|dc.description.abstract||Organic synthesis has a rich history, ever since Friedrich Wohler’s synthesis of urea from ammonium cyanate in 1828 that gave birth to this important discipline of science. While organic synthesis has found many application and witnessed many triumphs in improving the quality of life, it is the creative instinct, reminiscent of art, associated with this discipline that holds special appeal. This creative flair finds its best expression in the domain of natural product synthesis, which has witnessed spectacular advances and attainments, particularly in the second half of the 20th century .
Many natural products exhibit wide range of biological activities and thus provide useful leads for drug discovery. But very often, they are available only in minute quantities from natural sources and access to them poses threat to biodiversity conservation. Therefore, it is sometimes necessary to synthesize bioactive natural products to obtain requisite quantities and build diversity around their scaffold to explore their therapeutic potential. Thus, natural product synthesis provides an opportunity to organic chemists not only to demonstrate their creativity and intellectual ability but also render available materials for possible application in human health and wellbeing. It is not surprising that the study of the chemistry, biology and synthesis of natural products has emerged as one of the most flourishing and rewarding frontiers in modern science.
The efforts delineated in this thesis are the continuation of our research group’s long standing interest directed towards the total synthesis of structurally complex and biologically promising natural products. The present thesis entitled “Towards the total
synthesis of polycyclic polyprenylated acyl phloroglucin (PPAP) natural products : garsubellin A and hyperforin” is organized in two parts.
PART A : Towards the total synthesis of garsubellin A and PART B : Towards the total synthesis of hyperforin.
PART A : TOWARDS THE TOTAL SYNTHESIS OF GARSUBELLIN A Part A deals with the studies towards the total synthesis of garsubellin A 1, a polycyclic polyprenylated acyl phloroglucin (PPAP) natural product. In the year 1997, garsubellin A was isolated from the wood of garcinia subelliptica by Fukuyama and coworkers. Structurally, garsubellin A belongs to a small but rapidly growing class of natural products characterized by the presence of a highly oxygenated and densely functionalized bicyclo[3.3.1]nonan-2,4,9-trione core embellished with more than one hydrophobic prenyl side chain. Apart from its enchanting structural architecture, garsubellin A 1 also exhibits promising biological activity. Neurodegenerative diseases like Alzheimer’s have been attributed to the deficiencies in the level of neurotransmitter acetylcholine (ACh). Any inducer of the enzyme choline acetyltransferase (ChAT), which is responsible for the biosynthesis of acetylcholine has therapeutic potential in the area of neuro degeneration. Preliminary investigations have indicated that 1 enhances in vitro choline acetyltranferase (ChAT) activity in P10 rat septal neuron cultures by 154% at 10 μM concentration. In view of this unique bioactivity and complex structural architecture garsubellin A 1 attracted immediate attention from the synthetic organic chemistry community. We too, entered this arena being inspired by the complexity and the biological activity of 1. It is relevant to mention that garsubellin A 1 is a typical example of PPAP class of natural products like nemorosone 2, clusianone 3, hyperpapuanone 4, enervosanone 5 and garcinol 6, to name a few, Chart 1
Chart 1 (Fig)
Garsubellin A 1 Nemorosone 2 Clusianone 3
Our first generation synthetic strategy towards garsubellin A 1 is depicted in Scheme 1. Retrosynthetically, it was envisaged that the functionalization pattern in the bicyclic core of garsubellin A 1 could be accessed from an appropriately functionalized
bicyclo[3.3.1]nonene-9-one 7. Bicyclic nonane(-)-7, in turn could be obtained from di-prenylated cyclohexanone derivative (+)-8 through regio-selective allylation followed by Kende cyclization. Chiral cyclohexanone derivative (+)-8 was to be prepared through a series of chemical transformation from campholenic aldehyde (+)-9, readily available from monoterpene chiron (-)-(α)-pinene 10, Scheme1.
