Synthetic approaches to syn-diol containing biologically active compounds /

The deoxy derivative of pancratistatin 1.10 was prepared in good yield through the use of a [4+2] Diels-Alder cycloaddition and Bischler-Napieralski cyclization approach. The Bischler-Napieralski cyclization was shown to yield two additional side products 2.9, 2.10, however, under slightly modified hydrolysis conditions, the tetracyclic product 2.11 was obtained exclusively in greater than 84% yield. Initial screening of the di-hydroxylatgd derivative, and the other complementary pair analogue 1.10' previously prepared in our laboratories gave interesting results. Both of these compounds were shown to exhibit cytostatic activity; the mono-alcohol was marginally active while the di-hydroxylated analogue proved to be more potent although one to two magnitudes less potent than pancratistatin itself Human tumour cell line assay results indicated that the di-hydroxylated derivative exhibited selective cytotoxic inhibition in the following cell lines: non-small cell lung cancer line NCI-H226 (ED50 - 0.65 ^g/mL), leukemia cell lines CCRF-CEM (ED30 = 0.55 Hg/mL) and HL-60(TB) (ED50 = 0.89^ig/mL). Our results demonstrated that the pharmacophore is not a mono-alcohol, and that the minimum pharmacophore contains the hydroxyl group at the C4 position in addition to either, or both, of the hydroxyl groups present at C2 and C3.' The minimum pharmacophore has been narrowed to only three possibilities which are current synthetic targets in several research groups. The controlled Grignard addition to the tartaric acid derived bis-Weinreb amide 1.25 afforded a direct entry to a host of 1,4-diflferentiated tartaric acid derived intermediates (2.12-2.18). This potentially usefiil methodology was demonstrated through the efficient synthesis of the naturally occurring lactone 2.23, which bears the inherent syn-dio\ subunit. Based on this result, a similar approach to the synthesis of syn-dio\ bearing natural products looks very promising? A direct 2,3-diol desymmetrization method using TIPS-triflate was shown to be effective on the selective differentiation of Z,-methyl tartrate (and diisopropyl tartrate). The mono-silyl-protected intermediates 2.31 also proved to be useful when they were selectively differentiated at the 1,4-carboxyl position (2.35, 2.36) through the use of a borohydride reducing agent. Furthermore, the mono-silyl-protected derivative underwent periodate cleavage affording two synthetically useful a,P-unsaturated esters 2.43, 2.44, with one of esters being obtained via a silyl-migration method.''

Synthetic studies on prostaglandins: "synthesis of a new thia-PGEI analog"

A PGE1 analog, namely (±)-trans-2-(6'-carbomethoxyhexyl)-3- (E-3"-thia-1 "-octene)-4-hydroxycyclopentanone 71, has been prepared for the first time. Towards the synthesis of this compound, several synthetic approaches aimed at the preparation of the required acetylenic and E-halovinylic sulfides as building blocks were investigated. Among all the methods examined, it appeared evident that the best route to ethynyl n.pentyl sulfide 81 is via a double dehydrohalogenation of the corresponding 1,2-dibromoethyl sulfide with sodium amide in liquid ammonia. In addition, the isomerically pure E-2-iodoethenyl n.pentyl sulfide 85 is conveniently prepared in high yield and stereoselectivity by hydrozirconation-iodination of the terminal ethynyl sulfide 81. The classical hydroalumination and hydroboration reactions for the preparation of vinyl halides from alkynes gave only small yields when applied as methods towards the synthesis of 85 . The building block 2-(6'-carbomethoxyhexyl)-4-hydroxy-2- cyclopentenone (±)-1 carrying the upper side-chain of prostaglandin E 1 was prepared by a step-wise synthesis involving transformations of compounds possessing the required carbocyclic framework (see scheme 27). The synthesis proved to be convenient and gave a good overall yield of (±)-1 which was protected as the TH P-derivative 37 or the siloxy derivative 38. With the required building blocks 81 and 37 in hand, the target 1S-thia-PGE1 analog (±)-71 was prepared via the in situ higher cuprate formation-conjugate addition reaction. This method proved to be convenient and stereospecific. The standard cuprate method, involving an organocuprate reagent generated from an isolated vinyl iodide, did not work well in our case and gave a complicated mixture of products. The target compound will be submitted for assessment of bio log ical activity.

A G. C./M. S. study of the reaction and decomposition products of pentafluorophenyl-grignard and lithium reagents

Decomposition and side reactions of, and the synthetic use of, pentafluorophenylmagnesium bromide and pentafluorophenyllithium have been investigated using G,C9/M.S, techniques• Their reactions with reagents such as CgF^X (X - H, F, CI, Br, 1), C6F4X2 (X - H, CI)f C6F3C13, C6H6. (CgX5)3P (X = H, F), (C6X5)3P=0 (X = H, F), (CgX5)Si (CH3)3 (X = H, F) and (CH0K SiCl , n = 1,2, in ether or ether/n-hexane were studied• In addition to the principal reaction of synthetic use, namely the replacement of a halogen by a pentafluorophenyl group, two types of side reactions were observed* These were (i) intermolecular loss of LiF via a nucleophilic substitution, and (ii) intramolecular loss of LiF, followed by the addition of either inorganic salts such as lithium or magnesium halides, or organometal compounds such as organolithium or organo-Grigaard* G.C«/M.S. techniques were routinely employed to study complicated reaction mixtures. Although mass spectrometry alone has disadvantages for the identification of isomers, deduction of the most probable pathway often helps overcome this problem.