Conformational constraint in natural product synthesis

Abstract

Natural products that exhibit significant biological activity often possess complex molecular structures such as caged frameworks, strained motifs, inherent instability, and many stereogenic carbon centers, etc. Achievement of those total syntheses always requires the powerful methodologies and judicious strategies to fulfill the stereochemical requirements of the target compounds. Building on our successful stereo-controlled syntheses, we have established the concept of conformational constraint, which renders the approach of reactants under a controlled manner during the bond-forming process through the best orbital overlap. Important factors that affect the proper orientation of substrates are (i) acyclic allyl strain, (ii) stereoelectronic effect, (iii) chelation control, etc. Established methodologies include (i) heteroatom directed conjugate addition for diastereoselective C–C bond formation, (ii) 100% α-selective C-glycosidation by using alkynyl-silane, (iii) cobalt acetylene chemistry for medium-size ring formation, followed by its functional group transformation. The author has named such total concept as conformational constraint and has illustrated it with the finished examples of total syntheses. These examples are taken from maytansine, okadaic acid, tautomycin, tetrodotoxin, ciguatoxin, etc.

Conformational Constraint for Natural Product Synthesis

1. Introduction

This review article deals with the important factors that control the stereochemistry emerging during the total synthesis of complex natural products. The author’s research group has successfully achieved stereoselective, target-oriented syntheses of various natural products. Representative examples are presented in Fig. 1, which illustrates the terpenoid vernolepin 1-1,¹⁾ the ansa-macrolide maytansine 1-2,²⁾ the polyoxygenated compounds okadaic acid 1-3³⁾ and tautomycin 1-4,⁴⁾ the cage molecule tetrodotoxin 1-5,^5,6) and the gigantic polyether ciguatoxin 1-6.⁷⁾ All of them contain many stereogenic carbons. Some other syntheses, including those of maringamide X,⁸⁾ the cyclic ionophore cereulide,⁹⁾ the heterocyclic compound coelenterazine,¹⁰⁾ are not discussed in this review due to different bioorganic purposes. For example, dehydro-coelenterazine has no stereogenic carbon, and its bioluminescence mechanism elucidates an important stereochemical issue. Specifically, it covalently binds with cysteine-390 of the photoprotein and an incorrect stereochemistry binding the chromophore prevents light emission and even gives fatal damage to the active center.¹¹⁾ Stereochemistry is an indispensable aspect not only for organic synthesis but also for bioorganic chemistry field. This article focuses on the topic of stereochemical control, established through our synthetic studies. Herein, we use a new term conformational constraint as the concept to render the substrates into proper orientation, and enforce the best molecular orbital overlap during the bond-forming process. For better understanding of conformational constraint, the related partial reaction sequences are extracted instead of the complete total synthesis pathways.

Figure 1.

(Color online) Natural products to be discussed on stereochemical control.

2. Acyclic allyl strain and design of heteroconjugate addition

The elements of conformational constraint are i) acyclic allyl strain (AAS), ii) dipole-dipole interaction, iii) stereo-electronic momentum, etc. Two reacting moieties can approach each other under the influence of chelation control, stereo-electronic effect, reaction kinetics, steric bulk, and other factors. These elements play significant roles not only in the ground state but also in the transition state, where bond-forming/breaking occurs through the best orbital overlap. Figure 2 illustrates a simple oxy-ketone 2-1 (compound 1 in Fig. 2), to which nucleophile (Nu) approaches either from top or bottom face of the olefinic plane (2-1a). The stereogenic carbon exhibits relative conformations around C1/C2 bond, allowing the Nu to add from either direction. In this case, Felkin’s rule has been widely accepted for predicting the major stereoisomeric product. In a polar solvent, substrate 2-1a undergoes Nu attack from the anti-direction due to C→O polarization at the transition state. The chelation effect, on the other hand, is known as shown in 2-1b to alter the stereo-chemical outcome.¹²⁾ We have refined and optimized this system aiming to achieve much higher stereoselectivity.

Figure 2.

(Color online) Addition reaction modes: 1,2- 1,4- and heteroconjugate addition.

Let us now consider a conjugate addition (on paper) to 2-2, in which a tri-substituted olefin is inserted between C1–C4 (relative to 2-1). Its dominant conformation must be 2-2a, where the olefinic plane must include all nine atoms except the C4-OR and C4-Me groups. The orientation of C4 stereogenic carbon should bring the smallest H atom into the co-planarity due to allyl strain (A-strain), potentially leading to better selectivity compared to the case of 2-1. Further improvement (by making stronger A-strain) has been designed as 2-3 by using C2-TMS and C2-sulfone groups to promote conjugate addition with a more pronounced A-strain effect as in 3a. In fact, addition of MeLi•LiBr in THF solvent gave exclusively (100%) syn-product 2-4 under chelation mechanism. We named this method as ‘heteroconjugate addition’ and later as ‘heteroatom directed conjugate addition’ (HADCA).

Heteroconjugate addition from 3-1 was originally designed to afford syn-product during maytansine synthesis.¹³⁾ Later it has been improved to elicit anti-products such as 3-3 and 3-4 as shown in Fig. 3. Such a versatile method enable us to explore switching the syn/anti diastereo-selectivity under various sequence of reactions.¹⁴⁾

Figure 3.

(Color online) Diastereo-switching heteroconjugate addition giving syn or anti products.

During the course of searching factors that influence syn/anti selectivity, we found that stereo-electronic effect enhances the rate that favors the formation of anti-isomer. This effect is derived from the polarized C→O (2-3a) making anti manner of Nu-addition much faster.¹⁵⁾ It was new for us to accelerate the rate for anti-addition which allow us to change the selectivity without diminishing the reaction rate (e.g. by steric hindrance). Figure 3 illustrates the transition states governed by chelation control (3-1a) and stereo-electronic effect (3-1c), the phenomenon we term “diastereo-switching” on the basis of AAS. The contribution through 3-1b is minimal, as it represents only a minor component. Back to Fig. 2, 2-3a, the ‘chelation face’ and ‘stereoelectronic face’ is switchable (by changing polarity of the solvent), allowing the Nu to approach the conjugated sulfonyl group and selectively affords syn or anti product, respectively. As a conclusion of the polarized C→O bond, stereoelectronic effect¹⁶⁾ has been defined in much wider criteria.

3. Maytansine synthesis based on syn-selective heteroconjugate addition

On the basis of HADCA, we performed a retro-synthetic analysis of maytansine (4-1) (highlighted in a square in Fig. 4, 4-2–4-4). The ansa-macrolactam moiety known as maytansinol (3-OH), features a 19-membered ansa-ring with seven stereogenic carbon atoms. Our synthetic strategy is anticipated to induce all the six stereogenic carbons by starting from C-7. In our retrosynthesis, 4-2 emerges by disconnections at the C2–C3 bond (retro-aldol), and C11–C12 bond (retro-Wittig). This compound can be accessible from tetrahydropyrane-ring (4-3), which could be produced using HADCA chemistry from methyl-ethyl glycoside 4-4.

