Keywords

20.1 Introduction

Sugars are involved in a wide range of biochemical pathways [1]. Precise synthesis of the related natural products is essential to elucidate the biological phenomena and to develop pharmaceuticals based on natural products. Because sugars have multiple hydroxy groups and their structures are diversified by the position and the number of substitutions of the hydroxy groups, “distinction of hydroxy groups” is inevitably the key to their synthesis. Conventionally, protection/deprotection sequence in accordance with the original reactivity of the multiple hydroxy groups has been required for the distinction of them. Although protection/deprotection strategy has already been established as a reliable strategy for the synthesis, it generally leads to complicated synthetic routes and low efficiency for the total synthesis.

Ellagitannin is one of the hydrolyzable tannins containing the general structure of glucose esterified by gallic acid derivatives (Fig. 20.1) [2]. There are more than one thousand natural products contingent on the position and number of modifications by gallic acid derivatives and the mode of oligomerization. Since it has been reported that ellagitannins show various attractive biological activities, such as antiviral and immunostimulatory activity, synthetic studies on ellagitannins have been actively performed [2,3,4]. The conventional synthesis has been developed based on the protection/deprotection strategy as described above, which led to almost half of the synthetic processes being devoted to protection and deprotection steps (Fig. 20.1, route A).

Fig. 20.1
A chart presents the chemical structures of strictinin, pterocarinin, cercidinin A, coriariin A, and punicafolin. Route A yields protected glycosyl donor. Route B yields D glucose.

Structures of selected examples of ellagitannins and their synthetic strategy

In contrast to the protection/deprotection strategy, we have planned an unconventional strategy for the synthesis of ellagitannins. An efficient and short-step total synthesis of ellagitannins would be enabled by direct and sequential site-selective functionalization of d-glucose (Fig. 20.1, route B) [5]. The motivation was derived from the catalytic site-selective acylation reaction that we had reported in 2007 (Scheme 20.1) [6,7,8]. With organocatalyst 6, acylation takes place highly selectively at the C(4) of β-d-glucopyranoside. The important point is that the reaction proceeds selectively at an intrinsically less reactive secondary hydroxy group in a catalyst-controlled manner, even in the presence of the primary hydroxy group. Molecular recognition process between the substrate and the catalyst through multiple H-bonding interactions seems to be critically involved in achieving the catalyst-controlled selectivity (Fig. 20.2). We envisioned that the method for the selective functionalization of the desired hydroxy group of the sugar moiety streamlines the synthetic scheme by excluding protection and deprotection steps. Herein we describe our efforts for developing the protocol for the direct functionalization of a particular hydroxy group of glucose. The practical utility of the proposed synthetic strategy was demonstrated by the total syntheses of some ellagitannins.

Scheme 20.1
A reaction presents the conversion of beta glucopyranoside to 4 O acylate 8 through i P r C O whole twice O and catalyst 6.

Organocatalytic site-selective acylation of β-d-glucopyranoside

Fig. 20.2
A reaction schematic. It begins with the conversion of strictinin to 1, 4, 6 trigallate through oxidative phenol coupling, followed by the formation of beta glycoside through acylation and glucose through stereoselective glycosylation.

Proposed model for the transition state structure of site-selective acylation promoted by catalyst 6