Scheme 1 (Fig)
Retrosynthetic analysis for garsubellin A
The choice of α-pinene, available in both the enantiomeric forms, as the starting material provided the opportunity to devise the first enantiospecific approach toward garsubellin A 1 and offered opportunity to address the issue of absolute configuration of the natural product and its siblings
(-)-α-pinene 10 was reconstructed into (+)-campholenic aldehyde 9 through epoxidation and Lewis acid mediated fragmentation as described in literature. OsO4 mediated dihydroxylation of (+)-9 followed by Wittig olefination furnished (+)-11 as a a single diastereomer. Compound (+)-11 was subjected to oxidative cleavage in the presence of NaIO4 to furnish a keto-aldehyde which upon base mediated aldol cyclization furnished cyclohexenone (+)-12. Luche reduction in (+)-12 was stereoselective and the resulting allylic alcohol was subjected to a stereospecific orthoester Claisen rearrangement (Johnson modification) to deliver (+)-13. Base promoted hydrolysis of ester (+)-13 furnished the carboxylic acid which under standard iodolactonization protocol afforded iodolactone (+)-14. Reductive deiodination in (+)-14 in presence of TBTH led to a lactone which upon stereoselective DIBAL-H reduction delivered lactol (_)-15. Lactol (-)-15 was bearing the masked aldehyde functionality which was required to install the second prenyl side chainIsopropylidlene Wittig olefination of the lactol (-)-15 proceeded as planned to give cyclohexanol (+)-16 with two prenyl sub units at the desired positions, Scheme 2. Oxidation in (+)-16 with PCC led to cyclohexanone (+)-8 and further allylation employing NaH was stereoselective and furnished a single diastereomer (+)-17 to set the stage for Kende cyclization. To execute the Kende cyclization, cyclohexanone derivative (+)-17 was transformed to its TMS enol ether and subjected to Pd+2 mediated Kende cyclization protocol to furnish(-)-7 in moderate yield, Scheme 3.
Scheme 3 (Fig) (+)-8 (+)-17 (-)-(-)-7
Reagents and conditions : i) PCC, DCM, 0 oC, 1 h, 98 % ; ii) NaH, allyl bromide, THF, 60 oC, 4 h, 70 % ; iii) LDA, TMSCl, THF, -78 oC, 1 h ; iv) Pd(OAc)2, CH3CN-DCM, rt, 12 h, 30 % (over two steps) .
Having demonstrated an enantiospecific route to the bicyclo[3.3.1]nonan-9-one core (-)-7 present in garsubellin A 1 from(-)-α-pinene, efforts were directed to build the oxyfunctionalization pattern present in the natural product. But various oxidative maneuvers on (-)-7 were not successful and we did not succeed in introducing the key enone functionality by employing allylic oxidation. As a result, we had to explore an alternative synthetic approach that could provide a short access to the core structure present in polyprenylated acyl phloroglucin natural products with appropriate functionalization.
Our second generation approach depicted in scheme 4, emanated from commercially available dimedone. Retrosynthetically, it was envisaged that garsubellin A 1 could be elaborated from an appropriately functionalized bicyclo[3.3.1]nonan-9-one derivative 20 which in turn could be accessible from enol lactone 19. The enol lactone 19 could be made from dimedone 18 in only three carefully crafted steps, Scheme 4.
Scheme 4 (Fig)
Retrosynthetic analysis for garsubellin A Sequential addition of methyl acrylate and prenyl bromide to dimedone 18 in the presence of DBU led to the formation of 21 in a single-pot operation. Acid catalyzed hydrolysis of 21 delivered carboxylic acid 22 and was transformed into enol lactone 23 following standard enol lactonisation protocol. Quenching the kinetic enolate derived from 23 with prenyl bromide introduced the second prenyl group to deliver 24 in a stereoselective manner. Enol lactone 24 was subjected to DIBAL-H reduction to trigger the retro-aldol/re-aldol cyclization cascade. In the event, the anticipated bicyclo[3.3.1]nonane diol 25 was realized and its structure was secured through regioselective derivatization to a crystalline monoacetate 26 and X-ray crystalstructuredetermination,Scheme 5.