Figure 4.

(Color online) Maytansin synthesis based on heteroconjugate addition.

Our first total synthesis of maytansinol was achieved in racemic form, starting from acrolein dimer 4-5, which was converted to vinyl-silyl-sulfone 4-7. The olefin-plane adopts AAS, and the chelated MeLi with methoxyethyl glycoside 4-7a represents the first practical example of conformational constraint, leading to 4-8 in 100% syn-selectivity. Exclusive selectivity was also observed with epoxide 4-9 to yield 4-10.²⁾ In the asymmetric total synthesis of maytansinol (Fig. 5), 5-9 was obtained from D-mannose 5-2 via the key reaction involving 4-11, executed with complete syn-selectivity to give 4-12. Each of these HADCA products was then converted to the key intermediates 5-3 and 5-4, respectively. The final steps for both intermediates were identical, leading to the total synthesis of maytansinol in both racemic and chiral forms 5-9, as shown in Fig. 5.²⁾

Figure 5.

(Color online) Total synthesis of Maytansine in racemic and chiral form.

4. Substrate control in epoxidation to give syn or anti epoxide

The key intermediate (5-3, racemic), was converted to allyl alcohol 5-6. Epoxidation with m-CPBA exclusively resulted in the undesired stereochemistry (see Table 1). This result is attributed to the conformational constraint (the C3–C6 plane) and chelation control with the peracid between C3-OH and C7-O atoms, directing epoxidation to occur on the chelation face (below) as illustrated in Fig. 6A. Attempts to perform Sharpless’s asymmetric epoxidation¹⁷⁾ condition using (−)-tartrate on racemic 5-6 did not yield enantiotopic epoxide at all but instead yielded diastereotopic epoxide. This epoxide was conducted under substrate control (not reagent control), which was a key issue for racemate synthesis.

Table 1.

Epoxidation of allyl strained olefins

Figure 6.

(Color online) Conformational constraint of epoxidation giving opposite diastereomers.

Allyl strain, caused by the tri-substituted olefin and stereogenic carbon C-6, forces the orientation of C-7, which carries oxygen atom capable of capturing Ti(OiPrⁱ)₄ by chelation. As a result, TBHP should react from the non-blocked opposite face, as illustrated in Fig. 6B. (In this case, presence of the chiral source only resulted in kinetic resolution when the reaction was interrupted before completion.) The stereochemistry of Ti-TBHP mediated epoxide (without any chiral source) was chemically proven to be opposite diastereomer to the one with m-CPBA epoxidation product. The driving force in discriminating between the olefine faces was the difference in chelation power (see Table 1, Fig. 6).¹⁸⁾

5. Effects of remote oxygen atoms to aldol reaction on epoxy-aldehyde

This speculative mechanism has been rationalized through our comparative experiments using both the tri-substituted substrates (T1-1a,b) and similar compounds listed in Table 1. The di-substituted allyl alcohols (T1-4a–c) exhibit no allyl strain; however, a chelation effect appears to play a role. In fact, m-CPBA produced exclusive syn-epoxides (T1-4b,c) (1:>25), whereas Ti-TBHP did not seem to exhibit a strong influence. The observed difference is due to the assistance provided by A-strain when comparing with the tri-substituted olefin (T1-1a,b). It is likely that chelation-blocking mechanism in Ti-TBHP along with the role of remote oxygen atoms, contributes to the selectivity, as evidenced by the better ratio of T1-1b (>25:1) compared to T1-1a (10:1). The roles of these remote oxygen-ligands (on R-chain) continues, as seen in aldol reaction (Table 2).

Table 2.

Aldol reaction between the epoxy aldehydes with ethyl acetate

The next topic of conformational constraint is in the aldol reaction between the epoxy aldehyde 5-7b and the lithium salt of ethyl acetate, yielding (3S)-5-8. The selectivity of aldol reactions with related epoxy-aldehydes is compared in Table 2. A simple T2-1 showed almost no selectivity (3:2, Table 2, entry 1), but T2-2 (entry 2) gave a better result (6:1). This implies the involvement of the β-oxygen atom. The diastereoisomeric epoxy-aldehyde T2-3 (entry 3) gave ratio of 3:2, suggesting the non-involvement of the β-oxygen. In contrast, the aldol reaction of T2-4 having three additional remote oxygen atoms (entry 4), exhibited exclusive selectivity (50:1) to provide T2-4a.¹⁸⁾ A good selectivity (6:1) of aldol reaction, leading to the formation of 5-8, was also observed in the total synthesis of maytansine.²⁾

6. Conformational constraint of epoxide with neighboring stereogenic carbon(s)

A simple tri-substituted epoxide 7-1, in general, must be the first case to consider conformational constraint, since the rigid structure around epoxy ring is similar to the allyl strain observed in olefins (compare with Fig. 2: 2-2a R=R’=H, R’’=Me). As shown in Fig. 7, epoxy-aldehyde 7-1 is served as an example for its conformational energy diagram, calculated by rotating C4–C3 axis from 0° to 360°. The calculated result (anti-clockwise rotation) is shown in the graph, where two high energy peaks (III and V) arise due to the eclipsing interaction of C2-Me with C4-OMe or C4-Me, respectively. The low energy valleys (II) and (IV) represent the stable conformers. The lowest energy orientation (IV) positions C4-H and C2-Me more or less in a plane, and this conclusion is similar to the conclusion drawn from acyclic allyl strain of olefin as demonstrated in Fig. 2. The corresponding ball-stick models I–V are depicted in Fig. 7. These calculations were recently conducted by S. Hannongbua, P. Saparpakorn and M. Isobe (unpublished result).

Figure 7.

Energy diagram and conformational constraint of epoxy-aldehydes.

The transition state models based on such energy minimum conformation 7-IV, are illustrated in 7-2–4, following a similar ‘plane model’ by positioning C4-Me, C4, C5 and C6 in the plane, while C3 and C5-H are fixed and positioned slightly above the plane. A IV-equivalent conformation (obtained by rotating C5–C6 axis) places C6-H more or less in this plane. The middle drawing in Fig. 7 (7-3) shows the Li enolate of EtOAc during the aldol reaction of T2-2 at the transition state of entry 2. The conformational constraint from C3 through C7 facilitates the Li-chelation [SiO-Li-O(=CH)], bringing C7-O and C3=O into close proximity, thereby directing the enolate to approach from re-face of the aldehyde. In 7-4 (Fig. 7), the nucleophile (EtOAc anion) undergoes a re-face attack on the aldehyde from an open space, to form the C3(S) product. The C7-(Si)O-Li cannot chelate with the aldehyde oxygen in 7-4, unlike in 7-3, due to the presence of remote oxygen atoms (similar to the epoxidation of 5-6). Of course, such differences are unpredictable without knowing the experimental results. However, now with the conformational constraint, we can understand such mechanism.