20.2 Total Synthesis of Strictinin

Strictinin (1) was isolated by Okuda et al. in 1982 from the leaves of Casuarina Stricta (Casuarinaceae) [9, 10] (Scheme 20.2). Here, 1 possesses a galloyl group at C(1)-OH and hexahydroxydiphenoyl (HHDP) group with S axial configuration at C(4)- and C(6)-OH of d-glucose. Extensive studies on the biological activities of 1 indicated the potential utility of 1 for therapeutic applications, including antiallergic and immunostimulating agents [11,12,13,14,15,16,17,18]. Khanbabaee [19] and Yamada [20] reported pioneering studies on the total synthesis of 1, focusing on the introduction or construction of HHDP moiety stereoselectively. Our retrosynthetic analysis of 1 based on the sequential site-selective functionalization strategy is described in Scheme 20.2. We expected that oxidative phenol coupling between the galloyl groups at the C(4) and C(6) of 9 would construct the HHDP moiety of 1 with high diastereoselectivity, according to Yamada’s precedent [20]. Coupling precursor 9 was planned to be synthesized by the introduction of two galloyl groups with adequate protective groups at C(4)- and C(6)-OHs into β-glucopyranoside 10 in a catalyst-controlled and substrate-controlled manner, respectively. Stereoselective glycosylation of gallic acid derivative using unprotected glucose would allow us to commence the scheme for the streamlined total synthesis of strictinin (1) without the protection of glucose hydroxy groups.

Scheme 20.2
A reaction schematic. It begins with conversion of strictinin to 1, 4, 6 trigallate through oxidative phenol coupling followed by formation of beta glycoside through acylation and glucose through stereoselective glycosylation.

Retrosynthetic analysis of strictinin (1), MOM = methoxymethyl

Investigation for the first step, stereoselective glycosylation of unprotected glucose, was summarized in Table 20.1. In 1979, Grynkiewicz reported that the Mitsunobu reaction of glucose with phenol successfully provided the phenol glycoside (α/β = 1/8) [21]. Referring to the procedure, treatment of glucose and gallic acid derivative with diisopropyl azodicarboxylate (DIAD) and triphenylphosphine in N,N-dimethylformamide (DMF) provided the desired glycoside in 60% yield (entry 1). However, the α:β ratio was not satisfactory (α/β = 50/50). To improve the stereoselectivity, the effects of solvent were investigated. The reaction using tetrahydrofuran (THF) as a solvent dramatically increased the β-selectivity, while the yield of the glycoside significantly decreased (entry 2, 17% yield, α/β = 1/99). Product analysis indicated that the major side product was 1,6-diacylated product, which was derived from further Mitsunobu reaction at C(6)-OH of the β-glycoside. 1,4-dioxane was found to be the best solvent for our purpose to give the β-glycoside 10 in 64% yield, although glucose was scarcely soluble in 1,4-dioxane (entry 3). Finally, the use of excess amounts of glucose and Mitsunobu reagents improved the reaction yield to provide the glycoside 10 in 78% yield (entry 5).

Table 20.1 Optimization of Mitsunobu glycosylation using glucose

To elucidate the origin of the high stereoselectivity, a mechanistic analysis was performed. To begin with, we did not pay attention to the configuration of the anomeric carbon of commercial glucose. In the course of the mechanistic study, we recognized that commercial d-glucose is supplied as an almost pure α-form in most cases (Fig. 20.3). Selective crystallization of α-anomer of glucose is supposed to take place during the manufacturing of commercial d-glucose [22], although there was no description of the anomeric ratio on the label of the commercial reagent which we had employed at the initial study. To verify the possibility of the inversion of the anomeric stereogenic center, the reactions were performed using partially anomerized glucose (Fig. 20.4). Benzoylation of α-glucose under Mitsunobu condition in 1,4-dioxane gave the β-glycoside with high stereoselectivity. An increase in the β-anomer content in the starting d-glucose led to an increase in the α-anomer ratio of the product. In addition to these results, the 13C kinetic isotope effect experiments [23, 24] convinced us that Mitsunobu glycosylation in dioxane proceeds via a direct SN2 mechanism to give the inversion product, while the reaction in DMF gave almost 1:1 mixture of the α- and β-glycoside via an SN1 mechanism [25].

Fig. 20.3
A spectral graph plots the N M R spectrum of commercial D-glucose, with peaks labeled as H O D or H 2 O and D M S O, and a chemical structure of D-glucose indicating a J-value of 3.6 hertz.

