Synlett 2024; 35(05): 543-551
DOI: 10.1055/a-2021-7993
account
Biomimetic Synthesis

Investigations on Biomimetic Dimerization in Natural Product Synthesis

Fan Zhang
,
Chongchong Chen
,
Xiangdong Hu

We are grateful for financial support from the National Natural Science Foundation of China (22071192, 21772153) and the Shaanxi Provincial Science and Technology Department (2019JM-151).
 


Abstract

Biomimetic dimerization is a fascinating pathway to natural product synthesis. By using structurally inferior monomers, complex molecular architectures can be readily established with distinct efficiency and elegance. In this Account, our investigation on biomimetic dimerization in natural product synthesis has been summarized, which includes our synthetic exploration of linderaspirone A, bi-linderone, parvistemin A, (±)-diperezone, scabellone B, and spiroxins A/C/D.

1 Introduction

2 Biomimetic Dimerization in the Synthesis of Linderaspirone A and Bi-linderone

3 Biomimetic Dimerization in the Synthesis of Parvistemin A and (±)-Diperezone

4 Biomimetic Dimerization in the Synthesis of Scabellone B

5 Dimerization Investigation in the Synthesis of Spiroxins A/C/D

6 Conclusion


#

Biographical Sketches

Zoom Image

Fan Zhang was born in Shaanxi Province, China in 1998. She received her BSc degree from Hainan Normal University in 2016, and then joined Prof. Xiangdong Hu’s research group at Northwest University. Her current research focuses on synthesis-oriented methodology.

Zoom Image

Chongchong Chen was born in Anhui Province, China in 1997. He received BSc degree from Huaibei Normal University (2015–2019), and then joined Prof. Xiangdong Hu’s research group at Northwest University. His current research focuses on natural product synthesis.

Zoom Image

Xiangdong Hu was born in Ningxia Province, China in 1975. He obtained his BSc degree from Lanzhou University (1993–1997), After three years of work at Lanzhou Chemical Industry Company (1997–2000), he pursued his PhD degree at Lanzhou university under the supervision of Prof. Yong-Qiang Tu (2001–2006). Then he worked as a postdoctoral researcher in Dr. Roman Manetsch’s group at University of South Florida (2006–2009). He started his independent career in the department of Chemistry & Material Science at Northwest University, China in 2009. His current research interests include the total synthesis of natural products and the development of synthetic methodology.

1

Introduction

For decades, the biosynthetic pathways of natural products have been demonstrated to be a very attractive field to the organic synthesis community. Looking into this field, one will meet innumerable highly impressive cases of constructing super complex molecular architectures through efficient and elegant biosynthetic pathways.[1] It is stunning what a wonderful synthesis wisdom stores in nature. Among the kaleidoscopic pathways for assembling natural products under physiological environments, dimerization is a particular pattern, which catches more and more attention in recent years. And it should be noted that about 15–20% of the entire natural products might be formed in a dimerization manner.[2]

From the perspective of synthesis, the dimerization of two structurally inferior monomers possesses a distinct superiority in the efficiency and elegance of reaching molecular complexity, which therefore endows biomimetic dimerization with much more charming features and makes it a popular fashion in the field of natural product biomimetic synthesis. To date, the progress and achievements of biomimetic dimerization in natural product synthesis have been well documented in several elegant reviews.[2a] [3] In this Account, we hope to summarize investigations on biomimetic dimerization in natural product synthesis carried out in our laboratory.