After the enolate approaches from the open space as in 7-4, the stereoelectronic effect becomes particularly important. In the alignment orientation (7-4) (with red arrow on epoxide C4→O and between the π-orbitals of aldehyde from re-face), the nucleophile/anti-π-orbital of aldehyde/epoxide can enhance reactivity due to the stereoelectronic effect. The advantage of a highly functionalized substrates T2-4 and 5-7b in the total synthesis resulted in C3(S)-OH stereoselectivity of (50:1) and (6:1), respectively. We proposed this mechanism after completing the total synthesis, as analyzed through the conformational constraint concept.

We continued to develop heteroconjugate addition (Fig. 8) on 8-1, by trapping the anionic intermediate (8-1b) after addition of MeLi through chelation (8-1a). When this anion (8-1b) was trapped with phenylselenenyl chloride, it gave 8-2 with a carbon atom bonded with three different atoms (Si, S and Se). Subsequent oxidation of this carbon to COOH via sila-Pummerer rearrangement afforded Prelog-Djerassi lactone 8-3 (Fig. 8).¹⁹⁾

Figure 8.

(Color online) Prelog-Djerassi lactone via HADCA and sila-Pummerer rearrangement.

7. Second generation of HADCA via C-glycosidation with silylacetylene

In the first generation, heteroconjugate addition was originally designed for the synthesis of maytansine and further developed during okadaic acid synthesis. As for the second generation HADCA, we had first improved the preparation in 3 steps (Fig. 9); (i) C-glycosidation of triacetyl D-glucal²⁰⁾ 9-1 with PhS(TMS)-acetylene, followed by (ii) hydrosilylation and (iii) oxidation to afford 9-5. The C-glycosidation mechanism starts from Ferrier type-1 reaction into oxocarbenium ion 9-2, where π-orbital of silyl-acetylene overlaps with the α-π-orbital (destined to axial orientation) of the oxocarbenium ion. Desilylative neutralization of the sulfur-stabilized cation 9-3 provided the alkynyl group in 100% α-orientation, affording 9-4. Cobalt-catalyzed hydrosilylation of phenylsulfinyl acetylene 9-4 proceeded with high regio-selectivity to give Z-vinyl(silyl)sulfide, which was further oxidized to silyl-sulfonyl-Z-olefin 9-5. This method is broadly applicable to other glycals, providing α-selective acetylenes followed by cobalt-catalyzed hydrosilylation (Fig. 9).²¹⁾

Figure 9.

(Color online) C-Glycosidation of silylacetylene to D-glycals to afford 100% α-orientation.

The chiral vinyl-silyl-sulfone was prepared from a pyranose derived C6-aldehyde via Peterson olefination, followed by oxidation.^2,3) This method produced E/Z-regio-isomeric mixture, which was used without separation, since both isomers exhibited high syn-selectivity through α-chelation (10-1 C4-OSiR₃ or without C4-OH). When anti-selectivity was required, 10-1(C4-OH) was used along with Grignard reagent as a nucleophile, leading to the formation of anti 10-3 via β-chelation.²²⁾ This diastereo-switching method was successfully applied in our total synthesis of okadaic acid (Fig. 10).³⁾

Figure 10.

(Color online) HADCA can provide either enantio- and diastereo-isomers from D-glucose.

Applying this method to tetra-O-acetyl-D-glucal 10-4 gave 10-6 in four steps via 10-5, as shown in Fig. 10. The isomerization of its α-axial alkynyl group was explored to obtain the β-equatorial epimer 10-9 through cobalt complex and acid treatment, followed by the same reaction sequence afforded 10-10. Substrates 10-6 and 10-10 possess absolute configurations at C1, and can undergo syn/anti regio-switching addition through α/β-chelation either with ring-O or C2-O anion. This method enable us to explore a new conceptual approach for preparing both enantiomers from D-glucose. Figure 10 outlines general conversion of the functional groups; specifically, the C6-OR is convertible to terminal olefin via the iodide (C6-I) treatment with zinc metal. This transformation was employed in tautomycin total synthesis.⁴⁾

One of the applications of the current method demonstrates the synthesis of two unnatural isomers (Fig. 11). Compounds 11-1 and 11-2 are the hybrid protein phosphatase inhibitors designed from type-1 selective tautomycin 1-4 and type-2 selective okadaic acid 1-3, respectively. These compounds contain enantiomeric spiro-moieties on the right segments of tautomycin (Fig. 11), both of which were prepared from D-glucose.²³⁾ This synthesis has contributed to bioorganic assignment that the type-selective inhibition depends on the absolute configuration of spiro-moiety.²⁴⁾

Figure 11.

(Color online) Hybrid compounds of OKA and TTM synthesized from D-glucose.

The second generation HADCA was further extended in various aspects related to the syntheses of natural products tautomycin 1-4 and ciguatoxin 1-6. In the earlier stage of CTX synthesis, the C-glycosidation of pentopyranose-glycals, such as 12-1, was studied, providing 1,4-anti product 12-4 (Fig. 12).²⁵⁾ The stereochemical process occurs under stereoelectronic effect on the carbenium ion 12-2, which forces the nucleophile to approach in an vinylogous anti manner to the axial acetoxy group (12-3), thereby affording 12-4. The stereochemical outcome of this reaction arises from conformational constraint, which differs from hexose-glycals (such as Fig. 9) due to the absence of the C6 carbon. As a result, the axial C4-OAc conformation 12-3 becomes the more stable one to interact with the olefin π-orbitals. (Opinions about this mechanism, both favored²⁶⁾ and disfavoured²⁷⁾ have been reported.) This reaction enabled us to prepare stereochemically authentic samples of ciguatoxin.²⁸⁾

Figure 12.

(Color online) C-glycosidation of pentopyranose-glycal to 1,4-anti-alkyne.