Scheme 5 (Fig) 18 21 22 23 26 25 24 Reagents and conditions : i) a) DBU, methyl acrylate, THF, rt, 3 h ; b) DBU, prenyl bromide, THF, rt, 3 h, 70 % (over two steps) ; ii) conc.HCl, acetone-H2O, 50 oC, 12 h, 87% ; iii) NaOAc, Ac2O, 140 oC, 1 h, 92 % ; iv) LHMDS, prenyl bromide, THF, 78 oC,1h,62% ; v) DIBAL-H, DCM, 0 oC, 2 h, 52 % ; v) Ac2O, Et3N, DMAP, DCM, 0 oC, 3 h, 92 % .
At this stage, attention was turned to address the issue of installation of the C-7 prenyl group on the bicyclic skeleton by employing a similar strategy. Towards this end, the methyl enol ether derivative of dimedone was subjected to prenylation under kinetically controlled condition to furnish 27 in excellent yield and was converted to xiv28 in a four steps sequence. Apart from acid catalysed hydrolysis of 27 to afford the cyclohexa-1,3-dione derivative, the other three steps were exactly similar to those depicted in Scheme 5. DIBAL-H reduction of 28 led to the desired structural reconstitution along with the concomitant reduction of the bystander carbonyl group to afford a mixture of bicyclic diols 29. The diol mixture was regioselectively protected as acetate and diastereomeric hydroxy group was oxidized by PCC to deliver bicyclic diketo acetate 30, Scheme 6.
Scheme 6 (fig)
Reagents and conditions : i) TiCl4, MeOH, 0 oC-rt, 1 h, 85 % ; ii) LDA, prenyl bromide, THF, 78 oC-rt, 12 h, 90 % ; iii) conc.HCl, acetone-H2O, 12 h, 86 % ; iv) a) DBU, methyl acrylate, THF, rt, 3 h b) DBU, prenyl bromide, THF, rt, 3 h, 49 % (over two steps) ; v)
conc.HCL, acetone-H2O, 60 oC, 12 h, 85 % ; vi) NaOAc, Ac2O, 140 oC, 1 h, 69 % ; vii) DIBAL-H. DCM, 0 oC, 2 h, 46 % ; viii) Ac2O, Et3N, DMAP, DCM, 0 oC, 1 h, 76 % ; ix) PCC, DCM, rt, 2 h, 82 % .
In this model study, we installed the C-7 prenyl group but the stereochemistry at this center was found to be epimeric as compared to the natural product garsubellin A 1. Thus, our strategy needed further rectification. Moreover, in the approach depicted in Scheme 6, an additional unwanted oxy functionality at C-6 was also generated.
A refined second generation approach was thus devised. Accordingly, dimedone 18 was elaborated to a phenyl thio ether derivative 31 following the literature procedure. Compound 31 was transformed to 33 via kinetic prenylation product 32 followed by 1,3-transposition of the carbonyl functionality. Enone double bond of 33 was reduced and the resulting cyclohexanone derivative was prenylated under kinetically controlled conditions to afford 34. Michael addition of methyl acrylate to 34 completed the sequential geminal alkylation and generation of the key quaternary centre. Since, the stereochemistry of alkylation in 34 was determined by the pre-existing prenyl group, the sequence of prenylation and Michael addition in 34 can be harnessed for the installation of requisite stereochemistry. Ester hydrolysis followed by enol lactonisation gave 19 which was subjected to DIBAL-H reduction to deliver the bicyclo[3.3.1]nonane derivative 35 as an epimeric mixture through retro-aldol/re-aldol reaction cascade. Oxidation in 35 with PCC furnished the diketo compound 20 which was no longer bearing any excess oxy functionality. Enone functionality was introduced in 20 to furnish 36 by employing Saegusa’s protocol, Scheme 7.
