The synthesis of 7,9-dimethoxy-3-propyl-3,4-dihydro-1 H -benzo[ g ]isochromene-1,5,10-trione: A potential monomer for the synthesis of the natural product xylindein

A novel synthesis of a 3-propyl-substituted-benzo[ g ]isochromene quinone, a potential monomer of the natural product xylindein, was accomplished in 9 steps (overall yield of 8.2%) from 2,4-dimethoxybenzaldehyde. Key steps included the use of a cross-metathesis reaction in which ethyl-3-allyl-4-(benzyloxy)-1,6,8-trimethoxy-2-naphthoate was converted into ethyl-4-(benzyloxy)-1,6,8-trimethoxy-3-(4-oxopent-2-enyl)-2-naphthoate as a mixture of ( E) - and ( Z) -isomers. Following an oxa-Michael addition reaction, a racemic mixture of the desired product, 5-(benzyloxy)-7,9,10-trimethoxy-3-(2-oxopropyl)-3,4-dihydro-1 H -benzo[ g ]isochromen-1-one, the basic xylindein lactone skeleton, was obtained.


Introduction
Anhydrofusarubin lactone (1) and 5-methoxy-3,4-dehydrosemixanthomegnin (2) are examples of lactonecontaining aromatic compounds.Both compounds belong to an expanding class of quinone-containing compounds (Figure 1). 1 Anhydrofusarubin lactone (1) 2 is found in several fungi such as Nectria haematococca and Fusarium solani, while 5-methoxy-3,4-dehydrosemixanthomegnin (2) 3 has been isolated from Paepalanthus latipes Silveira (family Eriocaulaceae).Related aromatic compounds, but not quinones, are paepalantine (3) and 9-O-methyl paepalantine (4), as well as the dimer of paepalantine (5), have been isolated from a related Eriocaulaceae species Paepalanthus bromelioides 4 and Paepalanthus vellozioides. 5A more complex, related dimeric-fused lactone-containing aromatic compound possessing a stereogenic center at C-3 is xylindein (6), produced by fungi from genus Chlorociboria aeruginosa (Figure 2).This pigment is known for the green staining of wood.The artificial coloring of wood with C. aeruginosa is a patented process, and oak staining using this procedure is used in decorative woodworking such as Tunbridge ware. 6The structure of xylindein was elucidated, independently, by Todd et al. 7 and by Edwards and Kale. 8More recently, the absolute configuration and tautomeric structure of xylindein have been described. 9An interesting potential application of xylindein is that it has been investigated for use as a sustainable organic agent for optical-and electronic-property applications. 10wo research groups have attempted the synthesis of xylindein.Giles et al. 11 .and Green et al. 12 have postulated that the possible biosynthetic precursor to xylindein was the lactone-containing quinone 7 (Figure 2).While the racemic quinone 7 could be synthesized in very poor yields, the authors concluded that "new synthetic strategies to the elusive -lactone 7 will have to be sought." 12ill and Donner 13 postulated that the naphthol 8, or even the lactone deficient aromatic compound 9, might be better building blocks for the formation of xylindein. 14The use of these building blocks was postulated on the basis of previous literature.In this research, the extended quinone, xylaphin, was prepared from a pyran-containing aromatic compound rather than a lactone precursor.Naphthol 8 was synthesized by Gill using Staunton-Weinreb annulation methodology, while 9 was synthesized by Donner using the Diels-Alder reaction as a key step.None of these approaches, however, allowed for the synthesis of the target compound, xylindein.
In our laboratories, we have explored the synthesis of lactone-containing aromatic compounds.This has resulted in the synthesis of 5-methoxy-3,4-dehydrosemixanthomegnin (2) and 9-O-methylpaepalantine (5), a methoxy derivative of paepalantine (3). 15Hence, we believed that we would be able to synthesize the quinone-containing lactone 7.In principle, dehydration of 7 should result in the formation of xylindein.