Conformational constraint theory prompted us to confirm the stereochemical process involved in forming cyclobutane ring for solanoeclepin A (see Section 11), as shown in Fig. 13. Starting from D-arabinal 13-1, we followed 1,4-anti C-glycosidation (13-2). Catalytic hydrosilylation gave 13-3 and HADCA of lithium acetylide to 13-4 afforded the cyclobutene 13-5 in one pot under complete stereo-control. The process involves α-chelation (13-A) and bond-rotation (with keeping configuration) of the lithium carbanion (13-B), which opens the epoxide at C2 (13-C) by overlapping with the quasi-axial anti-bonding orbital to form the cyclobutane ring 13-D (X-ray crystallographic analysis).²⁹⁾ The high stereocontrol of carbanion intermediates is often retained its configuration in THF solvent by chelating with Li ion.³⁰⁾ These experiments afforded a bicyclo[4.2.0]octan 13-5 as a model study for constructing a tricyclo[5.2.1.0^1,6]decene framework in solanoeclepin A synthesis (see Section 11).

Figure 13.

(Color online) Cyclobutane ring-closing via anion-relay under conformational constraint.

8. Prostaglandin synthesis by anion relay after HADCA

Conformational constraint theory was further extended to an ‘anion-relay’ type synthesis aimed at the one-pot synthesis of prostaglandin (Fig. 14). The chiral starting material (R)-14-1 reacted with the chiral vinyl-lithium, generated in situ from 14-2, to afford the cyclopentanone 14-3 in one pot. The A-strain of 14-1 (14-A) facilitates Nu attack through chelation, leading to the exclusive formation of the syn adduct 14-B. Subsequent bond rotation allows a Brook rearrangement to shift from C-Si to O-Si (14-C), giving rise to dianion 14-D. Using carbanion with pure lithium ion stops at 14-D before cyclization to 14-E, affording linear products in 50–60% yields after work-up. The corresponding di-deuterated product was obtained by quenching 14-D with D₂O.³¹⁾ However, when the same reaction was conducted in the presence of NaBr, which facilitates partial exchange of Li cation, the ring closure proceeded (14-E) to form the keto-sulfone enolate 14-F and ultimately led to the isolation of cyclopentanone product 14-3. In this case, only minimal alkylation of 14-F with 14-4 took place (until once isolate 14-3). After workup, 14-3 underwent the same alkylation with allylbromide 14-4, followed by deprotection, to provide methyl ester of prostaglandin E2 14-5.³²⁾

Figure 14.

(Color online) Anion relay synthesis of prostaglandin E2 methyl ester.

9. Principal strategy for tetrodotoxin synthesis

Tetrodotoxin (TTX) 15-1 was elegantly synthesized in racemic form by Y. Kishi in 1972, starting from Diels-Alder cycloaddition followed by judicious introduction of all the necessary parts.³³⁾ We planned to synthesize TTX in an optically active form by employing chiral dienophile levoglucosenone (15-2) to construct the chiral cyclohexene 15-4. In 1987, a preliminary work was done with all the necessary stereogenic carbons except for the nitrogen functionality, yielding 15-5.³⁴⁾ The structures of 15-1 and 15-5 are drawn from the same angle to facilitate the comparison of the cyclohexane skeletons, although the actual conformation seems to be 15-6. The introduction of a nitrogen atom dramatically changed the compound’s chemical behavior compared to the one without nitrogen. It took a long time for us before we succeeded in this approach for total synthesis of TTX in 2003–4.^5,6)

On the basis of this result, we designed the retrosynthesis as shown in Fig. 15, where only two key points of conformational constraint are selected here: (1) epoxide opening (15-7) and (2) [3,3]-sigmatropic rearrangement (retro-Overman rearrangement)³⁵⁾ (15-9→15-8). Compound 15-10 can be access via Diels-Alder cycloaddition from levoglucosenone (15-2), which can be prepared in 0.1 kg-lab-scale through acidic pyrolysis of cellulose.³⁶⁾

Figure 15.

(Color online) TTX N-introduction via conformational constraint Overman rearrangement.

‘Opening of an epoxide ring’ on a 6-membered ring of 15-7, in general, provides di-axial product. The conformational constraint mechanism dictates that the nucleophile coordinates with quasi-axial-antibonding orbital and opens the quasi-equatorial C–O bond with S_N2 inversion, resulting in a di-axial orientation.

The other important point for TTX synthesis is the Overman rearrangement. Although it seems not necessarily appropriate by the author himself to introduce N-atom into sterically more congested direction. Our reason why we planned to employ the Overman rearrangement was driven by the conformational constraint imposed by the typical allyl-strain of 15-9a. The presence of an exo-cyclic tri-substituted olefin renders its conformation as shown in 15-9b, in which the bulky acetonide moiety adopts an axial orientation.³⁷⁾ This rearrangement should facilitate the N-atom to approach the bulky acetonide group from the anti-direction. This strategy was first examined with 15-9a,b. As anticipated, the allyl-strain helped the ground state conformation, favoring 15-9b over 15-9a, similar to the transition state. The sigmatropic equilibrium involving 15-8a is dynamic, because it is destined to invert to its more stable conformation di-equatorial 15-8b. Following this strategy, we converted it to the cyclic guanidine 15-11. This approach ultimately enable the total synthesis of tetrodotoxin in 2004 and many deoxy-TTX analogs (1999–2015) as potential biosynthetic precursors by Nishikawa.³⁸⁾

During such great progress toward deoxy-TTX synthesis before the achievement of our total synthesis of TTX,⁶⁾ we examined the second strategy for TTX (16-1). This alternative strategy aimed to generate chiral cyclohexanes with more functional groups at the earlier stages, in contrast to the previous Diels-Alder strategy. The initial retrosynthesis of 16-2, still based on Overman rearrangement (as realized in 17-9–11),³⁸⁾ was planned using D-glucal tetraacetate 16-10 as the starting material, as shown in Fig. 16.

Figure 16.

(Color online) Second synthetic plan of TTX and Claisen-Aldol strategy.

The critical C6–C7 bond formation (16-5,6) was planned on methylketone (16-7) after Claisen rearrangement (16-8) (Although not illustrated here, a [2,3]-Wittig rearrangement could generate the necessary framework as well).³⁹⁾ The 3D structures 16-5a–8a are drawn from the same angle as TTX 16-1 to facilitate comparison. However, along the way, we gave up the Overman rearrangement (16-2) because it did not proceed as expected. Thus, we unfortunately obliged re-planning. The critical introduction of the N-atom at C8a was re-designed from 16-5, after confirming that the configuration of C5-OH of 16-6 was opposite to that of natural TTX, and its inversion at this stage was difficult to perform.

