Biosynthesis of Pyrrolizidine Alkaloids

The work presented in this thesis is concerned with the biosynthesis of the necic acid portions of pyrrolizidine alkaloids and has been divided into five main areas: (a) the development of a general route to a variety of 13C- and 2H- labelled isoleucines; (b) the synthesis of a specifically labelled isoleucine; (c) the synthesis of labelled 2-aminobutanoic acid; (d) feeding experiments with plants and transformed root cultures which produce pyrrolizidine alkaloids; (e) the synthesis of new and known synthanecines for biosynthetic studies and investigation of their anti-tumour activity. (a) Development of a General Route to Isoleucine Isoleucine has been shown to be specifically incorporated into the acid portions of several pyrrolizidine alkaloids. Some acids with ten carbon atoms are formed from two isoleucine units. In the development of a synthetic route to isoleucine it would be desirable to assemble the target compound by the stepwise addition of one and two carbon units. This would allow the same route to furnish a range of isoleucines, isotopically labelled with 2H or 13C at different positions. We decided to modify a published route to valine. However this synthesis required many stages, a few of which were low yielding, making this scheme uneconomic as an all-purpose route to labelled isoleucines. DL-[6-2H3]isoleucine hydrochloride (A) together with the allo-racemate was the only labelled amino acid prepared using this procedure. (b) Synthesis of a Specifically Labelled Isoleucine In the biosynthesis of the acid portion of dicrotaline (B) in Crotalaria lachnosema, isoleucine labels four out of the six carbons. A route was developed to [3-2H]isoleucine hydrochloride (C) and the allo-racemate in an attempt to discover which of the two methylene groups of 3-hydroxy-3-methylglutaric acid is provided by isoleucine. (c) Synthesis of Labelled 2-Aminobutanoic acid 2-Aminobutanoic acid is also a biosynthetic precursor to several necic acids. A short synthesis of this precursor from diethyl acetamidomalonate was developed from literature procedures. Samples of D[3,4-13C2)]-2-aminobutanoic acid hydrochloride (D) and DL[3,4-2H5]-2-aminobutanoic acid (E) were made. (d) Feeding Experiments Feeding experiments were carried out on two plants, Senecio pleistocephalus and Crotalaria lachnosema, and on two transformed root cultures, Senecio vulgaris and Emilia flammea. C. lachnosema produces dicrotaline (B) and 13-O-acetyldicrotaline (F). The crude alkaloid mixture isolated from this plant was converted wholly into 13-O-acetyldicrotaline, and a full 1H and 13C NMR assignment and conformational analysis of this compound (F) were performed. However all feeding experiments on this plant species failed to show any incorporation of stable isotopes into the alkaloid above natural abundance. Successful feeding experiments were carried out with S. pleistocephalus which produces rosmarinine (G). Deuterium (E) and 13C-13C doubly labelled 2-aminobutanoic acid (D) were incorporated into rosmarinine (G) labelling two pairs of carbon atoms, namely C-13 and C-19 plus the C-20 and C-21 positions and equal labelling was observed in both halves of the acid portion. Senecionine (H) from S. vulgaris was also labelled after feeding experiments with 2- aminobutanoic acid. Again the C-13 and C-19 plus C-20 and C-21 positions were labelled with the amino acid incorporated into both halves of the necic acid equally. The root culture E. flammea produces two alkaloids, senecionine (H) and emiline (I). However in the feeding experiment with 2-aminobutanoic acid, only senecionine and a small amount of its geometrical isomer integerrimine (J) were observed. The labelling patterns for senecionine were the same as previously observed with material isolated from S. vulgaris. [diagram] (e) Formation of Synthanecines Synthanecine A (K) and synthanecine B (L) were prepared using known procedures. These synthanecines were fed to S. vulgaris root cultures, but no new alkaloid analogues could be detected. The use of 1,3-dipolar cycloaddition reactions was investigated as a method for the manufacture of synthanecines. The use of trimethylamine N-oxide (M) and lithium diisopropylamide with a variety of alkenes proved unsuccessful. In an alternative procedure N-benzyl-N-(trimethylsilylmethyl)aminomethyl methylether (N) did undergo the desired cycloaddition with diethyl fumarate, and the product was subsequently converted into a novel synthanecine (O)