School of Biomolecular and Physical Sciences, Griffith University, Nathan QLD 4111, AUSTRALIA.
Furo- and pyranoquinolinone containing natural products have been isolated from the Rutaceae family since the 1960's1 and 1970's2 and have shown to exhibit a range of biological activities such as antifungal, antibacterial, antiviral, antimicrobial, antimalarial, insecticidal, antineoplastic, antidiuretic, antiarrythmic and sedative properties. Due to this, synthetic strategies have been developed to afford some of the selected natural products. In the course of extracting natural products at Natural Product Discovery (Brisbane, QLD, Australia), one natural product, 1.01, was isolated from the species Euodia asteridala (Rutaceae) and was of interest because of the ongoing screening of that class of molecules and the synthetic challanges involved. The natural product 1.01 consists of a furoquinolinone core with a cyclopropane ring fused onto the C2-C3 carbons of the furan ring, and a gem dimethyl functionality off the quarternary carbon of the cyclopropane ring. This thesis explores a total synthesis approach to compound 1.01, with focus on a new synthetic strategy to afford the furoquinolinone core structure as a key intermediate.
In Chapter one, syntheses of other natural products are shown together with precedent synthetic strategies of the furo- and pyranoquinolinone core structures. The overall approach for the synthesis of the natural product 1.01 is outlined incorporating the new approach for synthesising the furoquinolinone core structure. Related chemistries are referenced, discussed and taken into account for all the proposed synthetic steps.
In Chapter two, the N-methyl protected form of the furoquinolinone core structure, being the N-methyl key intermediate 5-methylfuro[3,2-c]quinolin-4(5H)-one (1.30), was synthesised through a Heck mediated C-C coupling. The optimised Heck coupling reaction conditions included the use of the less common palladium catalyst, palladium oxide, together with additive, base and a polar solvent to afford the desired N-methyl key intermediate 1.30, in a 89 % yield. Using the corresponding bromo and chloro precursors for the synthesis of 1.30, showed a relative reactivity trend of I>Br>Cl. A second key intermediate, the 0-methyl protected furoquinolinone core structure 4-methoxyfuro[3,2-c]quinoline (2.01), was synthesised via an N-SEM protected derivative 5-((2- ( trimethylsily l)ethoxy )methyl)furo [3 ,2-c] quinolin-4( 5H)-one (2.02). Using the Heck type coupling conditions developed for substrate 1.30, the N-SEM protected material 2.02 was afforded in a yield of 87 %. Deprotection of 2.02 and concomitant conversion to the desired methyl enol ether protected substrate 2.01, was achieved in a yield of 63 % over three steps.
In Chapter three, reactivity of the exo double bond of the furan ring in the furoquinolinone key intermediates, 1.30 and 2.01, was investigated with the diazo reagents 2-diazopropane, ethyl 2-diazopropanoate and dimethyl diazomalonate. No reaction was observed when investigating the reaction of 2-diazopropane with either substrate 1.30 or 2.01. However, the desirable cyclopropane ring carbon skeleton was installed when reacting ethyl 2-diazopropanoate with substrate 1.30 or 2.01 giving cyclopropanated products ethyl 5,7-dimethyl-6-oxo-5,6b,7,7atetrahydro- 6H-cyclopropa[ 4,5]furo[3,2-c ]quinoline-7-carboxylate (3.01) and ethyl 6-methoxy- 7-methyl- 7, 7 a-dihydro-6bH-cyclopropa[4,5]furo[3,2-c]quinoline- 7- carboxylate 3.02, respectively. Reaction of the N-methyl substrate 1.30 with dimethyl diazomalonate, gave no detectable cyclopropanated material, but interestingly two novel insertion products dimethyl 3-(3-methoxy-2- ( methoxycarbony 1)-3-oxoprop-1-enyl)-5-methyl-4-oxo-4,5-dihydrofuro [3 ,2- c ]quinoline-2,2(3H)-dicarboxylate (3.03) and tetramethyl 3-(2,2- bis(methoxycarbonyl) -5-methyl-4-oxo 2,3,4,5tetrahydrofuro[3,2-c]quinolin-3- yl)cyclopropane-1,1,2,2-tetracarboxylate (3.04) were obtained. Structure 3.03 showed addition of two dimethyl malonate carbenes, while the minor component's structure 3.04 showed addition of three dimethyl malonate carbenes. While reaction of the 0-methyl substrate 2.01 with dimethyl diazomalonate gave no cyclopropanated material, a pyrano substrate dimethyl 5-methoxy-2H-pyrano[3,2-c]quinoline-2,2-dicarboxylate (3.05) was isolated instead. This product showed that addition of one dimethyl malonate carbene had occurred. Synthesis of a third key intermediate containing a methyl ester functionality linked to the C-3 position of the furan ring of 1.30, was attempted by using the newly developed synthetic approach. This gave however only low yields of the desired product methyl 5- methyl-4-oxo-4,5-dihydrofuro[3,2-c]quinoline-3-carboxylate (3.08). The synthesis of this third key intermediate 3.08 as a whole was therefore discarded. For the investigation of the exo double bond of the furan ring of substrates 1.30 and 2.01 to be complete, electrophilic aromatic substitution reactions such as; bromination, Vilsmeier Haak formylation and Friedel Crafts acetylation, were conducted. The N-methyl protected substrate 1.30 showed high reactivity towards bromination conditions and two products, a mono-brominated 2-bromo-5-methylfuro[3,2-c]quinolin-4(5H)-one (3.12) and a bis-brominated 2,8-dibromo-5-methylfuro[3,2-c]quinolin-4(5H)-one (3.13) product, were isolated in a ratio of 65:35. After optimisation, the mono-brominated substrate 3.12 was selectively synthesised and isolated in a 94% yield. The other two selected reagents showed poor reactivity towards the N-methyl substrate 1.30. The 0-methyl substrate 2.01 showed poor site selectivity under bromination conditions and the protecting group, the methyl enol ether, was cleaved to give a mixture of products. However, the mono-brominated product 2-bromo-5-methylfuro[3,2-c]quinolin-4(5H)-one (3.14) was isolated in a 23 % yield. The other two selected reagents showed even lower site selectivity and converted the 0-methyl key intermediate 2.01 to its precursor in excellent yields (100% conversion using formylation, and 80% conversion using acetylation conditions).