The stereochemistry of C5-OH should be derived by the epoxidation of 16-6b from its si-face. This prediction was analyzed based on conformational constraint; (1) the C4a–C5 bond is restricted (with C4a-H and OMe lying in the same plain) due to allyl strain (16-6b) (2) a remote chelation effect with m-CPBA (similar to the case in Fig. 6, Table 1). In fact, C5(S)OH was predominantly obtained in a 7:1 ratio, as analyzed.⁵⁾

The exocyclic olefin of 17-1 was prepared as planned. The corresponding imidate adopted its conformation as 17-2; however, it gave 17-3 through [1,3]-sigmatropic acetamide product but did not provide 17-4 [3,3]-product. This result strongly contrasted with the Overman rearrangements, such as the transformation of 17-9 to 17-11, which proceeded to give [3,3]-product.³⁸⁾ This outcome was unlikely due to steric hindrance of C5-OBz group for approaching imino group, as the corresponding C5-epimer also failed to give [3,3]-product. We revisited 17-1 to use C5-OH (17-5) through an intramolecular conjugate addition of carbamate-N to unsaturated ester 17-6. The N-atom was introduced at C8a position of 17-7a and then re-protected as 17-8 as shown in Fig. 17.⁵⁾

Figure 17.

(Color online) Changing strategy of N-introduction via intramolecular conjugate addition.

Conformational constraint is of particular significance, even in cases where they necessitate to change from the original plan to an alternative one. Another example is shown in Fig. 18, concerning the introduction of C9(R)-OH in 18-4 from 18-3. Lactone 18-3, obtained via epoxide opening of 18-2, could not be converted under possible conditions to the oxylactone 18-4. An important solution was experimentally discovered that enolate 18-5 from aldehyde 18-2 afforded the Z-vinyl ether 18-6 via the epoxide opening. Further dihydroxylation and oxidation to ketolactone 18-7 was followed by hydride reduction from the open space approach to provide C9(R)-OH in 18-4. Notably, Z-enolate configuration in 18-5 was not an equilibrium mixture of regio-isomers, but it was formed exclusively during heating 18-2 with DBU. This geometry was confirmed by isolating the corresponding Z-O-silyl ether (J_9,10 = 7.5 Hz). This outcome may arise from the mechanism as speculated as 18-2a, which is not predictable based on conformational constraint, but it is conceivable.⁵⁾

Figure 18.

(Color online) Introduction of C9-OH of TTX.

The last part of the total synthesis of TTX focused on the formation of the ortho ester and cyclic guanidinium moieties from the lactone 18-4 (→19-1) as shown in Fig. 19. The glycol 19-1 was cleaved to the corresponding aldehyde, which then spontaneously cyclized to guanidine-aminal 19-2, a protected form of (−)-TTX. For deprotection, 19-2 was first converted to methyl-O,N-acetal 19-3a. Deprotection process via 19-4 afforded the same equilibrium mixture in acidic media as natural product TTX 19-5–7. We achieved the first asymmetric total synthesis of (−)-tetrodotoxin in 2003.⁵⁾

Figure 19.

(Color online) First asymmetric total synthesis of tetrodotoxin.

The total synthesis of TTX and its deoxy-analogs on the basis of Overman rearrangement has also achieved, which has been now linking to the biosynthesis of TTX.³⁸⁾ In parallel, ultra-micro analytical studies of tetrodotoxin were also done in our laboratories. This investigation included Field Desorption MS, nano-LC-ESI-QTOF mass analysis, and deuterium exchange method. Most of these studies have been published as the company technical reports by the instrument suppliers.⁴⁰⁾

10. Total synthesis of ciguatoxin and conformational constraint

The structure of ciguatoxin (CTX) was reported by T. Yasumoto group in 1989,⁴¹⁾ and the first total synthesis was elegantly achieved by M. Hirama group in 2006.⁴²⁾

During the past decades, we have achieved several other total syntheses of different classes of natural products under highly stereochemical control, and learned how to realize the principle of bond forming reactions under the control in mind. This author has proposed that conformational constraint has continued to play a pivotal role for stereo-control, even after the achievement of those total syntheses. We have continued to pursue the total synthesis aiming at new classes of natural products, with the purpose to search for potentially new principle of conformational constraint. Ciguatoxin 1-6, CTX serves as a good example for this purpose; namely, gigantic molecule, CTX 1-6 comprises 13 ether rings and a four-carbon chain connecting at one end C5. It features contiguous C–C bonds from C1 through C55 with 33 stereogenic carbon atoms, five Me groups, and six OH groups. It seems very difficult to achieve the total synthesis even by assembling all existing methodologies. We need a challenge for exploring new principle for conformational constraint from this target oriented total synthesis. As we had already learned from previous synthetic works, synthetic plan had often been obliged to change through the experimental process, so that the retrosynthesis should be brief and flexible upon facing unexpected problems.

CTX 1-6 is retro-synthetically disconnected into smaller fragments, and the first disconnection would be around the middle of such big molecule into right and left segments. It should be followed by further disconnections into smaller segments. This means that we need multiple rounds of ‘coupling’ and ‘medium size ether ring-formation’.⁴³⁾ In the past, we experienced unusual problem in coupling reaction during okadaic acid 1-3 synthesis. Specifically, some carbanion and an aldehyde survived in a vessel for long time without undergoing coupling. We discovered that this problem caused by strong Li⁺-O chelation, due to high purity of Li⁺, stopped the anticipated coupling reactions.^3,44) Such problems have been solved, in general, by introducing a catalytic amount of Na⁺ and/or changing solvent (vide supra). Utilization of alkynyl groups for the coupling offers a better alternative, since they react either under basic (e.g. 20-1a, ) or acidic (20-1b, , ) conditions (Fig. 20) or even using transition metals (Pd) as a catalyst to obtain 20-2.

Figure 20.

(Color online) Central strategy using acetylene-cobalt complex for CTX; for coupling, medium ring cyclization, and transformation to i) olefin, ii) vinylsilane, iii) allyl alcohol, or iv) ketone.

We have established C–C bond forming reactions applicable to CTX synthesis, one of which is silyl-acetylene chemistry during α-selective C-glycosylation (Fig. 9, Fig. 12). The coupled alkynyl product 20-2 can be converted into its dicobalthexacarbonyl complex 20-3, which undergoes C–O bond formation under acidic condition. This method enables stereoselective cyclization for closing medium size (7, 8, 9-membered) ether rings for CTX through Nicholas cation intermediate 20-3.⁴²⁾ In this case, the concept of conformational constraint of 20-3 renders the reactants, OH and the cation, into close proximity due to a ‘push-in effect’ by the bulky six CO-ligands. The syn/trans stereochemistry of the cyclized ether ring of 20-4 is always exclusive because the bulkier group adopts the position which allows to bring small H inside.⁴⁵⁾ After cyclization, the cobalt complex 20-4 can be converted into either an olefin 20-5, allyl alcohol 20-6, or sometimes to ketone 20-7, depending on the needs of the next ring formation (e.g. 20-8 of CTX). No such reaction was known at all before our study on CTX.

