Stereocontrolled Routes to 4-Methoxypentadienoates for Use in Natural Product Synthesis

Abstract

Mild and efficient routes to (E,E)- and (E,Z)-4-methoxypentadienoic acid esters from readily accessible γ,δ-epoxydienoates are described. The crucial epoxide methanolysis can be carried out in stereocomplementary ways by the use of either acid-mediated or palladium-catalysed procedures, the latter procedure proving ­preferable in most cases.

JBIR-23 (1) and JBIR-24 (2) are two novel natural products which were isolated in 2009 from Streptomyces sp. AK-AB27 and shown to possess promising anticancer activity against an aggressive form of lung cancer, malignant pleural mesothelioma (Figure [1]).[ ¹ ] The only structurally related natural products are cuevaene A (3) and B (4) which were reported in 2000 by Gräfe et al.[ ² ] also having been isolated from Streptomyces and shown to possess moderate activity against Gram-positive bacteria. Compounds 1–4 all contain the same (E,Z,E)-4-methoxy-6-methylheptatrienoic acid side chain which, to the best of our knowledge, is unique to this small group of natural products.

(±)-Cuevaene A (3) has recently been synthesised by Liu et al.[ ³ ] and by our own group[ ⁴ ] with both synthetic routes employing a conventional Wittig-based procedure to install the alkenes in a stepwise manner. Whilst successful, this linear strategy was rather lengthy and the Horner–Wadsworth–Emmons reaction using methyl(dimethoxyphosphoryl)(methoxy)acetate employed to install the key methoxy enol ether proved rather capricious.

As part of our ongoing interest in the total synthesis of natural products 1–4, we have investigated alternative, more efficient procedures to prepare building blocks for the construction of methoxylated polyene side chains. Herein we report two complementary and straightforward procedures for the stereocontrolled conversion of dienoates 5 into a range of (E,E)- and (E,Z)-4-methoxypentadienoates (E,E)-6 or (E,Z)-6 (Scheme [1])

The starting point for this investigation was a report by Watt et al.[ ⁵ ] in 1983 in which they converted γ,δ-epoxy­dienoate 7 into ethoxy dienoate 9 (Scheme [2]).

The acid-promoted ring opening of epoxide 7 with ethanol afforded ethoxy alcohol 8 (stereochemistry not defined) in good yield. Treatment of alcohol 8 with mesyl chloride and subsequent elimination of the intermediate mesylate with t-BuOK delivered, in moderate yield, the desired 4-ethoxy-diene 9. Watt et al.[ ⁵ ] did not comment on the stereoselectivity of this process, but the reported data indicates that only one product was formed, suggesting that the process was diastereoselective. Given these results, we envisioned that the analogous epoxide ring opening using methanol, rather than ethanol, should deliver the methoxydienes required for our natural product studies. In this publication, we describe the successful implementation of this strategy.

The initial studies were carried out using (E)-trans-11a, readily obtained by modification of a literature procedure[ ^6a ] from methyl hepta-(2E,4E)-dienoate (E,E)-10 [ ^6b ] (Scheme [3]). Treatment of the epoxide (E)-trans-11a with methanol and concentrated sulfuric acid, delivered the expected methoxy alcohol as a single diastereomer, believed to be (E)-anti-12a based on S_N2 ring opening. Subsequent mesylation under standard conditions and elimination using DBU, which in our hands proved to be more efficient than t-BuOK, delivered the desired (E,E)-methoxy-diene (E,E)-13a (containing at most 8% of the E,Z isomer) in good yield. The stereochemistry of (E,E)-13a was confirmed by NOE NMR spectroscopic experiments (see later) and is consistent with S_N2 epoxide ring opening followed by a DBU-promoted E2 elimination as shown in Scheme [3].

