Review
Studies on Comprehensive Total Synthesis of Natural and Pseudo-Natural Products for Drug Discovery
2024 Volume 72 Issue 5 Pages 422-431
Details
2024 Volume 72 Issue 5 Pages 422-431
Natural products are important for the development of pharmaceuticals and agrochemicals; thus, their synthesis and medicinal chemistry research is critical. Developing a total synthesis pathway for natural products confirms their structure and provides the opportunity to modify the structure in a targeted manner. A simple modification of a single oxidation step can increase the biological activity, or the complexity of the molecule can alter the property. Herein, we discuss the asymmetric total synthesis of dihydroisocoumarin-type natural products, the creation of novel antibacterial compounds through partial structural modification, and the development of antioxidants with high activity and low toxicity through dimerization strategies.
Natural products have significantly impacted the world, particularly in the development of pharmaceuticals and agrochemicals, and they are considerably important. Over 40% of small molecule pharmaceuticals are developed from natural product related compounds.1) The author has taken an interest in natural products that exhibit strong and intriguing biological activities and their total synthesis. Additionally, pseudo-natural products are synthesized by the structural modifications of natural products. This approach is aimed at obtaining compounds with strong biological activities or different biological activities from the parent natural products. For example, by constructing a chemical library of natural and pseudo-natural compounds based on resorcylic acid lactones, we have successfully obtained new pseudo-natural compounds with promising biological activities.2) Additionally, the author has synthesized skeletal complex alkaloids, which have not been previously synthesized, using a tandem cyclization reaction, thereby laying the foundation for evaluating biological activities.3) Furthermore, the author is also developing chemical biology tools by understanding the fluorescent properties of novel fluorescent molecules with unique structures and expanding into the development of fluorescent labeling reagents for small molecules, including natural products.4–6) Herein, we focus on the research of dihydroisocoumarin-type natural products, eurotiumides.7–10) We describe the establishment of asymmetric total syntheses and elucidate the absolute structures of eurotiumides A, B, F, and G, and we discuss the medicinal chemistry research of pseudo-natural products derived from eurotiumide A by structural modification.
Eurotiumides are dihydroisocoumarin-type natural products that were isolated from a gorgonian-derived fungus, Eurotium sp. XS-200900E6, in 2014.11) Seven species of eurotiumides (A–G) have been reported (Fig. 1).
The structural common motif of eurotiumides A–E is the 4-methoxy-isochroman-1-one skeleton, which is a rare structure in nature. Eurotiumide A (1), which was originally assigned the trans configuration of H3/H4, was reported to exhibit potent antimicrobial activities against Vibrio anguillarum, Vibrio parahaemolyticus, Staphylococcus epidermidis, Staphylococcus aureus, Bacillus cereus, and Escherichia coli. Eurotiumide B (2), which was initially assigned the cis configuration of H3/H4, was reported to be a potent antifouling agent against the larval settlement of the barnacle Balanus amphitrite. Furthermore, eurotiumides C (3) and D (4) were reported to have antimicrobial and antifouling activities. These natural products were isolated as racemates, the enantiomeric separations for 1, 2, 3, 4, and 5 were conducted by chiral HPLC, and their absolute structures were speculated by the electronic circular dichroism (ECD) spectra. Conversely, eurotiumides F (6) and G (7) have a different skeleton from eurotiumides A–E. They have a dihydropyran ring structure, formed from the prenyl side-chain and an adjacent phenolic hydroxy group, and a methylacetal moiety at the C1 position instead of a lactone carbonyl group. These compounds were isolated as racemic mixtures. However, unlike eurotiumides A–E, their absolute stereochemistry could not be determined because of the difficulty in resolving them by chiral HPLC. In recent years, the emergence of resistant bacteria, exemplified by methicillin-resistant Staphylococcus aureus (MRSA), has intensified the demand for new antimicrobial agents. Furthermore, the prevention of marine fouling organisms, such as barnacles attaching ship hulls, is critical to improving efficiency and preventing the spread of invasive species through maritime transport. The development of environmentally friendly antifouling agents derived from natural sources is highly demanded. Eurotiumides, particularly eurotiumide A (1) and B (2), have piqued our interest because of their potential as promising antimicrobial and antifouling agents that can address social demands. Additionally, the elucidation of the absolute structures of eurotiumides F (6) and G (7) was considered academically significant. As part of our expansion into drug discovery, we have been engaged in the scientific study of eurotiumides.
