Synthesis of Novel Macroheterocycles Through Intramolecular Oxidative Coupling of Furanoid β-Ketoesters

Furanocembranoids are an intriguing class of marine diterpenoids characterized by a C₁₄ cembrane skeleton with an embedded furan heterocycle.^[1] Derived from these structures are various more advanced secondary metabolites arising through biooxidation of their furan moieties into more complex cembranoids which often undergo transannular transformations involving carbon–carbon and/or carbon–oxygen bond formations.^[1] Bielschowskysin (1),^[2] verrillin (2),^[3] and plumarellide (3)^[4] are three such molecules (Figure 1a). Their origins can be traced back to exo-enol ether/cyclic hemiacetal precursors 4, 5, and 6, respectively, as shown in Figure 1a. Methods for the synthesis of the latter precursors are, therefore, crucial to testing such biosynthetic hypotheses and to synthetic strategies toward these natural products. In addition to the cytotoxic and antimalarial properties that many furanocembranoids exhibit,^[1a] these macrocyclic exo-enol ether/cyclic hemiacetal structural motifs (see A, Figure 1b) may be of interest to chemical biology and medicinal chemistry as novel molecular diversity whose biological properties remain largely unexplored.

Figure 1.

a) exo-Enol ether/cyclic acetal derived furanocembranoids (1, 2, and 3) and their hypothetical biosynthetic precursors (4, 5, and 6) and b) exo-enol ether/cyclic hemiacetal structural motif A.

As part of a program directed toward the synthesis of complex furanocembranoids, we recently reported^[5] a synthesis of a bielschowskysin model system that featured an intermolecular oxidative coupling^[6,7] of a furan and a β-ketoester as a means to form the exo-enol ether/cyclic acetal functionality.^[8] Herein we describe the intramolecular version of this reaction which can be fine-tuned to provide either the monomeric or dimeric macrocyclic systems. A salient feature of this novel macrocyclization is its tolerance for various functional groups and structural motifs, including certain nitrogen and oxygen heterocycles.

Scheme 1 summarizes the mechanistic rationale^[5a,7] for the designed ring closure of furan β-ketoester substrates (I) to the desired macrocyclic exo-enol ether/cyclic acetals (VIII) via postulated intermediate species II–VII, in the presence of cerium ammonium nitrate (CAN) as an oxidant and methanol as a nucleophile.

Scheme 1.

Mechanistic rationale for the oxidative macrocyclization of furan β-ketoester I to macrocycle VIII.

Table 1 shows our investigations to optimize conditions for the formation of the desired macrocyclic exo-enol ether/cyclic acetal system 8 from the corresponding substrate, furan β-ketoester 7. Under the initially employed conditions of adding a slight excess of oxidant (i.e. CAN) to a solution of substrate 7 in methanol (0.05 M, entry 1) at 0 °C, it was striking to observe the formation of the dimeric 22-membered macrocyclic product 9 in 45% yield (ca. 1:1 diastereomeric ratio) as the only isolated product. Carrying out the reaction under high dilution conditions (i.e. 0.001 M, entry 5) led to the isolation of the desired monomer 8 in 25% yield as a single olefinic isomer. Based on the observation that these reactions were extremely rapid, we reasoned that slow addition of the substrate to a solution of CAN in methanol would achieve even higher dilution conditions, thereby increasing the yield of 8. We therefore resorted to a syringe pump addition mode of substrate 7 to a methanolic solution of CAN. Indeed, under such extremely high dilution conditions (0.05 M methanolic solution of 7 added to a 0.23 M methanolic solution of CAN, at 0 °C, over 1.5 h) cyclic product 8 was obtained in 55% yield (entry 6). Several other commercially available cerium(IV) complexes, as well as other known single electron oxidants, failed to surpass CAN as the reagent of choice in terms of efficiency (entries 3–4, 7–11). It is noteworthy that the manganese(III),^[9] iron(III),^[10] and cobalt(II)^[11] complexes shown in entries 8–11 failed to produce any of the monomeric or dimeric macrocycles.

Table 1.