1H NMR spectrum of commercial d-glucose

Fig. 20.4
A chemical reaction of glucose being converted into glycoside 13, along with a table that provides data on the yield and alpha by beta ratio of the product under different conditions, highlights the efficiency and outcomes of various solvent and ratio combinations.

Analysis of stereochemical course depending on solvents, Bz = benzoyl

We then investigated the second step, the organocatalytic C(4)-OH selective acylation of the β-glycoside 10 (Table 20.2). Acylation of 10 catalyzed by 6 with anhydride 14 under the previously optimized conditions [7, 8] was sluggish to give the desired 4-O-gallate in low yield (entry 1, 18% yield). The use of 2,4,6-collidine as a part of the solvent (CHCl3/2,4,6-dollidine = 9/1) afforded 15 in a much better yield (entry 2, 83% yield). This is probably because a large amount of collidine contributes to avoiding protonative deactivation of the catalyst by the in situ generated carboxylic acid from anhydride 14, even though the basicity of collidine is significantly lower than that of the catalyst (Fig. 20.5). Finally, the yield of the desired 1,4-digallate 15 was improved to be 91% in a reaction with a substrate concentration of 0.04 M (entry 3).

Table 20.2 Optimization of organocatalytic C(4)-OH selective acylation
Fig. 20.5
A chemical reaction presents the regeneration of catalyst through 2, 4, 6 collidine, with p K a values indicated for two resulting compounds.

Regeneration of the active catalyst by proton exchange process between the catalyst and the excess weak base

In the third step, the selective galloylation of the C(6)-OH of 15 was examined (Scheme 20.3). Initially, the third step was not assumed to be difficult because of the intrinsically high reactivity of the primary hydroxy group. However, a considerable amount of examination was required to achieve a satisfactory selectivity. The introduction of a galloyl group to the C(6)-OH was accomplished by treatment with gallic acid derivative 16 and 2-chloro-1,3-dimethylimidazolium chloride (DMC) to give 1,4,6-trigallate 17 in 72% yield. The second and third steps, the introduction of galloyl groups at the C(4)-OH and C(6)-OH, were successfully accomplished in a one-pot procedure through the activation of gallic acid 16, in situ generated from anhydride 15, by the addition of DMC to the reaction medium after the estimated completion of the C(4)-OH acylation (Scheme 20.4). The one-pot transformation was applicable to gram-scale synthesis of the 1,4,6-gallate 17.

Scheme 20.3
A reaction schematic presents the conversion of compound 15 to 17 by treatment with D M C, D M A P and C H C l 3, collidine, pyridine.

Site-selective acylation of C(6)-OH of 1,4-digallate 15

Scheme 20.4
A reaction schematic. It begins with conversion of compound 10 to compound 15 and compound 17.

One-pot procedure for acylation of C(4)- and C(6)-OHs

Synthetic scheme toward strictinin (1) was summarized in Scheme 5. Based on our original protocol for the sequential site-selective functionalization of glucose, the key intermediate, 1,4,6-trigallate 17, was obtained by only 2 steps from d-glucose. The precursor 18 for the stereoselective oxidative phenol coupling was obtained by hydrogenolytic removal of the Bn groups of 17. On the treatment of 18 with CuCl2 and butylamine, the oxidative coupling proceeded smoothly to construct the HHDP group with the desired S axial configuration, as expected from Yamada’s report [20]. Finally, global deprotection of MOM groups under acidic conditions provided strictinin (1). By virtue of the sequentially selective modification of glucose –OHs, total synthesis of 1 was achieved in 5 overall steps from d-glucose [26]. The extremely short-step total synthesis stems from avoiding protective groups for glucose.

Scheme 20.5
A reaction schematic. It begins with conversion of alpha D glucose to compounds 10, 17 and 18 to strictinin.