# 2

Biomimetic Dimerization in the Synthesis of Linderaspirone A and Bi-linderone

Our investigation of biomimetic dimerization began with the synthesis of linderaspirone A (1) and bi-linderone (2). Compounds 1 and 2 are disclosed by Liu and co-workers from the root of Lindera aggregata (Sims) Kosterm., which belongs to a well-known traditional plant species being used as an analgesic and antispasmodic medicine source in China and Japan.[4] Notably, both 1 and 2 demonstrated significant bioactivity against glucosamine-induced insulin resistance in trials with HepG2 cells, and therefore, promising lead compounds for the development of new medications for type II diabetes. Structurally, 1 and 2 possess two unique spiro-cyclopentenedione frames with highly congested eight- or six-membered-ring skeletons. It was proposed that both 1 and 2 are generated through biogenetic dimerization pathways using methyl linderone (3) as the monomer, which was found in the same plant. As shown in Scheme [1, a] stepwise [4+4] cycloaddition would provide 1 in a very efficient manner (path A). For the formation of bi-linderone, the radical intermediate 4 could be formed first, and then afford intermediate 6 through a sequential ring-closing/ring-opening radical isomerization process. The following radical dimerization could be the biosynthetic route to 2 (path C). In a refined version proposed by Wang and co-workers, the radical isomerization from 4 could be another pathway to 1 (path B).[5]

Zoom Image
Scheme 1 The biosynthetic proposal for linderaspirone A and bi-linderone

Due to unique frames and promising bioactivity features, the synthesis of 1 and 2 obtained the attention of organic synthetic chemists. Liu and co-workers, the discoverers, reported a one-step biomimetic approach to 1 and 2 through a 2,2′-azobisisobutyronitrile-promoted dimerization of 3.[6] With the employment of a photochemical [2+2] cycloaddition/Cope rearrangement and photochemical [2+2] cycloaddition/radical rearrangement, Wang and co-workers accomplished a different version of biomimetic total synthesis of 1 and 2.[5] Our study commenced with the synthesis of 3 based on a ring expansion of cyclobutenedione developed in our laboratory.[7]

Zoom Image
Scheme 2 Biomimetic dimerization in total synthesis of linderaspirone A and bi-linderone

Commercially available dimethyl squarate (7) was treated with bromoketone 8 under the excess equivalent of lithium bis(trimethylsilyl)amide (Scheme [2]). The resulting Darzens reaction will generate the epoxide intermediate 9, which will afford linderone (10) through a ring expansion and subsequent enolization process. The following methylation finished the synthesis of 3 in only two steps with an overall yield of 57%. During the investigation of biomimetic dimerization of 3, it was observed that both 1 and 11 were generated under the irradiation of a metal halide light at neat conditions. Interestingly, we found that the oxygen atmosphere exhibited a distinct effect on the formation of 1 with a higher yield than the control experiment under argon. For the formation of 1, the first step would be the generation of 11 through a [2+2] photocycloaddition. Then, the subsequent Cope rearrangement would enable the formation of 1.[5] As an interesting plus, it was disclosed that 2 can be readily obtained from heating 1 in p-xylene. We proposed that a thermal Cope rearrangement and radical ring-opening/rearrangement sequence of 1 will lead to the formation of 2, which stands as a supplementary biosynthetic pathway to 2.[7a]

Besides, we were curious about another plausible biosynthetic pathway to 2 through a heterodimerized Diels–Alder reaction, which has been demonstrated as a common transformation in the biosynthesis of natural products.[1k] [8] As shown in Scheme [3], we wondered about the possibility that the Diels–Alder reaction of 13 and 14 might be another biosynthetic route to 2. Therefore, the synthesis of 14 was executed in our laboratory.[9] Compound 17 was readily prepared from commercially available 15 and 16 under basic conditions. With the application of the Darzens/ring expansion strategy developed in our laboratory, 17 was treated with 7 in the presence of LiHMDS. The subsequential methylation afforded 18 with the structure unambiguously assigned by X-ray crystallography.

The second methylation was enabled by the employment of diazomethane, which led to the generation of desired 14 in 70% yield over three steps. Out of our expectation, compound 14 presents very poor stability. It can transform into highly congested spiro compound 20 through a 6π-electrocyclization when it was kept in a neat state or in deuterated chloroform. Meanwhile, silica gel can significantly promote this transformation and led to the formation of 20 in quantitative yield. Besides, under thermal conditions, 18 can also take place the 6π-electrocyclization and gave similar product 19. The molecular structures of both 19 and 20 were demonstrated by X-ray crystallography. Interestingly, the processes of 18 and 14 exhibited a rare case of remarkable difference of 6π-electrocyclization in not only the reactivity but also the products. As the result, it is in a very low possibility that the Diels–Alder reaction of 13 and 14 would be a feasible biosynthetic pathway to 2.