Screening of conditions.^[a]

Entry	Oxidant	Conc.	Temp.	Time	Yield[%]
1	Ce(NO3)6(NH4)2	0.05M	0 °C	15 min	45 (9)[b]
2	Ce(NO3)6(NH4)2	0.05M	−78-0 °C	60 min	42 (9)[b]
3	Ce(SO4)2·8H2O	0.05 M	0 °C	30 min	11 (9)[b]
4	Ce(NH4)4(SO4)4	0.05 M	0 °C	15 min	12 (9)[b]
5[d]	Ce(NO3)6(NH4)2	0.001 M	0 °C	60 min	25 (8)[c]
6[e]	Ce(NO3)6(NH4)2	0.025 M	0 °C	90 min	55 (8)[c]
7	Ce(OTf)4	0.05 M	0-25 °C	18 h	n.d.[f]
8	Mn(OAc)3	0.05 M	0-60 °C	48 h	n.d.[f]
9	Mn(OAc)3/Cu(OTf)2	0.05M	0-60 °C	18 h	n.d.[f]
10	(Cp2Fe)PF6	0.05M	0-25 °C	18 h	n.d.[f]
11	Co(OAc)2	0.05M	0-60 °C	18 h	n.r.[g]

^aReactions were performed on ca. 50 mg scale

^byield refers to spectroscopically pure material (ca. 1:1 mixture of diastereomers)

^cyield refers to spectroscopically and chromatographically homogenous material

^dthe substrate in methanol (0.01 M) was added into a solution of CAN in methanol (0.001 M)

^ethe substrate in methanol (0.05 M) was added via syringe pump over 1.5 h to a solution of the oxidant at 0 °C

^fnot detected by TLC or crude ¹H NMR analysis

^gno reaction observed.

In order to explore the generality and scope of this macrocyclization reaction, we constructed a range of substrates and subjected them first to the optimized conditions for monomer formation, and then for dimer formation. Table 2 summarizes the results of these investigations leading to a variety of macrocyclic systems containing the exo-enol ether furanoid structural motif. Not surprising, but of interest, was the failure of substrate 10 to furnish any of the corresponding 9-membered ring monomer (i.e. 21, entry 1) under the high dilution conditions, undoubtedly due to the severe strain and transannular interactions associated with this medium-sized ring. Other medium-sized rings (10–13-membered) notable for their intransigence to be formed (i.e. 8, 22–24, entries 2–5) were generated, however, in low to respectable yields as shown in Table 2. Gratifyingly, compound 8 yielded crystals suitable for X-ray crystallographic analysis [m.p. = 93 – 95 °C (CHCl₃), see ORTEP, Figure 2], thus confirming its connectivity as well as its E enol ether configuration.^[12] Larger rings (i.e. 14- and 15-membered) were formed in even higher yields, as we expected (compounds 25 and 26, 71% and 76%, respectively, entries 6 and 7). The introduction of an acetylenic bond in the aliphatic tether of the precursor leading to the 13-membered ring (i.e. 24, 53°% yield, entry 5) had a detrimental effect on the cyclization yield, leading to product 28 (11%, entry 9), presumably due to the strain associated with the acetylenic linkage and/or oxidation of the methylene between the furan and acetylene moieties. However, the acetylene functionality had no significant effect on the cyclization leading to the 15-membered macrocycle 29 (75% yield, entry 10) as compared to the saturated chain counterpart (i.e. 26, 76% yield, entry 7). Similarly, the introduction of a cis olefinic bond within the aliphatic linker as shown in substrate 19 (entry 11) had little effect on the efficiency of the ring closure (71% yield of 30 vs. 76% yield of 26, entry 7). Additional oxygenation in the backbone was also tolerated, as evidenced by the formation of acetoxy macrocycle 31 (72% yield, entry 12).^[13]

Table 2.

Synthesis of monomeric exo-enol ether/cyclic acetals.^[a]

Entry	Substrate	Product	Yield [%][b]
1			0
2			21
3			55
4			57
5			53
6			71
7			76
8[c]			40
9			11
10			75
11			71
12[d]			72[e]

^aReactions were performed on ca. 50 mg scale as follows: a methanolic solution of the substrate (0.05 M) was added via syringe pump over ca. 1.5 h to a 0 °C methanolic solution of CAN (4.5 equiv, 0.23 M)

^byields refer to spectroscopically and chromatographically homogenous material unless otherwise noted

^cthe substrate was added as a solution in methanol:CH₂Cl₂ 5:3 (for solubility purposes)

^dreaction was performed on 35 mg of substrate

^eyield refers to spectroscopically pure material (ca. 3:2 mixture of diastereomers).

Figure 2.

X-ray derived ORTEP of exo-enol ether 8. Thermal ellipsoids at 30% probability.