In the synthesis of eurotiumides, we set a common synthetic intermediate having a cis-4-methoxy-isochroman-1-one skeleton, compound 13, as our first synthetic target (Chart 1). As 1 and 2, as well as 6 and 7, are diastereomers at the C3 and C4 positions, the desired compounds can be accessed by the hydrolysis of the lactone moiety of 8 and 11. Thereafter, the intramolecular Mitsunobu reaction is conducted to afford the precursors of (+)-1 and 6. We considered that 8 could be obtained by the bromination of 13 and a Pd-catalyzed coupling reaction to introduce the prenyl side-chain. At this stage, the variation of the coupling partner could allow for the diversification of this site, laying the base for the construction of the future compound library. Dihydropyran compounds, 11, could be obtained from dimethylpropargyl derivatives, 10, via thermal cyclization reaction. The methyl acetal moiety would be formed by the reduction of the lactone, followed by acidic treatment in methanol with 11 and 12, respectively. It is anticipated that the stereochemistry at the C1 position would converge to that of the natural products. We envisioned that the common synthetic intermediate, 13, could be prepared from aryl bromide, 14, via Pd-catalyzed C1 insertion/lactonization cascade reaction. Furthermore, 14 would be synthesized from trans alkene, 15, via Shi asymmetric epoxidation, followed by the epoxide ring-opening reaction with MeOH.
First, we synthesized a common intermediate 13, having a cis-4-methoxyisochroman-1-one skeleton (Chart 2). Starting from commercially available 2,5-dihydroxybenzaldehyde and 1-hexanol, we prepared the aldehyde 1612) and PT-sulfone 17 in two steps; thereafter, they were subjected to the condition of the Julia–Kocienski olefination to afford alkene 18. The desired epoxide 19 was obtained in good yield by applying the Shi asymmetric epoxidation,13) and the enantiomeric excess was determined as 86%. Next, the treatment of 19 using 10 equiv. of MeOH with BF3–OEt2 as the Lewis acid at −40 °C afforded the ring-opening product 21 as a single stereoisomer. The enantiomeric excess of 21 was increased from 86 to 94% by a simple filtration and recrystallization. The structure of 21 was confirmed by X-ray crystallographic analysis to have the cis configuration of H3/H4. In a typical epoxide ring-opening reaction at the benzylic position, stereo inversion via SN2 mechanism or stereo mixture via SN1-like process would have occurred. However, this time, MeOH was introduced at the benzylic position with stereo retention. The reason for this high stereoselectivity is as follows. When the epoxide of 19 is activated by BF3–OEt2 at low temperature, ortho-quinone methide intermediate is formed with the oxygen atom at the ortho-position, and owing to the allylic 1,3-strain effect, the stable conformation of this opened intermediate should exist as 20. Thereafter, MeOH attacks the benzylic position from the opposite side of the alkyl side-chain. Having prepared the precursor of the 4-methoxyisochroman-1-one compounds 21, we attempted the Pd-catalyzed C1 insertion/lactonization cascade reaction. The conventional C1 insertion reaction condition under CO atmosphere did not give good result. After several trials, we finally found that the Pd-catalyzed fluorocarbonylation reaction condition using N-formylsaccharin (22) as the CO source, as reported by Manabe and Konishi et al.,14–16) was effective for this cascade reaction. After optimizing the reaction condition, the desired dihydroisocoumarin derivative 23 was obtained in excellent yield, along with a small amount of the mono methoxymethyl (MOM) derivative 24 corresponding to the common intermediate. Thus, following the C1 insertion/lactonization cascade reaction, the MOM group adjacent to the lactone carbonyl was selectively removed by the one pot addition of MgBr2.17) This procedure successfully afforded the desired common intermediate 24 in 89% yield. Bromine atom was successfully introduced to 24 with N-bromosuccinimide, and the subsequent MOM protection of a phenolic hydroxy group in one pot afforded another intermediate 25 in good yield. Fortunately, we obtained a good quality single crystal of 25, and