# 3

Biomimetic Dimerization in the Synthesis of Parvistemin A and (±)-Diperezone

Our next investigation on biomimetic dimerization was explored in the synthesis of parvistemin A (21) and (±)-diperezone (22).[10] Stemona parviflora C. H. Wright belongs to the well-known Stemona genus, which has been widely applied in Chinese traditional medicines for the treatment of respiratory diseases, including bronchitis and tuberculosis. During the study on Stemona parviflora Wright, Bringmann and co-workers reported the discovery of parvistemin A (21) in 2007.[11] Diperezone (22) was discovered first by Garcia and Mendoza from the roots of Perezia afumani var. oolepis in 1982,[12] and then by Roussis and co-workers from the Caribbean soft coral, Pseudopterogorgia rigida in 2013.[13] Structurally, both 21 and 22 possess the same core biquinone skeleton. The plausible biosynthetic approach to 21 was proposed by Bringmann, the discoverer. As shown in Scheme [4a], the structurally similar compound stilbostemin B (23), which was isolated from the same natural source, will be the reasonable starting material. The oxygenation process will lead to the formation of triphenol 24. Through the radical intermediate 25, the corresponding phenol-oxidative coupling reaction will secure the generation of 21 in the end. As a note, Moody and co-workers proved the feasibility of this oxidative phenolic coupling process using FeCl3·SiO2 conditions in their total synthesis of 21.[14] Interestingly, the biosynthetic proposal to 22 had been demonstrated in different pathways. Joseph-Nathan and co-workers disclosed that 22 can be generated from the treatment of the homologous natural product perezone (26) with boron trifluoride.[15] Compound 27 might be formed during the process, which will finally establish 22 after aerobic oxidation (Scheme [4b]).

Zoom Image
Scheme 3 Attempt on the synthesis of bi-linderone by heterodimerized Diels–Alder reaction
Zoom Image
Scheme 4 Plausible biosynthetic pathway to parvistemin A and diperezone

Our biomimetic synthesis of 21 and 22 commenced with the same starting material, squaric diisopropyl ester (28). As shown in Scheme [5], the addition of methyl lithium and following acidic treatment generated compound 29, which was submitted to the treatment of 30. The resulting adduct went through a four-electron electrocyclic ring-opening/6π-electrocyclization cascade under thermal conditions in toluene, which is the ring expansion process of cyclobutenones developed by Moore[16] and Liebeskind.[17] After the removal of the isopropyl group, precursor 31 was obtained. The critical biomimetic phenol-oxidative coupling reaction was explored under various oxidative conditions, including DDQ, PhI(OAc)2, CAN, and FeCl3/KOH. The best result for the biomimetic synthesis of 21 was recorded with the employment of K3Fe(CN)6/KOH. For the synthesis of 22, the same protocol was applied. With the employment of 32 and 34, the diphenol compound 35 was readily prepared. Due to the inferior stability, the product from the removal of the isopropyl group of 35 was submitted for the subsequent phenol-oxidative coupling directly. It is noteworthy that, rather than the condition of K3Fe(CN)6/KOH, the best result was recorded from FeCl3/KOH, which afforded (±)-diperezone (22) and meso-diperezone (36) in a 1:1 mixture.

Zoom Image
Scheme 5 Biomimetic dimerization in total synthesis of parvistemin A and (±)-diperezone

# 4

Biomimetic Dimerization in the Synthesis of Scabellone B

In 2011, Copp and co-workers disclosed four marine meroterpenoids, scabellones A–D (3740), from ascidian Aplidium scabellum in New Zealand.[18] During the biological valuation, it had been demonstrated that scabellone B (38) possesses selective antimalarial activity toward Plasmodium falciparum (at an IC50 of 4.8 μM) without apoptosis effect on human neutrophils. Therefore, 38 presents a promising potential for the development of new medicines for malaria. In terms of molecular structure, 3740 exhibited an unique benzo[c]chromene-7,10-dione frame. Later on, Copp and co-workers reported the synthesis of 3739 through a bioinspired dimerization based on an O2–CuCl–pyridine-promoted phenol-oxidative coupling process.[19] Regarding the biosynthetic route to these meroterpenoids, we proposed a refined version in Scheme [6]. The dimerization of 41 will be enabled by a phenol-oxidative coupling process, which could lead to the formation of biquinone 42. The subsequent approach to scabellones A–D could be varied in three different pathways. In path A, a particular enolization will lead to the formation of intermediate 43. Then scabellone B will be generated through an oxo-6π-electrocyclization. Scabellone A could be produced through the oxo-6π-electrocyclization of intermediate 44, which could be formed by a different enolization as shown in path B. Regarding the generation of scabellones C and D, the enolization process of scabellone A will provide intermediate 45.