Scheme 7 (Fig)
Reagents and conditions : i) LDA, prenyl bromide, THF78 oCrt, 12 h, 95 % ; ii) LAH, THF, rt, 1 h ; iii) HgCl2, CH3CN-H2O (5:1), 60 oC, 1 h, 64 % (over two steps) ; iv) NiCl2, NaBH4, MeOH, 0 oCrt, 1 h, 94 % ; v) LDA, prenyl bromide, THF, -78 oC, 72 % ; vi) methyl acrylate, KOtBu, C6H6, 30 min, 71 % ; vii) KOH, MeOH, H2O, 60 oC, 1 h, 93 % ; viii) NaOAc, Ac2O, 140 oC,1h,79%;ix)DIBAL-H,DCM,0 oC,2h,67%;x)PCC,DCM,rt,1h,95%;xi)Et3N, DMAP, TMSOTf, DCM, 0 oC, 1 h ; xii) Pd(OAc)2, CH3CN, 6 h, 60 oC, 59 % (over two steps).
Nucleophilic epoxidation in 36 led to a single epoxide and the α-epoxy ketone was converted to β-hydroxy ketone 37 through reductive cleavage in the presence of NaSeH. The intent was to oxidize 37 to a 1,3-diketo compound and to use the highly enolizable 1,3-diketo functionality for selective functionalization of the bridgehead prenyl group to generate the tetrahydrofuran ring present in garsubellin A 1.
Scheme 8 (Fig)
Reagents and conditions : i) H2O2, NaOH, MeOH, 0 oC, 1 h, 85 % ; ii) PhSePhSe, NaBH4, EtOH, 0 oC, 1 h, 71 % ; iii) PCC, DCM, rt, 1 h, 76 % ; iv) Et3N, DMAP, TMSOTf, DCM, 0 oC, 1 h ; v) Pd(OAc)2, CH3CN, 60 oC, 6 h, 55 % (over two steps) ; vi) PPTS, MeOH, 0 oC, 30 min, 88 % .
Oxidation of 37 in the presence of PCC directly and quite unexpectedly furnished 38 having a tetrahydrofuran ring. Compound 38 was converted to 39 in a three steps xvi sequence involving Saegusa’s protocol followed by deprotection of tertiary OTMS group, Scheme 8.
This was a welcome outcome as prenylation in 39 at C-3 position could lead us to the advanced intermediate of Danishefsky and a formal total synthesis of garsubellin A 1. However, the crystal structure of compound 38 revealed that the stereochemistry at C-18 was epimeric compared to that of the natural product. At this stage, we decided to revert back to our original proposition to install 1,3-dicarbonyl functionality by oxidation of 37. Many additional efforts in this direction did not bear fruit. Repeated failure to generate the requisite 1,3-dicarbonyl functionality forced us to look at alternatives and it was decided to explore Effenberger cyclization to achieve our desired 1,3-diketo functionality in direct way and in a much shorter sequence. To adopt this shorter sequence, we went back to compound 34 which was transformed to its TBS enol ether 40 and was exposed to malonyl dichloride under carefully controlled conditions to afford a non separable mixture of regioisomers 41 and 42. This mixture of regioisomers was converted to their methyl enol ethers 43 and 44 and could be readily separated to furnish the two regioisomers in equal amounts, Scheme 9.
Scheme 9 (Fig)
Reagents and conditions : i) Et3N, DMAP, TBSOTf, DCM, 0 oC, 98 % ; ii) malonyl
dichloride, DCM, 10 oC, 24 h ; iii) TMS-CHN2, Et2O, 0 oC, 1 h, 31 % (over two step,
43:44=1:1) ; iv) PTSA, HC(OMe)3, MeOH, 50 oC, 48 h, 67 % .
The structures of two isomers 43 and 44 could be assigned by comparing with the spectra reported recently in the literature. Among these two diastereomers, 44 was useful for selective functionalization of the bridgehead prenyl group but the formation
of 43 was not fully unwelcome as it could be isomerized to 44 under thermodynamic conditions, Scheme 9.
