Transforming terpene feedstock into polyketide architecture

Abstract

The Cu-catalyzed synthesis of skipped 1,4-dienes from allylic acetates and vinyl-Grignard reagents is key to bidirectional modifications of acyclic terpene acetates. As a result, trisubstituted double bond containing subunits can be readily transferred into complex polyketides from inexpensive bulk terpenes.

---

The focus on target-oriented synthesis of biologically active molecules is shifting toward more economic¹ and sustainable chemical processes.² Among many approaches³ to quantify synthetic efficiency, it has been recently suggested⁴ along the lines of Hendrickson's (1975) definition⁵ to use the quotient (construction reactions + strategic redox reactions)/total number of reaction steps as a yardstick for the “ideality” of a given synthesis. We feel that this definition undervalues the importance of minimizing construction steps, i.e. the overall number of C–C-bond forming reactions. Our group's research is dedicated to making larger substructures within natural products⁶ available from terpene feedstock,⁷fermentation or catalytic build-up reactions,⁸e.g. telomerizations.⁹

As a step toward this goal, we have embarked on the use of oxidized monoterpenes in conjunction with the development of efficient C–C bond forming reactions. Following this loose strategic blueprint, we recently completed the synthesis of englerin A¹⁰ from the epoxide of the readily available monoterpene nepetalactone (Fig. 1).

Fig. 1 Terpene epoxides and substructure analysis within ripostatin B.

In this communication, we would like to disclose a strategy for the conversion of epoxygeranyl acetate (1a) and related molecules into 1,4-dienes and representative subsequent conversions into polyketide building blocks using organocatalytic and transition metal catalyzed reactions.

In contrast to terpene biosynthesis which has a sudden increase in complexity through cationic cyclization events, polyketide complexity grows in a rather sequential fashion by two- and three-carbon building blocks. Instead of mimicking the intricate biosynthetic machinery of polyketide synthases, a potential shortcut lies in the possibility to disconnect polyketides into larger fragments. When pondering over the retrosynthesis of ripostatin B,¹¹ a potent RNA polymerase inhibitor,¹² we realized that 22 out of its 30 carbon atoms could—in principle—be derived from two molecules of epoxygeranyl acetate (blue), phenyl- and vinylmagnesium bromide (red) as starting materials. The key transformation in such an approach would be the coupling of the Grignard reagents with the allylic acetates.

While the substitution of allylic acetates with phenyl- and alkylmagnesium halides using Li₂CuCl₄ was precedented by Gansäuer et al.,¹³ surprisingly, apart from particular examples of Pd-catalyzed couplings with vinylstannanes,¹⁴ no general and scalable method for the synthesis of 1,4-dienes^15,16 (skipped dienes) from allylic acetates has been reported to our knowledge.

The use of those previously established catalytic conditions (Li₂CuCl₄ in THF) with vinylmagnesium bromide as the nucleophile met limited success as only 23% of the desired substitution product 2a was detected. After extensive optimization,¹⁷ we found that using copper(I)-bromide dimethylsulfide complex (20 mol% in THF) afforded 70% of the desired substitution product. The yield could be further improved by using 1.5 equivalents of vinylmagnesium bromide with copper(I)-iodide (20 mol%) as catalyst in a solvent mixture of tetrahydrofuran/dimethylsulfide (10 ∶ 1) giving rise to 88% of the skipped 1,4-diene 2a (Scheme 1).

Scheme 1 Optimized conditions for the synthesis of 1,4-dienes.

In order to probe the generality (functional group tolerance, integrity of the trisubstituted double bond configuration) of this reaction, we applied our best conditions to the vinyl substitution of a variety of allylic acetates (Table 1). Both epoxygeranyl and epoxyneryl acetate afforded the expected 1,4-dienes with complete retention of the configuration of the allylic double bond (entries 1 and 2) in good to excellent yields. In this instance, the epoxide serves as a handle in subsequent transformations, readily converted to an aldehyde moiety upon treatment with sodium periodate. In addition, their parent derivatives (entries 3 and 4) reacted smoothly to give the desired trienes. In the presence of non-allylic acetates, only those in the allylic position are substituted in good to moderate yield while the others remain unaffected (entries 5 to 8). Furthermore, the presence of free alcohols is well tolerated at the expense of one additional equivalent of Grignard reagent (entries 9 and 10).

