Degree Name

Doctor of Philosophy


School of Chemistry


The polyhydroxylated alkaloids are natural heterocycles that contain one or more basic nitrogen atoms and several hydroxy substituents. The hyacinthacine alkaloids are polyhydroxylated 3-hydroxylmethylpyrrolizidines which were named after the plant family, Hyacinthaceae, or bluebell family from which they were isolated. Since the structure of first hyacinthacine alkaloid was published in 1999, nineteen hyacinthacines have been reported, these are hyacinthacines A1-A7, B1-B7, C1-C5, two hyacinthacine A1 derivatives and one hyacinthacine A2 derivative. The hyacinthacine alkaloids show weak to moderate activity against glycosidases. Up to the end of 2013, twelve different hyacinthacine alkaloids, A1, A2, A3, A6, A7, B1, B2, B3, B7, C1, C3 and C5, have been synthesized. These syntheses confirmed the structures of seven of these alkaloids but showed that the original structures assigned to hyacinthacines B7, C3 and C5 were incorrect.

The main aims of this PhD project were to synthesize 5-epi-hyacinthacine B5 and to compare its NMR spectroscopic data with those of natural hyacinthacine B7 and to synthesize hyacinthacine B5 to confirm its structure and absolute configuration.

The total syntheses of 5-epi-hyacinthacine B5 and hyacinthacine B5 were designed to start with commercially available (S)-4-penten-2-ol. In Chapter 2, the preparation of the 1,2-anti amino alcohol 155 using the Petasis boronic acid Mannich reaction is described using a modification of the previous synthetic method that was developed in our group. The new synthetic route, involving the Sharpless ADH of the terminal alkene 151, followed by the regioselective oxidation of the resulting diol 190, and then the Petasis boronic acid Mannich reaction gave 155 in 72% yield over 3 steps. This was a significant improvement over the previous method that proceeded in 24% overall yield over 3 steps.

The base induced E2-elimination-SN2-cyclization of the N-Boc mesylate 176, that was designed to be employed to invert the configuration of the pro-C7 centre, worked efficiently in a model study. Unfortunately, the steric hindrance caused by the α-methyl group of 155 prevented its N-Boc protection. Consequently our synthetic route was changed to follow our previous work on the total synthesis of hyacinthacine B3 with the aim of inverting the stereochemistry at the pro-C7 centre after the pyrrolizidine ring was formed.

The total synthesis of 5-epi-hyacinthacine B5 was accomplished in 15 steps and in 5.58% overall yield form commercially available (S)-4-pentan-2-ol. The 1H and 13C NMR data of 5-epi-hyacinthacine B5 were markedly different to those of natural hyacinthacine B7. In addition, the specific rotation of the natural product ([α] D - 4.4, c 0.20, H2O) and synthetic compound ([α] 25/D + 18.5, c 0.14, H2O) were different in magnitude and opposite in sign. This synthesis confirmed that the structure and absolute configuration of natural hyacinthacine B7 did not match with that of 5-epi-hyacinthacine B5.

The total synthesis of hyacinthacine B5 is described in Chapter 3 starting with the bicyclic oxazolidinone intermediate 160 used in the total synthesis of 5-epi-hyacinthacine B5. This synthesis was accomplished in 14 steps and in 2.21% overall yield form commercially available (S)-4-pentan-2-ol. This total synthesis also produced, hyacinthacine B4 in 15 synthetic steps and in 1.11% overall yield, and 7a-epi-hyacinthacine B3 in 14 synthetic steps and in 2.03% overall yield, as side products. The 1H and 13C NMR chemical shifts of these compounds were in very close agreement with those of natural hyacinthacines B5 and B4, respectively as well as the specific rotation results. These comparisons indicated that the proposed structures and absolute configurations of hyacinthacine B5 and B4, that were published, were the correct ones.

A study of the relationship between the configurations of the methyl group at C5 and the hydroxy group at C7 on the 13C NMR chemical shifts of natural hyachithacine B3, natural hyachithacine B5, synthetic 5-epi-hyacinthacine B5, synthetic hyacinthacine B7 and natural hyacinthacine B7 was made. This comparison showed that the configuration at C5 affected the 13C NMR chemical shifts of C3, C5 and C9. While the configuration at C7 affected the 13C NMR chemical shifts of C7 and C7a. Using this information we suggested that natural hyacinthacine B7 should have an α-methyl group at C5 and an α-hydroxy group at C7, similar to that of natural hyacinthacine B5. We concluded that natural hyacinthacine B7 is most likely to be hyacinthacine B5, however without access to the authentic samples we cannot make an unequivocal conclusion.

In Chapter 4, an attempted synthesis of hyacinthacine B4, starting from 3,5-di-O-benzyl- α,β-L-xylofuranose, is described. Due to time constraints only the first four steps of this synthesis were achieved.