Of course, the main targets for the natural product study were 4-methoxyhepta-(2E,4Z)-dienoates and therefore it was important to establish the stereoselectivity of this ­sequence (Scheme [4]). Therefore, methyl hepta-(2E,4Z)-di-enoate (E,Z)-10 [ ^6a ] was converted into epoxide (E)-cis-11a by treatment with MCPBA.[ ^6b ] Subsequent acid-catalysed methanolysis to give (E)-syn-12a followed by mesylation and treatment with DBU delivered predominantly the expected product diene (E,Z)-13a (E,Z/E,E = 88:12). We attribute the stereochemical leakage to a competing E1cb pathway; in all cases the E/Z-mixtures were inseparable using conventional chromatography.

NMR spectroscopy and NOE experiments were employed to confirm the structures of dienes (E,E)- and (E,Z)-13a (Figure [2]) as shown in Scheme [5] and Table [1].

Having established successful conditions for the formation of dienes (E,E)-13a and (E,Z)-13a, we went on to investigate the scope of the process in terms of the terminal substituent (Scheme [5] and Table [2]).

Table 2 Acid-Mediated Epoxide Ring-Opening Sequence Leading to 4-Methoxypentadienoates 13 ^a–c

Entry	Epoxide	Alcohol	Diene
1	(E)-trans-11a	(E)-anti-12a 66%, 2 h	(E,E)-13a 76%, 5 h E,E/E,Z = 92:8
2	(E)-cis-11a	(E)-syn-12a 76%, 1 h	(E,Z)-13a61%, 2 h E,Z/E,E = 88:12
3	(E,trans-11b	(E)-anti-12b 80%, 2 h	(E,E)-13b 56%, 48 h E,E/E,Z = 72:28
4	(E)-cis-11c	14 33%, 0.5 h	not attempted
5	(E)-trans-11d	(E)-anti-12d 63%, 16 h	failed

^a Epoxide-opening reactions were performed with concd H₂SO₄ in 0.1 M MeOH at r.t. for the indicated time.

^b Mesylation reactions were performed with MsCl (1.5 equiv) and Et₃N (3.0 equiv) in 0.1 M CH₂Cl₂ at r.t. for 2 h.

^c Elimination reactions were performed with DBU (1.2 equiv) in 0.1 M CH₂Cl₂ at r.t. for the indicated time.

However, although it proved possible to vary the type of terminal aliphatic substituent [from ethyl (Table [2], entries 1 and 2) to cyclohexyl (Table [2], entry 3), for example], this sequence became problematic with functionalised aliphatic and aromatic substrates (Table [2], entries 4 and 5). Thus, unsurprisingly, acid-sensitive groups are not well-tolerated in the acid-catalysed ring-opening process; ­epoxide (E,cis)-11c containing a TBS-protected primary alcohol underwent deprotection and further degradation in addition to the expected methanolysis (Table [2], entry 4). A different problem was encountered with the phenyl analogue (E,trans)-11d; in this case, the desired epoxide ring opening proceeded smoothly to deliver alcohol (E,anti)-12d in good yield, but on treatment with MsCl/DBU no elimination was observed (Table [2], entry 5), possibly for steric reasons.

In order to overcome the above problems, particularly with acid-sensitive substrates, alternative, milder procedures were considered for the epoxide ring-opening step. In fact, such a procedure was readily available, as in 2008 Miyashita’s group reported that treatment of γ,δ-epoxy­dienoate epoxides with B(OR)₃ and catalytic Pd(PPh₃)₄ produced alkoxy alcohols stereoselectively and in excellent yield; thus, reaction of epoxide 15 with B(OMe)₃ gave methoxy alcohol 16 stereoselectively and in excellent yield, presumably via the intermediacy of the π-allyl palladium(II) species 17 (Scheme [6]).[ ⁷ ]

We therefore decided to investigate this palladium-­catalysed epoxide-opening procedure in combination with the DBU-mediated elimination in order to expand the range of 4-methoxypentadienoates that could be accessed. This study was successful, and the results are summarised in Table [3].

