Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 Sep 21.
Published in final edited form as: Org Lett. 2012 Sep 10;14(18):4738–4741. doi: 10.1021/ol3019488

Oxyanionic Sigmatropic Rearrangements Relevant to Cyclooctadienone Formation in Penostatins I and F

Matthew J Jansma 1, Thomas R Hoye 1,*
PMCID: PMC3501534  NIHMSID: NIHMS407195  PMID: 22963036

Abstract

graphic file with name nihms407195u1.jpg

The results of several experiments designed to probe the energetic viability of a reaction path for generation of penostatins I (3) and F (4) via spontaneous [3,3]-sigmatropic rearrangement are reported. In particular, the enolate derived from the 2-vinyl-6-acyldihydropyran 8-cis gave cyclooctadienone 12 via facile anionic oxy-Claisen rearrangement, demonstrating the feasibility of such an event.

The penostatins A (1), B (2), I (3), and F (4) are four (of nine total) members of a polyketide-derived family of secondary metabolites that have been isolated from the microorganism Penicillium sp. OUPS-79, a fungus that was originally separated from the marine alga Enteromorpha intestinalis1,2 (Scheme 1). Each of 1/2 and 3/4 are tetraepimeric. That is, each member of a given pair possesses the same absolute configuration at the C5 carbinol stereogenic center but the opposite absolute configurations at the remaining four stereogenic centers. This configurational relationship among members of the same natural product family is certainly rare, if not unprecedented.

Scheme 1.

Scheme 1

Open in a new tab

Hypothesis for the biosynthesis of penostatin I (3) and penostatin F (4).

We have hypothesized that 3 and 4 are biosynthetically derived from 1 and 2 (or their C9 epimers). This mechanistic thinking is also evident from synthetic studies carried out in the Snider2a and Barriault2c labs. Specifically, 1 or 2 could, via keto-enol tautomerization [KET], generate 5 or 6 (Scheme 1). Embedded within each of these enol tautomers is an allyl vinyl ether (AVE) subunit that bears a donor atom at C1. Spontaneous [3,3]-sigmatropic rearrangement of 5 or 6 would then directly generate the bicyclo[5.3.1]undecenone core of 3 or 4, respectively. Paquette and co-workers have extensively studied the ring expansion of architecturally similar 2-alkylidene-6-alkenyl pyrans to generate cyclooct-4-enones via thermal Claisen rearrangement.3

In 1981 a theoretical treatment of the aliphatic Claisen rearrangement by Carpenter and Burrows qualitatively demonstrated that π-donor substituents positioned at C1 (as well as C2 and C3) of simple AVEs should result in rate enhancement.4 The prescience of this analysis was soon realized when, in 1985, Koreeda and Luengo reported the discovery of the “anionic oxy-Claisen” rearrangement.5 These authors demonstrated, among other things, that the [3,3]-sigmatropic rearrangement of the potassium and sodium enolates derived from α-(allyloxy)propiophenone occurred with remarkable ease (t1/2 < 0.1 h and 2.4 h, respectively, at −23 °C).

The implications of the Koreeda study vis-à-vis the hypothesis proposed in Scheme 1 were intriguing. Might it be the case that (the enolates corresponding to) 5 and 6, present at low (but finite) concentration, undergo a pair of biosynthetic, anionic oxy-Claisen rearrangements to give rise to (the alkoxides precursors of) 3 and 4 at ambient temperature? We reasoned that this possibility could be addressed by interrogating the reactivity of the enolates derived from the dihydropyrans 7 and 8 (Scheme 2). These simplied analogs retain many of the key structural attributes of 1 and 2 but have the advantage of being much more readily accessible. The intent was to explore the behavior of 7 or 8 under basic conditions to learn if the enolate 9 or 10 would undergo [3,3]-sigmatropic rearrangement to produce the penostatin-like cyclooctadienone 11 or 12. We describe here the observations made through this model study.

Scheme 2.

Scheme 2

Open in a new tab

Proposed model study.