HADCA reaction between 21-1 and lithium acetylide 21-2 provided 21-3 through β-chelation mechanism involving C44-OLi (21-1a) as shown in Fig. 21. Compound 21-3 was converted to the dicobalt-hexacarbonyl complex and subsequently treated with BF₃•OEt₂ to give cyclization product 21-4 as a single syn/trans isomer. The cobalt and sulfonyl group of 21-4 were successively removed with tin-hydride and sodium amalgam to give the K-ring olefin 21-5 with an anti-Me(C58) configuration.⁴⁶⁾ The cobalt complex was found generally removable to give olefin using n-Bu₃SnH or NaHPO₂ in addition to Wilkinson catalyst.⁴⁷⁾

Figure 21.

(Color online) HADCA coupling and HIJK-ring of CTX.

This reaction sequence was further examined under various possible combinations to find best stereoselectivity of JKLM ring system as shown in Fig. 22. The precursor for LM-ring was prepared from D-glucose through steps 22-1’ to 22-5’ (conjugate addition of cuprate followed by enolate trapping with MeI in all axial manner to 22-2’),⁴⁸⁾ and used as the nucleophile acetylide 22-6’ in the reaction with 22-1 to give 22-2. The cobalt assisted cyclization proceeded as in the previous sequence, leading to K-ring olefin 22-5 and then 22-7. At this stage, we confirmed the stereochemical purity, and found out an imperfect selectivity with a ratio of 3.6:1 (22-3 68% anti-Me, 22-4 19% syn-Me) caused by HADCA reaction with lithium acetylide of 22-6’. Although our previous studies suggested Grignard reagent (Mg⁺⁺) for the ‘β-chelation’,²⁰⁾ applying this condition unexpectedly resulted in no addition product 22-2 at all. The other problem of forming the K-ring at C49 (Nicholas cation) involved potentially competition between C44-OH and C52-OR (R=TBS, Ac). Cyclization to 22-3 required higher concentration (1.72 mM BF₃•OEt₂) than the previous cases (21-4 0.16 mM), which led to deprotection of C55-OPMB (22-3,4). Reductive cobalt decomplexation to 22-5 was carried out using tin-hydride in the presence of bis-TMS-acetylene, which effectively captured the cobalt species. After re-protection and oxidation, the terminal olefin was converted to methyl-ketone via Wacker oxidation 22-6. Dihydroxylation of the inner olefin conducted in the presence of C58-SO₂Ph, exhibited better stereoselectivity than the de-sulfonylated Me derivative (for example 21-5). Then, the spiro LM-ring 22-7 cyclized under thermodynamically stable form as the classic stereo-electronic effect.⁴⁸⁾

Figure 22.

(Color online) Synthesis of the JKLM-ring fragment of ciguatoxin.

The right-hand segment was completed as shown in Fig. 23, with all the carbon atoms in place after the en-yne coupling between 23-1 (derived from D-glucose) and 23-2 to give 23-3.⁴⁹⁾ The 8-membered I-ring was subsequently cyclized via the cobalt-assisted method to give syn/trans configuration 23-4. This was then converted to enone 23-5 guided by conformational constraint considerations.⁵⁰⁾ Namely, the I-ring readily adopts the 23-5a conformation due to the conjugated π-orbitals and polarization of C36→O bond (shown by the two arrows). As a result, the lone pair electrons of C33-O preferably attack in an anti-mode to give 23-6 with trans-H/I-ring under stereo-electronic control. Finally, the β-hydroxybenzoate 23-6 was converted to 23-10, completing the right-hand segment (Segment R), ready for coupling with the lithium acetylide of the left-hand segment (Segment L).

Figure 23.

(Color online) HIJKLM-ring synthesis along the line on the conformational constraint.

Next key point of conformational constraint is the stereoselective formation of Me-57 in the α-orientation from the endo-olefin 23-8 to 23-9 (see also Fig. 24). The essential stereochemistry lies in how to convert the ketone 23-7 to α-methyl group 23-9 with high selectivity. A model HIJ-ring system having exo-olefin was directly reduced with diimide (HN=NH) to predominantly afford α-methyl product (3.4:1).⁵¹⁾ To enhance this selectivity, we designed the following experiment from conformational constraint point of view.

Figure 24.

(Color online) Conformational flipping of 8-membered I ring in endo/exo-cyclic double bond.

The conceptually significant formation of Me57 on 8-membered I-ring was independently confirmed using 6/8/6-tricyclic compounds 24-1-endo-a and 24-2-exo-a. The conformations of these models are contrastive providing the open space where the reducing agent would approach in opposite way, as indicated by two arrows as shown in Fig. 24. However, standard hydrogenation using Pd or Pt catalysts, gave a mixture of 24-3 (α-Me) and 24-4 (β-Me) isomers.⁵²⁾ Such results are due to simultaneous isomerization between 24-A and 24-B, catalyzed by Pd/Pt tending from endo- into exo-olefin. In contrast, kinetic hydrogenation using Crabtree’s Ir catalyst exclusively afforded 24-3 (α-Me) from 24-1-endo.⁵³⁾

As shown in Fig. 25, the syn/trans cyclization from 25-1 to D-ring 25-2 was achieved following the previously described reaction pathway, and the next step then involved its conversion to allyl alcohol 25-6, for which we developed essential 3-step conversion. First, regiospecific hydrosilylation of 25-2 to 25-3 was followed by epoxidation to give 25-4. Subsequently, Ohira-Bestman alkynylation was introduced at this stage for E-ring construction, affording 25-5. The epoxy-silane to allyl alcohol rearrangement was then explored with BF₃ to afford allyl-alcohol 25-6. The stereochemical process of epoxide opening (25-5a) involves the hydride shift (25-5b) and desilylation of the β-cation (25-5c). This process was repeated from 25-6 to achieve E-ring formation, and implemented by Kira.⁵⁴⁾

Figure 25.

(Color online) Epoxy-silane rearrangement for D/E-ring of CTX.

Transformation of an acetylene dicobalt complex into ketone 20-7 (Fig. 20) was one of the most difficult part in CTX synthesis. Figure 26 illustrates our first approach for E’FGH-ring synthesis through the ketone formation. The coupling between acetylene 26-1 and aldehyde 26-2 (as a representative of the two major segments), was followed by the cobalt complex strategy (26-4), where treatment with toluene sulfonic acid* provided F-ring syn/trans cyclization product 26-5. [*When using BF₃•OEt₂, we observed an unwanted benzylic hydride transfer to the Nicholas cation 26-4 as a side reaction.] The key reaction converting 26-5 to ketone 26-6 originated from some side products in our earlier experiments. Specifically, reductive decomplexation condition unexpectedly afforded ketone 26-6. It was converted to the F-ring of 26-7, in which we observed slow conformational equilibrium as 26-7a,b similar to CTX (Fig. 26).⁴⁹⁾ The source of oxygen is air (rather than water), and through our systematic experiments, we found that the regio-selectivity was best conducted by changing the ligands of cobalt-complex to DPPM (Ph₂P-CH₂-PPh₂).⁵⁵⁾

Figure 26.

