Doctor of Philosophy
Department of Chemistry, Faculty of Science
Batenburg-Nguyen, Tam-Dan, Synthesis of novel compounds based on reticuline scaffold for new drugs discovery, Doctor of Philosophy thesis, Department of Chemistry, Faculty of Science, University of Wollongong, 2005. http://ro.uow.edu.au/theses/1190
This thesis examines the synthesis of a library of benzyl- and bisbenzylisoquinolines (BBI) derivatives based on the reticuline motif. These compounds were assessed for their; i) cytotoxicity on 3 cancer cell lines, ii) activity on HIV-infected cells, iii) antibacterial activity, and iv) CNS receptor binding affinities. Chapter 2 describes the employment of palladium-catalysed Stille, Heck and Sonogashira coupling reactions to synthesise a library of BBI derivatives. 2’-Vinyl- (67), 2’-allyl- (68) and 2’-iodo (58) derivatives of racemic, N-TFA protected, norlaudanosine were used as the key building blocks in this investigation. The key 2’-vinyl- and 2’-allyl-norlaudanosine derivatives 67 and 68, respectively were readily prepared from palladium-catalysed Stille coupling reactions of the 2’-iodonorlaudanosine derivative 58 and vinyl- or allyl-tributylstannane. The Heck coupling reactions between the 2’-vinyl-norlaudanosine derivative 67 and the 2’-iodo- norlaudanosine derivative 58 gave not only the desired stilbene BBI derivative 65 but also the unexpected 1,1-disubstituted regioisomer 69. This unexpected regioisomer was a result of the electron rich nature of both stating materials that favoured a cationic palladium intermediate. The best Heck coupling reaction conditions involved the use of Pd(OAc)2, DMG, NaOAc and NMP at 130 0C. These conditions gave the highest yield and the best regioisomer selectively in favour of the BBI derivative 65. Fortunately these regioisomers were readily separated by triturating the product mixture with methanol. The Heck coupling reaction between the 2’-allylnorlaudanosine derivative 68 and the aryl iodide 58 successfully afforded the three carbon tethered BBI derivative 66 in moderate yield. It was found, however, that these Heck coupling reaction conditions were only efficient with aryl iodide precursors. This was evident from the attempted intramolecular Heck coupling reactions on the aryl bromide precursor 89, to give the macrocyclic BBI derivative 88. The optimised Heck coupling reaction conditions failed to produce the desired product, while more traditional Heck coupling conditions gave the required product in poor yield (15 %). The unsaturated BBI derivative 65 and its regioisomer 69 were subjected to hydrogenation conditions over Pd/C under a hydrogen atmosphere. However, the regioisomer 69 was found to be too sterically hindered and did not undergo the hydrogenation reaction, while derivative 65 encountered solubility problems and only rac-65 underwent the hydrogenation reaction to give rac-80, leaving the less soluble meso-65 intact. The compounds rac-80 and meso-65 were readily separated by column chromatography. Chapter 2 also described the successful synthesis of the targeted acetylinic BBI derivative 63 via coupling of the 2’-ethynylbenzylisoquinoline derivative 84 with the aryl iodide 58, using a Pd/Cu catalysed Sonogashira coupling reaction, followed by N-TFA deprotection of the N-TFA 2’-ethynylbenzylisoquinoline derivative 83. The synthesis of a library of 2’-arylvinyl- and 2’-arylallyl-benzylisoquinolines derivatives using the optimised Heck coupling reaction conditions developed in Chapter 2 is described in Chapter 3. This set of compounds included benzylisoquinolines having either an exocyclic N,N-dimethylamino (92-103) or N-acetamido (104-107) substituent. A third group of compounds (108-111) in this set had the exocyclic amino or amido group completely excluded. It was found that the Heck coupling reaction of the 2’-vinyllaudanosine derivative 67 and the aryl iodides 118, 119, 131 and 135 afforded only one regioisomer, unlike the Heck coupling between 67 and 58 in Chapter 2, which gave the two regioisomers 65 and 69. The Heck coupling reactions between the 2’-allyllaudanosine derivative 68 and the aryl iodides 118, 119, 131 and 135 gave two regioisomers 115a,b; 116a,b; 129a,b and 137a,b, respectively, due to two possible sites of palladium hydride elimination. In Chapter 4, the use of the ruthenium-catalysed CM and RCM reactions toward the successful synthesis of the four carbon tethered BBI derivatives, 138-142, in both unsaturated and saturated forms (via hydrogenation reactions) was described. The synthesis of the analogous two and three carbon tethered BBI derivatives via this method proved less efficient. Chapter 5 reported the synthesis of a library of aminoalkyl benzylisoquinoline derivatives, incorporating both cyclic and acyclic amines (155-162). These analogues were obtained by a simple reductive amination methodology involving the reaction of commercially available amines with the aldehydes 186 and 187, which were generated from the 2’-vinyl- and 2’-allyllaudanosine derivatives 67 and 68, respectively. The initially planned pathway to one of these aldehydes involved the rearrangement of the epoxide 188, however this epoxide was too unstable under the reaction conditions and readily underwent ring opening with m-chlorobenzoic acid. An alternative pathway using oxidative cleavage of the diols 190 and 191, which were generated from dihydroxylation of the 2’-vinyl- and 2’-allyllaudanosine derivatives, 67 and 68, respectively, was found to be more successful for the synthesis of these aldehydes. Chapter 5 also described the synthesis of an additional class of aminoalkyl benzylisoquinoline derivatives, 163 and 164, containing a -amino alcohol moiety. Retro-synthetic analysis showed two possible synthetic pathways which were either via the ring opening of the cyclic sulfates 195 and 196 or via the nucleophillic displacement of the tosylates 197 and 198 with an amine nucleophile. The latter pathway proved more successful and afforded the -amino alcohol derivatives 163 and 164, however, the yields of these reactions should be optimised in future studies. The synthesis of the benzylisoquinoline derivatives containing a nine- and ten- membered heterocylic ring, 165-167, was also described in Chapter 5. The synthesis of these analogues was initially attempted via the intramolecular reductive amination reaction between an aldehyde moiety at the C2’ position of 219 and its free isoquinoline amino group. However, the synthesis of the aldehyde moiety via the hydrolysis of its protected diacetal form was very difficult; therefore an alternative synthesis was developed. This method involved an intramolecular nucleophilic displacement of the chloride of the -chloroacetamides 214 and 215 by the free isoquinoline amino moiety. This method successfully afforded the nine- and ten-membered ring benzylisoquinoline derivatives 165 and 167 in moderate yields (42-57 %). Lithium aluminium hydride reduction of 165 gave the corresponding cyclic diamino derivative 166 in high yield. Some of the benzyl- and bis-benzylisoquinoline derivatives reported in Chapters 2-5 were sent for biological testing for their cytotoxicity on 3 cancer cell lines, activity on HIV-infected cells, their antibacterial activity and CNS receptor binding affinities. The BBI derivatives showed higher activity on cancer cell lines than the corresponding benzylisoquinoline derivatives. Various BBI and benzylisoquinoline derivatives have showed promising CNS-receptor binding affinities, especially for 5HT receptors and more prominently on the 5-HT1B, 5-HT2A and 5-HT7 receptors. At this stage, a clear structure-activity trend could not be discerned and the mode of action of these analogues was not clear. Further results on the awaiting analogues may help to develop pharmacophore models for CNS active compounds in the future, and eventually, allow the design and development of more selective and potent ligands.