Keywords

16.1 Introduction

Lingzhi, a fungus of the genus Ganoderma, which is widely distributed over tropical and subtropical latitudes in Asia, has been widely used as a traditional Chinese medicine for the treatment of cancer, hypertension, and asthma, particularly in China, Korea, and Japan [1]. In 2014, Cheng and co-workers reported the isolation of cochlearol B (1) in its racemic form from Ganoderma cochlear [2] (Fig. 16.1). In the same year, Qiu and co-workers isolated ganocin A (2) in its racemic form from the same fungus [3]. These two natural products are meroterpenoids, which have closely related fused pentacyclic structures. Natural product 1 possesses a substituted cyclobutane ring instead of the substituted tetrahydrofuran ring of 2 in its framework. With regard to biological activities, the (−)-enantiomer of 1, which was obtained after chiral HPLC separation, exhibited potent inhibitory activity against p-Smads, while the (+)-enantiomer was inactive [2]. Thus, (−)-1 is a potential lead for renoprotective agent. On the other hand, only anti-acetylcholinesterase activity study was examined for 2. However, no activity was observed [3]. We assumed that the enantiomer of 2 having the same configuration as (−)-1 would inhibit p-Smads. Accordingly, efficient synthetic methods are required to access these structurally intriguing natural products in order to evaluate their biological activities accurately.

Fig. 16.1
Two chemical structures of cochlearol B and ganocin A with their equivalent chemical structures. Cochlearol B has a benzene ring bonded to tetrahydropyran and a six-membered ring. Ganocin A has a benzene ring bonded to tetrahydropyran and a six-membered ring. There is a C O C bond.

Structures of cochlearol B (1) and ganocin A (2)

Full size image

To date, four groups have reported the total synthesis of these two natural products. In 2020, Zhao and co-workers reported the first total synthesis of racemic 2 [4], and in 2021, our group reported the first total synthesis of racemic 1 [5]. Subsequently, Schindler and co-workers reported the total synthesis of the (+)-1 by optical resolution in 2022 [6]. Further, in 2022, Hao and co-workers reported the total syntheses of racemic 1 and 2 [7]. In 2023, we reported the enantioselective total syntheses of (−)-1 and (+)-2, which are expected to show p-Smads inhibitory activity [8]. Here, we describe the enantioselective total syntheses of (−)-1 and (+)-2, together with the racemic total synthesis of (±)-1.

16.2 Retrosynthetic Analysis of Cochlearol B (1)

In order to develop a synthetic route that can be applied to the total synthesis of the optically active forms of 1 and 2, we first performed the retrosynthetic analysis of the racemic form of 1. The retrosynthetic analysis of 1 is shown in Scheme 16.1. The α,β-unsaturated aldehyde moiety, which seems to be the most labile in 1, should be incorporated in 3 at the final stage. One of the key steps in this synthesis is the intramolecular [2+2] photocycloaddition of 4. We envisioned that the pentacyclic framework of 3 could be formed by photo irradiation of 4. The tricyclic structure of 4 was expected to be accessed by the phenolic oxidative cyclization of phenol 5. Incorporation of the alkenyl chain could be realized by the Nozaki-Hiyama-Kishi (NHK) reaction of iodide 6, which could be prepared from commercially available 7 in four steps, including a copper-catalyzed coupling reaction.

Scheme 16.1
A chemical reaction pathway. Cochlearol B with double bond saturates by alpha, beta-unsaturated aldehyde moiety incorporation. The bridge system breaks by 2 plus 2 photo cycloaddition, followed by phenolic oxidative cyclization, incorporation of alkenyl chain, and coupling reaction to give 1 benzyloxy 3 iodobenzene.

