Facile synthesis of cyclopentenone B1- and L1-type phytoprostanes

Phytoprostanes (PhytoPs) represent non-enzymatic metabolites of α-linolenic acid (ALA), the essential omega-3 polyunsaturated fatty acid (PUFA) derived from plants. PhytoPs are present in the plant kingdom and represent endogenous mediators capable of protecting cells from oxidative stress damages in plants. Recently, it was found that such metabolites are present in cooking oil in high quantities, and also that B1-PhytoPs protect immature neurons from oxidant injury and promote differentiation of oligodendrocyte progenitors through PPAR-γ activation. We report a novel and facile synthesis of natural 2,3-substituted cyclopentenone PhytoPs, 16-B1-PhytoP, and 9-L1-PhytoP. Our strategy is based on reductive alkylation at the 2-position of 1,3-cyclopentanedione using a recent protocol developed by Ramachary et al. and on a cross-coupling metathesis to access conjugate dienone system. In conclusion, this strategy permitted access to B1- and L1-PhytoPs in a relative short sequence process, and afford the possibility to easily develop analogs of PhytoPs.


Introduction
Non-enzymatic production of biologically relevant compounds seems rather scarce in nature. One of the reasons could be the lack of researchers willing to accept the idea that auto-oxidative formation of metabolites could possess a primordial role in biology, and therefore to unravel their hidden properties. While this idea is only starting to grow in the field of oxidative stress and the production of non-enzymatically formed metabolites of polyunsaturated fatty acid (PUFA), it is still rather uncommon. In 1990, Morrow and Roberts discovered isoprostanes (IsoPs), compounds produced in human fluids and tissues by non-enzymatic free radical peroxidation of arachidonic acid (C20:4 n-3, AA) in phospholipid membranes (Morrow et al., 1990). These compounds later became the "gold standard" biomarker of lipid oxidative stress (Kadiiska et al., 2005), but also showed many relevant biological activities (Jahn et al., 2008;Galano et al., 2015). In 1998, Mueller and coworkers discovered that peroxidation of α-linolenic acid in plant (C18:3 n-3, ALA) lead to dinor-isoprostanes, later called Phytoprostanes (PhytoPs) (Parchmann and Mueller, 1998;Jahn et al., 2010). They also represent a standard of oxidative stress in plants , and are considered, like the jasmonate compounds, to be able to activate genes implicated in the detoxification response, and many other response signals to protect the plants . PhytoPs are also relevant to human, particularly to human diet, being present in vegetable oil (Karg et al., 2007;Collado-González et al., 2015a,b), and it was demonstrated that their level in humans can increase after high consumption of ALA (Barden et al., 2009). Recently, 2,3-substituted cyclopentenone PhytoPs such as 16-B 1 -PhytoPs showed protective effects on immature neurons from oxidant injury and on the differentiation of oligodendrocyte progenitors through PPAR-γ activation (Minghetti et al., 2014).
A few syntheses of such 2,3-substituted cyclopentenone phytoprostanes are reported in the literature (El Fangour et al., 2005;Schmidt and Boland, 2007;Mikołajczyk, 2009, 2011;Vázquez-Romero et al., 2009, 2013Beretta et al., 2015), and our group developed the first ones in 2005, 9-L 1 -PhytoP and 16-B 1 -PhytoP (El Fangour et al., 2005). Later, Boland and coworkers developed a very elegant and rapid synthesis of such compounds and analogs based on a 1,3cyclopentanedione functionalization and Heck or Sonogashiratype coupling reactions (Schmidt and Boland, 2007). We believed that a rapid and facile unexplored approach could also be carried out based on a cross-metathesis strategy between conjugated 3-vinyl-cyclopentenone derivatives such as A and enone or allyl unit partners to introduce the unsaturated side chain. Derivatives such as A could be obtained from 2-alkylation of 1,3cyclopentanedione followed by introduction of the conjugated dienone unit (Scheme 1). Interestingly, large variation at the 2position could be achieved using the elegant Ramachary protocol for reductive alkylation of 1,3-cyclopentanedione (Ramachary and Kishor, 2008). Here we described the rapid and facile syntheses of 2,3-substituted cyclopentenone PhytoPs, 16-B 1 -PhytoP and 9-L 1 -PhytoP based on the novel approach described above.