In Chapter four, the ester functionality of the cyclopropanated material 3.01 was reduced to a neopentyl alcohol 7-(hydroxymethyl)-5,7-dimethyl-5,6b,7,7atetrahydro- 6H-cyclopropa[ 4,5]furo[3,2-c]quinolin-6-one (4.01). Preliminary results when attempting to deoxygenate the alcohol, gave ring opened rearrangement products 2-isopropyl-5-methylfuro[3,2-c]quinolin-4(5H)-one (4.04) and 2- isopropenyl-5-methylfuro[3,2-c]quinolin-4(5H)-one (4.05) being isomers of the natural product almeine. The product 4.04 was a result of a cyclopropane collapse to give an isopropene unit off the C-2 carbon of the furoquinolinone core structure. Further, the product 4.05 consisted of an isopropanyl unit off the C-2 carbon. A study was undertaken using other literature methods to give the gem dimethyl functionality under various conditions, however after numerous attempts the desired product was not afforded. It was discovered though that using either thioacetic acid or thiobenzene as nuchleophiles under Mitsunobu conditions, the neopentyl alcohol functionality was able to be displaced to give the corresponding thio esters S-[(5, 7-dimethyl-6-oxo-5, 6b, 7, 7 a-tetrahydro-6Hcyclopropa[ 4,5 ]furo [3,2-c]quinolin- 7-yl)methy l]ethanethioate (4.06) and 5,7- dimethyl- 7-[(phenylsulfanyl)methyl]-5,6b, 7, 7a-tetrahydro-6Hcyclopropa[ 4,5]furo[3,2-c]quinolin-6-one (4.09) in a 74 and 65% yield, respectively. Several attempts to reductively cleave the thio esters to give the gem dimethyl functionality failed. Finally, in an alternative strategy to install the gem dimethyl functionality, it was envisaged the neopentyl alcohol could first be oxidised to the corresponding aldehyde followed by reduction to give the desired product. When this strategy was attempted, only complex mixtures were observed with indication of a ring opened byproduct.
In Chapter five, a new synthesis strategy was elaborated on to give the natural compound 1.01. This strategy incorporated a vinyl alkylation of an 0-methyl protected compound 4-hydroxy-2-methoxyquinoline-3-carbaldehyde (5.04) prepared by literature methods, using l-bromo-z-methylpropene, This was envisaged to be followed by a novel tandem cyclisation/cyclopropanation reaction to furnish the natural product carbon scaffold. Several attempts to install the vinyl ether onto 5.04 failed, and therefore the strategy was modified. The phenol type alcohol (5.04) was first alkylated with an appropriate allyl group (l-bromo-zmethylprop- 2-ene) to give compound 2-methoxy-4-[(2-methylprop-2-en-1-yl) oxy]quinoline-3-carbaldehyde (5.13), which was then envisaged to be isomerised to the corresponding vinyl ether 2-methoxy-4-[ (2-methylprop-1- enyl)oxy]quinoline-3-carbaldehyde (5.03). Even though several attempts were made to isomerise the allyl ether to the corresoponding vinyl ether 5.03 using literature methods, no product was able to be observed. The outlined synthetic strategy was however, proven to be sound since the tandem cyclisation/cyclopropanation reaction for the model case 5.13, gave the desired product 2-methoxy-9a-methyl-1, 1 a,9,9a-tetrahydrocyclopropa[ 4,5]pyrano[3 ,2-c]quinoline (5.14).