Further efforts are on to achieve the selective functionalization of the bridgehead prenyl side arm in 44 which will lead to the formation of tetrahydro-furan ring with correct stereochemistry at C-18 centre. In summary, we have demonstrated the first enentiospecific approach towards garsubellin A 1 starting from ()--pinene. We have also delineated a protocol for the construction of the bicyclo[3.3.1]nonan-9-one framework present in garsubellin A 1 in 5 steps with overall yield of 17 % starting from dimedone. By using the same strategy
We have synthesized the 18-epi-tricyclic core present in garsubellin A 1. PART B : TOWARDS THE TOTAL SYNTHESIS OF HYPERFORIN The use of St. John’s Wort in the treatment of mild to moderate depression has been known for long time. The prominent constituent of the St. John’s Wort is hyperforin 45, a polycyclic polyprenylated acyl phloroglucin natural product (PPAP), which is also responsible for the biological activity. Structurally,
Hyperforin is characterized by a highly oxygenated and densely substituted bicyclo [3.3.1]nonan -2,4,9-trione core embellished with several prenyl and one homoprenyl subunit.
Chart 2 (Fig)
The major structural difference between garsubellin A 1 and hyperforin 45 is with respect to the C-8 quaternary centre. Garsubellin A 1 is embodied with a gem-dimethyl group at C-8 centre whereas in hyperforin 45 bears a stereogenic C-8 centre with a methyl and a homoprenyl substituents. Probably because of this stereogenic C-xviii 8 centre, hyperforin and its analogues (Chart 2) constitute thorny synthetic challenges and remains to this day defiant to chemical synthesis. But, the remarkable biological as well as structural features associated with this molecule make hyperforin 45 as a tempting target for total synthesis. So far, very little attention has been paid towards it’s total synthesis and no one has yet addressed the crucial issue of setting up the C-8 stereogenic centre.
For our synthetic approach towards hyperforin 45, it was crucial to identify a starting material in which the key C-8 quaternary centre was pre-installed. Towards this end, the cyclohexane 1,3-dione derivative 53 was identified as the synthon and our synthetic strategy towards hyperforin was designed in a manner that could give expression to an experiences in the quest for garsubellin A. Hyperforin 45 synthesis was sought to be accomplished from fully embellished bicyclic ketone 51. Its precursor enol lactone 52 could readily be transformed to bicyclic ketone 51 via DIBAL-H promoted retro-aldol/re-aldol cascade cyclization pathway. It has already been demonstrated that the type of enol lactone like 52 could be accessed from cyclohexane-1,3-dione derivative 53 through appropriate stereocontrolled chemical steps, Scheme 10.
Scheme 10 (Fig)
Retrosynthetic strategy for hyperforin Since the cyclohexane-1,3-dione derivative 53 has not been reported in the literature, a straight forward multi-gram access to it from commercially available citral 54 was devised. Citral 54 was transformed to -unsaturated ketone 55 viaMeLi addition and MnO2 oxidation. Tandem Michael addition-Claisen condensation in 55 with diethyl malonate delivered a cyclic intermediate 56 which upon further decarboxylation furnished the requisite 1,3-dicarbonyl compound 53 in moderate yield, Scheme 11.
One-pot tandem Michael addition with methyl acrylate and prenylation in 1,3-diketone 53 in the presence of DBU led to a readily separable mixture of diastereomers 57 and 58 in a ratio of 1.2 :1, Scheme 11.
Scheme 11 (Fig)
Reagents and conditions : i) MeLi, 0 oC, Et2O, 82 %, ii) MnO2, rt, 12 h, 72 % ; iii) CH2(CO2Et)2, NaOEt, EtOH, 60 oC, 12 h; iv) KOH, MeOH, 60 oC, 72 h, 58 % (2 steps), v) a) DBU, methyl acrylate, THF ; b) prenyl bromide, rt, 6 h, 48% (over two steps); (57 : 58 = 1.2 :1)
It was decided to first proceed with the diastetreomer 58. The diastereomer 58 washydrolyzed to the carboxylic acid and the acid was further elaborated to enol lactone 59. At this stage, it was felt useful to introduce an additional prenyl sub-unit that would eventuate at the C-3 position in hyperforin 45. This was readily realized through kinetic deprotonation of 59 and a prenyl bromide quench to furnish 60 through 1,3-stereoinduction attributable to the pre-existing bridgehead prenyl group. Chemoselective DIBAL-H reduction of the lactone moiety in 60 initiated the thermodynamically controlled retro aldol/re-aldonization cascade to furnish the desired bicyclo[3.3.1]nonan-9-one scaffold 61 in a moderate yield, Scheme 12.
