Doctor of Philosophy
School of Chemistry
Ritthiwigrom, Thunwadee, The structure of Uniflorines A and B and the total synthesis of Casuarine, Australine and their epimers, Doctor of Philosophy thesis, School of Chemistry, University of Wollongong, 2010. http://ro.uow.edu.au/theses/3134
The polyhydroxylated alkaloids uniflorines A and B were isolated in 2000 from the leaves of the tree Eugenia uniflora L. The common name for this tree is Surinam Cherry and the water soluble extracts of its leaves have been used as an antidiabetic agent in Paraguayan traditional medicine. Uniflorines A and B showed moderate activity in inhibiting the -glucosidases, rat intestinal maltase (IC50 values of 12 and 4.0 M, respectively) and sucrase (IC50 values of 3.1 and 1.8 M, respectively). Uniflorines A and B were deduced from NMR analysis to have pentahydroxyindolizidine structures. In 2004 Davis and Pyne reported the synthesis of the proposed structure of uniflorine A. Unfortunately, the NMR spectroscopic data of this synthetic material did not match with those of natural uniflorine A. Davis and Pyne then assumed that the structure of uniflorine A must be a diastereomer of the initially proposed structure. Several other diastereomers of the proposed structure, that were epimeric in the A-ring, have been synthesised by other researchers. One remaining A-ring diastereomer was the C-2 epimer of the proposed structure. In Chapter 2 of this thesis we report the synthesis of this compound in 11 synthetic steps and in 0.5% overall yield from L-xylose. However the NMR spectroscopic data for this synthesic compound did not match with those of the natural product. We then re-examined the NMR spectroscopic data of the natural product and revised the structures of uniflorines A and B from initially proposed pentahydroxyindolizidines to 1,2,6,7-tetrahydroxy-3-hydroxymethylpyrrolizidines. Uniflorine B was the known alkaloid casuarine, while uniflorine A was tentatively assigned as 6-epi-casuarine. This was confirmed by the synthesis of the enantiomer of 6-epi-casuarine and then 6-epi-casuarine itself. These syntheses are reported in Chapters 2.2 and 2.3. The total synthesis of uniflorine A (6-epi-casuarine) was achieved in 11 steps and in 13% overall yield from L-xylose. The NMR spectroscopic data of this synthetic compound matched with those of the natural product uniflorine A. Thus we had successfully determined the correct structures of uniflorines A and B. Glycosidase inhibitor testing of uniflorine A at 143 M showed it had 94-97% inhibition against the -D-glucosidases of Saccharomyces cerevisiae and Bacillus sterothermophilus and the amyloglucosidase of Aspergillus niger. The IC50 values were only determinated for the two aforementioned -D-glucosidases and were found to be modest at 34 and 28 M, respectively. In addition, we describe a flexible method for the diastereoselective total synthesis of several natural and unnatural polyhydroxylated indolizidines and pyrrolizidines from a common precursor. The synthesis of the alkaloid casuarine, which was obtained in total of 13 synthetic steps and in 8% overall yield from L-xylose, is described in Chapter 3. A key step in this synthesis was a regioselective epoxide ring-opening reaction with hydrogensulfate ion. This reaction secured the correct configurations at C-6 and C-7 of the target molecule. In Chapter 4 we describe the successful synthesis of australine in a total of 14 steps and 6% overall yield from from L-xylose. Key steps in this synthesis were a regioselective epoxide ring-opening reaction with LiAlH4 followed by a Mitsunobu reaction that secured the correct configuration C-7 of the target molecule. The synthesis of the natural product 3-epi-casuarine was completed in 13 steps and in 0.4% overall yield. This synthesis required an inversion of configuration at C-3’ of the butyl side chain which was achieved using the Mitsunobu reaction. The low overall yield was due to a low yielding epoxide ring-opening reaction due to a competing intramolecular epoxide ring-opening reactions involving the 3--hydroxymethyl substituent. Natural 3-epi-australine was obtained in total of 14 synthetic steps and in 2% overall yield, all from L-xylose. This synthesis required an inversion of configuration at C-3’ of the butyl side chain which was achieved using the Mitsunobu reaction. Key steps in this synthesis were a regioselective epoxide ring-opening reaction with LiAlH4 followed by a Mitsunobu reaction that secured the correct configuration C-7 of the target molecule. From this work a number of novel unnatural indolizidine and pyrrolizidine compounds were also obtained as side products. Some of these compounds were screened against 10 different glycosidases at 800 g/mL. Unfortunately, none showed strong inhibition with only four compounds showing approximately 40-50% inhibition at this relative high concentration.
02Whole.pdf (4126 kB)