A straightforward, five-step sequence was employed to prepare the substrates 7-cis and 8-cis (Scheme 3). The stannylene ketal6 derived from hexa-1,5-diene-3,4-diol (13) (ca. 1:1 meso + d,l)7 was alkylated with tert-butyl bromoacetate (Bu4N+Br−, PhH, reflux) to deliver the isomeric dioxanones 14-trans/cis together with minor amounts of their acyclic, hydroxy ester progenitors. This mixture was driven to the desired products upon treatment with TFA (PhH, reflux).8 Separation of these isomers (medium pressure liquid chromatography on silica gel) provided 14-trans (>95% purity) and 14-cis (ca. 90% purity), the relative configurations of which were assinged by analysis of vicinal 3JH5,H6 coupling values.9

Scheme 3.

Scheme 3

Open in a new tab

Preparation of the model substrates 7-cis and 8-cis.

Lactone 14-trans10 was then subjected to Ireland–Claisen11 rearrangement to effect dioxanone-to-dihydropyran reorganization12,13 (Scheme 3). In the event, the intermediate silyl enol ether 15-trans rearranged smoothly (PhMe, 100–110 °C8,14) to deliver the TBS ester 16-cis. It should be noted that isolation of this latter, acid-sensitve substance required the use of TMS-functionalized silica gel.15 The Weinreb amide 18-cis then emerged in moderate overall yield from a two-step procedure that involved (i) exposure of 16-cis to an excess (3 equiv) of the Vilsmeier reagent16 and (ii) amidation of the resulting crude acid chloride 17-cis under standard conditions. Finally, the styrenyl-(7-cis) and phenyl-(8-cis) ketones were uneventfully produced upon acylation of either β-lithiostyrene17 or phenylmagnesium bromide with the amide 18-cis.

When ketone 7-cis was exposed to K2CO3 in MeOH for 90 min at 60 °C, the only isolable product was the symmetric cycloheptadiene derivative 20 (Figure 1, Panel A), derived from [2,3]-Wittig rearrangement18 of the enolate 9 (Panel B, entry 1). This product was again formed, this time within minutes at room temperature, upon addition of 7-cis to a solution of tetra-n-butylammonium hydroxide (Bu4N+HO−) in i-PrOH/MeOH (entry 2).

Figure 1.

Figure 1

Open in a new tab

Oxyanionic rearrangements of enolates 9 and 10, derived from 7-cis and 8-cis, respectively. A: Rearrangement of 9 via competitive [2,3]-Wittig (to 19) vs. Claisen/Cope (to 23) pathways. B: Table of results. C: Rearrangement of 10 via a Claisen pathway.

We then attempted to trap the silyl enol ether derived from 9 (M = Na) with the intent of subsequently examining its thermal reactivity. Deprotonation of 7-cis with NaHMDS in THF at −78 °C followed by treatment with TBSCl prior to work-up again gave rise to 20, but now along with a second unexpected product, the conjugated cyclooctadienone 24 (Figure 1, Panel B, entry 3). Its structure was established through extensive 1-D (1H and 13C) and 2-D (1H–1H COSY, HMQC, and HMBC) NMR experiments and analysis. Although this product was formed as essentially a single diastereomer, we have not unambiguously established its relative configuration.

Formation of 24 can be explained by the following sequence, which is a pathway that is a consequence of the fortuitous positioning of the (E)-styrenyl appendage in 7-cis (Figure 1, Panel A). The cyclooctadienone 21 arising from anionic oxy-Claisen rearrangement of 9 contains yet another (all-carbon) vic-ππ subunit. It bears an aromatic substituent at C119 and an electron donating oxido at C3.20 [3,3]-Sigmatropic (anionic oxy-Cope20,21) rearrangement of 21 gives the isomeric cyclooctadienone 22. The prototropic shift indicated within 22 produces the isomeric and presumably more stable dienolate 23. Silylative quenching then gives rise to the observed dienol ether 24.22

In light of this proposed mechanism, we judged that the product derived from anionic oxy-Claisen rearrangement might be isolable if the styrenyl unsaturation were removed from 7-cis. Thus, the phenyl ketone 8-cis was subsequently studied (Figure 1, Panel C). Indeed, deprotonation of this compound with NaHMDS in THF at −78 °C followed by warming gave rise to the cyclooctadienone 12 as the major product23 (Figure 1, Panel B, entry 4). Notably, this rearrangement took place within 10 min at 0 °C.

