Ga(III)-Catalyzed Cycloisomerization Approach to (±)-Icetexone and (±)-Epi-Icetexone

Abstract

A Ga(III)-catalyzed cycloisomerization reaction provides expedient access to a benzannulated cycloheptadiene bearing a cyano group, which has been applied to the syntheses of several icetexane diterpenoids including icetexone and epi-icetexone. Key to the synthesis is a novel insitu generated diazene rearrangement.

The icetexane family of diterpenoids is a group of [6-7-6] tricyclic natural products, which are believed to arise in Nature from an oxidation-induced rearrangement of the structurally-related abietane diterpenoids.¹ Biosynthetic modification of the abietane skeleton has led to a wide variety of icetexane natural products that possess unique structures as well as biological activity.² We became interested in the synthesis of members of the icetexane family as a prelude to a systematic investigation of their biological activity. Previously, we reported the syntheses of several members including salviasperanol (1, Figure 1), abrotanone (2) and 5,6-dihydro-6α-hydroxysalviasperanol (3).³ In this communication, we report the extension of these studies to the formal syntheses of icetexone (5) and epi-icetexone (6).

Figure 1.

Selected icetexane natural products

Icetexone and epi-icetexone have been isolated from several Mexican Salvia plants including Salvia ballotaeflora and Salvia gillessi.⁴ Of these two natural products, icetexone has been shown to possess trypanocidal activity against the trypomastigote form of Trypanosoma cruzi, the parasite that causes Chagas’ disease.⁵ In 2009, Majetich and Grove reported the first syntheses of 5 and 6,⁶ which exploited a cyclialkylation to forge the seven-membered ring,⁷ as well as a late stage iodolactonization to construct the bridging lactone.

In principle, the related natural products anastomosine (4), icetexone (5) and epi-icetexone (6) could arise from hydroxyl-directed oxygenation of the β-methyl group at C-19 (see 3 ) followed by a series of oxidations and dehydrations. However, the anticipated instability of potential alkoxy radical intermediates,⁸ as well as uncertainties regarding the stereoselectivity of the C-H functionalization process, dissuaded us from this line of inquiry.

Encouraged by our previous synthetic studies on the icetexane diterpenoids, we were drawn to the use of benzannulated cycloheptadienes (e.g., 8) as key precursors to 5 and 6 (see Scheme 1). Importantly, a cyano group at C-4 could serve as a masked carboxylic acid group that would be unveiled at a late stage. Benzannulated cycloheptadiene 8 could in turn be prepared from alkynyl indene 9 via a Ga(III)-catalyzed cycloisomerization.

Scheme 1.

Retrosynthetic Analysis of 5 and 6

The incorporation of a cyano group as a part of the alkynyl indene substrate for the cycloisomerization reaction has not been previously demonstrated. Thus, the present plan would seek to advance the substrate scope of this seven-membered ring forming cycloisomerization reaction.

In this communication, we present the realization of this synthetic strategy, which has led to the formal syntheses of (±)-icetexone (5) and (±)-epi-icetexone (6).

Our efforts commenced with the preparation of iodide 17 as outlined in Scheme 2. The sequence started with 3-bromo-1-propanol (10), which was converted to TIPS ether 11 under standard conditions. Bromide 11 served as an electrophile for the alkylation of methyl cyanoacetate, the product of which yielded 12 upon subsequent methylation. The ester group of 12 was then selectively reduced to provide the corresponding alcohol (13). Swern oxidation of 13 followed by homologation of the resulting aldehyde with the Ohira-Bestmann reagent (14) gave alkyne 15. Subsequent transformation of 15 to iodide 17 proceeded via a sequence involving silyl ether cleavage to yield 16, followed by conversion of the hydroxy group to the corresponding mesylate and displacement with sodium iodide.

Scheme 2.

Synthesis of alkyl iodide 17

At this stage, the synthesis of indanone 22 (Scheme 3), a precursor to the other coupling partner required for the construction of 9, was pursued.⁹

Scheme 3.

Synthesis of indanone 22

The sequence began with known benzyl alcohol 18,^6,7 which was oxidized under Parikh-Doering conditions¹⁰ to afford aldehyde 19. Horner-Wadsworth-Emmons homologation of 19 afforded enoate 20, which following hydrogenation and saponification gave acid 21. At this stage, the carboxylic acid group was converted to the corresponding acid chloride, and an ensuing Friedel-Crafts acylation gave indanone 22.

With indanone 22 in hand, we next investigated its conversion to indene substrate 9 as outlined in Scheme 4. Claisen condensation of 22 and dimethyl carbonate provided a β-ketoester, which upon alkylation with 17 gave alkyne 23. A subsequent saponification/decarboxylation sequence afforded indanone 24. Selective reduction of the carbonyl group in the presence of the nitrile group was accomplished with NaBH₄ and was followed by elimination of the resulting hydroxyl group to afford indene 9.

Scheme 4.

