structure ( Figure 8). As the nucleophile and leaving group were totally the same alcohols. The C-O (breaking bond) and C-O b (forming bond) bond distances at the transition state were not significantly different. However, the slightly shorter new C-O b (2.51 Å) bond had a higher preference than the C-O bond. The short H a -O bond (0.97 Å) also supported the fact that the -OCD3 of 2β-10 tends to leave after deuteration by TFA-d1. All of the illustrated evidence reasonably demonstrates the transformation of 2β-10 to 2α-10 occurring through a SN2-like transition state [13]. Our finding is strikingly different from that of Johansson [12], who proposed that the epimerization of furanosides proceeded through an S N 1-like mechanism without experimental data to support this claim. We also confirmed a transformational possibility of 2β-10 by bond distance observation of the structure at the transition state, using the PCM/B3LYP/6-31+G(d,p) method to optimize the transition state structure ( Figure 8). As the nucleophile and leaving group were totally the same alcohols. The C-O (breaking bond) and C-O b (forming bond) bond distances at the transition state were not significantly different. However, the slightly shorter new C-O b (2.51 Å) bond had a higher preference than the C-O bond. The short H a -O bond (0.97 Å) also supported the fact that the -OCD 3 of 2β-10 tends to leave after deuteration by TFA-d 1 . All of the illustrated evidence reasonably demonstrates the transformation of 2β-10 to 2α-10 occurring through a S N 2-like transition state [13]. According to the above results, the mechanism of ether furofuran lignans formation could be deduced as the oxocarbenium ion being first generated by the protonation of the hemiacetal center of samin through an SN1-like mechanism to furnish both α-and βproducts in an equal ratio. During the reaction progress, the β-product transformed to an α-product via the SN2-like transition state contributed by an acid catalyst (Scheme 3

Evaluation of α-Glucosidase Inhibition and Antioxidant Activity
All furofuran lignans were evaluated for rat intestinal α-glucosidase inhibition as well as antioxidation against DPPH and ABTS radicals ( Table 3). The thioether furofuran lignans 2α-7f, 2β-7f, and 2α-7o showed comparable inhibitory effects against αglucosidases (11.8-16.2 mM) and free radicals (0.20-0.93 mM) while other thioether furofuran lignans and ether furofuran lignans were not active. Remarkably, the bioactive lignans comprised free phenolic moiety that may be involved in exerting an inhibitory effect. The results in this work are consistent with previous observations that the number of free phenolic groups is critical for exhibiting the inhibition [8][9][10]. Additionally, the presence of two tert-butyl groups in 2α-7o as bulky groups does not alter the inhibitory effects, therefore confirming that free phenolic moiety principally plays an important role in the inhibition.
1.56 ± 0.5 0.14 ± 0.8 a IC 50 or SC 50 values represent as mean ± SD of three determinations. b Not determined.

Enzyme Kinetic Study
In order to gain further insight into how these furofuran lignans interact with rat intestinal maltase and sucrase, the inhibition modes of 2α-7f and 2α-7o, the representative inhibitors, were analyzed by a kinetic study. In the case of 2α-7f, the Lineweaver-Burk plot of maltase and sucrase (Figures 9 and 10) showed a series of straight lines, all of which intersected in the second quadrant. Kinetic analysis subsequently showed that V max decreased with elevated K m in the presence of increasing concentrations of 2α-7f. This behaviour suggests that 2α-7f inhibited maltase and sucrase in a mixed-type manner comprising two different pathways: competitive and noncompetitive. The observed result was elaborated by the simultaneous formation of an enzyme-inhibitor (EI) and enzyme-substrate-inhibitor (ESI) complexes in competitive and noncompetitive manners, respectively (Scheme 4). We further investigated the pathway by which 2α-7f was preferentially preceded through determining the dissociation constants of the EI (K i ) and ESI (K i ) complexes (Table 4). Apparently, the secondary plots for maltase (Figure 9b,c) and sucrase (Figure 10b,c) show K i values less than the K i values ( Table 4), indicating that 2α-7f was predominantly bound to maltase and sucrase (EI) rather than forming the ESI complex.

Nonlijingcompetitive
For 2α-7o, the inhibitory mechanisms against maltase and sucrase (Figures 11 and  12) were also examined using the above methodology. The active 2α-7o also inhibited maltase via a mixed-type inhibition ( Figure 11). On the other hand, the Lineweaver-Burk plots of 2α-7o against sucrase ( Figure 12) showed a series of straight lines, all of which intersected on the X-axis. Kinetic analysis showed that Vmax decreased with unchanged Km in the presence of increasing concentrations of inhibitor. This behaviour suggested that 2α-7o was a noncompetitive inhibitor against sucrase. The kinetic parameters of the active compounds are summarized in Table 4.

