`LETTERS
`
`2005
`Vol. 7, No. 8
`1489-1491
`
`A Pauson- Khand Approach to the
`Synthesis of Ingenol
`
`Jeffrey D. Winkler,* Esther C. Y. Lee, and LaToya I. Nevels
`
`Department of Chemistry, UniVersity of PennsylVania,
`Philadelphia, PennsylVania 19104
`
`winkler@sas.upenn.edu
`
`Received January 18, 2005
`
`ABSTRACT
`
`Pauson- Khand cyclization of dioxanone photoadduct 21 leads to the formation of a single product in good yield. However, retro-aldol
`fragmentation of the pentacyclic cyclopentenone 22 leads to the formation of 23, with cis C-8/C-10 intrabridgehead stereochemistry, unlike the
`target compound ingenol 1, which possesses C-8/C-10 trans intrabridgehead stereochemistry.
`
`Scheme 1
`
`The therapeutic importance of C-3 esters of ingenol 1 and
`the dearth of exploration of structure-activity relationship
`data for this class of compounds make the development of
`efficient pathways for the synthesis of ingenol and analogues
`an important goal. Of particular note in the synthesis of
`ingenol is the establishment of the C-8/C-10 trans intra-
`bridgehead stereochemistry, which is critical for the biologi-
`cal activity of 1. In 2002, we reported the first total synthesis
`of racemic 1, in which the trans intrabridgehead stereochem-
`istry was established via intramolecular dioxenone photo-
`addition. The total synthesis proceeded in 42 steps from
`commercially available starting materials in an overall yield
`of 0.042%.1 Since that
`time,
`two other total syntheses
`have appeared by: Tanino and Kuwajima (2003) and Wood
`(2004), which proceeded in ca. 45 and 38 steps, respectively.2
`In an effort to develop a more efficient approach to the
`synthesis of ingenol, we have examined the strategy outlined
`in Scheme 1 for the synthesis of 1, in which the C-8/C-10
`intrabridgehead stereochemical relationship is established via
`
`(1) Winkler, J. D.; Rouse, M. B.; Greaney, M. F.; Harrison, S. J.; Jeon,
`Y. T. J. Am. Chem. Soc. 2002, 124, 9726-9728.
`(2) a) Tanino, K.; Onuki, K.; Miyashita, M.; Nakamura, T.; Takahashi,
`Y.; Kuwajima, I. J. Am. Chem. Soc. 2003, 125, 1498-1500. (b) Nickel,
`A.; Maruyama, T.; Tang, H.; Murphy, P.; Greene, B.; Yusuff, N.; Wood,
`J. L. J. Am. Chem. Soc. 2004, 126, 16300-16301.
`
`10.1021/ol050103s CCC: $30.25
`Published on Web 03/17/2005
`
`© 2005 American Chemical Society
`
` P. 1
`
`UT Ex. 2029
`SteadyMed v. United Therapeutics
`IPR2016-00006
`
`
`
`Scheme 2
`
`We therefore turned our attention to sulfide 14 as a
`protecting group for the offending diene functionality in 10
`(Scheme 4). Oxidative elimination of 15, the photoadduct
`obtained from 14, would then lead to the formation of 11.
`Conjugate addition of isobutylthiol to 10 gave 14). While
`
`Scheme 4
`
`Pauson-Khand cyclization of 4 to give 3. The A-ring cyclo-
`pentenone moiety in retroaldol product 2 would then be used
`to complete the synthesis of 1. The Pauson-Khand substrate
`4 should be available by the intramolecular dioxenone
`photocycloaddition of 5. We envisioned that the C-11 methyl
`group (ingenol numbering) and the gem-dimethylcyclopro-
`pane in 5 would be derived from 6, the preparation of which
`has been described from (+)-carene.3 We report herein the
`results of our model study for this new reaction sequence.
`To determine the viability of the route outlined in Scheme
`1, we examined the irradiation of 10 (Scheme 2) as a model
`system for the photocycloaddition of methylene dioxenone
`5 (Scheme 1). The synthesis of 10 is outlined in Scheme 2.
`Unsaturated aldehyde 7 was prepared in a one-pot procedure
`by Swern oxidation of 7-octen-1-ol followed by reaction of
`the intermediate aldehyde with Eschenmoser’s salt.4 Reac-
`tion of 7 with the conjugate base of tert-butyl acetate then
`gave 8, which on MnO2 oxidation afforded ketoester 9.
`Exposure of 9 to dioxenone-forming conditions (TFAA,
`TFA, Ac2O, Me2CO) led to the formation of the dioxenone
`photosubstrate 10 in 75% yield. However, irradiation of 10
`(3.0 mM in 10% Me2CO/MeCN, 450 W Hanovia mercury
`lamp, 3 h) resulted only in the recovery of unreacted 10
`without formation of the desired photoadduct 11.
`While we have shown that irradiation of 12 leads to the
`formation of 13 in good yield (Scheme 3),5 irradiation of a
`
`Scheme 3
`
`irradiation of 14 does lead to the formation of the desired
`photoadduct 15, the irradiation of the corresponding sulfoxide
`16 (Scheme 5), obtained by reaction of 14 with m-CPBA
`
`Scheme 5
`
`1:1 mixture of 10 and 12 led to the formation of none of the
`desired photoadduct 13, a result that is consistent with
`quenching of the dioxenone triplet (of both 10 and 12) by
`the diene moiety present in 10.
`
`(3) Satoh, T.; Kaneko, Y.; Okuda, T.; Uwaya, S.; Yamakawa, K. Chem.
`Pharm Bull. 1984, 32, 3452-3460.