Results and Discussion
We have previously described the synthesis of the trioxygenated naphthalene 11 16a as well as related PIFAmediated methodology 16b for the synthesis of the correctly substituted 1,4,6,8-tetraoxygenated naphthalene nucleus 12 of the xylindein monomer.We envisaged that the ester at C-2 and the allyl substituent at C-3 of the naphthalene 12 would allow for possible synthetic routes for the synthesis of the desired isochromenone moiety of 10.
2,4-Dimethoxybenzaldehyde was treated under Stobbe reaction conditions with diethyl succinate in the presence of t-BuOH and t-BuOK to afford the intermediate half-ester, which was subjected to Ac 2 O and NaOAc to afford the required naphthalene, ethyl 4-acetoxy-6,8-dimethoxynaphthalene-2-carboxylate, along with a minor product, t-butyl-4-acetoxy-6,8-dimethoxy-2-naphthoate (see experimental).Following a further synthetic step described in the literature, 16a the desired trioxygenated naphthalene 11 containing a C-allyl substituent was furnished in good yield.Using the PIFA-mediated Kozlowski conditions 17  The next step in the synthesis was the construction of the appropriate 3-propyl-substituted chromanone ring via a cross metathesis reaction of 13 with methyl vinyl ketone in the presence of the Grubbs II catalyst.This resulted in the formation of the α,β-unsaturated carbonyl compound 14 as a mixture of (E)-and (Z)isomers.The Michael acceptor 14 was exposed to BF 3 •OEt 2 at 0 C to give the desired isochromene-1-one 15 in good yield.Treatment of 14 with a 10% potassium hydroxide solution in a mixture of water and ethanol resulted in the formation of anthracene 16 in a poor yield of 20%. 18he next step required the removal of the ketone carbonyl of 15.The thioacetal 17 was readily formed by treatment of 15 with ethanedithiol under acidic conditions as shown in Scheme 2 (structure confirmed by Xray crystallography 19 ), along with a minor product, which indicated that one of the methyl ether groups was no longer present, and had resulted in the formation of naphthol 17a.Our basis for suggesting that the 10hydroxynaphthalene 17a had formed was that a signal for the hydroxyl H was observed at δ 12.99 in the 1 H NMR spectrum, indicating hydrogen bonding to the adjacent lactone carbonyl.
The conversion of the thioacetal 17 to liberate the propyl substituent was the next step.As the thioacetal 17 also contained an O-benzyl ether, we believed that reaction conditions could be manipulated to efficiently afford the trimethoxynaphthol, 5-hydroxy-7,9,10-trimethoxy-3-propyl-3,4-dihydro-1H-benzo[g]isochromen-1one that would facilitate the oxidation step to the required naphthoquinone 10.Thioacetal 17 was exposed to Raney nickel in refluxing ethanol for 72 h.Much to our surprise, not only was the thioacetal no longer present, but the methoxy substituent at C-10 that we had previously introduced was no longer evident.As expected, however, under the reaction conditions, the O-benzyl substituent was no longer present, and the C-5 naphthol, 5-hydroxy-7,9,-dimethoxy-3-propyl-3,4-dihydro-1H-benzo[g]isochromen-1-one (18) had formed.Evidence for this was that no hydrogen-bonded phenol was observed in the 1 H NMR spectrum of 18, and the proton at C-10 of the 1 H NMR spectrum which appeared at δ 8.65 seemed to be incorrect for the C-10 naphthol.In fact, if the regioisomer, 10-hydroxy-7,9-dimethoxy-3-propyl-3,4-dihydro-1H-benzo[g]isochromen-1-one, had been formed, the 1 H NMR spectrum would show a more shielded singlet at about δ 7.This was corroborated by examining the 1 H NMR of a related product, 10-hydroxy-7,9-dimethoxy-3-methyl-1Hbenzo[g]isochromen-1-one, where the proton attached to C-5 was observed at 6.92 in the 1 H NMR spectrum. 20Therefore, the isochromenone 18 containing a 5,7,9-trihydroxynaphthalene had been formed.The 1 H NMR spectrum showed one aromatic singlet at δ 8.65, as well as two meta-coupled doublets at δ 6.93 (1H, d, J 1.9 Hz) and 6.47 (1H, d, J 1.9 Hz) while, in the 135 DEPT spectrum, there were three aromatic C-H signals at δ 119.78, 98.13 and 91.5.Further evidence for the formation of the naphthol 18 was obtained from HRMS (calcd.for C 18 H 21 O 5 (M+H) + 317.1390, found 317.1390).Undeterred by this result, we exposed the naphthol 18 to CAN which furnished the desired quinone 10 in good yield.