Retrosynthetic analysis of cochlearol B (1)

Full size image

16.3 Synthesis of Diketone 15

Initially, we conducted a coupling reaction between the commercially available iodide 7 and cyclohexan-1,3-dione 8 using a catalytic amount of CuI and proline to obtain enol 9 in 93% yield (Scheme 16.2) [9]. Following this, the Appel-type reaction of enol 9 proceeded smoothly to give iodide 6 in 89% yield [10]. To prepare the metal reagent in the next step, the carbonyl group in 6 was protected as 1,3-dioxolane to produce dioxolane 10 in 90% yield. The resulting dioxolane 10 was treated with tBuLi or Mg to convert into a lithium or a Grignard reagents, but the addition reaction with ketone 11 did not proceed. This was probably because of the bulky structures around the reaction sites. To solve this problem, we employed the NHK reaction [11]. The coupling between iodide 6 and aldehyde 13 by the NHK reaction afforded alcohol 14 in 78% yield. Subsequently, the IBX oxidation of 14 produced diketone 15 in 90% yield [12].

Scheme 16.2
A reaction pathway. An iodide reacts with cyclohexan-1, 3-dione to give an enol. The carbonyl group of the iodide does not react and a dioxolane is produced. This dioxolane gives a Grignard reagent. The reaction between iodide and aldehyde gives alcohol. This reacts with I B X and produces diketone.

Synthesis of diketone 15

Full size image

16.4 Synthesis of Phenol 5

Next, we focused on the incorporation of the methyl group. Diketone 15 was reacted with a methyl magnesium bromide, resulting in the selective methylation of the carbonyl group in the six-membered ring, to give the undesired 17 in 59% yield (Scheme 16.3). Based on this result, the reactivity of the carbonyl group in the six-membered ring toward nucleophiles was anticipated to be higher than that of the carbonyl group in the side chain. Accordingly, to obtain enone 5, we planned the synthesis such that the methyl group is incorporated into the carbonyl group in the side chain after reducing the more reactive carbonyl group in the six-membered ring, followed by oxidation. Consequently, the treatment of diketone 15 with sodium borohydride produced alcohol 18 regioselectively in 78% yield. Next, the methylation of 18 afforded diol 19 in 98% yield. For the phenolic oxidative cyclization, the benzyl group of 19 was cleaved by the lithium-naphthalene system, producing phenol 20 in 85% yield [13]. Subsequent Swern oxidation of 20 proceeded to afford phenol 5 in 89% yield [14].

Scheme 16.3
A reaction pathway. A diketone reacts with N a B H 4 and produces an alcohol. It reacts with M e M g B r and gives diol. The keto group converts into M e and O H. It reacts with L i, naphthalene, and gives a substituted phenol. If the diketone reacts with M e M g B r, O H converts into M e and O H.

Synthesis of phenol 5

Full size image

The regioselectivity of the reduction of diketone 15 was next investigated. The conformational analysis of 15 was performed using DFT calculations (Scheme 16.4). In the conformation shown in Scheme 16.4, nucleophilic attack to the carbonyl group in the side chain was probably blocked by the benzene ring and the hydrogens in the cyclohexanone ring. Thus, it is likely that the carbonyl group in the six-membered ring, which is less hindered, reacted selectively to produce 18.

Scheme 16.4
A reaction pathway. A diketone reacts with N a B H 4 and converts into a product. The keto group attached to the cyclohexene of the reactant is converted into the O H group. It presents the ball and stick model of the optimized structure of diketone. It has steric hindrance by aryl and alkyl groups.

Proposed mechanism of the regioselective reduction of diketone 15

Full size image

16.5 Phenolic Oxidative Cyclization and Intramolecular [2+2] Photocycloaddition

With 5 in hand, we investigated the phenolic oxidative cyclization—the key reaction of this synthesis, to construct the tricyclic framework (Table 16.1) [15]. Initially, we expected to obtain hydroquinone derivative 4 as the main product. However, quinone hemiacetals 21 and 22 were obtained as a diastereomeric mixture, and 4 was not obtained under any conditions. First, the treatment of phenol 5 with iodobenzene diacetate (PIDA) in hexafluoro-2-propanol (HFIP) afforded a complex mixture of unknown compounds (Table 16.1, entry 1). In contrast, the oxidative cyclization did not proceed in dichloromethane (DCM) (entries 2 and 3). The reaction in the presence of bis(trifluoroacetoxy)iodobenzene (PIFA) in DCM or iodosobenzene (PhIO) in HFIP afforded a complex mixture of unknown compounds (entries 4 and 5). However, when 5 was treated with PIDA (1.2 eq) in HFIP/DCM (1/50), the desired tricyclic compounds were produced (entry 6). Eventually, the phenolic oxidative cyclization of 5 by the treatment with PIDA (5.0 eq) in HFIP/DCM (1/50) at − 78 to − 40 °C successfully afforded tricyclic compounds 21 (62% yield) and 22 (14% yield) (entry 7) [15].