the recrystallization improved the enantiomeric excess from 86 to 99%. Furthermore, we confirmed the absolute structure of 25 by the X-ray crystallographic analysis (CCDC 1821161). Thus, we established a reliable synthetic route for preparing 24 and 25. Finally, after the introduction of a prenyl side-chain to 25 by the Stille coupling reaction, all MOM groups were removed under acidic condition (6 M HCl aq in MeOH), affording the proposed structure of (+)-eurotiumide B (2). However, the proton NMR of the synthetic product did not match that of (+)-eurotiumide B as originally proposed by Wang and colleagues.11) The various spectral data of our synthetic product were completely consistent with those reported for isolated (−)-eurotiumide A (1). Furthermore, for compound 26, which has the trans configuration of H3/H4, via hydrolysis, followed by the Mitsunobu reaction, all spectral data, including proton NMR, were identical with those of (+)-eurotiumide B (2) reported in the isolation paper.11) Based on these results, we conclude that the actual structure of eurotiumide A, which was originally reported to have the trans configuration of H3/H4, is a cis isomer, and similarly, eurotiumide B has been clarified to have the trans configuration of H3/H4 through this total synthesis. The comparison of the X-ray crystallographic results and optical rotation data between the synthetic and the reported one allowed us to elucidate the relative and absolute stereo configurations of eurotiumides A and B.
During the total syntheses of eurotiumides A and B, which possess a 5,8-dihydroxy-dihydroisocoumarin skeleton, we found that they exhibited light blue fluorescence in various organic solvents upon UV irradiation (Fig. 2a). Owing to this fluorescent property of these molecules, they are applicable as natural fluorescent probes for the study of chemical biology, particularly to observe the localization of eurotiumides in cells. As a preliminary proof of concept, we confirmed that eurotiumide A can be used as a natural fluorescent probe using B. cereus, which is one of the pathogens of food poisoning. Consequently, we observed the living B. cereus with fluorescence (Fig. 2b).
Observing them under bright field and through a 4′,6-diamidino-2-phenylindole filter upon excitation at 345 nm.
Next, we focused on the asymmetric total syntheses of eurotiumides F (6) and G (7). First, we constructed a dihydropyran ring, which is a characteristic structure of 11 and 12, from the synthetic intermediate 13 which could be supplied in large scale via 10. According to our initial plan shown in Chart 1, we first introduced the dimethyl propargylic residue on the phenolic hydroxy group of 24 following the method reported by Godfrey et al.18) to afford the cyclization precursor 27 in good yield (Chart 3). After several attempts, we found that the desired thermal cyclization occurred in dichlorobenzene at 160 °C to obtain the dimethylpyran 28 in quantitative yield.19–21) Interestingly, this thermal cyclization proceeded more rapidly under microwave irradiation condition (200 °C, 300 W) and the reaction completed within 10 min. The lactone moiety of 28 was reduced to lactol with diisobutylaluminium hydride (DIBAL), yielding 29 as a single product. Observing the reaction in detail, we found that the ratio of diastereomers at C1 positions was approximately 1 : 1 in the reduction step, and the subsequent treatment of crude product with aqueous Roschelle salt solution converged them to the thermodynamically stable product, 29. Lastly, we treated 29 with 6 M aqueous HCl in MeOH to remove the two MOM groups and convert the C1 hydroxy group to a methoxy group, affording the product 7′ whose spectral data including the proton and carbon NMR data were identical with those of reported eurotiumide G (7).11) However, the X-ray crystallographic analysis of synthetic eurotiumide G (7′) revealed the relative stereochemistry between H1 and H4 as a trans configuration, not the cis configuration reported in the isolation paper. Thus, we revised the structure of the originally reported eurotiumide G as illustrated in Chart 3 and determined the stereochemistry of (+)-eurotiumide G (7′) as (1R,3S,4S).