`
`1490
`
`(-78 (cid:176)C, 97% yield, as a ca. 1:1 ratio of sulfoxide
`diastereomers), gave a cleaner reaction and higher yields.
`Irradiation of 16 led to the formation of a ca. 1:1 mixture
`of diastereomeric photoadducts 17. Oxidation of the mixture
`of diastereomeric products to a single sulfone (m-CPBA, 72%
`yield) confirmed that the photocycloaddition of 16 proceeded
`with a unique sense of induction from the C-10 stereocenter.
`
`(4) Takano, S.; Inomata, K.; Samizu, K.; Tomita, S.; Yanase, M.; Suzuki,
`M.; Iwaubuchi, Y.; Sugihara, T.; Ogasawara, K. Chem. Lett. 1989, 1283-
`1284.
`(5) Winkler, J. D.; Hey, J. P.; Hannon, F. J. Heterocycles 1987, 25, 55-
`60.
`
`Org. Lett., Vol. 7, No. 8, 2005
`
` P. 2
`
`UT Ex. 2029
`SteadyMed v. United Therapeutics
`IPR2016-00006
`
`
`
`The stereochemical outcome of the photocycloaddition of
`16 can be attributed to allylic strain effects. Selective
`formation of 17 is consistent with reaction via the conforma-
`tion shown in A [Scheme 6; R ) CH2S(O)i-Bu], in which
`
`Scheme 6
`
`the C-10 hydrogen eclipses the dioxenone ring. The structure
`of 18, the sulfone derived from 17, was confirmed by X-ray
`crystallographic analysis. Heating sulfoxide photoadduct 17
`to 160 (cid:176)C in quinoline led to the formation of the desired
`methylene photoadduct 11, the formal product of [2 + 2]
`cycloaddition of 10 (Scheme 4) in good yield.
`The Pauson-Khand substrate 21 was then prepared via
`alkylation of the conjugate base of 11 (LDA, THF, DMPU,
`-78 (cid:176)C) with 3-trimethylsilylpropargyl bromide 19 to give
`20, followed by desilylation with TBAF (THF, 100%) to
`give 21 (Scheme 7). Reaction of 21 with Co2(CO)8 and 4 Å
`
`Scheme 7
`
`molecular sieves in toluene at room temperature for 2 h
`followed by slow addition of a suspension of trimethylamine
`N-oxide dihydrate in toluene at 0 (cid:176)C led to the formation of
`22 as a single diastereomer in 60-70% yield.6 It
`is
`noteworthy that the Pauson-Khand reaction of 21 in the
`
`presence of the trimethylamine N-oxide dihydrate was
`considerably more efficient than the reaction using anhydrous
`trimethylamine N-oxide. This pronounced difference could
`be attributed to the attenuation of the nucleophilicity of the
`hydrated amine oxide ligand, which could retard decom-
`plexation of the initially formed cobalt-alkyne complex.7
`The structure and stereochemistry of 22 was confirmed
`by X-ray crystallographic analysis, which revealed that it
`did not contain the requisite C-8/C-10 relative stereochem-
`istry for the synthesis of ingenol. Retro-aldol fragmentation
`of 22 led to the formation of 23, with cis intrabridghead
`stereochemistry, which was verified by X-ray crystallo-
`graphic analysis. While the fragmentation product was
`initially formed as a single C-6 epimer (C-6(cid:226) ester as shown
`in 23), prolonged exposure of 23 to the basic reaction
`conditions (K2CO3, MeOH) led to the formation of a mixture
`of C-6 epimeric products.
`While the C-8/C-10 intrabridghead stereochemical rela-
`tionship in 22 is established in the Pauson-Khand reaction
`of 21, that relationship is indirectly established in 21, since
`the propargyl moiety in 21 can only approach the C-10
`exocyclic methylene from the (cid:226)-face as shown to give 22.
`In the retrosynthetic plan outlined in Scheme 1,
`the
`C-8/C-9 ring fusion stereochemistry in 4 is trans, which
`forces the approach of the propargyl moiety in 4 to the
`R-face of the C-10 methylene,
`thereby generating the
`requisite C-8/C-10 trans intrabridgehead stereochemistry
`shown in 3. However, irradiation of 16 led to the exclusive
`formation of the cis-fused bicyclo[5.2.0]nonane moiety as
`shown in 17 (Scheme 5). The successful implementation of
`the retrosynthetic plan in Scheme 1 therefore depends on
`the preparation of a trans-fused photoadduct or its equivalent
`from 16. Studies directed toward the construction of the
`requisite trans-fused photoadduct are currently in progress,
`and our results will be reported in due course.
`
`Acknowledgment. We would like to thank the National
`Institutes of Health (CA40250), GlaxoSmithKline, Amgen,
`and Merck for their generous support of this program.
`
`Supporting Information Available: Spectral data and
`experimental procedures for 8-11, 14-18, and 20-23 and
`X-ray data for 18, 22, and 23. This material is available free
`of charge via the Internet at http://pubs.acs.org.
`
`OL050103S
`
`(6) Perez-Serrano, L.; Casarrubios, L.; Dominguez, G.; Perez-Castells,
`J. Org. Lett. 1999, 1, 1187-1188.
`(7) (a) Shen, J.; Shi, Y.; Gao, Y.; Shi, Q.; Basolo, F. J. Am. Chem. Soc.
`1988, 110, 2414-2418. (b) Shojaie, A.; Atwood, J. D. Organometallics
`1985, 4, 187-190.
`
`Org. Lett., Vol. 7, No. 8, 2005
`
`1491
`
` P. 3
`
`UT Ex. 2029
`SteadyMed v. United Therapeutics
`IPR2016-00006