The first point to note is that the Miyashita palladium-catalysed sequence has the opposite stereoselectivity to the acid-mediated variant. Thus, under acid-mediated conditions, epoxide (E)-trans-11a gives diene (E,E)-13a and (E)-cis-11a gives diene (E,Z)-13a (Table [2], entries 1 and 2) whereas under palladium-catalysed conditions, epoxide (E)-trans-11a gives diene (E,Z)-13a and (E)-cis-11a gives diene (E,E)-13a (Scheme [7, ]Table [3], entries 1 and 2). The rationale for this stereocomplementarity is straightforward – with acid catalysis, epoxide ring opening by methanol occurs by an S_N2 mechanism whereas with palladium catalysis a ‘double inversion’ takes place via a π-allyl palladium(II) intermediate giving the opposite diastereomer of the methoxy alcohol.

This stereocomplementarity is valuable where only one diastereomeric epoxide is readily available; for example, the cyclohexyl system (E)-trans-11b gives (E,E)-13b with the acid-mediated route (Table [2], entry 3) and (E,Z)-13b with the palladium-mediated route (Table [3], entry 3). Again, we attribute the stereochemical leakage to a competing E1cb pathway.

It should also be noted that, when comparisons can be made (Table [3], entries 1–3), the combined yields for the palladium-catalysed sequence are comparable or slightly higher than those for the acid-mediated sequence. The main advantage of the palladium-catalysed route, however, is that the mild nature of the procedure leads to a considerable expansion of its scope being applicable to several acid-sensitive substrates (Table [3],entries 4 and 5). A further point of interest is that (E)-syn-12d undergoes the expected E2 elimination to give diene (E,Z)-13d (Table [3], entry 7) whereas the diastereomeric alcohol (E)-anti-12d proved resistant to elimination (Table [2], entry 5). This slightly surprising observation presumably reflects the enhanced steric hindrance in the latter case (antiperiplanar elimination positions the bulky acrylate group gauche to both the mesylate and phenyl substituents).

Table 3 Pd(0)-Catalysed Epoxide Ring-Opening Sequence Leading to 4-Methoxypentadienoates 13 ^a–c

Entry	Epoxide	Alcohol	Diene
1	(E)-trans-11a	(E)-syn-12a 82%, 1 h	(E,Z)-13a 61%, 2 h E,Z/E,E = 92:8
2	(E)-cis-11a	(E)-anti-12a 80%, 1 h	(E,E)-13a 76%, 2 h E,E/E,Z = 88:12
3	(E)-trans-11b	(E)-syn-12b 80%, 2 h	(E,Z)-13b 62%, 2 h E,Z/E,E = 83:17
4	(E)-cis-11c	(E)-anti-12c 82%, 1 h	(E,E)-13c 90%, 2 h E,E/E,Z = >95:<5
5	(E)-trans-11c	(E)-syn-12c 57%, 1 h	(E,Z)-13c 69%, 2 h E,Z/E,E = >95:<5
6	(E)-cis-11e	(E)-anti-12e 52%, 2 h	(E,E)-13e 84%, 7 h E,E/E,Z = >95:<5
7	(E)-trans-11d	(E)-syn-12d 84%, 1 h	(E,Z)-13d 72%, 2 h E,Z/E,E = >95:<5

^a Epoxide-opening reactions were performed with Pd(PPh₃)₄ (10 mol%) and B(OMe)₃ (1.2 equiv) in 0.2 M THF at r.t. for the indicated time.

^b Mesylation reactions were performed with MsCl (1.5 equiv) and Et₃N (3.0 equiv) in 0.1 M CH₂Cl₂ at r.t. for 2 h.

^c Elimination reactions were performed with DBU (1.2 equiv) in 0.1 M CH₂Cl₂ at r.t. for the indicated time. PT = 1-phenyl-1H-tetrazo-5-yl.

In conclusion, we developed two methods, one acid-mediated and the other palladium-catalysed, for the synthesis of 4-methoxypentadienoic esters from readily available epoxide starting materials. The two methods exhibit complementary stereoselectivity to deliver the resulting dienes predominantly as single isomers.[8] [9] [10] [11] [12] We are currently exploiting the methodology described herein in natural product synthesis.

