“Oleocanthal” is specifically deacetoxydialdehydic ligstroside aglycone, which exists as a single isomer (enantiomer). The (-)-enantiomer is the natural product and has the following chemical formula:
http://www.google.com/patents/EP2583676A1?cl=en
The Trustees of The University of Pennsylvania,
Monell Chemical Senses Center,
Russell S. J. Keast, Qiang Han, Amos B. Smith Iii, Gary K. Beauchamp, Paul A. S. Breslin, Jianming Lin,
-
In 1993, Montedoro and co-workers reported the isolation of a new class of phenolic compounds (1-4), including the dialdehydic and aldehydic forms of ligstroside (5) and oleuropeine (6) from virgin olive oils (Montedoro, G. et al. (1993) J. Agric. Food Chem. 41:2228-2234) (See Figure 1 for structures). These phenolic compounds comprise important minor constituents of virgin olive oils that have been implicated in the organoleptic characteristics including bitterness, pungency, and astringency (Andrewes, P. et al. (2003) J. Agric. Food Chem. 57:1415-1420 ).
-
In addition, these agents have been suggested to contribute to the oxidative stability of virgin olive oil and as such are associated with health benefits of olive oils, specifically their antioxidant/anticancer activities (Owen, R.W. et al. (2000) Food Chem. Toxicology 38:647-659; Owen, R.W. et al. (2000) Eur. J. Cancer 36(10):1235-1247; Baldioli, M. et al. (1996) J. Am. Oil Chem. Soc. 73(11):1589-1593; Manna, C. et al. (2002) J. Agric. Food Chem. 50(22):6521-6526).
-
Similar structural features have been reported in the constituents of the Jasminum (Somanadhan, B. et al. (1998) Planta Medica 64:246-50; Takenaka, Y. et al. (2002) Chem. & Pharm. Bull 50(3):384-389) and related plant species (Takenaka, Y. et al. (2002) Phytochemistry 59(7):779-787). It has been shown that both ibuprofen and a Mediterranean diet (i.e., high in olive oil) both decrease the risk/incidence for breast and lung cancer.
-
In 2003, Busch and co-workers at Unilever Research and Development Vlaardingen (The Netherlands) identified deacetoxydialdehydic ligstroside aglycone as a principal contributor to the potent pungent (burning) sensation at the back of throat associated with high quality virgin olive oils (Andrewes, P. et al. (2003) J. Agric. Food Chem. 57:1415-1420). Studies at Firmenich, Inc., reached the same conclusion (Firmenich, Inc. study). The structure of 1 was assigned,
employing a series of 1 and 2D NMR experiments (Andrewes, P. et al. (2003) J. Agric. Food Chem. 57:1415-1420), in conjunction with comparison to literature data (Montedoro, G. et al. (1993) J. Agric. Food Chem. 41:2228-2234). The absolute stereochemistry remained undetermined. That 1 was responsible for the strong pungent (burning) sensation at the back of the throat was based on an extensive series of HPLC fraction analysis, omission analysis and correlation, and hydrolysis studies, in conjunction with human sensory studies. Andrewes et al., however, acknowledged that “a coelution compound causing the burning sensation” could not be eliminated without completing a synthesis of 1, which they stated to be “extremely challenging.”
EXAMPLES
Example 1: Isolation of deacetoxydialdehydic ligstroside aglycone “Oleocanthal”A. Synthesis of Oleocanthal
-
Retrosynthetically, we envisioned both enantiomers of (1) to derive from the enantiomeric forms of cyclopentanediols (7) via oxidative cleavage of the diol moiety (Scheme 1). The requisite cyclopentanediols (7) in turn would be prepared from cyclopentanones (+)- and (-)-(10), via alkylation to introduce stereoselectively the side chain from the convex face, followed by stereoselective Wittig ethylnation and removal of the acetonide moiety (Scheme 1).
(5) Initially (+)- and (-)-cyclopentanones (10) were prepared via the sulfoximine and/or enzymatic protocols introduced and developed by Johnson (Johnson, C.R. and T. Penning (1988) J. Am. Chem. Soc. 110:4726-4735; Johnson, C.R. (1998) Acc. Chem. Res. 31:333-341). Although effective on modest scale (10-100mg), the requirement for gram quantities of the oleocanthals demanded that we secure for more scalable routes to (10). Towards this end, we optimized a hybrid of synthetic approaches (Moon, H. et al. (2002) Tetrahedron: Asym. 13(11):1189-1193; Jin, Y. et al. (2003) J. Org. Chem. 68(23):9012-9018; Yang, M. (2004) J. Org. Chem. 69(11):3993-3996; Palmer, A. et al. (2001) Eur. J. Org. Chem. 66(7):1293-1308; Paquette, L. and S. Bailey (1995) J. Org. Chem. 60:7849-7856) as outlined in Scheme 2. Importantly, both enantiomers of (10) could be prepared in multi-gram quantities in 7 steps, with an overall efficiency of 40% from inexpensive D-(-)-ribose. Key elements of both sequences entailed vinyl Grignard addition to the enantiomers of aldehyde (12), followed in turn by ring closing metathesis (RCM), PCC oxidation and hydrogenation (Scheme 2).