Conclusions
Using 2,4-dimethoxybenzaldehyde as the starting material, we have been able to develop a novel synthesis of a 3-propyl-substituted benzo[g]isochromenone quinone, a potential monomer of the natural product xylindein, in 9 steps with an overall yield of 8.2%.Currently, we are exploring the utilization of this quinone as the monomer unit for the assembly of xylindein.As the methoxy substituent at C-10 was unexpectedly lost in the conversion of 17 into 18, attempts are being made to conduct the same synthetic steps commencing with compound 13, lacking the methoxy substituent, however, at C-1.

Experimental Section
General. 1 H NMR and 13 C NMR spectra were recorded on a Bruker AVANCE 300 spectrometer.All chemical shift values are reported in parts per million (ppm) referenced against TMS which is given an assignment of zero ppm.Coupling constants (J-values) are given in Hertz (Hz).All mass spectroscopy data were collected on a Waters Acquity UPLC system coupled to a Waters HDMS G1 QTOF mass spectrometer.UPLC settings: Analytical column: BEH C18 150  2.1 mm; Column temperature: 60 °C; Mobile phase: 90% water (0.1% formic acid): 5% acetonitrile; Flow rate: 0.4 mL/min.MS settings: Mode: VTOF; Ionisation: ESIPos and ESINeg; Scan range: 100-1000 Da; Scan speed: 0.1 second; Run time: 10 mins.Infrared spectra were recorded on a Bruker Tensor 27 standard system spectrometer with diamond ATR attachment.Macherey-Nagel Kieselgel 60 (particle size 0.063-0.200mm)was used for conventional silica gel column chromatography with various EtOAc and hexane mixtures as the mobile phase.TLC was performed on aluminum-backed Macherey-Nagel Alugram Sil G/UV254 plates pre-coated with 0.25mm silica gel 60.The Bruker D8 VENTURE PHOTON CMOS area detector diffractometer, equipped with a graphite monochromated Mo Kα 1 radiation (50 kV, 30 mA), was used to collect all the intensity data.The program SAINT+, v. 6.02 21 was used to reduce the data, and the program SADABS was used to make corrections to the empirical absorptions.Space group assignments were made using XPREP 21 on all compounds.In all cases, the structures were solved in the WinGX 22 Suite of programs by direct methods using SHELXS-97 23 and refined using full-matrix least-squares/difference Fourier techniques on F 2 using SHELXL-97. 23All non-hydrogen atoms were refined anisotropically.All C-H hydrogen atoms were placed at idealized positions and refined as riding atoms with isotropic parameters 1.2 or 1.5 times those of the 'heavy' atoms to which they are attached.Diagrams and publication material were generated using ORTEP-3, 24 and PLATON. 25Experimental details of the X-ray analyses are given for t-butyl-4acetoxy-6,8-dimethoxy-2-naphthoate in the experimental section and for 5-(benzyloxy)-7,9,10-trimethoxy-3-(2-methyl-1,3-dithiolan-2-yl)methyl-3,4-dihydro-1H-benzo[g]isochromen-1-one (17) in the experimental section.

Figure 2 .
Figure 2. Postulated monomeric aromatic building blocks for the assembly of xylindein.