Table 16.1 Investigation of phenolic oxidative cyclization
Full size table

We then focused on the next key reaction—the intramolecular [2+2] photocycloaddition (Scheme 16.5). First, we investigated the conversion of the quinone hemiacetal 21 obtained as the main product to 4. As expected, the conversion of the quinone hemiacetal moiety to phenol was difficult. Consequently, Luche reduction at − 78 °C afforded 4 in a low yield of 37% [16]. Subsequently, 4 was subjected to intramolecular [2+2] photocycloaddition upon irradiation with a mercury lamp [17]. Contrary to our expectations, the desired cyclized compound 3, as shown in the retrosynthetic analysis, was not produced; instead, a mixture of unknown compounds was obtained. On the other hand, in 21—the major product of the phenolic oxidative cyclization, the double bond of the cyclohexenone ring, which was the reaction center for the intramolecular [2+2] photocycloaddition, was conjugated with the electron-withdrawing quinone hemiacetal ring. Therefore, the reactivity for the intramolecular [2+2] photocycloaddition was thought to be much higher than that of the double bond of the cyclohexenone ring of 4. As expected, the intramolecular [2+2] photocycloaddition of 21 proceeded smoothly to form four- and five-membered rings simultaneously, affording pentacyclic 23 in 74% yield. On the other hand, pentacyclic 24 was not formed from diastereomer 22. We assumed that the steric repulsion between the hydroxy group of the quinone hemiacetal and alkenyl chain inhibited the intramolecular [2+2] photocycloaddition. Moreover, it was found that 21 and 22 existed in equilibrium in DCM at room temperature, resulting in a mixture with a ratio of 3:1 to 4:1.

Scheme 16.5
3 reaction pathways of a substituted phenol in the presence of P I D A in H F I P. It produces three compounds, a hydroquinone derivative, and two tricyclic diastereomer compounds. The hydroquinone derivative does not react under h nu. Photocycloaddition of a tricyclic compound gives pentacyclic.

Phenolic oxidative cyclization and intramolecular [2+2] photocycloaddition

Full size image

16.6 Completion of the Total Synthesis of (±)-Cochlearol B (1)

The completion of the total synthesis of cochlearol B (1) is depicted in Scheme 16.6. Luche reduction of pentacyclic 23 proceeded smoothly to afford phenol 3. We spent a lot of time for the introduction of the α,β-unsaturated aldehyde moiety. After extensive experimentation, the following route was found to give the best result. Pivaloyl protection of phenol 3 followed by treatment with Bredereck’s reagent 26 afforded enaminone 27 in 89% yield from 3 [18]. After the conversion of enaminone 27 to triflate 28 using triflic anhydride [19] and the subsequent reduction of 28 by triethylsilane in the presence of tetrakis(triphenylphosphine)palladium (0), α,β-unsaturated aldehyde 29 was obtained in 64% yield from 27 [20]. Finally, cleavage of the pivaloyl group furnished (±)-cochlearol B (1) in 94% yield. This first total synthesis of 1 was achieved via the longest linear sequence of 16 steps in 9% overall yield.

Scheme 16.6
A reaction pathway. A pentacyclic reacts with N a B H 4. The product reacts with P i v C l. O H converts into O P i v. It reacts with a reagent and T f 2 O. It gives C H O. When it reacts with E t 3 S i H and O T f converts into H. On oxidation, it gives cochlearol B. It has a ball and stick model.

Completion of the total synthesis of (±)-cochlearol B (1)

Full size image

16.7 Enantioselective Total Synthesis of (−)-Cochlearol B (1)

Having succeeded in the total synthesis of (±)-1, we attempted to synthesize an optically active form of 1. According to reports, only the (−)-enantiomer of 1 exhibits inhibitory activity against p-Smads. In order to accurately evaluate the biological activities of any synthesized compounds, it is necessary to synthesize the compound in optically active forms. We planned the total synthesis of 1 in its optically active form, which could be utilized for the synthesis of 2 in its optically active form. To develop efficient enantioselective synthetic routes for 1 and 2 in their optically active forms, the introduction of a chiral center by the enantioselective reduction of 15 was thought to be effective.