In summary, as part of our studies designed to probe the hypothesis that penostatin F (3) and penostatin I (4) arise via spontaneous [3,3]-sigmatropic rearrangements of their progenitors 1 and 2, respectively (Scheme 1), the homologous 2,6-disubstituted dihydropyran model substrates 7-cis and 8-cis were prepared by five-step synthetic sequences emanating from the dioxanone 14-trans. The rearrangements of each were studied under a variety of basic conditions. The observed formation of cyclooctadienone 12 via a rapid [3,3]-sigmatropic (anionic oxy-Claisen) rearrangement serves as evidence that the pathways shown in Scheme 1 that interconnect 1 with 3 and 2 with 4 are energetically viable.

Supplementary Material

1_si_001

Acknowledgments

This investigation was supported by a grant awarded by the U.S. Department of Health and Human Services (National Institute of General Medical Sciences, GM-65597).

Footnotes

Supporting Information Available. Experimental procedures, characterization data, and copies of 1H- and 13C NMR spectra for all isolated compounds (44 pages).

References

  • 1.Isolation and structure determination of all members of the penostatin family: Takahashi C, Numata A, Yamada T, Minoura K, Enomoto S, Konishi K, Nakai M, Matsuda C, Nomoto K. Tetrahedron Lett. 1996;37:655–658.Iwamoto C, Minoura K, Hagishita S, Nomoto K, Numata A. J Chem Soc, Perkin Trans 1. 1998:449–456.Iwamoto C, Minoura K, Oka T, Ohta T, Hagishita S, Numata A. Tetrahedron. 1999;55:14353–14368.
  • 2.Synthetic studies toward penostatins A, B, and F: Snider BB, Liu T. J Org Chem. 1999;64:1088–1089.Snider BB, Liu T. J Org Chem. 2000;65:8490–8498. doi: 10.1021/jo000850x.Barriault L, Ang PJA, Lavigne RMA. Org Lett. 2004;6:1317–1319. doi: 10.1021/ol049680r.Total synthesis of penostatin B: Fujioka K, Yokoe H, Yoshida M, Shishido K. Org Lett. 2012;14:244–247. doi: 10.1021/ol203021c.
  • 3.E.g.: Kinney WA, Coghlan MJ, Paquette LA. J Am Chem Soc. 1984;106:6868–6870.Paquette LA, Philippo CMG, Yo NH. Can J Chem. 1992;70:1356–1365.and references cited therein; Paquette LA, Wang T-Z, Vo NH. J Am Chem Soc. 1993;115:1676–1683.
  • 4.Burrows CJ, Carpenter BK. J Am Chem Soc. 1981;103:6984–6986. [Google Scholar]
  • 5.Koreeda M, Luengo JI. J Am Chem Soc. 1985;107:5572–5573. [Google Scholar]
  • 6.David S, Thieffry A, Veyrières A. J Chem Soc, Perkin Trans 1. 1981:1796–1801. [Google Scholar]
  • 7.(a) Hekmatshoar R, Yavari I, Beheshtiha YS, Heravi MM. Monatsh Chem. 2001;132:689–691. [Google Scholar]; (b) Trost BM, Aponick A. J Am Chem Soc. 2006;128:3931–3933. doi: 10.1021/ja0578348. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Burke SD, Sametz GM. Org Lett. 1999;1:71–74. doi: 10.1021/ol9905506.In this report, a single enantiomer of 14-trans was prepared by a route that commenced with D-mannitol.
  • 9.The vicinal 3JH5,H6 values observed for 14-trans and 14-cis (9.0 and 4.5 Hz, respectively) are in good agreement with those reported for related structures by Burke and co-workers; see ref 12d
  • 10.As a matter of experimental convenience, the higher purity lactone 14-trans (which gives 16-cis upon Ireland–Claisen rearrangement12,13) was typically taken forward. Thus the ketones 7-cis and 8-cis were used in the majority of the studies described here. However, the analogous series of isomeric intermediates derived from 14-cis (i.e., 7-trans, 8-trans, 16-trans, and 18-trans) have also been prepared and fully characterized; see the Supporting Information for full details.