Acknowledgment

We are grateful to Ms. Heather Fish (NMR), Dr. Karl Heaton (MS), and Dr. Graeme McAllister (microanalysis) for technical assistance. We would also like to thank the University of York and Elsevier for postgraduate support (P.G.E.C.) and to the Society of Chemical ­Industry for additional Scholarship funding.

• References and Notes

• 1 Motohashi K, Hwang J.-H, Sekido Y, Takagi M, Shin-ya K. Org. Lett. 2009; 11: 285
  CrossrefSearch in Google Scholar
• 2 Schlegel B, Groth I, Gräfe U. J. Antibiot. 2000; 53: 425
  Search in Google Scholar
• 3 Chen Y, Huang J, Liu B. Tetrahedron Lett. 2010; 51: 4655
  CrossrefSearch in Google Scholar
• 4 Craven PG. E, Taylor RJ. K. Tetrahedron Lett. 2012; 53: 5422
  CrossrefSearch in Google Scholar
• 5 Voyle M, Kyler KS, Arseniyadis S, Dunlap NK, Watt DS. J. Org. Chem. 1983; 48: 470
  CrossrefSearch in Google Scholar
• 6a Tsuboi S, Masuda T, Takeda A. J. Org. Chem. 1982; 47: 4478
  CrossrefSearch in Google Scholar
• 6b Uchida K, Ishigami K, Watanabe H, Kitahara T. Tetrahedron 2007; 63: 1281
  CrossrefSearch in Google Scholar
• 7 Yu X.-Q, Yoshimura F, Ito F, Susaki M, Hirai A, Tanino K, Miyashita M. Angew. Chem. Int. Ed. 2008; 47: 750
  CrossrefSearch in Google Scholar
• 8 General Procedure for the Acid-Promoted Epoxide Ring Opening To a stirred solution of epoxide (1.0 equiv) in MeOH (0.1 M) was added a catalytic amount of concd H₂SO₄. The resulting solution was stirred at r.t. until complete as monitored by TLC. After this time, the reaction mixture was quenched with brine and extracted with EtOAc. The combined organic layers were dried (MgSO₄), filtered, and concentrated in vacuo. The resulting crude material was purified by flash column chromatography to afford the desired alcohol.
• 9 General Procedure for the Pd(0)-Mediated Epoxide Ring Opening To a stirred solution of epoxide (1.0 equiv) and B(OMe)₃ (1.2 equiv) in THF (0.2 M) at r.t. was added Pd(PPh₃)₄ (10 mol%). The resulting solution was stirred at r.t. until complete as monitored by TLC. After this time, the reaction mixture was quenched with NaHCO₃ (sat. aq solution) and diluted with EtOAc. The aqueous layer was extracted with EtOAc (3 ×). The combined organics were washed with brine, dried (MgSO₄), filtered, and concentrated in vacuo. The resulting crude material was purified by flash column chromatography to afford the desired alcohol.