-
Alkylation of (+)- and (-)- cyclopentanone (10) with methyl bromoacetate was then anticipated to proceed from the less hindered convex face of the bicyclic skeleton to install the side chain in a stereoselective fashion. Initial attempts however to alkylate (-)-(8) with methyl bromoacetate employing LDA in the presence of HMPA furnished only a complex mixture containing only trace amounts of (-)-(16). Neither addition of Cu(I) (Johnson, C.R. and T. Penning (1988) J. Am. Chem. Soc. 110:4726-4735) reportedly to suppress side reactions, nor the use of the corresponding tin enolate [generated by treatment of (-)-(10) in THF with LDA, followed by HMPA and tributyltin chloride (Suzuki, M. et al. (1985) J. Am. Chem. Soc. 107:3348; Nishiyama, H. et al. (1984) Tetrahedron Lett. 25:223)] improved the situation. Alkylation of the zinc enolate of (-)-(10) [generated by treatment of (-)-(10) in THF with 1.1 eq. LHMDS, followed in turn by HMPA (3.0 eq.) and dimethyl zinc (Morita, Y. et al. (1989) J. Org. Chem. 54:1787-1788) (1.0 eq.)] with methyl bromoacetate, however consistently furnished (-)-(16) in 55-60% yield as a single diastereomer (this reaction was fairly clean except some baseline materials. Using t-butyl bromoacetate instead of methyl bromoacetate did not improve the yield) (Scheme 3).
-
Wittig ethylnation of (-)-(16) was next achieved with ethyltriphenylphosphine bromide. Best results were obtained employing LDA as the base at -45°C. Although excellent stereoselectivity (ca., 10:1 E:Z) favoring the E-isomer (-)-(17) was achieved, the yield was only modest (42%), presumably due to the ease of enolization of (-)-(16) (Edmunds, M. “The Wittig Reaction” In MODERN CARBONYL OLEFINATION, Takeda, Ed., John Wiley & Sons, New Jersey, 2004). Interestingly, the stereoselectivity varied dramatically with reaction temperature. At 0°C, the E:Z selectivity was 3.3:1, while at room temperature the selectivity was 1.6:1. Assignment of the E geometry of the olefin was based on NMR NOE analysis (Scheme 4).
-
Hydrolysis of ester (-)-(17) (LiOH/THF/H2O) next afforded acid (-)-(18), which was subjected to Mitsunobu esterification (Mitsunobu, O. (1981) Synthesis 1-28) with 4-hydroxyphenethyl alcohol to furnish phenol (-)-(19) in 92% yield. As expected, the Mitsunobu reaction proceeded with complete chemoselectivety at the primary hydroxyl (Appendino, G. et al. (2002) Org. Lett. 4:3839-3841). Completion of the synthesis of (-)-oleocanthal (1) was then achieved via liberation of the vicinal diol moiety (4N HCl/acetonitrile), followed by oxidative cleavage (NaIO4); (-)-oleocanthal (1) was identical in all respects (e.g., 1H and 13C NMR, IR and HRMS) with an authentic sample isolated from virgin olive oil, the latter possessing spectral data identical to that reported in the literature (Montedoro, G. et al. (1993) J. Agric. Food Chem. 41:2228-2234). The structural assignment of (1) was also confirmed by COSY NMR analysis. Synthetic (-)-(1) displayed a small negative optical rotation ([α]25D -0.78, c = 0.9, CHCl3) identical to that obtained from a sample isolated from virgin olive oil ([α]25 D -0.9, c = 2.0, CHCl3). Thus the stereochemistry of (-)-oleocanthal (1) is 3S, 4E. The enantiomer of the natural product (+)-(1) was prepared via a similar reaction sequence beginning with (+)-(10) to furnish (+)-1 ([a]25 D +0.73, c = 0.55, CHCl3) (Scheme 5).
-
In summary, an effective, scalable synthesis of both enantiomers of oleocanthal (1) has been achieved, each in 13 steps (7 % overall yield) from inexpensive (D)-(-)-ribose, requiring only 6 chromatographic separations. The structural similarity of oleocanthal to a number of related natural products (Somanadhan, B. et al. (1998) Planta Medica 64:246-50; Takenaka, Y. et al. (2002) Chem. & Pharm. Bull. 50(3):384-389; Takenaka, Y. et al. (2002) Phytochemistry 59(7):779-787) suggests that the synthetic approach presented here should also be applicable to their construction.
-
Figure 3 shows the synthetic scheme of (-)-oleocanthal.
-
Figure 4 shows the synthetic scheme of (+)-oleocanthal.