Table 1 Investigation of the substrate scope

Entry	Substrate	Product	Yield^a (%)
a Isolated yields.
1			88
2			69
3			76
4			90
5			54
6			52
7			66
8			47
9			58
10			78

---

In order to showcase the bidirectional modification of epoxydiene 2a and its potential use in the synthesis of polyketides, we converted it into the C5–C15 fragment of ripostatin B (Scheme 2). The four-step synthetic sequence commenced with the cleavage of the epoxide moiety to yield dienal 3. MacMillanet al.¹⁸ recently reported the elegant conversion of aldehydes into epoxides using SOMO organocatalysis. Applying this methodology, we were able to isolate the terminal epoxide 5 in an unoptimized 56% yield and with 95% ee. Epoxide opening¹⁹ (5 → 6) followed by the oxidative removal of the dithiane²⁰ afforded hydroxyketone 7 in good yield.

Scheme 2 Reagents and conditions: (a) NaIO₄, THF/H₂O (2 ∶ 1), 20 °C; (b) (1) 4·TFA, K₂S₂O₈, LiCl, Cu(TFA)₂·xH₂O, acetonitrile, water, 4 °C, (2) NaBH₄, 0 °C, (3) NaOH, EtOH, H₂O, 20 °C; (c) 2-methyl-1,3-dithiane, n-BuLi, THF, 20 °C; (d) PhI(TFA)₂, CaCO₃, MeOH/acetonitrile (9 ∶ 4), 20 °C.

Interestingly, the ripostatin C12–24 subunit 9 bearing a skipped arene-ene moiety can be prepared in a similar approach from epoxygeranyl acetate 1a (Scheme 3). Substitution with phenylmagnesium chloride is followed by periodate cleavage of the epoxide. In order to show the general potential of combining the allylic substitution with an organocatalytic aldol reaction with acetone,²¹ we obtained the quite advanced fragment 9 in just three synthetic steps from 1a. The major byproduct 10 can be converted to 9 with significantly higher enantioselectivity than for the direct aldol reaction using an oxa-Michael addition developed by List et al.²²

Scheme 3 Reagents and conditions: (a) Li₂CuCl₄, PhMgCl, THF, 0 °C; (b) NaIO₄, THF/H₂O (2 ∶ 1), 20 °C; (c) D-proline, i-PrOH, acetone, 20 °C; (d) 11, Cl₃CCO₂H, H₂O₂, dioxane, 32 °C, then P(OEt)₃.

Finally, we want to provide an example for the further functionalization of the terminal alkene moiety in 2a. Apart from possible hydroborations, oxidative cleavage and cross-metathesis reactions, we want to use the hydroformylation reaction for the introduction of a free aldehyde functionality on one end of the aliphatic chain while the epoxide on the other terminus remains a masked aldehyde (Scheme 4). Using 2 mol% of Rh(acac)(CO)₂ and Xantphos²³ as the ligand, we obtained the unbranched aldehyde 12 in virtually quantitative yield. We think that this model compound represents a useful and orthogonally functionalized fragment for the natural product dolabelide D.²⁴

Scheme 4 Reagents and conditions: (a) Rh(acac)(CO)₂, Xantphos, toluene, H₂/CO (20 bar), 60 °C.

In conclusion, we have developed a simple protocol for the Cu-catalyzed substitution of allylic acetates with vinyl Grignard reagents that is tolerant towards a variety of functional groups. A possible ramification of this method is the reduction of construction steps in the total synthesis of polyketides. In particular, the efficient conversion of terpene acetate derivatives into polyketide building blocks is a harbinger of their increasing importance as cheap and renewable resource in complex molecule synthesis.