(Color online) Ketone synthesis after Co-assisted coupling.

During our ligand-exchange program, DPPM (Ph₂P-CH₂-PPh₂) was identified, among others, with specific character. It is the only diphosphine ligand capable of ligating P-Co in a one-to-one fashion, forming five-membered ring structure (26-8). We propose the following mechanistic pathway to show the regioselectivity, as illustrated in 26-8–13.⁵⁴⁾ One of the two apical CO ligands is eliminated upon heating to give 26-9, then oxygen displaces forming peroxide 26-10. At this point, one of the two highly congested L-C–Co bonds rearranges to expand from a three-membered ring (26-10) to a four-membered ring (26-11). This process is favored over the rearrangement of the less congested S-C–Co bonds. After reductive elimination of cobalt to give olefin 26-12, the carbonyl group ultimately ends up on the L-side as 26-13.⁵⁵⁾

The final coupling between the 2 major Segments L (acetylene 27-1) and R (aldehyde 27-2), along with the following total synthesis, is shown in Fig. 27. The essential steps included the conversion of acetylene-dicobalt complex 27-3 into the regiospecific ketone 27-4. Details of these steps were 1) the coupling reaction, 2) cobalt assisted cyclization under mild acidic condition. To this end, there were very limited conditions due to the presence of acid-sensitive functional groups, such as 5,6-spiroketal (LM-ring), the potential benzyl-hydride migration (see 26-4), and the presence of diallyl ethers. These factors posed the potential problem under acidic treatment for F-ring cyclization, which were solved under the optimized conditions to give 27-7.

Figure 27.

(Color online) Coupling Segment-L and -R and total synthesis of ciguatoxin.

To avoid experimental sample loss by spending the late-stage compounds (27-7→27-8), we used the stock compounds such as ABC,^28,56) BCDEF,⁴⁵⁾ EFGH,⁵⁵⁾ HIJ,⁴⁹⁾ JKLM,⁴⁸⁾ etc. These compounds served for the stability tests to confirm suitable reaction conditions toward total synthesis of CTX. Protecting groups for C54-OH on M-ring were changed due to the fact that this spiro-ketal portion was acid labile but it can survive as its acylates (C54-O-Bz or -OAc) under mild acid conditions. The synthesis of A-ring and side chain had already been studied under both acidic and basic conditions, as mentioned in Fig. 20 and Fig. 28 (28-4a,b). However, the coupling and cyclization was risky to handle in small scale. We employed the third method of Pd-mediated en-yne coupling between 28-1 and 28-5 as representatives of 27-6 and 27-7 with C54-OAc.⁵⁶⁾ However, the tri-pivalate cobalt complex 28-2 did not provide cyclization product 28-3 upon treatment with BF₃•OEt₂ in CH₂Cl₂. Changing acid to TMSOTf led to a loss of pivalic acid (28-2a), but no cyclization took place. Addition of THF as a co-solvent into this reaction mixture gave 28-3. Such solvent effect suggests a speculative involvement of the counter pivalate anion. Notably, the similar coupling product 28-6 did not encounter such problem to give 28-3 at all. Applying these conditions between 27-6 and 27-7, enabled the formation of 27-8 with the A-ring cyclization after the double bond migration. Reductive decomplexation was achieved using sodium hypophosphite,⁵⁷⁾ and hydrolysis of the acetates led to the completion of the total synthesis of ciguatoxin (27-9). These conclusive steps were successfully modified and implemented by Hamajima to conclude the total synthesis of CTX.^7,58)

Figure 28.

(Color online) The 3rd generation method for A-ring cyclization.

Let us summarize the new methodology using the alkynyl groups, which have played significant role for ciguatoxin synthesis. Figure 29 shows a special illustration to explain the roles of alkynyl group. Among the seven medium size ether-rings (A, D, E, F, G, I, K), all except G were constructed using coupling and ring cyclization through the cobalt complex methodology. After the cyclization, for D and E-ring synthesis, epoxy-silane rearrangement was developed from their cobalt-complex intermediates (Fig. 25). After cyclization of F-ring, the Co-complex was transformed to ketone by new oxidative conversion (ketone synthesis), which was regio-selective through ligand exchange with DPPM. This ligand exchange also effected regio-selectivity in hydrosilylation process.⁵⁹⁾ The C29-ketone was spontaneously cyclized with C34-OH to form a ketal, which was further reduced to complete the G-ring synthesis.

Figure 29.

(Color online) Summary of CTX synthesis on acetylene cobalt chemistry.

The acetylene (C37–C38, ethyne) was connected at each side with two fragments (C30–C36, C40–C45), both of which were derived from D-glucose via Pd-mediated cross coupling and acetylide-aldehyde docking, respectively. The eight-membered ether I-ring was cyclized between C36 cation and C42-OH via Co-ligand push-in effect. The trans 6/8 H/I-ring cyclization proceeded under stereo-electronic control from the specific conformation of the I-ring. Furnishing C39(R)-Me at C57 was achieved via kinetically controlled Ir-catalyzed hydrogenation from the endo-olefin under conformational constraint. The formation of L-ring, which contains all five stereogenic carbons induced from D-glucose, was completed at the same time with M-ring via C47–C48 acetylene-reductive decomplexation with tin-hydride to olefin, followed by asymmetric dihydroxylation. In parallel, the left end moiety was commenced from both enantiomers, synthesis from D-glucose, as a precursor for B-ring.^28,56) A new stereo-electronic effect, which produced 1,4-anti products, was discovered during the course of C-glycosydation of pentose-glycals with bis-trimethylsilylacetylene. This new method allowed us to establish the configuration at C2(S)-OH.^28,56) Regarding the cobalt-assisted A-ring cyclization, we ultimately established three different coupling methods as well as new decomplexation using tin-hydride or sodium hypophosphite.^47,57) These are the results based on the pivotal concept of conformational constraint during the synthesis of ciguatoxin.

11. Partial synthesis of solanoeclepin A

The studies toward solanoeclepin A (30-1) involved high stereoselectivity based on the principle of conformational constraint. Tanino successfully achieved the first total synthesis of solanoeclepin A (30-1).⁶⁰⁾ While we have not yet completed our own total synthesis, some successful results have been reported on the basis of conformational constraint. Toward our retrosynthetic segments 30-2, and 30-3, we synthesized bicyclic compound 30-4 in optically active form. The acetylene group in 30-4 was used for coupling with cyclohexene 30-5, which represents the Right Segment, to form 30-6 as shown in Fig. 30.