Scheme 12 (Fig)
Reagents and conditions : i) conc.HCl, acetone-H2O, 60 oC, 88 % ; NaOAc, Ac2O, 140 oC, 1 h, 75 % ; iii) LDA, prenyl bromide, THF, 78 oC,1h,51%;iv)DIBAL-H,DCM,0 oC,2, 54 % .
This was a very satisfying outcome as in 61 we not only had correctly installed the C-8 stereogenic centre but also adequate functionality in the two bridges for further elaboration to the target structure. However, oxidation of 61 to the triketone and introduction of the C-7 prenyl group through α-prenylation proved unproductive.
Recourse was further taken to a modified strategy to install the C-7 prenyl group. According to this plan, the 1,3-dicarbonyl moiety in 53 was readily elaborated to the methyl enol ether and the prenylation of this compound under kinetically controlled conditions led to a diastereomeric mixture of enol ethers 62 and 63 (1.2:1). As anticipated, there was only marginal diastereoselection during this alkylation, but the two stereogenic centres corresponding to C-7 and C-8 of the target structure were now duly installed. Among two diastereomers, the right one was picked up and elaborated towards the target structure. Compound 63 was converted to enol lactone 64 in a four steps sequence as described previously. Enol lactone 64 was reduced with DIBAL-H to trigger the desired structural rearrangement and furnish a mixture of diastereomers. PCC oxidation of the diol readily afforded the tricarbonyl compound 65, Scheme 13.
Scheme 13 (Fig)
Reagents and conditions : i) TiCl4, MeOH, 0 oC -rt, 90%, ii) LDA, prenyl bromide, THF, -78 oC -0 oC, 12 h, 90%; iii) conc.HCl, acetone, rt, 12 h, 83%; iv) a) DBU, methyl acrylate, THF; b) prenyl bromide, rt, 6 h, 45%; v) conc.HCl, acetone-H2O, 50 oC, 12 h, 87%; vi) NaOAc, Ac2O, 140 oC, 1 h, 70%; vii) DIBAL-H, DCM, 0 oC, 2 h, 41%; viii) PCC, DCM,0 oC-rt,1h,70%.
From structural perspective, the C-7 prenyl (exo-) and C-8 homoprenyl (endo-) are trans-in hyperforin 45 but quite interestingly in guttiferone A 48 and hypersampsone F 49, the C-7 prenyl is endo-and the C-8 homoprenyl is exo-, although their trans relationship is retained. Therefore, the stereochemical disposition of prenyl and homoprenyl groups at C-7 and C-8 centre respectively in 65 matched with the basic skeleton of guttiferone A 48 and hypersampsone F 49 instead of hyperforin 45. In order to realize the requisite C-7, C-8 stereochemistry of the target molecule hyperforin 45, it was considered essential to manipulate the sequence of single pot tandem dialkylation and once again we ventured to build upon our experience in context of the synthesis of garsubellin A 1. Accordingly, 53 was elaborated to the ethyl enol ether derivative 66 and its prenylation under kinetically controlled conditions afforded a readily separable mixture of diastereomers 67 and 68 (1.2:1). Following Prenylation, one of the diastereomers 68 was subjected to DIBAL-H reduction to furnish a mixture of allylic alcohol and this mixture on acid catalysis furnished the enone 69. Enone double bond of 69 was stereoselectively and regioselectively reduced to afford tri-alkylated cyclohexanone 70. LDA mediated prenylation on 70 went in a straightforward manner to deliver a mixture of diastereomers which underwent Michael addition to methyl acrylate in presence of KOtBu to furnish 71 as a single diastereomer. To realize the requisite stereochemistry at C-7 and C-8 stereogenic centre of the natural product hyperforin 45, it was inevitable to carry out prenylation first on 70 followed by Michael addition of acrylate in a stepwise manner to afford 71, Scheme 14.