  • 11.Ireland RE, Mueller RH. J Am Chem Soc. 1972;94:5897–5898. [Google Scholar]
  • 12.(a) Burke SD, Armistead DM, Schoenen FJ. J Org Chem. 1984;49:4320–4322. [Google Scholar]; (b) Burke SD, Armistead DM, Fevig JM. Tetrahedron Lett. 1985;26:1163–1166. [Google Scholar]; (c) Burke SD, Schoenen FJ, Murtiashaw CW. Tetrahedron Lett. 1986;27:449–452. [Google Scholar]; (d) Burke SD, Armistead DM, Schoenen FJ, Fevig JM. Tetrahedron Lett. 1986;27:2787–2801. [Google Scholar]; (e) Burke SD, Schoenen FJ, Nair MS. Tetrahedron Lett. 1987;28:4143–4146. [Google Scholar]; (f) Burke SD, Lee KC, Santafianos D. Tetrahedron Lett. 1991;32:3957–3960. [Google Scholar]; (g) Burke SD, Buchanan JL, Rovin JD. Tetrahedron Lett. 1991;32:3961–3964. [Google Scholar]; (h) Burke SD, Piscopio AD, Kort ME, Matulenko MA, Parker MH, Armistead DM, Shankaran K. J Org Chem. 1994;59:332–347. [Google Scholar]
  • 13.(a) Büchi G, Powell JE., Jr J Am Chem Soc. 1967;89:4559–4560. [Google Scholar]; (b) Büchi G, Powell JE., Jr J Am Chem Soc. 1970;92:3126–3133. [Google Scholar]; (c) Danishefsky S, Funk RL, Kerwin JF., Jr J Am Chem Soc. 1980;102:6889–6891. [Google Scholar]
  • 14.Angle SR, Breitenbucher JG, Arnaiz DO. J Org Chem. 1992;57:5947–5955. [Google Scholar]
  • 15.E.g., ref 14 and Overman LE, Angle SR. J Org Chem. 1985;50:4021–4028. (the protocol used for the preparation of the TMS-functionalized silica gel used here is described in the Supporting Information)
  • 16.Wissner A, Grudzinskas CV. J Org Chem. 1978;43:3972–3974. [Google Scholar]
  • 17.Neumann H, Seebach D. Tetrahedron Lett. 1976;17:4839–4842. [Google Scholar]
  • 18.Thomas AF, Dubini R. Helv Chim Acta. 1974;57:2084–2087. [Google Scholar]
  • 19.Gentric L, Hanna I, Huboux A, Zaghdoudi R. Org Lett. 2003;5:3631–3634. doi: 10.1021/ol035283p. [DOI] [PubMed] [Google Scholar]
  • 20.Evans DA, Golob AM. J Am Chem Soc. 1975;97:4765–4766. [Google Scholar]
  • 21.(a) Lee E, Shin IJ, Kim TS. J Am Chem Soc. 1990;112:260–264. [Google Scholar]; (b) Lee E, Lee YR, Moon B, Kwon O, Shim MS, Yun JS. J Org Chem. 1994;59:1444–1456. [Google Scholar]
  • 22.We considered an alternative mechanism wherein 21 would arise by [1,2]-migration of a C–C bond within 19. This type of reorganization has been considered previously; see: Kirchner JJ, Pratt DV, Hopkins PB. Tetrahedron Lett. 1988;29:4229–4232. A control experiment was performed in which 20 was reexposed to NaHMDS under the conditions used to generate 24 from 7-cis. Following reaction quench with TBSCl, none of 24 was observed, ca. 20% of 20 was recovered, and a by-product that was shown to be a monosilylated dimer was isolated in ca. 19% yield, the structure of which was not fully delineated.
  • 23.The β,γ-unsaturated cyclooctadienone 12 was also seen to form slowly upon exposure of 8-cis to aqeuous base (i.e., 0.1 M NaOH, THF, rt, 20 d, ca. 7% conversion to 12). Under these conditions, the epimers 8-cis and 8-trans10 (ca. 0.9:1) were observed (1H NMR analysis) as the major components of the crude mixture in addition to the rearrangement product 12. This implies that enolate 10 was reprotonated (from both diastereotopic faces) more rapidly than it underwent [3,3]-sigmatropic rearrangement to 25. Moreover it suggests that 12 is more stable than one of its isomeric α,β-unsaturated cyclooctadienone isomers under these apparently equilibrating conditions.

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1_si_001
NIHMS407195-supplement-1_si_001.pdf (6.4MB, pdf)

RESOURCES