Total synthesis of strictinin (1)

20.3 Total Synthesis of Tellimagrandin II and Pterocarinin C

The sequential site-selective functionalization strategy established for the total synthesis of strictinin (1) was then applied to the synthesis of tellimagrandin II (19) and pterocarinin C (2) (Fig. 20.6). Tellimagrandin II (19) [27], isolated from Tellima grandiflora in 1976 by Wilkins and Bohm, is a 4,6-HHDP-type ellagitannin, showing potent antiviral activity [28, 29]. Pterocarinin C (2) is a regioisomeric natural product of tellimagrandin II, possessing an HHDP group at the C(2)- and C(3)-OHs of glucose. Pterocarinin C (2) was first isolated from the leaves of Tibouchina semidecandra by Okuda et al. [30] and reported to show neuroprotective activity [31]. The total syntheses of 19 and 2 were achieved by Feldman [32] and Khambabaee [33], respectively. In both cases, protected d-glucose derivatives (20, 21), in which C(4)- and C(6)-OHs are differentiated from C(2)- and C(3)-OHs by proper protective groups, were employed for the total syntheses. In contrast, we envisioned that the application of our strategy for the sequential functionalization of C(1)-OH, C(4)-OH, and C(6)-OH allowed us to accomplish the total syntheses of 19 and 2 by almost the same synthetic scheme without protection of glucose –OHs.

Fig. 20.6
A chart presents the chemical structures of tellimagrandin and pterocarinin C. Previous report uses the former to form compounds 20 and 21. Our synthesis uses the latter to form glucose.

Structures and retrosynthesis of tellimagrandin II and pterocarinin C

The synthetic schemes of 19 and 2 were described in Scheme 20.6. The β-glycoside 10 was prepared by direct glycosylation of unprotected d-glucose. The differently protected galloyl groups (G2) were introduced at C(4)- and C(6)-OHs in a similar manner for the synthesis of strictinin (1). Introduction of all-MOM-protected galloyl groups (G1) at C(2)- and C(3)-OHs followed by deprotection of the benzyl groups of G2 provided coupling precursor 22. Oxidative HHDP construction and global deprotection of MOM groups gave tellimagrandin II (19) in 6 overall steps. Similarly, the site-selective introduction of G1 and G2 groups into glycoside 10 provided the precursor 23 for the oxidative construction of the 2,3-HHDP group. As we had expected, total synthesis of pterocarinin C (2) was also achieved by the oxidative coupling of 23 and the acidic deprotection of the MOM groups [34].

Scheme 20.6
A chart presents 2 reaction schematics. Both begin with compound 10. In one, it is followed by compound 22 and tellimagrandin 2. In the other, it is followed by compound 23 and pterocarinin C.

Total synthesis of tellimagrandin II (19) and pterocarinin C (2) EDCI = 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride

In nature, 19 is proposed to be produced without protective groups by sequential enzymatic reactions (Scheme 20.7a) [35]. β-Glucogallin, derived from enzymatic glycosylation of uridine 5′-diphosphate (UDP)-glucose, is sequentially converted to β-pentagalloyl glucose by several acyltransferases. Surprisingly, site-selective construction of the 4,6-HHDP group from 25 was accomplished by a particular oxidase to furnish tellimagrandin II (19). Unexpectedly, our synthetic scheme became similar to the biosynthetic pathway (Scheme 7b). Direct stereoselective glycosylation of glucose provided the first intermediate, β-glycoside, and the second intermediate, a pentagalloyl glucose derivative, generated by sequential site-selective galloylation of the β-glycoside. The similarity in both synthetic schemes seems to be closely related to the high efficiency of the synthesis.

Scheme 20.7
Two chemical synthesis pathways. A. Biosynthetic pathway begins with U D P glucose followed by beta glucogallin, pentagalloylglucose and tellimagrandin 2. B. Current pathway begins with glucose followed by compound 10, penta galloyl glucose derivative and tellimagrandin.