For the synthesis of eurotiumide F (6), we transformed the cis configuration of H3/H4 positions to their trans configuration. The hydrolysis of 28, followed by the Mitsunobu inversion, afforded the H3/H4 trans product 30 (Chart 4). DIBAL reduction was employed to convert 30 to 31, and we treated 31 with 6 M aqueous HCl in MeOH for the synthesis of eurotiumide F. The stereochemistry of 31 was confirmed by the nuclear Overhauser effect spectroscopy (NOESY) experiment. However, contrary to our expectation, (−)-eurotiumide G (7′) was obtained instead of eurotiumide F. This could be because, first, the methylacetal formation occurred under acidic condition in MeOH to afford compound 32. Thereafter, the subsequent epimerization at the C4 position of 32 afforded compound 33. Finally, the MOM deprotection afforded (−)-eurotiumide G (7′). The density functional theory calculation supported that compound 33 was slightly more thermodynamically stable than 32.22) Based on the previous experimental results, the epimerization at the C4 position readily occurs when the adjacent aromatic ring is electron-rich under strong acidic condition, which allows for the removal of the MOM group. Therefore, we considered that this epimerization could be prevented by first removing the MOM group of 30, which has a lower electron density than 31. As we expected, the asymmetric total synthesis of eurotiumide F (6) was achieved by the treatment of 30 with 6 M aqueous HCl to remove the MOM group, affording 34, followed by the DIBAL reduction and the subsequent weak acid treatment (0.4 M aqueous HCl in MeOH). The NOESY experiment confirmed the stereochemistry at the C1, C3, and C4 positions, observing the correlation shown in Chart 4, and the results supported the originally proposed structure by Wang and colleagues. The optical rotation of 6 was positive, and the absolute structure of (+)-eurotiumide F (6) was determined to be (1S, 3R, 4S).
Eurotiumide A (1′) was reported as a potent antimicrobial agent against Staphylococcus epidermidis, Bacillus cereus, Vibrio anguillarum, and Escherichia coli. However, a comprehensive structure–activity relationship (SAR) study has not been conducted. The multidrug-resistant bacterial pathogens are still serious threats, and the development of novel and effective antimicrobial drugs against numerous pathogenic bacteria, including MRSA, is important for public health.23–25) Therefore, we chose eurotiumide A (1′) as a lead compound to develop a new antimicrobial agent and conducted the SAR study of 1′. In performing the chemical transformation of 1′ for the SAR study, multiple potential sites for modification were determined. Considering synthetic accessibility, we chose the side-chain of the aromatic ring as the modification site and constructed a chemical library of the side-chain derivatives of eurotiumide A.
Our compounds and synthetic design are displayed in Chart 5. The side-chain of the moiety was introduced in the late stage of synthesis from key intermediates 24 or 25 having a dihydroisocoumarin skeleton.7) We classified the residues to be introduced as side-chains into three categories: Type A encompasses hydrocarbon chains with hydrogen, alkyl, and aryl moieties (36–42); Type B comprises heteroatoms or alkyl groups containing heteroatoms (43–48); and Type C consists of halogen atoms (49–52). The cross coupling reactions and the subsequent chemical transformations would allow the syntheses of Types A and B derivatives from 25. The halogenated derivatives classified as Type C were synthesized through the direct halogenation of intermediate 24. Furthermore, we synthesized a series of derivatives using the racemic forms of 24 and 25 with a primary focus on the rapid construction of the chemical library.
We synthesized the derivatives of Type A (Chart 6). A nonsubstituted derivative 36 was obtained from 24 by removing the two MOM groups under acidic condition. An isopentyl derivative 38 was transformed from eurotiumide A (1′) by the catalytic hydrogenation in quantitative yield. The alkyl group and aromatic rings, such as methyl, vinyl, phenyl, and biphenyl, were introduced with 25 by the Stille coupling and the Suzuki–Miyaura cross coupling, and the subsequent acidic treatment afforded 37, 39, 41, and 42 in good yields, respectively. An ethynyl derivative 40 was synthesized from an aldehyde derivative 44a (vide infra) by the Seyferth–Gilbert homologation with Ohira–Bestman reagent (53) and the subsequent deprotection of diMOM groups to afford the desired ethynyl derivative 40 in 68% yield. Next, we focused on preparing Type B derivatives (Chart 7).