16.8 Enantioselective Synthesis of Diol (+)-19

Enantioselective reduction, one of the key reactions of this enantioselective synthesis, was carried out on the common intermediate diketone 15 in racemic total synthesis of 1 (Scheme 16.7). The Corey-Bakshi-Shibata (CBS) reduction was selected for the enantioselective reduction of the common intermediate 15 to incorporate the chiral center, which would control all the subsequent stereogenic centers. Treatment of 15 with borane dimethylsulfide complex in the presence of (S)-CBS catalyst proceeded smoothly to furnish chiral alcohol (+)-18 in 73% yield with 93% ee [21]. The absolute configuration of (+)-18 was determined by comparing the calculated and experimental circular dichroism (CD) spectra (Fig. 16.2). High regioselectivity was achieved in this reduction, similar to that observed in the synthesis of (±)-1 using NaBH4. Subsequent incorporation of a methyl group in (+)-18 produced diastereoselectively diols (+)-19a (74% yield) and 19b (24% yield). The proposed mechanism of this diastereoselective synthesis is depicted in Scheme 16.8.

Scheme 16.7
A reaction pathway. A diketone reacts with B H 3 S M e 2 and gives an alcohol. The keto group of the cyclohexene converts into O H. When it reacts with M e M g B r, E t 2 O, it gives 2 isomeric compounds. The keto group of the chain converts into M e and O H.

Enantioselective synthesis of diol (+)-19

Full size image
Fig. 16.2
A line graph of del epsilon versus wavelength. It plots the variations of the expected spectrum for synthetic 18 O B z and the calculated spectrum for R 18 O B z. The spectrum for synthetic begins from (0, 0), and R 18 O B z begins from (0, negative 38). The lines overlap and fluctuate with each other.

Comparison of calculated CD spectrum of (R)-18-OBz (blue) to experimental spectrum of synthetic 18-OBz (red)

Full size image
Scheme 16.8
A reaction pathway. An alcohol reacts with M e M g B r. The product has an O M g O bridge bond. The M e ion attacks the keto group and it converts into M e and O H groups. It gives a diol. The O H and M e are added to the chain moiety.

Proposed mechanism of the diastereoselective methylation of (+)-18

Full size image

To determine the absolute configuration, (+)-18 was converted to benzoyl ester 18-OBz. The calculated spectrum of (R)-18-OBz was obtained using TD-DFT calculations at the TDDFT-CAM-B3LYP/6-311G + (d,p) level with the solvent model density for MeCN, implemented in the Gaussian 16 program package [22]. The calculated spectrum of (R)-18-OBz (blue) agreed well with the experimental spectrum (red).

Our proposed mechanism of the diastereoselective methylation is shown in Scheme 16.8. One equivalent of methyl magnesium bromide deprotonated the hydroxy group of (+)-18, forming a salt bridge of Mg2+ through chelation. The methyl anion then attacked from the less-hindered side of the carbonyl group to afford (+)-19a with the desired stereochemistry.

16.9 Synthesis of Quinone Hemiacetal (−)-21

The optically active (+)-19a was treated with lithium-naphthalene system to remove benzyl protection, affording diol (+)-20a in 94% yield (Scheme 16.9). Next, Swern oxidation of (+)-20a furnished enone (−)-5 in 89% yield. Phenolic oxidative cyclization of (−)-5 using PIDA produced tricyclic quinone hemiacetals (−)-21 in 62% yield and (−)-22 in 14% yield.

Scheme 16.9
A reaction pathway. An optically active compound reacts with L i, naphthalene. O B n group is converted into O H. When it is treated with D M S O, the O H of cyclohexene converts into keto group. On treatment with P I D A, 2 products are obtained. These are isomers. The tetrahydropyran is formed.