Encouraged by these results, we proceeded to examine the tolerance of heterocycles within our substrates to the reaction conditions. Scheme 2 shows the results of macrocyclization of butenolide-containing substrates 32a (R = Me) and 32b (R = tBu). We anticipated that ring closure of these substrates might be more challenging than their saturated counterparts (see 13, Table 2, entry 5), due to an additional ring within the carbon framework. We were pleased to find in a preliminary experiment that ring closure of 32a occurred, furnishing product 33a as a ca. 2:3 mixture of α- and β-methoxy isomers, of which the major β-isomer [33a(β)] was isolated in 15% yield. The latter compound crystallized from chloroform in suitable crystals for X-ray crystallographic analysis (m.p. = 188 – 190 °C, see ORTEP, Scheme 2b),^[12] which allowed the relative configuration of the newly generated stereocenter to be assigned as shown in 33a(β).^[14] Switching the ester from methyl to tert-butyl in the substrate (32b) led to an increased overall yield of the product [33b, 40%, ca. 1:2 α:β, chromatographically separated, (Scheme 2a)].

Scheme 2.

a) Macrocyclic exo-enol ether/cyclic acetal formation of butenolide furan β-ketoesters 32a and 32b and b) X-ray derived ORTEP representation of 33a(β). Thermal ellipsoids at 30% probability.

Scheme 3 shows the results of the ring closure of triazole-containing substrate 34 under both the high and low dilution conditions. Impressively, the high dilution conditions (syringe pump addition of substrate into CAN) led to the expected 15-membered macrocyclic system 35 in 91% yield. More strikingly, the high concentration conditions also led to the monomeric macrocycle 35 (75% yield) with no dimer being detected. Apparently, the overwhelming rigidification effect of the triazole ring on the backbone of precursor 34 renders rapid and exclusive ring closure to afford 15-membered ring macrocycle 35 in preference to intermolecular coupling. The latter compound (i.e. 35) crystallized from Et₂O:CH₂Cl₂ (10:1, m.p. = 142–144 °C) and was subjected to X-ray crystallographic analysis (see ORTEP, supporting information), thereby confirming its connectivity, enol ether configuration, and triazole-ring integrity under the reaction conditions.^[12]

Scheme 3.

Monomer formation of triazole substrate 34 under both monomer and dimer conditions

Having recognized the propensity of this reaction to form dimeric macrocyclic structures under certain conditions (i.e. 7→9, Table 1, entry 1), we opted to explore the generality of the dimerization process as a means to produce new molecular diversity. As seen from Table 3, substrates 7 and 10–16 reacted with CAN under high concentration conditions at 0 °C to afford dimeric macrocycles 9 and 36–42 (sizes 18-, 20-, 22-, 24-, 26-, 28-, 30-, and 40-membered) in yields ranging from 23–45%. Products 9 and 36–42 were obtained as ca. 1:1 diastereomeric mixtures.^[13] Gratifyingly, 26-membered macrocyclic dimer 39 crystallized from chloroform (m.p. = 130–132 °C), yielding crystals suitable for X-ray crystallographic analysis (see ORTEP, Figure 3) that reveal its anti arrangement and E configuration of the two enol ether olefinic bonds.^[12]

Table 3.

Synthesis of dimeric exo-enol ether/cyclic acetals.^[a]

Entry	Substrate	Product	Yield [%][b]
1	10		31
2	11		45
3	7		45
4	12		37
5	13		42
6	14		25
7[c]	15		26
8[d]	16		23

^aReactions were performed on ca. 50 mg scale as follows: CAN (4.5 equiv) was added to a 0 °C methanolic solution of the substrate (0.05 M) in one portion

^byields refer to spectroscopically pure material (ca. 1:1 mixture of diastereomers)

^creaction was performed on 100 mg of substrate

^dthe substrate was dissolved in methanol:CH₂Cl₂ (5:2) for solubility purposes.

Figure 3.

X-ray derived ORTEP of dimer 39-anti. Thermal ellipsoids at 30% probability.

In conclusion, a new cyclization method was developed that renders macrocycles equipped with the exo-enol ether/cyclic acetal structural motif, and their dimeric counterparts, readily available from simple furanoid β-ketoester substrates. The range of products obtained through this novel cyclization include heterocycles that are relevant to biosynthetic hypotheses and total synthesis strategies for the cembranoid family of natural products as well as chemical biology and medicinal chemistry investigations.