Zoom Image
Scheme 6 Biosynthetic proposal to scabellones A–D

The oxo-Michael reaction of 45 and subsequent aerobic oxidation could stand as a reasonable pathway to 39 and 40. Commercially available 46 was the starting material of our synthesis (Scheme [7]).[20] The sequenced introduction of TMS and geranyl groups gave compound 47 in 78% overall yield. With the removal of the TMS group, the key precursor 41 was obtained readily. The biomimetic dimerization of 41 was facilitated by a phenol-oxidative coupling under the conditions of ammonium ceric nitrate (CAN). For the synthesis of scabellone B (38), the proposed oxo-6π-electrocyclization (path A in scheme [6]) was checked. To our great pleasure, the employment of Et3N afforded 38 as the only product and in 95% yield. The formation of scabellone A (37), scabellone C (39), and scabellone D (40) were observed under other conditions. It is noteworthy that the proposal of generation of 39 and 40 from 37 had been proven by the treatment of 37 with CuCl in deuterated chloroform, which afforded 39 in 56% and 40 in 15% yield. As a result, the total synthesis of scabellone B (38) had been achieved in 5 steps with a total yield of 31%.

Zoom Image
Scheme 7 Biomimetic dimerization in the synthesis of scabellone B

# 5

Dimerization Investigation in the Synthesis of Spiroxins A/C/D

In 1999, McDonald and co-workers from Wyeth-Ayerst Research disclosed marine fungi metabolites spiroxins A–E (4852) from fungal strain LL-37H248, which exists in a soft orange coral found in Dixon Bay, Vancouver Island, Canada.[21] The bioactivity tests discovered that 48 displayed antitumor activity toward human ovarian carcinoma and antibacterial activity against Gram-positive bacteria. It is noteworthy that the biological property could derive from the cleavage of single-stranded DNA. Compounds 4852 are structurally combined by two naphthoquinone epoxide units. There are six or seven stereocenters in the distorted cage-like skeleton forged by a carbon–carbon bond and a spiroketal frame. The unique molecular structures and the intriguing biological features attracted much attention from organic synthetic chemists. The first report on the total synthesis of racemic 50 was from Imanishi and co-workers using a TBAF-activated Suzuki–Miyaura cross-coupling reaction.[22] Suzuki and co-workers achieved the asymmetric total synthesis of 50 and both enantiomers of 48 based on the development of a stereospecific intramolecular photoredox reaction of naphthoquinone derivatives.[23]

The biosynthetic route to spiroxins A–E had not been proposed. We anticipate that, under specific acidic conditions, the dimerization of 53 and ent-53 could take place, which might rescue the formation of spiroxin C (50) through the release of a molecule of water (Scheme [8]). The selective reduction of the right-side carbonyl of 50 could afford spiroxin D (51). And the chlorination and oxidation processes of 50 might be a reasonable pathway to spiroxin A (48). Then, chlorination of 48 will generate spiroxin B (49). Spiroxin E (52) would be the product of the selective reduction of right-side carbonyl in 49. During our investigation on the synthesis of spiroxins (Scheme [9]), the dimerization of 53, which was obtained from the epoxidation of commercially available juglone (54), was checked under acidic conditions. However, there was no observation of the formation of 50 with the employment of various acids including AlCl3, BF3·Et2O, ZnCl2, Sc(OTf)3, Cu(OTf)2, Zn(OTf)2, Ni(OTf)2, and B(C6F5)3.