Synthesis of quinone hemiacetal (−)-21

Full size image

16.10 Completion of Enantioselective Total Synthesis of (−)-Cochlearol B (1)

Optically active tricyclic (−)-21 was subjected to intramolecular [2+2] photocycloaddition to afford pentacyclic (−)-23 bearing a four-membered ring in its framework in 74% yield (Scheme 16.10). Then, Luche reduction of quinone hemiacetal (−)-23 produced hydroquinone (+)-3 in 84% yield. Pivaloyl protection of (+)-3, followed by dimethylaminomethylenation with Bredereck’s regent afforded (−)-27 in 88% yield over two steps. Treatment of (−)-27 with trifluoromethanesulfonic anhydride and subsequently with triethylsilane in the presence of a Pd (0) catalyst produced α,β-unsaturated aldehyde (−)-29 in 84% yield from (−)-27. Finally, treatment of (−)-29 with potassium carbonate in methanol furnished (−)-cochlearol B (1) in 94% yield.

Scheme 16.10
A reaction pathway. An optically active tricyclic compound converts into a pentacyclic. It reduces in the presence of N a B H 4. When treated with P i v C l, D M A P, the O H converts into O P i v. It converts into alpha, beta-unsaturated aldehyde. On oxidation, it gives cochlearol B.

Enantioselective total synthesis of (−)-cochlearol B (1)

Full size image

16.11 Retrosynthetic Analysis of (+)-Ganocin A (2)

Having accomplished the total synthesis of (−)-cochlearol B (1), we aimed for the enantioselective total synthesis of ganocin A (2). Retrosynthetic analysis of the optically active form of 2, which is based on the synthetic strategy of (−)-1, is shown in Scheme 16.11. To introduce the α,β-unsaturated aldehyde unit, the same sequence as that in the synthesis of 1 was adopted. Compounds 1 and 2 are structurally different in that 1 bears a cyclobutane ring, while 2 bears a tetrahydrofuran ring. To construct the pentacyclic framework of 30, we envisioned the acid-mediated cascade cyclization of 21, which efficiently afforded the cyclopentane ring and tetrahydrofuran ring in a single step (route 1). Additionally, we expected that after the bromohydration of 21, sequential radical cyclization, -reduction, and -benzylic oxidation would furnish pentacyclic skeleton (route 2). Compound 21 is an optically active common synthetic intermediate in the synthesis of (−)-1.

Scheme 16.11
A reaction pathway. Ganocin A undergoes incorporation of alpha, beta-unsaturated aldehyde moiety. It gives a bridged compound. It undergoes benzylic oxidative cyclization, radical cyclization, reduction, and bromohydration. Acid-mediated cascade cyclization also gives the product. It has tetrahydropyran.

Retrosynthetic analysis of (+)-ganocin A (2)

Full size image

16.12 Attempted Acid-Mediated Cascade Cyclization

We commenced the synthesis of (+)-2 from (−)-21 which is also an optically active intermediate in the synthesis of (−)-1. First, we proceeded to construct the pentacyclic skeleton. The acid-mediated cascade cyclization depicted in Scheme 16.11 was investigated as route 1 (Table 16.2). If this cascade cyclization could be realized, cyclopentane ring, tetrahydrofuran ring, and four stereogenic centers would be formed simultaneously. Treatment of (−)-21 with Fe2(SO4)3 as the Lewis acid produced unknown compounds (entry 1) [23]. Thus, lanthanoid triflates that could be used in water were examined [24]. Contrary to expectations, spiro-compound 33, which was generated by the Michael addition of the hydroxy group formed by the ring opening of the cyclic hemiacetal of (−)-21, was afforded (entries 2 and 3). Moreover, when (−)-21 was treated with 6 M HCl aq., 33 was produced in 90% yield (entry 4). Based on the above results, we concluded that route 1 was difficult to complete.