• 10 General Procedure for the Mesylation and Elimination Sequence To a stirred solution of alcohol (1.0 equiv) in CH₂Cl₂ (0.15 M) at r.t. was added MsCl (1.5 equiv) followed by Et₃N (3.0 equiv). The resulting solution was stirred at r.t. for 2 h. After this time, the reaction mixture was diluted with brine (20 mL). The aqueous layer was extracted with CH₂Cl₂ (2×). The combined organic layers were dried (MgSO₄), filtered, and concentrated in vacuo. The resulting crude mesylate (1.0 equiv) was redissolved in CH₂Cl₂ (0.1 M), and DBU (1.2 equiv) was added at r.t. The resulting solution was stirred at r.t. until complete as monitored by TLC. After this time, the reaction mixture was quenched with 10% HCl and diluted with H₂O and CH₂Cl₂, and the layers were separated. The aqueous layer was extracted with CH₂Cl₂ (2 ×). The combined organic layers were washed with brine, dried (MgSO₄), filtered, and concentrated in vacuo. The resulting crude material was purified by flash column chromatography to afford the desired diene.
• 11 Methyl (2E,4E)-4-Methoxyhepta-2,4-dienoate [(E,E)-13a] IR (film): ν_max = 1712 (C=O), 1645 (C=C) cm^–1. ¹H NMR (400 MHz, CDCl₃): δ = 1.03 (3 H, t, J = 7.5 Hz), 2.20–2.28 (2 H, m), 3.57 (3 H, s), 3.74 (3 H, s), 4.98 (1 H, t, J = 7.9 Hz), 6.21 (1 H, d, J = 15.0 Hz), 7.45 (1 H, d, J = 15.0 Hz) ppm. ¹³C NMR (100 MHz, CDCl₃): δ = 15.2, 19.7, 54.3, 59.8, 111.9, 117.6, 135.0, 150.0, 167.4 ppm. MS: m/z = 171 [MH]⁺, 193 [MNa]⁺. ESI-HRMS: m/z calcd for C₉H₁₅O₃: 171.1016; found [MH⁺]: 171.1013 (1.5 ppm error).
• 12 Methyl (2E,4Z)-4-Methoxyhepta-2,4-dienoate [(E,Z)-13a] IR (film): ν_max = 1715 (C=O), 1639 (C=C) cm^–1. ¹H NMR (400 MHz, CDCl₃): δ = 1.04 (3 H, t, J = 7.6 Hz), 2.26 (2 H, q, J = 7.6 Hz), 3.62 (3 H, s), 3.76 (3 H, s), 5.39 (1 H, t, J = 7.6 Hz), 6.02 (1 H, d, J = 15.5 Hz), 7.03 (1 H, d, J = 15.5 Hz) ppm. ¹³C NMR (100 MHz, CDCl₃): δ = 13.4, 19.3, 51.4, 59.8, 116.0, 130.3, 140.9, 152.7, 167.4 ppm. MS: m/z = 171 [MH]⁺, 193 [MNa]⁺. ESI-HRMS: m/z calcd for C₉H₁₅O₃: 171.1016; found [MH⁺]: 171.1013 (1.5 ppm error).