Materials and Methods
All reactions requiring anhydrous conditions were conducted in oven-dried glassware (120 • C for 12 h minimum), syringes, and needles with magnetic stirring under nitrogen unless mentioned otherwise. Anhydrous THF and Et 2 O were obtained from the Innovative Technology PureSolv ™ PS-400 solvent purification system. Other solvents and reagents were used as obtained from the supplier (Acros or Aldrich) unless otherwise noted. Lipase acrylic resin from Candida antarctica ≥5000 U/g, recombinant, expressed in Aspergillus niger was purchased from SCHEME 1 | Retrosynthetic analyses of 1 and 2 from 1,3-cyclopentanedione.
Sigma-Aldrich. Reactions were monitored by TLC using plates precoated with silica gel 60 (Merck). Reaction components were visualized by using a 254 nm UV lamp, treatment with acidic p-anisaldehyde stain followed by gentle heating. Organic layers were dried with anhydrous MgSO 4 unless otherwise stated. Column chromatography was performed by using silica gel 40-63 µm. Optical rotations were measured with a Jasco P2000 digital polarimeter at 20 • C. Concentrations c reported for the optical rotation data are given in g/100 mL. Infrared data are reported as wavenumbers (cm −1 ). ES-MS data were obtained by positive electrospray ionization methods. 1 H NMR spectra were obtained at 300 or 500 MHz on Bruker spectrometers (AMX300 or Avance 500 MHz). The spectra were recorded in CDCl 3 (internal reference at δ = 7.26 ppm) unless otherwise noted. The 1 H NMR spectra are reported as follows: chemical shift in ppm [multiplicity, coupling constant(s) J in Hz, relative integral]. The multiplicities are defined as follows: br., broad; m, multiplet; AB, AB system; s, singlet; d, doublet; t, triplet; or combinations thereof. Selected 13 C NMR spectra were recorded by using a Jmodulated sequence, and the central peak of the CDCl 3 triplet was used as the internal reference (δ = 77.16 ppm) and MeOD (fixed at δ = 49.0 ppm). The NMR spectra were assigned by homonuclear ( 1 H-1 H) and heteronuclear ( 1 H-13 C) correlation spectroscopy (COSY45, HSQC, HMBC). Infrared spectra were obtained using Perkin-Elmer Spectrum One Spectrometer. They were reported as wavenumber (cm −1 ) of significant peaks.
The mixture was used directly in the next step. To a solution of contaminated 16-oxo-16-B 1 -PhytoP methyl ester 16 in MeCN (8.5 mL) and phosphate buffer (8.5 mL, pH = 7) CALB (36 mg, 20% w/w) was added. The reaction mixture was stirred 5 h. After filtration, the crude mixture was extracted with EtOAc (3 × 50 mL). The organic layers were washed with brine (2 × 25 mL) and dried over anhydrous MgSO 4 . The solvents were removed under vacuum. The crude of the reaction was purified under silica gel chromatography (100/0 to 90/10: CH 2 Cl 2 /MeOH) to give 80.6 mg of 16-oxo-

Result and Discussion
Our initial target was 9-L 1 -PhytoP, and while the ethyl group at the 2-position of 1,3-cyclopentanedione could be introduced using the Ramachary protocol (Ramachary and Kishor, 2008), we started with commercial and affordable 2ethyl-1,3-cyclopentanedione (Scheme 2). Protection of one of the ketone groups was done by O-alkylation with commercial TMSCHN 2 in a MeOH/Et 2 O solution to give the corresponding enol ether 3. The latter is then treated with a solution of vinyl magnesium bromide in THF to provide the desired conjugated dienenone system 4 in 61% yield over 2 steps (Fisher et al., 1988). Having in hand 4, the synthesis of racemic coupling partner 5 is easily achieved by allylic oxidation of commercial methyl 10-undecenoate with SeO 2 and tBuOOH in refluxing CH 2 Cl 2 overnight in 61% yield (Scheme 3). The corresponding enone compound 6 was obtained by Swern oxidation sequence in 79% yield.
Enzymatic resolution of 5 was carried out using Candida antartica lipase B (CALB) with vinyl acetate in hexane on gram scale. After overnight conversion, non-reactive alcohol and acetate compounds are separated by silica gel to afford (S)-5 in 99% ee and 45% yield and (R)-5-Ac in 46% yield and 93% ee. (R)-5-Ac can be submitted to a second enzymatic resolution, after methanolysis, CALB conversion and methanolysis to afford (R)-5 in an improved 99% ee in 76% overall yield from (R)-5-Ac. Absolute configurations are determined using the corresponding mandelic ester derivatives (Chataigner et al., 1998).
Cross-metathesis (CM) between diene (or electron deficient diene) and alkene (or electron deficient alkene) are nowadays quite widespread in the literature (Grubbs, 2004), however CM between conjugated dienone and allylic alcohol, or even more complex electron deficient enone substrate have never been attempted to the best of our knowledge (Wojtkielewicz, 2013).
To access 9-oxo-9-L 1 -PhytoP, which we believe is a potential metabolite of 1, we decided to perform a CM between cyclopentenone 4 and enone 6. Reaction conditions proved to be smoother than before as only one batch of 2.5 mol% of HG-II catalyst in refluxing CH 2 Cl 2 for 5 h afforded the corresponding coupling product with complete consumption of starting materials. Purification was only performed to remove catalyst because substrate 8 slowly decomposed over silica gel. We also noted that dimers of 4 and 6 were formed in greater amount than from the previous CM with allylic alcohol 5, even if we did not intend to recover them for further SCHEME 6 | LiTMP dimerization of 1,2-epoxybutane.
What remained to be accomplished was the metathesis CM between 13 and enantiomerically enriched penten-3-ol or corresponding penten-3-one. Resolution of penten-3-ol was already described in the literature by Sharpless asymmetric epoxidation (Moslin and Jamison, 2006), however we could not achieve the level of enantiopurity reported (66% vs. >90% ee). While no report for lipase enzymatic resolution of penten-3ol was reported, we screened a panel of commercially available lipases but did not reach a significant level of enantiopurity either, which could certainly be explained due to the relatively small difference between the vinyl and the ethyl groups. However, it is well known that CM can also be performed starting with the homodimer of the olefin partner (Finnegan et al., 2006). Fortunately, a very simple procedure to access SCHEME 7 | Synthesis of 16-B 1 -PhytoP, its enantiomer and 16-oxo-B 1 -PhytoP.

Conclusion
In this paper, we have developed a rapid and flexible synthetic strategy for 2,3-substituted cyclopentenone phytoprostanes which can permit development of synthetic analogs from commercially available materials within a few steps. This strategy can compete with the Boland's one in terms of the number of steps and flexibility with a limitation in terms of yield in the CM compare to the Heck reaction (higher yield of coupling and scale), however no protecting group of the side chains was required in our strategy.