Comparison between a biosynthetic pathway and b our synthetic route for tellimagrandin II (19)

20.4 Total Synthesis of Cercidinin A

Cercidinin A (3) (Fig. 20.7) was isolated from the fresh bark of Cercidiphyllum japonicum by Nishioka in 1989 [36]. After a revision of the first proposed structure [37], the revised structure 3 with the 3,4-HHDP group was confirmed to be correct through the total synthesis by Yamada’s group [38]. For the synthesis of 3, the differentiation of C(3)- and C(4)-OHs from C(2)- and C(6)-OHs is essential. However, differentiation of the two secondary hydroxy groups at C(2) from at C(3) has never been accomplished so far by our strategy for the sequential site-selective functionalization described above (Scheme 20.8). To synthesize cercidinin A (3), it was necessary to further develop the site-selective acylation strategy.

Fig. 20.7
Left. Chemical structure of cercidinin A. Right. A chart presents a structure that marks the distinction of C 2 and C 3 O H.

Structures of cercidinin A and the issue in synthesis

Scheme 20.8
A reaction schematic. Glucose is converted into 1, 4 digallate through glycoylation and C 4 O H acylation. 1, 4 digallate is converted into 4, 6 H H D P ellagitannin, strictinin, tellimagrandin 2 and 2, 3 H H D P ellagitannin, and pterocarinin C. 1, 4 digallate is converted into 3, 4 H H D P ellagitannin and cercidinin A.

Outline of synthesis of cercidinin A based on sequential site-selective functionalization

In the synthesis of strictinin (1), tellimagrandin II (19), and pterocarinin C (2), selective acylation of the primary hydroxy group was accomplished based on its intrinsic high reactivity among the three free hydroxy groups of the 1,4-digallate (Scheme 20.3) [26, 34]. Actually, acylation took place selectively on the C(6)-OH simply by using condensation agent DMC. On the other hand, by treatment with acid anhydride 14, 15 underwent preferential acylation at C(3)-OH (Scheme 20.9).

Scheme 20.9
A reaction schematic presents the conversion of compound 15 to 3 O acylate, 2 O acylate and 6 O acylate.

C(3)-OH preferential acylation of 1,4-digallate

Unexpectedly, under these conditions, the primary C(6)-OH was totally unreactive. The dramatic reactivity change of the three hydroxy groups of the 1,4-digallates may be attributed to the effects of the counteranion of the reactive catalytic intermediate [39,40,41]. The counteranion of acylpyridinium salts (ArCOO) could possibly form dual H-bonds with C(2)- and C(3)-OHs, resulting in the selective acylation of these hydroxy groups in the presence of the primary C(6)-OH (Fig. 20.8).

Fig. 20.8
A reaction schematic begins with reaction between 4, dimethylamino, pyridine with compound 14 to obtain acyl pyridinium carboxylate and 3 O acylate.

Proposed rational explanations for relatively high reactivity of C(3)- and C(2)-OHs

Having been able to distinguish the C(2)-OH from C(3)-OH of the 1,4-digallate 15, we worked on the total synthesis of cercidinin A (3) (Scheme 20.10). The oxidative coupling reaction of phenol 28, prepared via galloylation of the free hydroxy groups at C(2) and C(6) of 26 followed by deprotection of the benzyl groups, successfully took place to give coupling product 29 with the desired R configuration as a single diastereomer. However, a serious problem arose during the final deprotection step. Under usual acidic conditions (HCl in i-PrOH/THF), 3 was not obtained because of the degradation of 29 and uncompleted partial deprotection. Under these circumstances, we noticed Sajiki’s report that deprotection of the acid-sensitive protective groups was feasible under the conditions of hydrogenation with Pd/C in MeOH or EtOH [42]. According to the report, 29 was subjected to the hydrogenation conditions in CHCl3/MeOH (1/1) to afford cercidinin A (3) in 63% yield via removal of the MOM groups with minimal degradation of 29. Thus, total synthesis of cercidinin A (3) was also completed without protection of the hydroxy groups of the glucose moiety [43].

Scheme 20.10
A reaction schematic begins with compound 26 converting to 28 and 29 followed by cercidinin A.