Scheme 4.
Putative inhibitory mechanism of 2α-7f and 2α-7o against rat intestinal α-glucosidases. E, S, I, and P represent enzyme, substrates (maltose and sucrose), inhibitors (2α-7f and 2α-7o), and glucose, respectively. For 2α-7o, the inhibitory mechanisms against maltase and sucrase (Figures 11 and 12) were also examined using the above methodology. The active 2α-7o also inhibited maltase via a mixed-type inhibition ( Figure 11). On the other hand, the Lineweaver-Burk plots of 2α-7o against sucrase ( Figure 12) showed a series of straight lines, all of which intersected on the x-axis. Kinetic analysis showed that V max decreased with unchanged K m in the presence of increasing concentrations of inhibitor. This behaviour suggested that 2α-7o was a noncompetitive inhibitor against sucrase. The kinetic parameters of the active compounds are summarized in Table 4.

Chemicals
All moisture-sensitive reactions were carried out under a nitrogen atmosphere solvents were distilled prior to use. High resolution electrospray ionisation mass spe (HRESIMS) were recorded with a Bruker microTof spectrometer (Billerica, MA, USA and 13 C NMR spectra were recorded (CDCl3 and CD3OD as solvents) at 400 and MHz, respectively, on a Varian Mercury + 400 NMR and a Bruker (Avance) 400 N spectrometer. Chemical shifts are reported in ppm downfield from TMS or solvent due. Thin layer chromatography (TLC) was performed on precoated Merck (Rahw NJ, USA) silica gel 60 F254 plates (0.25 mm thick layer) and visualized using anisaldehyde reagent. Column chromatography was performed using Merck silica 60 (70-230 mesh) and Sephadex LH-20.

Chemicals
All moisture-sensitive reactions were carried out under a nitrogen atmosphere. All solvents were distilled prior to use. High resolution electrospray ionisation mass spectra (HRESIMS) were recorded with a Bruker microTof spectrometer (Billerica, MA, USA). 1 H and 13 C NMR spectra were recorded (CDCl 3 and CD 3 OD as solvents) at 400 and 100 MHz, respectively, on a Varian Mercury + 400 NMR and a Bruker (Avance) 400 NMR spectrometer. Chemical shifts are reported in ppm downfield from TMS or solvent residue. Thin layer chromatography (TLC) was performed on precoated Merck (Rahway, NJ, USA) silica gel 60 F 254 plates (0.25 mm thick layer) and visualized using p-anisaldehyde reagent. Column chromatography was performed using Merck silica gel 60 (70-230 mesh) and Sephadex LH-20.
The isolated lignan 2β-7f used for structure characterization and bioactivity evaluation was obtained through multiple synthesis using the procedure described above. The combined crude reaction mixtures were further purified to obtain isolated 2β-7f and 2α-7f in pure form.

α-Glucosidase Inhibitory Activity
The α-glucosidase inhibitory activity against rat intestinal maltase and sucrase was determined according to our previous report [14]. The crude enzyme solution prepared from rat intestinal acetone powder (Sigma, St. Louis, MO, USA) was used as a source of maltase and sucrase. Rat intestinal acetone powder (1 g) was homogenized in 30 mL of 0.9% NaCl solution. After centrifugation (12,000× g × 30 min), the aliquot was subjected to an assay. The synthesized compounds (1 mg/mL in DMSO, 10 µL) were added with 30 µL of the 0.1 M phosphate buffer (pH 6.9), 20 µL of the substrate solution (maltose: 2 mM; sucrose: 20 mM) in 0.1 M phosphate buffer, 80 µL of glucose assay kit (SU-GLLQ2, Human), and 20 µL of the crude enzyme solution. The reaction mixture was then incubated at 37 • C for 10 min (for maltose) and 40 min (for sucrose). Enzymatic activity was quantified by measuring the absorbance of quinoneimine formed (500 nm) using the Bio-Rad 3550 microplate reader (Hercules, CA, USA). The percentage inhibition was calculated by [(A 0 − A 1 )/A 0 ] × 100, where A 1 and A 0 are the absorbance with and without the sample, respectively. The IC 50 value was deduced from a plot of percentage inhibition versus sample concentration and acarbose was used as a positive control. The experiment was performed in triplicate.

Kinetic Study of α-Glucosidase Inhibition
For kinetic analysis of the active compound, α-glucosidases and active compounds were incubated with increasing concentrations of maltose (0.5-8 mM) and sucrose (5-80 mM). The type of inhibition was determined by the Lineweaver-Burk plot. For calculation of K i and K i values, the slope and intercept from the Lineweaver-Burk plot were replotted vs.
[I], which provided the secondary plot.

