The Synthesis of Pyrrolobenzothiadiazepines and Related Compounds, and a Study Into the Reaction of Cyclopropenones with 2-Vinyl-1-Azetines and other Cyclic Imines.

The focus of this thesis is split between two projects, the first of which describes the synthesis of novel analogues of pyrrolobenzodiazepines (PBDs). The established biological activity of the PBDs make such analogues worthy targets. Sulfone analogues(I and II)of the natural products fuligocandin B (III) and fuligocandin A (IV) were the first target molecules, as fuligocandin B has shown promising activity against leukaemia cells.

The initial approach to these pyrrolobenzothiadiazepine (PBTD) analogues was based on using the Eschenmoser episulfide contraction as a key step in the synthesis. Whilst this approach proved unsuccessful, analogue II was eventually synthesised via an intermediate diazoketone based method. This route, summarised below, also led to work beginning on the synthesis fuligocandin homologues (V, for example) in a similar manner.

The synthesis of analogue I was attempted using several methods, however these proved ineffective due to low recoveries and unoptimized reaction conditions.

The circumdatin series of natural products was also a target for the development of routes to PBTD analogues. Attempts to synthesise the sulfonyl analogue of circumdatin H(VI) via a range of N-arylations on a PBTD backbone were unsuccessful, before finding success by utilising a copper-catalysed N-arylation to the quinazolino moiety.

As an extension to the synthesis of analogue VI, two new homologues (VII, X = S and X = (CH2)2) of the PBD the natural product circumdatin H (VII,X = CH2) were successfully synthesised through an N-arylation and subsequent aza-Wittig imine formation.

With an effective route to novel PBTDs and PBD homologues, a selection of compounds was selected and tested in vitro against human colon carcinoma cells. Most of the results from these tests were inconsequential, however one compound tested showed impressive selectivity towards the cancer cells and has prompted further research for the future.

The second focus of this thesis explores the reaction between symmetrical diarylcyclopropenones and 4-vinyl-1-azetines, and a subsequent aza-Cope rearrangement. It was observed that when the reaction was carried out at room temperature, the intermediate cycloadduct VIII was isolated along with aza-Cope rearrangement product, the azabicyclo[4.2.1]nonene IX. It was also observed that when carried out at a higher temperature, only azabicyclo[4.2.1]noneneIX was observed. This reaction was carried out using a number of azetines and cyclopropenones, resulting in the isolation of twenty azabicyclo[4.2.1]nonenes, and five intermediate cycloadducts.

Five azabicyclo[4.2.1]nonenes were subjected to biological testing against human ovarian carcinoma cells and it was found that two of these compounds showed promising effects on both cisplatin sensitive and resistant carcinoma cells.

Expanding upon this, cyclopropenones were reacted with the polycyclic imineX, and it was observed that a dimerised compound (XII) was produced, as opposed to the expected cycloadduct (XIa).This was further explored with other cyclopropenones to similar effect. A mechanism is proposed involving the formation of a captodative radical (XIb) at the bridgehead.