Table 16.2 Attempted acid-mediated cascade cyclization
Full size table

16.13 Intramolecular Radical Cyclization and Benzylic Oxidative Cyclization

The key steps in route 2 are the intramolecular radical cyclization and benzylic oxidative cyclization. To obtain the precursor for the key steps, tricyclic (−)-21 was subjected to bromohydration using N-bromosuccinimide to produce bromohydrin (−)-32 in 93% yield [25] (Scheme 16.12). To our delight, the next intramolecular radical cyclization using tributyltinhydride and AIBN proceeded smoothly to produce pentacyclic compound (−)-34 in 63% yield [26], in a one pot manner via the formation of a cyclopentane ring, reduction of quinone hemiacetal, and subsequent cyclic hemiacetal formation. Compound (−)-35 was obtained in 95% yield upon pivaloyl protection of (−)-34, and intramolecular benzylic oxidative cyclization of (−)-35 upon treatment with ceric ammonium nitrate (CAN) successfully furnished pentacyclic compound (+)-36 in 92% yield [27].

Scheme 16.12
A reaction pathway. A compound with tetrahydropyran moiety reacts with N B S. B r is added to the chain. On treatment with n B u 3 S n H, it gives a pentacyclic. Quinone hemiacetal reduces. A cyclic hemiacetal is formed. O H is replaced by O P i v. It reacts with CAN and gives a pentacyclic compound.

Intramolecular radical cyclization and benzylic oxidative cyclization

Full size image

16.14 Completion of the Enantioselective Total Synthesis of (+)-Ganocin A (2)

With pentacyclic (+)-36 in hand, we finally aimed to incorporate the α,β-unsaturated aldehyde moiety (Scheme 16.13). Following the synthesis of 1, (+)-36 was treated with Bredereck’s reagent, affording enaminone (+)-37 in 93% yield. (+)-37 was converted to triflate (+)-38 in 95% yield, with the formation of the α,β-unsaturated aldehyde moiety using triflic anhydride. Subsequent reduction of triflate (+)-38 using triethylsilane in the presence of a Pd (0) catalyst furnished (−)-39 in 98% yield. Finally, methanolysis of (−)-39 furnished (+)-ganocin A (2) in 99% yield. Here, the enantioselective total synthesis of (+)-2 was accomplished via the longest linear sequence of 17 steps in 9% overall yield.

Scheme 16.13
A reaction pathway. A pentacyclic with C O H bond is converted to a compound that incorporates the alpha, beta-unsaturated aldehyde moiety. The triflate reacts with E t 3 S i H, E t 3 N giving a compound. It is oxidized in the presence of K 2 C O 3 and gives ganocin A.

Completion of the enantioselective total synthesis of (+)-ganocin A (2)

Full size image

16.15 Determination of the Absolute Configuration of (+)-Ganocin A (2)

The absolute configuration of synthetic (+)-ganocin A (2) was determined by comparing the calculated and experimental circular dichroism (CD) spectra (Fig. 16.3). The calculated spectrum of (4aS,5S,7aR,12bR)-2 was obtained using TD-DFT calculations at the TDDFT-B3LYP/6-31G(d,p) level with the solvent model density for MeCN, implemented in the Gaussian 16 program package [22]. The calculated spectrum of (4aS,5S,7aR,12bR)-2 (blue) agreed well with the experimental spectrum (red).

Fig. 16.3
A line graph of del epsilon versus wavelength plots the variations of expected spectrum for synthetic plus ganocin A and calculated spectrum for 4 a S, 5 S, 7 a R, 12 b R ganocin A. The spectrum for synthetic begins from (0, 15) and 4 a S, 5 S, 7 a R, 12 b R begins from (0, 25). Both fluctuate.

Comparison of calculated CD spectrum of (4aS,5S,7aR,12bR)-2 (blue) to experimental spectrum of synthetic (+)-ganocin A (2) (red)

Full size image

16.16 Conclusion

The first total synthesis of (±)-cochlearol B (1) and the catalytic enantioselective total syntheses of (−)-cochlearol B (1) and (+)-ganocin A (2) were achieved. This is the first study to report the synthesis of the (−)-enantiomer of 1 which has been reported to show potent inhibitory activity against p-Smads. This is also the first study to report the enantioselective total synthesis of (+)-2.

These syntheses were accomplished through the NHK reaction, phenolic oxidative cyclization, enantioselective CBS reduction, intramolecular [2+2] photocycloaddition, intramolecular radical cyclization, oxidative benzylic cyclization, and Luche reduction. These total syntheses are expected to lead to new developments in the field of medicinal chemistry.