Zoom Image
Scheme 8 Our biosynthetic proposal to spiroxins A–E

All tests led to either the decomposition of 53 or no clear transformation observed. Our successful approach to 50 started from the enantioselective epoxidation of 56, which was readily prepared from the treatment of 54 with 55 in the presence of Ag2O.[24] With the application of asymmetric phase-transfer catalyst 57, which was initially developed by Corey[25] and Lygo,[26] respectively, 58 can be obtained in 80% ee and 96% yield. And the reaction can be executed on 10 gram scale, which facilitated significantly further transformations. The recrystallization can increase the ee value of 58 over 99%. The chemoselective and diastereoselective reduction of bottom-side carbonyl of 58 and following protection afforded 59. Starting from 54 again, a three-steps routine operation delivered the bromide 60. The treatment of 59 and the lithium reagent from 60 was followed by the debenzylation operation, which led to the formation of 61. The vital spiroketal unit was constructed through a phosphoric acid promoted oxidation process.[27] The resulting product 62 was subjected to the removal of TBS group, oxidation of the hydroxy, and epoxidation. Then, the asymmetric total synthesis of spiroxin C (50) was accomplished. Based on our biosynthesis proposal (Scheme [8]), the synthesis of spiroxin D (51) was checked with the reduction of 50. To our pleasure, the employment of DIBAL-H enabled the chemoselective and diastereoselective reduction of right-side carbonyl of 50 and accomplished the first total synthesis of 51. The previously unknown stereoconfiguration of the hydroxy was characterized through X-ray crystallographic analysis. For the synthesis of spiroxin A (48), the ortho-selective chlorination of the phenol unit in 50 was achieved by using 1,3-dichloro-5,5-dimethylhydantoin (DCDMH, 63). Finally, we completed the asymmetric total synthesis of 48 through the application of the oxime ester directed acetoxylation developed by Sanford and co-workers.[28]

Zoom Image
Scheme 9 Investigation on total synthesis of spiroxins A/C/D

# 6

Conclusion and Outlook

In conclusion, as a charming synthetic strategy, biomimetic dimerization attracted the attention of many organic synthetic chemists. In this Account, we summarized our investigation on biomimetic dimerization explored in the total syntheses of linderaspirone A, bi-linderone, parvistemin A, (±)-diperezone, scabellone B, and spiroxins A/C/D. There are two major benefits we collected in this journey. One is, through the imitation of the efficient and elegant establishment of molecular architectures under physiological conditions, to feel the amazing power and wisdom stored in nature. Another is the new insights into the biosynthetic pathway of natural products, which lead us to better understanding of the formation of natural products and related transformations.


#
#

Conflict of Interest

The authors declare no conflict of interest.

Acknowledgment

We appreciate all former and current group members for their contributions to the topic summarized in this Account.


Corresponding Author

Xiangdong Hu
Department of Chemistry & Material Science, Key Laboratory of Synthetic and Natural Functional Molecule Chemistry of Ministry of Education of China, Northwest University
Xi’an, Shaanxi Province, 710127
P. R. of China   

Publication History

Received: 07 December 2022

Accepted after revision: 28 January 2023

Accepted Manuscript online:
28 January 2023

Article published online:
28 February 2023

© 2023. Thieme. All rights reserved

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany


Zoom Image
Zoom Image
Zoom Image
Zoom Image
Scheme 1 The biosynthetic proposal for linderaspirone A and bi-linderone
Zoom Image
Scheme 2 Biomimetic dimerization in total synthesis of linderaspirone A and bi-linderone
Zoom Image
Scheme 3 Attempt on the synthesis of bi-linderone by heterodimerized Diels–Alder reaction
Zoom Image
Scheme 4 Plausible biosynthetic pathway to parvistemin A and diperezone
Zoom Image
Scheme 5 Biomimetic dimerization in total synthesis of parvistemin A and (±)-diperezone
Zoom Image
Scheme 6 Biosynthetic proposal to scabellones A–D
Zoom Image
Scheme 7 Biomimetic dimerization in the synthesis of scabellone B
Zoom Image
Scheme 8 Our biosynthetic proposal to spiroxins A–E
Zoom Image
Scheme 9 Investigation on total synthesis of spiroxins A/C/D