Total synthesis of cercidinin A

20.5 Total Synthesis of Punicafolin and Macaranganin

The final targeted natural products in this chapter are punicafolin (5) and macaranganin (30) (Fig. 20.9). Nishioka and co-workers reported the isolation of 5 from the leaves of Punica granatum in 1985 [44] and 30 from Macaranga tanarius in 1990 [45], respectively. Because of their characteristic 3,6-HHDP bridged structure, the pyranose ring is proposed to be in axial-rich conformation such as 1C4 conformation. The difference between the two natural products is the configuration of the axial chirality in the HHDP moiety. The R-isomer 5 shows the inhibitory activity of invasion of HT1080 fibrosarcoma cells [46], while S-isomer 30 exhibits the inhibitory effect of prolyl endopeptidase [47]. Due to their unique structural features, two challenging issues were identified for the synthesis: differentiation of the hydroxy groups and stereoselective formation of a 3,6-HHDP group with a less stable axial-rich conformer of glucose. Several examples emphasized the difficulties in the construction of the 3,6-HHDP bridge via the flipping process of the pyranose ring [48, 49].

Fig. 20.9
A chart presents the chemical structures of punicafolin, macaranganin, and beta glucose 4 C 1 and 1 C 4.

Structures of punicafolin and macaranganin

Yamada et al. reported an excellent strategy for the total synthesis of a 3,6-HHDP-type ellagitannin, (–)-corilagin (33) in 2008 (Scheme 20.11) [50]. They overcame the conformational problem by using the ring-opened intermediate 31. After the coupling reaction of 31, reconstruction of the pyranose ring led to the first total synthesis of 33. Under these backgrounds, we planned to construct the HHDP group directly from a pentagalloylglucose derivative without opening the pyranose ring.

Scheme 20.11
A reaction schematic. It begins with opening of pyranose ring of glucose leading to compound followed by oxidative coupling to obtain compound 32 followed by reconstruction of pyranose ring to obtain corilagin.

Yamada’s total synthesis of a 3,6-HHDP ellagitannin, (–)-corilagin

Our retrosynthetic analysis of punicafolin (5) is outlined in Fig. 20.10. We expected that the oxidative phenol coupling reaction of pentagalloylglucose derivative 34 could proceed via an unstable 1C4 conformation. The possibility was already suggested by Yamada in 2017 in the direct oxidative coupling reactions of the related pentagalloylglucose [51]. Inspired by the precedents, conformational analysis with molecular mechanics and density functional theory (DFT) calculation of β-glucose and pentabenzoylglucose was performed. The difference in the potential energy between the 4C1 and 1C4 conformers of β-glucose was found to be significant. On the other hand, to our surprise, the energy difference between those of pentabenzoylglucose was found to be only 1.0 kcal/mol. Natural bond orbital (NBO) analysis of both stable conformers suggested that the stronger anomeric effects in pentabenzoylglucose contribute to the relatively high stability of the axial-rich 1C4 conformer. The stronger anomeric effects [52,53,54,55] may result from the lowering of the energy level of the non-bonding σ* orbital of the C(1)-OBz bond by the electron-withdrawing group. Then, we decided to challenge the direct oxidative phenol coupling of the properly protected pentagalloylglucose derivative via the 4C1 to 1C4 ring flipping process of the pyranose ring.

Fig. 20.10
A reaction schematic. It begins with oxidative coupling with ring flipping of punicafolin to obtain compound 34 with 1 C 4 conformation and 4 C 1 conformation to further obtain glucose by sequential site selective galloylation.