DPPH Radical Scavenging
The radical scavenging activity was validated using the DPPH colorimetric method. Briefly, the synthesized compounds (20 µL) were added to 0.1 mM methanolic solution of DPPH (100 µL). The mixture was kept in the dark at room temperature in an incubator shaker for 15 min. The absorbance of the resulting solution was measured at 517 nm with a 96-well microplate reader. The percentage inhibition was calculated by [(A 0 − A 1 )/A 0 ] × 100, where A 0 is the absorbance without the sample, and A 1 is the absorbance with the sample. The SC 50 value was deduced from a plot of percentage inhibition versus sample concentration. Butylated hydroxytoluene (BHT) was used as the standard control and the experiment was performed in triplicate.

ABTS Radical Scavenging
The radical scavenging activity of synthesized compounds against ABTS•+ was carried out according to a procedure described previously [15]. Briefly, ABTS•+ radical cation was produced by mixing 10 mL of 7.4 mM ABTS with 0.5 mL of 2.6 mM potassium persulphate (K 2 S 2 O 8 ) for 16 h in the dark at room temperature. Before use, the ABTS•+ solution was diluted with ethanol to an absorbance of 0.70 ± 0.02 at 750 nm. The synthesized compounds (20 µL) were mixed with 80 µL of diluted ABTS•+ solution. After 2 h of incubation, the absorbance was read at 750 nm. The percentage inhibition was calculated by [(A 0 − A 1 )/A 0 ] × 100, where A 0 is the absorbance without the sample, and A 1 is the absorbance with the sample. The SC 50 value was determined from a plot of percentage inhibition versus sample concentration. Butylated hydroxytoluene (BHT) was used as the standard control and the experiment was performed in triplicate.

X-ray Crystallographic Analysis
Single crystal X-ray diffraction data were collected at 296(2) K on a Bruker X8 Prospector Kappa CCD diffractometer using an IìS X-ray microfocus source with multilayer mirrors, yielding intense monochromatic Cu-Ká radiation (ë = 1.54178 Å) for 2α-7f and on a Bruker X8 APEX II Kappa CCD diffractometer using graphite monochromatized Mo-Ká radiation (ë = 0.71073 Å) for 2β-7fBr. The structures were solved using SHELXTL XT 2013/1 [16], expanded using the difference Fourier method, and refined using full-matrix least squares on F2 with SHELXTL XLMP 2014/7 [17]. Absolute configurations of the two compounds were ambiguously determined with the estimated Flack parameters (x's) [18] that are statistically close to zero; the corresponding respective values are 0.040 (19) and 0.026 (12). A summary of selected crystallographic data for 2α-7f and 2β-7fBr is given in Table S2.

Computational Study
Complete geometry optimization was carried out with the density functional theory (DFT) calculations using the popular hybrid method (B3LYP) with the 6-31 + G(d,p) basis set. The methanol phase and the polarizable continuum model (PCM) calculations were performed using the Gaussian09 package (Gaussian, Inc.: Wallingford, CT, USA) [19] with default convergence criteria.

Conclusions
In conclusion, a new series of thioether and ether furofuran lignans were obtained by nucleophilic substitution between starting samin (5) and nucleophiles (thiols and alcohols) under acidic conditions. This synthetic approach provided 28 new furofuran lignans. Additionally, ether furofuran lignans displayed a diastereoselectivity higher than that of the thioether furofuran lignans. All synthesized compounds were fully characterized via different spectroscopic techniques (HRESI-MS and NMR) and single crystal X-ray diffraction. The reaction mechanisms of samin and alcohol nucleophiles were investigated for the first time. The NMR monitoring and DFT calculations suggested that the products were presumably formed through the S N 1-like mechanism by protonation of the hemiacetal center of samin (5) to generate the corresponding oxocarbenium ion as the intermediate. The subsequent reaction of this oxocarbenium ion with the alcohols then led to the observed only α-form either directly, or alternatively, by protonation of their epimer (β-form) through the S N 2-like transition state contributed by acid-catalysis. Furofuran lignans 2α-7f and 2α-7o, with a phenolic hydroxyl group, displayed remarkable inhibitory activities against rat intestinal maltase and sucrase as well as free radicals (DPPH and ABTS). Furthermore, the active 2α-7f and 2α-7o were selected as representatives for the kinetic analysis. The kinetic results revealed that 2α-7f and 2α-7o inhibited maltase and sucrase by mixed-type and noncompetitive inhibition. The present investigation provides fundamental clues to the structural motifs required for synthesizing a new series of improved inhibitors and a practical synthetic approach.
Funding: This project was financially supported by the Thailand Research Fund (RSA5880027) and the Faculty of Science, Chulalongkorn University (Sci-Super IV_61_003). W.W. is grateful to the Graduate School of Chulalongkorn University for a Postdoctoral Fellowship (Ratchadaphisek Sompot Fund). W.W., N.S., P.K., and P.P. are researchers at The Centers of Excellent in Natural Products (CENP), which is subsidized by Chulalongkorn University.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data presented in this work are available in the article and Supplementary Materials.

Conflicts of Interest:
The authors declare no conflict of interest.