Notes and references

1. T. Newhouse, P. S. Baran and R. W. Hoffmann, Chem. Soc. Rev., 2009, 38, 3010 and references therein.
2. P. Anastas and N. Eghbali, Chem. Soc. Rev., 2010, 39, 301.
3. (a) S. H. Bertz, J. Am. Chem. Soc., 1982, 104, 5801; (b) I. T. Horvath and P. T. Anastas, Chem. Rev., 2007, 107, 2167; (c) P. L. Fuchs, Tetrahedron, 2001, 57, 6855; (d) F. Qiu, Can. J. Chem., 2008, 86, 903.
4. T. Gaich and P. S. Baran, J. Org. Chem., 2010, 75, 4657.
5. J. B. Hendrickson, J. Am. Chem. Soc., 1975, 97, 5784.
6. For an example of this approach, see: C. F. Bender, F. K. Yoshimoto, C. L. Paradise and J. K. De Brabander, J. Am. Chem. Soc., 2009, 131, 11350.
7. K. A. D. Swift, Top. Catal., 2004, 27, 143.
8. F. M. A. Geilen, B. Engendahl, A. Harwardt, W. Marquardt, J. Klankermayer and W. Leitner, Angew. Chem., Int. Ed., 2010, 49, 5510.
9. A. Behr, M. Becker, T. Beckmann, L. Johnen, J. Leschinski and S. Reyer, Angew. Chem., Int. Ed., 2009, 48, 3598.
10. M. Willot, L. Radtke, D. Könning, R. Fröhlich, V. H. Gessner, C. Strohmann and M. Christmann, Angew. Chem., Int. Ed., 2009, 48, 9105.
11. Isolation: (a) H. Irschik, H. Augustiniak, K. Gerth, G. Höfle and H. Reichenbach, J. Antibiot., 1995, 48, 787; (b) H. Augustiniak, H. Irschik, H. Reichenbach and G. Höfle, Liebigs Ann., 1996, 1657; for a synthetic study, see: (c) C. Kujat, M. Bock and A. Kirschning, Synlett, 2006, 449.
12. D. Haebich and F. von Nussbaum, Angew. Chem., Int. Ed., 2009, 48, 3397.
13. A. Gansäuer, J. Justicia, A. Rosales and B. Rinker, Synlett, 2005, 1954.
14. (a) J. Justicia, J. E. Oltra and J. M. Cuerva, J. Org. Chem., 2004, 69, 5803; (b) L. Del Valle, J. K. Stille and L. S. Hegedus, J. Org. Chem., 1990, 55, 3019; (c) A. M. Echavarren and A. M. Castano, Tetrahedron Lett., 1996, 27, 6587; coupling with vinyltrifluoroborates: (d) G. W. Kabalka and M. Al-Masum, Org. Lett., 2006, 8, 11.
15. For an excellent overview, see: T. K. Macklin and G. C. Micalizio, Nat. Chem., 2010, 2, 638 and references therein.
16. (a) G. W. Kabalka and M. Al-Masum, Org. Lett., 2006, 8, 11; (b) A. Fürstner, O. Larionov and S. Flügge, Angew. Chem., Int. Ed., 2007, 46, 5545; (c) P. S. Diez and G. C. Micalizio, J. Am. Chem. Soc., 2010, 132, 9576.
17. See the ESI‡.
18. M. Amatore, T. D. Beeson, S. P. Brown and D. W. C. MacMillan, Angew. Chem., 2009, 121, 5223.
19. C. W. Wullschleger, J. Gertsch and K.-H. Altmann, Org. Lett., 2010, 12, 1120.
20. G. Stork and K. Zhao, Tetrahedron Lett., 1989, 30, 287.
21. B. List, P. Pojarliev and C. Castello, Org. Lett., 2001, 3, 573.
22. (a) C. M. Reisinger, X. Wang and B. List, Angew. Chem., Int. Ed., 2008, 47, 8112; (b) B. Vakulya, S. Varga, A. Csámpai and T. Soós, Org. Lett., 2005, 7, 1967.
23. A. M. Schmidt and P. Eilbracht, J. Org. Chem., 2005, 70, 5528.
24. (a) L. Grimaud, R. de Mesmay and J. Prunet, Org. Lett., 2002, 4, 419; (b) N. Desroy, R. Le Roux, P. Phansavath, L. Chiummiento, C. Bonini and J.-P. Genêt, Tetrahedron Lett., 2003, 44, 1763; (c) R. Le Roux, N. Desroy, P. Phansavath and J.-P. Genêt, Synlett, 2005, 429; (d) G. E. Keck and M. D. McLaws, Tetrahedron Lett., 2005, 46, 4911; (e) P. K. Park, S. J. O'Malley, D. R. Schmidt and J. L. Leighton, J. Am. Chem. Soc., 2006, 128, 2796.

---

Footnotes

† This article is part of the ‘Emerging Investigators’ themed issue for ChemComm.

‡ Electronic supplementary information (ESI) available: Synthetic procedures, optimization of reaction conditions and characterization of compounds. See DOI: 10.1039/c0cc02332a

---