Figure 30.

(Color online) Conformational constraint in the cobalt-assisted Prins-Ene reaction.

To achieve the stereoselective B-ring cyclization, acetylene-cobalt complex was employed in combination with Me₂AlCl, by using ‘push-in-effect’ of the 6-CO ligands to render pre-organization of 30-6 as 30-7. Intramolecular Prins-ene reaction of 30-6 was assisted by the Nicholas cation intermediate 30-8. Activation of the aldehyde carbonyl by a Lewis acid resulted in chelation with bridge oxygen atom on the side away from C2-OAc. This orientation led to the exclusive formation of product 30-9 with 10(S)-configuration, when metallic Lewis acid such as either R_nAlCl_m, TiCl₄, or SnCl₄ was used. In contrast, treatment with TMSOTf resulted in a mixture (2:1), highlighting that metal chelation (30-7) was the key in controlling the stereoselectivity. Regarding acetylene-dicobalt complex, the pushing-in-effect by the ligands facilitated the alignment of the two reactants (en-one) as shown in 30-10, leading to the highly stereoselective outcome. This reactivity is further supported by the stereo-electronic effect of the polarized C→O bond next to the carbonyl group (30-10). Specifically, C→O polarization assists in an anti-attack by enabling the overlap of the olefin’s π-orbital with carbonyl π-orbital at the transition state. Overall conformational constraint provided the exclusive formation of the C10(S) product 30-9 in 78% yield.⁶¹⁾

This effect was also applied to the tricyclo[5.2.1.0^1,6]decene framework of right segment 30-3 as shown in Fig. 31. The allyl-propargyl ether 31-1 underwent a stereoselective [2,3]-Wittig rearrangement (31-2) to form propargyl alcohol, which was subsequently converted to its dicobalt complex 31-3. Upon treatment with TMSOTf, the Nicholas cation emerged near the π-orbital of the allylsilane due to push-in effect of the bulky CO ligands, leading to π-π orbital overlap as shown in 31-4, which ultimately generated the tricyclic product 31-5. A tricyclo[5.2.1.0^1,6]decene frame work was synthesized by closing the cyclobutane ring via radical process from 31-9 to 31-10. These two examples represent basic studies contributing to the total synthesis of solanoeclepin A.⁶²⁾

Figure 31.

(Color online) Construction of tricyclo[5.2.1.0^1,6]decene framework via 2 different ways.

12. Summary

We have discussed the concept of conformational constraint as the pivotal principle for stereochemical control in the total synthesis programs on various categories of natural products. The technical terms, stereoelectronic effect, which has been revised or expanded from spiro-ketal analysis by Pier Deslongchamps since 1980th,¹⁶⁾ and allyl strain (A-strain), which has been focused on the stability of axial/equatorial substituents on exocyclic olefin as reported by Johnson, are discussed.⁶³⁾ Vernolepin synthesis serves as an early example where A-strain was applied in natural product synthesis.¹⁾ This effect was expanded to general acyclic systems during development of HADCA reaction in maytansine synthesis.²⁾ The restricted conformation of a stereogenic carbon locating at the allylic position must be followed by bond-forming process under limited direction. The reactants (substrate and reagent, nucleophile and electrophile) should approach toward the transition state as similar orientation as the restricted conformation (mostly in acyclic system). One of the key factors in this process is the chelation effect, which significantly enhances the reaction rate. The other factor is stereoelectronic effect, which lowers the potential energy by the molecular orbital overlap. The third factor is the size of surrounding moieties, such as push-in-effect from the cobalt-ligands. Even remote oxygen ligands play a critical role in epoxidation and aldol reactions. We have discussed these combined effects and results in the synthesis of okadaic acid, particularly for the acylic stereocontrol using A-strain and oxy-mercuration, as well as stereoelectronically controlled hydride reductions of ketones (though not included in this article). Additionally, Jim Marshall⁶⁴⁾ reported quite original idea on the use of allene with chelation control during tautomycin synthesis, which is also elegant.

In addition to the double bonds, let us take an epoxide as the next topic. The four substituents on the epoxide are as rigid as those on an olefin besides the bond angles. Taking epoxyaldehyde-aldol reaction in consideration, a neighboring stereogenic carbon can induce strong environmental differentiation in conformational constraint, further demonstrating the significance of these stereochemical principles. The significance of conformational constraint is clearly evident across the transition states in controlling the stereochemistry, that is often observed during critical bond formation in the total synthesis involving multiple stereogenic carbon atoms.

Non-standard abbreviation list

AAS

acyclic allyl strain

A-strain

allyl strain, CTX: ciguatoxin

HADCA

heteroatom directed conjugate addition

TBHP

tert-butylhydroperoxide, TMS: trimethylsilyl

TTX

tetrodotoxin

Profile

Minoru Isobe was born in Nagoya, Japan in 1944 and graduated from the School of Agriculture, Nagoya University, in 1969. He majored in bioorganic and natural product chemistry. He received his PhD degree in 1973 at Nagoya University under the guidance of Professor Toshio Goto, and he then moved to New York, U.S.A., to work with Professor Gilbert Stork as a postdoctoral fellow at Columbia University from 1973 to 1975. He became Associate Professor in the Laboratory of Organic Chemistry, School of Agriculture, Nagoya University in 1975. In 1991, he was appointed as Professor at this school until 2008. He was also co-appointed as Professor at the Advanced Research Institute of Nagoya University. He moved to the Chemistry Department of the National Tsing Hua University, Taiwan, and serving as NSC Chair Professor until 2014, and he then moved to the Chemistry Department of Sun Yat-sen University, Taiwan, until 2015. He spent 1 year (2015–2016) at Chulabhorn Research Institute (Bangkok, Thailand) as a Professor. He returned to Nagoya and has been appointed as Visiting Professor at Toyama Prefecture University. He has carried out pioneering work on the chemical synthesis of structurally complex natural products with special reference to stereochemical control. He also uncovered various bioorganic mechanisms of silkworm diapause (induction and termination), bioluminescence of oceanic squid (photoprotein), cereus toxin (food-borne illness), and bio-active peptides.

He has served on various committees of IUPAC and particularly as the Division President of Organic and Biomolecular Chemistry (2004–2007). He is the main founder of ACP (Asian Core Program, since 2005) to organize organic chemists in South/Ease Asia. He has received Royal Awards such as the Medal with Purple Ribbon (Japan, 2008), Princess Gold Medal (Thailand, 2012), the Order of the Sacred Treasure, Gold Rays with Neck Ribbon (Japan, 2018). He received the Nakanishi Prize from Japan/American Chemical Societies in 2024.

Conflict of interest

The author declares no conflicts of interest.