Synthesis of indene 9

On the basis of our previous studies, we first investigated the cycloisomerization of alkynyl indene 9 to cycloheptadiene 8 using Ga(III) salts (entries 1–6, Table 1). We found that GaCl₃, GaBr₃ or GaI₃ worked comparably in the cycloisomerization reaction with a 0.20 equiv catalyst loading at 40 °C in benzene (entries 1–3), resulting in ca. 40% conversion to 8 over 48 h. In the cases where GaI₃ was used to mediate the cycloisomerization, increasing the temperature to 65 °C (with 0.5 equiv of GaI₃; entry 4) or 80 °C (with 0.25 equiv of GaI₃; entry 5) led to complete conversion after 96 h. After a systematic investigation of various combinations of Ga(III) salts, solvents and temperatures, an optimal set of conditions was identified (0.25 equiv of GaCl₃, 100 °C, 48 h; entry 6), which gave 8 in 91% isolated yield. Interestingly, the use of PtCl₂ or InCl₃, which had been previously established as catalysts for enyne cycloisomerization,¹¹ returned only the starting material.

Table 1.

Cycloisomerization of 9

entry	catalyst	equiv	temp (°C)	time (h)	% Conversiona
1	GaCl3	0.2	40	48	41
2	GaBr3	0.2	40	48	43
3	GaI3	0.2	40	48	38
4	GaI3	0.5	65	96	100
5	GaI3	0.25	80	96	100
6	GaCl3	0.25	100b	48	100
7	PtCl2	0.2	40	72	NR
8	InCl3	0.2	40	120	NR

^aConversion of 9 based on ¹H NMR of crude reaction product following standard work-up.

^bReaction was run in toluene.

With benzannulated cycloheptadiene 8 in hand, our attention turned to the functionalization of the conjugated diene moiety. We first investigated the hydrolysis of the nitrile functional group. In this regard, we found that the transformation was best effected using Ghaffar and Parkins’ platinum catalyst (25, Scheme 5),¹² which was uniquely effective in giving primary amide 26 in 87% yield. Subjecting 26 to standard iodolactonization conditions, followed by IBX oxidation, led to the isolation of lactone 27, the structure of which was confirmed by X-ray crystallography (see ORTEP representation in Scheme 5). Although 27 was undesired with respect to our efforts toward icetexone and epi-icetexone, it provides a potential entry to the synthesis of anastomosine (4) and related natural products.

Scheme 5.

Initial iodolactonization studies

Following careful optimization studies, we established a route, as outlined in Scheme 6, which leads to the late-stage intermediates 7a and 7b for the synthesis of icetexone and epi-icetexone, respectively. The sequence begins with the diastereoselective epoxidation of cycloheptadiene 26 to afford 28 in 81% yield. Treatment of 28 with camphorsulfonic acid and tosylhydrazide in benzene at 80 °C gave a 2.5:1 mixture of 7a and 7b in a combined 42% yield. In this key reductive condensation event, the permethylated hydroquinone precursors to icetexone and epi-icetexone are formed in a single pot cascade transformation.

Scheme 6.

Synthesis of icetexone and epi-icetexone precursors 7a and 7b

Tetracycles 7a and 7b may arise via the postulated sequence outlined in Scheme 7. The transformation commences with protonation of the epoxide group of 28 and subsequent opening to afford allylic cation 29, which is trapped by tosylhydrazide to afford 30. This trapping step ultimately determines the ratio of 7a to 7b that is obtained (vide infra).¹³ From 30, the icetexone and epi-icetexone precursors 7a and 7b may arise via a sequence involving (A) cyclization with accompanying loss of a molecule of NH₃ to install the bridging lactone and (B) loss of p-toluenesulfinic acid to afford diazene 31, which undergoes a diazene decomposition¹⁴ with accompanying double bond transposition and loss of dinitrogen. The stereospecific nature of the diazene rearrangement leads to the observed mixture of 7a and 7b.

Scheme 7.

Proposed mechanism for the formation of 7a and 7b from 28

Support for the proposed formation of 7a and 7b from 28 was gained from the observations detailed in Scheme 8. Treating epoxide 28 with camphorsulfonic acid in wet CH₂Cl₂ yielded 32 as a single diastereomer.¹⁵ Subjection of 32 to camphorsulfonic acid and tosylhydrazide in benzene at 80 °C for 15 h gave 7b as the major product (>10:1 dr of 7b:7a). Presumably, attack of tosylhydrazide occurs with high diastereocontrol from the convex face of the rigid, polycyclic, allylic cation generated from 32 to give 33. Following the loss of p-toluenesulfinic acid, stereospecific diazene rearrangement yields 7 b as the major product. Because Majetich and Grove have previously employed 7b in the synthesis of epi-icetexone, this sequence completes a formal synthesis of this natural product. Furthermore, because 7a was also applied in the synthesis of icetexone by Majetich and Grove, we have also achieved a formal synthesis of this natural product.

Scheme 8.

Synthesis of 7b

In conclusion, we have applied a Ga(III)-catalyzed cycloisomerization reaction to the synthesis of a benzannulated cycloheptadiene bearing a cyano group that serves as a key intermediate in the synthesis of the icetexane diterpenoids icetexone and epi-icetexone. The synthesis features a late-stage cascade sequence that constructs the tetracyclic core of these natural products.

Because the syntheses of compounds closely related to 17 are known in enantioenriched form,¹⁶ the strategy outlined here should be readily applied to the enantioselective syntheses of 5 and 6. Our current efforts are focused on this pursuit and the application of cycloheptadiene intermediates such as 8 to the synthesis of other natural products.