Scheme 14 (Fig)
Reagents and conditions : i) TiCl4, EtOH, 0 oC-rt, 1 h, 83 % ; ii) LDA, prenyl bromide, THF, 78 oC-0 oC, 10 h, 92 % ; iii) DIBAL-H, DCM, 0 oC, 30 min ; iv) conc.HCl, acetone, 0 prenyl bromide, THF, 78 oC,4h,73%;vii)KOtBu, methyl acrylate, C6H6, rt, 15 min, 65 %
It was very satisfying to see that through this maneuver it was possible to install the quaternary centre in a requisite manner to secure the correct stereochemistry at C-7, C-8 present in hyperforin 45. Ester 71 was hydrolysed to carboxylic acid which was routinely transformed to enol lactone 72 under standard enol lactonisation condition. Enol lactone 72 was xxii exposed to DIBAL-H to trigger the retro-aldol/re-aldol reaction cascade and led to a diastereomeric mixture of bicyclic alcohols. This mixture of bicyclio[3.3.1]nonane based alcohols was routinely oxidized to bicyclic diketo compound 73 in good yield and further alkylated with prenyl bromide under kinetically controlled condition to afford 51 stereoselectively, Scheme 15.
Scheme 15 (Fig)
Reagents and conditions : i) KOH, MeOH, 60 oC, 78 %; ii) NaOAc, Ac2O, 140 oC, 1h, 69 %; iii) DIBAL-H, DCM, 0 oC, 2 h, 64%; iv) PCC, DCM, rt, 1 h, 87 %; v) LDA, prenyl bromide, THF, -78 oC, 2 h, 73 %. Overall, this was very pleasing outcome as we could construct the bicyclo[3.3.1]nonan-9-one framework 51 present in hyperforin 45 in which all the prenyl sub-units and the homoprenyl unit are duly installed. Our efforts to introduce the enone functionality in 51 through Saegusa and other related protocols was not successful. Therefore, it was mandatory to ponder over our strategy once again and tactically modify it.
Once again, it was decided to employ the Effenberger methodology. It was greatly advantageous to utilize Effenberger protocol because it was not only expected to shorten the synthetic sequence, but would also directly introduce the 1,3-dicarbonyl functionality in the substrate. Thus, we turned back to the diprenylated cyclohexanone derivative 70 and converted it to the TBS enol ether, which in turn, was exposed to malonyl dichloride under carefully controlled condition to afford an inseparable mixture of two regioisomers. This mixture of regioisomers was converted to the corresponding methyl enol ether derivatives 74 and 75 and these could be readily separable. Even though 74 and 75 were separated, it was difficult to identify them individually through spectroscopic techniques. Therefore, the mixture of isomers was thermodynamically equilibrated to a single isomer which was recognized as 75. This isomer which was serviceable for further elaboration to the natural product.
Thermodynamically more stable isomer 75 was subjected to kinetically controlled, LDA ediated prenylation to furnish 76 in fairly good yield, Scheme 16.In 76, we had arrived at very advanced precursor of the natural product hyperforin 45 and it was time to pass the baton to another colleague in the group to reach the summit.
Scheme 16 (Fig)
Reagents and conditions : i) Et3N, DMAP, TBSOTf, DCM, 0 oC, 1 h, 95 % ; ii) malonyl dichloride, DCM, 10 oC, 12 h ; iii) TMS-CHN2, Et2O, 0 oC, 1 h, 30 % (over two steps, 74:75=1:1) ; iv) PTSA, CH(OMe)3, MeOH, 50 oC, 48 h, 67 % ; v) LDA, prenyl bromide, _78 oC,1h,61%.
In short, construction of basic bicyclo[3.3.1]nonan-9-one framework present in hyperforin 45 has been outlined in which the crucial issue of setting up the C-8 stereogenic centre has been addressed for the first time. Installation of all prenyl and omoprenyl side chains present in the natural product was also achieved. Effenberger cyclization has been successfully employed to access an advance precursor in a shorter sequence but with the higher level of oxy functionalization.||en_US