An aldehyde derivative 44 was prepared from 39a in excellent yield by the sequential ozonolysis and deprotection. The reduction of the aldehyde moiety of 44a using sodium borohydride afforded a benzyl alcohol derivative 43a, and the subsequent acidic treatment afforded the desired alcohol 43. To introduce a nitrogen group at the benzylic position of 43a, the primary alcohol moiety was converted into a mesyl group (compound 53), and the subsequent nucleophilic substitution reaction with sodium azide afforded the protected azide derivative, 45a. The acidic treatment of 45a afforded the hydroxymethyl azide derivative 45 in moderate yield. The catalytic hydrogenation of 45a with triethylamine afforded the desired protected aminomethyl derivative 46a, and the removal of the diMOM groups afforded the aminomethyl derivative 46. The nitro derivative 47 was obtained by the sequential nitration reaction with HNO3 in AcOH and the deprotection of the diMOM groups. The reduction of the nitro group by the catalytic hydrogenation using platinum oxide afforded an aniline derivative 48. For the halogenated derivatives of Group C, the bromo derivative 51 was synthesized by the deprotection of the intermediate 25, and the chloro and iodo derivatives 50 and 52 were obtained by treating 24 with N-chlorosuccinimide and N-iodosuccinimide, respectively. Unfortunately, the fluoro derivative 49 could not be obtained in spite of several efforts.
With the initial set derivatives of eurotiumide A, the first antimicrobial activity screening against Gram-positive methicillin-sensitive Staphylococcus aureus (MSSA) and MRSA in 10 µM solutions of test compounds was conducted (Table 1). Eurotiumide A displayed medium efficacy against MSSA and MRSA. While almost all the compounds did not show the good antimicrobial activity against strains, the isopentyl derivative 38 and the iodo derivative 52 demonstrated antimicrobial efficacy against MSSA, which exceeded that of eurotiumide A. In particular, the isopentyl derivative 38 displayed good antimicrobial activity against even MRSA. Upon further detailed evaluation of the antimicrobial activity, we determined the half-maximal inhibitory concentration value of 38 against MSSA and MRSA to be 2.0 and 1.5 µg/mL, respectively. The structural difference between eurotiumide A and 38 was only the oxidation state of the side-chain. The recognition of the subtle difference was highly intriguing, and this result gave us the important insight into the development of potent antimicrobial agents. Further, we checked the antiproliferative activity of eurotiumide A, 38, and 52 against the human lung cancer (A549) cell line, and these compounds did not show the cytotoxicity in 10 µM.
| Strains | Eurotiumide A (1′) | 38 | 52 |
|---|---|---|---|
| Methicillin-susceptible S. aureus | — | 2.0 | 3.7 |
| Methicillin-resistant S. aureus | — | 1.5 | — |
Oxidative stress is constantly induced in biological systems and controls several life events.26,27) Thus, the collapse of the balance between oxidative and reductive environments increases the risk for a variety of diseases, such as carcinogenesis,28,29) autoimmune disease,30,31) and cardiovascular32) and neurodegenerative disorders.33–35) The representative chemical species that cause oxidative stress include reactive oxygen species (ROS), such as oxygen free radicals. An adequate amount of ROS in living cells are beneficial to their functions. However, an excessive amount of ROS oxidizes several biological constituents, such as lipids, sugars, nucleic acids, and proteins, resulting in cellular damage.36–39) For example, when intense oxidative stress occurs in the cerebral ischemia-reperfusion, the high level of ROS inflicts severe impairment in the microenvironment of brain cells.40) To prevent this damage, a rapid and potent radical scavenging agent, such as edaravone (54),41) is used as the therapeutic drug for cerebral ischemia-reperfusion42) (Fig. 3).
Therefore, the development of potent, fast-acting, and low toxicity antioxidants is still needed in clinical practice. A phenol moiety, particularly 1,4-hydroquinone, is a representative functional moiety that exhibits superior antioxidant properties and is often found in natural products. Eurotiumides have this functional moiety, and it is readily deducible that eurotiumides display the potent antioxidant activity. However, the antioxidant activity of eurotiumides has not been evaluated. Consequently, we verified the radical scavenging activity of eurotiumide A (1′) and developed a novel compound that had superior antioxidant properties. In the development of the new antioxidant agent based on eurotiumide A, a dimerized compound 55 was designed with the linkage of the C7–C7′ positions of 5,8-dihydroisochroman-1-one core. In addition to the increase in the number of Ar–O–H groups available for antioxidant activity, it was anticipated that the stabilization of the phenoxy radicals generated by the trapping of ROS through adjacent 1,4-hydroquinone or 1,4-quinone would lead to an enhancement in antioxidant activity.