• References and Notes

• 1 Motohashi K, Hwang J.-H, Sekido Y, Takagi M, Shin-ya K. Org. Lett. 2009; 11: 285
  CrossrefSearch in Google Scholar
• 2 Schlegel B, Groth I, Gräfe U. J. Antibiot. 2000; 53: 425
  Search in Google Scholar
• 3 Chen Y, Huang J, Liu B. Tetrahedron Lett. 2010; 51: 4655
  CrossrefSearch in Google Scholar
• 4 Craven PG. E, Taylor RJ. K. Tetrahedron Lett. 2012; 53: 5422
  CrossrefSearch in Google Scholar
• 5 Voyle M, Kyler KS, Arseniyadis S, Dunlap NK, Watt DS. J. Org. Chem. 1983; 48: 470
  CrossrefSearch in Google Scholar
• 6a Tsuboi S, Masuda T, Takeda A. J. Org. Chem. 1982; 47: 4478
  CrossrefSearch in Google Scholar
• 6b Uchida K, Ishigami K, Watanabe H, Kitahara T. Tetrahedron 2007; 63: 1281
  CrossrefSearch in Google Scholar
• 7 Yu X.-Q, Yoshimura F, Ito F, Susaki M, Hirai A, Tanino K, Miyashita M. Angew. Chem. Int. Ed. 2008; 47: 750
  CrossrefSearch in Google Scholar
• 8 General Procedure for the Acid-Promoted Epoxide Ring Opening To a stirred solution of epoxide (1.0 equiv) in MeOH (0.1 M) was added a catalytic amount of concd H₂SO₄. The resulting solution was stirred at r.t. until complete as monitored by TLC. After this time, the reaction mixture was quenched with brine and extracted with EtOAc. The combined organic layers were dried (MgSO₄), filtered, and concentrated in vacuo. The resulting crude material was purified by flash column chromatography to afford the desired alcohol.
• 9 General Procedure for the Pd(0)-Mediated Epoxide Ring Opening To a stirred solution of epoxide (1.0 equiv) and B(OMe)₃ (1.2 equiv) in THF (0.2 M) at r.t. was added Pd(PPh₃)₄ (10 mol%). The resulting solution was stirred at r.t. until complete as monitored by TLC. After this time, the reaction mixture was quenched with NaHCO₃ (sat. aq solution) and diluted with EtOAc. The aqueous layer was extracted with EtOAc (3 ×). The combined organics were washed with brine, dried (MgSO₄), filtered, and concentrated in vacuo. The resulting crude material was purified by flash column chromatography to afford the desired alcohol.
• 10 General Procedure for the Mesylation and Elimination Sequence To a stirred solution of alcohol (1.0 equiv) in CH₂Cl₂ (0.15 M) at r.t. was added MsCl (1.5 equiv) followed by Et₃N (3.0 equiv). The resulting solution was stirred at r.t. for 2 h. After this time, the reaction mixture was diluted with brine (20 mL). The aqueous layer was extracted with CH₂Cl₂ (2×). The combined organic layers were dried (MgSO₄), filtered, and concentrated in vacuo. The resulting crude mesylate (1.0 equiv) was redissolved in CH₂Cl₂ (0.1 M), and DBU (1.2 equiv) was added at r.t. The resulting solution was stirred at r.t. until complete as monitored by TLC. After this time, the reaction mixture was quenched with 10% HCl and diluted with H₂O and CH₂Cl₂, and the layers were separated. The aqueous layer was extracted with CH₂Cl₂ (2 ×). The combined organic layers were washed with brine, dried (MgSO₄), filtered, and concentrated in vacuo. The resulting crude material was purified by flash column chromatography to afford the desired diene.
• 11 Methyl (2E,4E)-4-Methoxyhepta-2,4-dienoate [(E,E)-13a] IR (film): ν_max = 1712 (C=O), 1645 (C=C) cm^–1. ¹H NMR (400 MHz, CDCl₃): δ = 1.03 (3 H, t, J = 7.5 Hz), 2.20–2.28 (2 H, m), 3.57 (3 H, s), 3.74 (3 H, s), 4.98 (1 H, t, J = 7.9 Hz), 6.21 (1 H, d, J = 15.0 Hz), 7.45 (1 H, d, J = 15.0 Hz) ppm. ¹³C NMR (100 MHz, CDCl₃): δ = 15.2, 19.7, 54.3, 59.8, 111.9, 117.6, 135.0, 150.0, 167.4 ppm. MS: m/z = 171 [MH]⁺, 193 [MNa]⁺. ESI-HRMS: m/z calcd for C₉H₁₅O₃: 171.1016; found [MH⁺]: 171.1013 (1.5 ppm error).
• 12 Methyl (2E,4Z)-4-Methoxyhepta-2,4-dienoate [(E,Z)-13a] IR (film): ν_max = 1715 (C=O), 1639 (C=C) cm^–1. ¹H NMR (400 MHz, CDCl₃): δ = 1.04 (3 H, t, J = 7.6 Hz), 2.26 (2 H, q, J = 7.6 Hz), 3.62 (3 H, s), 3.76 (3 H, s), 5.39 (1 H, t, J = 7.6 Hz), 6.02 (1 H, d, J = 15.5 Hz), 7.03 (1 H, d, J = 15.5 Hz) ppm. ¹³C NMR (100 MHz, CDCl₃): δ = 13.4, 19.3, 51.4, 59.8, 116.0, 130.3, 140.9, 152.7, 167.4 ppm. MS: m/z = 171 [MH]⁺, 193 [MNa]⁺. ESI-HRMS: m/z calcd for C₉H₁₅O₃: 171.1016; found [MH⁺]: 171.1013 (1.5 ppm error).