Retrosynthetic analysis of punicafolin and conformational analysis of the model compound of the key intermediate

Site-selective acylation of 1,4-digallate 35 with gallic acid anhydride 36 was investigated (Table 20.3). Digallate 35 was prepared by our established protocol including stereoselective Mitsunobu glycosylation and organocatalytic C(4)-OH acylation. DMAP-catalyzed acylation of 35 took place at the secondary hydroxy groups at the C(2)- and C(3)-OHs, with a slight preference for the C(2)-OH acylation (entry 1). With the expectation that H-bonding interactions between the substrate and catalyst affect the site selectivity [5, 56], the effects of catalyst 6 on the site selectivity of acylation were examined. However, the site selectivity was not improved (entry 2, 57% site selectivity). Then, its diastereomeric catalysts 39, ent-6, and ent-39 were examined to find that acylation catalyzed by 39 exhibited the highest site selectivity (entry 3, 70% site selectivity). To investigate the effects of the side chain of the catalyst, further screening of catalysts 4042 with the same configuration as that of catalyst 39 was performed (entries 6–8). The highest improvement of the site selectivity was observed in the case of catalyst 42 (entry 8, 75% site selectivity). Finally, treatment of 35 with excess amounts of 36 (2.2 eq.) slightly improved the yield of 2-O-acylate 37 and site selectivity (entry 9, 51% yield of 37, 78% site selectivity).

Table 20.3 Investigation for the C(2)-OH selective acylation of 1,4-digallate 35

With pentagalloyl glucose derivative 43 obtained by 2 steps sequence from 37, the challenging oxidative phenol coupling was investigated (Scheme 20.12). When 43 was treated with CuCl2 and butylamine, a standard protocol for the oxidative phenol coupling for the total synthesis of strictinin (1), only decomposition of 43 was observed. We then investigated the effects of the chiral ligands. In 2017, Quideau and Deffieux reported that sparteine acts as an efficient ligand for copper-mediated oxidative phenol coupling [57]. Fortunately, the CuCl2/sparteine system was found to be also effective for our purpose. The 3,6-HHDP bridge was successfully constructed by treatment of 43 with CuCl2 and (+)-sparteine. It is noteworthy that the axial configuration was completely controlled to be R under these conditions. In contrast, the use of (−)-sparteine instead of (+)-sparteine resulted in the formation of (S)-congener 45 as a single diastereomer. Thus, we developed a method for ligand-controlled stereoselective construction of the 3,6-HHDP bridge via the flipping process of the pyranose ring [58].

Scheme 20.12
A reaction schematic begins with compound 37 transitioning to compound 43 followed by R and S isomers 44 and 45. A table presents the yields of 44 and 45 for butylamine, plus and minus sparteine.

Stereodivergent construction of 3,6-HHDP bridge

Finally, deprotection of the MOM groups under hydrogenation conditions, as in the case of synthesis of cercidinin A (3), provided punicafolin (5) and macaranganin (30) (Scheme 20.13). Development of the catalytic C(2)-OH selective acylation and ligand-controlled stereoselective HHDP construction enabled us to achieve stereodivergent total synthesis of 5 and 30 from the common intermediate 43. Thus, the first total synthesis of complicated natural glycosides punicafolin (5) and macaranganin (30) has been achieved in only 7 steps from d-glucose [59].

Scheme 20.13
A reaction schematic. Compounds 44 and 45 result in punicafolin and macaranganin.

Endgame of total synthesis of punicafolin and macaranganin

20.6 Conclusion

We described our synthetic studies of ellagitannins based on a sequential site-selective functionalization strategy. The catalyst-controlled site-selective acylation led to the proposal of the non-conventional unique strategy. In the course of the studies on the total syntheses, we developed a method for stereoselective glycosylation using unprotected glucose and catalyst-controlled site-selective acylation of the desired position of 1,4-digallate. Although the protection/deprotection process has been considered inevitable for the synthesis of sugar-related compounds, the sequential site-selective functionalization strategy enabled to avoid protective groups for glucose throughout the total synthesis. The proposal of the novel retrosynthetic analysis and its realization in the actual total synthesis of natural products should contribute to the advancement of synthetic organic chemistry toward the dreams of truly protecting-group-free total synthesis.