Dimer 55 was synthesized as shown in Chart 8. Optically active 25 was prepared following our previously reported procedure.7) Boric acid pinacol ester 56 was synthesized from the bromo intermediate 25 (95% enantiomeric excess (ee)) by the Miyaura–Ishiyama boryration with Pd(PPh3)2Cl2, (BPin)2, and KOAc in 1,4-dioxane. The Suzuki–Miyaura cross coupling with 25 and 56 proceeded smoothly to afford the desired protected dimer 57 in 55% yield in 2 steps. Although the last step was only the removal of tetra MOM groups under acidic condition, only a small amount of 55 was obtained, and the major product was a diMOM protected compound 58. The relatively harsh acidic condition resulted in severe epimerization at the C4 and C4′ positions. Therefore, the diMOM protected compound 58 was isolated once and retreated with HCl to afford the desired dimer 55.
Having eurotiumide A (1′) and the dimer 55, their antioxidant activity was evaluated using 1,1-diphenyl-2-pycrylhydrazyl-radical scavenging assay43) at the 15 min time point (Fig. 4a). Edaravone (54) were used as a positive control. Eurotiumide A (1′) displayed the potent antioxidant activity as expected. Conversely, the synthetic dimer 55 exhibited more potent antioxidant activity than eurotiumide A and the control compounds, particularly at low concentrations. However, the observed antioxidant activity was lower than initially expected. We consider that this is attributed to the increased O–H bond dissociation energy, particularly in the fourth Ar–OH group, by the influence of the adjacent 1,4-quinone moiety. Further, we evaluated the cell antiproliferative activity of eurotiumide A and the dimer 55 against a murine colon carcinoma cell line (Colon-26) and a human myeloma cell line (RPMI8226), respectively (Fig. 4b). Eurotiumide A exhibited medium antiproliferative activity against Colon-26 and RPMI8226 cell lines (<25 µM). Interestingly, the dimer 55 did not display any cytotoxicity against two cell lines in the range of test concentrations (>50 µM). The mode of action and the target molecules in the cellular system are still unknown; we successfully developed the promising antioxidant agent that balanced potent antioxidant activity and low toxicity.
Natural products abound with compounds exhibiting intriguing biological activities, and there is a chance to obtain promising pseudo-natural products that could contribute to human health. We aim to create agents suitable for drug development by modifying the structures of naturally occurring substrates that exhibit promising biological activity. In this review, we introduced the asymmetric total syntheses of dihydroisocoumarin-type natural products, eurotiumides, their structural revision, and their preliminary application as natural chemical probes. Further, we developed a promising antimicrobial candidate by the one-point modification of the natural product and introduced the synthetic analog that exhibited potent antioxidant activity and low toxicity by dimerization. We are currently conducting research on the expansion of the chemical library based on the resorcylic acid lactones and the development of the novel analog based on the chippiine-type alkaloids, as well as further developments in the medicinal chemistry and chemical biology research of novel fluorescent molecules for drug development.
The research content of this review was conducted entirely in the laboratory of Prof. Kosuke Namba at Tokushima University. I would like to express my deepest gratitude to Prof. Namba for his invaluable advice and discussions regarding our daily research. I am deeply grateful to all members of Namba group (2014 to 2020) and all my collaborators whose names appear in the reference section for their dedicated efforts and contributions. The research was financially supported by JSPS KAKENHI (15K18830, 17K08365, 20K06939), the Kurita Water and Environment Foundation, the Japan Ecology Foundation, the Uehara Memorial Foundation, and Takeda Science Foundation. Tokushima University for their financial support of the Research Clusters program of Tokushima University (Nos. 1802001 and 1803003) and Research program of development of intelligent Tokushima artificial exosome (iTEX) from Tokushima University.
The author declares no conflict of interest.
This review of the author’s work was written by the author upon receiving the 2022 Pharmaceutical Society of Japan Award for Young Scientists.
