Synthesis of some furanone derivatives : putative quorum sensing or chitinase inhibitors

Bacteria are able to communicate through chemical signals. They can for instance \nestimate their population by the use of signalling compounds: this phenomenon is called \nquorum sensing. Some of these signalling compounds are derivatives of homoserine lactones. \nWhen their concentration reaches a certain level, bacterial genes are triggered, which leads to \nvirulence, bacterial film formation, and so on. Natural halogenated furanones 22 and 23 \nextracted from the marine alga Delisea pulchra proved to inhibit the quorum sensing. Other \nnatural furanones 24a-27a extracted from Streptomyces antibioticus TÜ 99 were also quorum \nsensing inhibitors. As the structures of these natural compounds were similar to the \nhomoserine lactones, additional furanones were synthesized in order to investigate their \nproperties towards the quorum sensing. \nFollowing the synthetic route published by Grossmann [11], the compounds 56a, 56b, \nand 58 were prepared. The key step of the synthesis, which is a condensation of the \nmenthylated furanone 38b and an aldehyde with LDA under kinetic conditions, was \noptimized. The biological activity of these three compounds, as well as of all intermediates, \nwas investigated. The tests were carried out with a mutant strain of Chromobacterium \nviolaceum, called CV026. \n \nSome of the synthetic furanones proved indeed to be weak quorum sensing inhibitors, \nhowever none of them was more active than the natural compound 22 [22]. In addition, some \nof these compounds were toxic for Chromobacterium violaceum. \nThe Grossmann method, followed by a reduction of the side chain, was also used for \nthe preparation of the flavour furanone 64, which had previously shown a weak inhibition of the pristinamycin production by Streptomyces pristinaespiralis. Further tests were carried out \nwith mammalian cells, but the compound 64 proved to be toxic [21]. \n \nOther biological tests showed that some menthylated furanones were weak chitinase \ninhibitors. Starting from the assumption that menthylated furanones could have structure \nanalogy with the most potent chitinase inhibitor allosamidin (75), the synthesis of \nglycosylated furanones was attempted. The main purpose was to replace the menthyl group by \na monosaccharide or a disaccharide in a very short and very cheap process. \nThe starting furanone 37 was treated with trichloroacetonitrile to give the \ntrichloroacetimidate 93, which was submitted to glycosylation following a “reversed type \nSchmidt glycosylation” procedure using TMSOTf as Lewis acid. The obtained acetylated \ncompounds 92, 96, and 97 were treated with guanidine. The deacetylated monosaccharides \n101 and 102 were obtained in low yields whereas 97 was not completely deacetylated. \n \nThe glycosylation afforded for each compounds 92, 96, and 97 a mixture of four \ndiastereomers. Only the diastereomers of 96 could be partially separated by chromatography \non silica gel. The structures of the a-anomers could not be completely elucidated, whereas the \nstructures of the b-anomers could be obtained from an X-ray structure determination of 96d. \nThe starting furanone 37 was also treated with (+)-isomenthol and (+)-menthol to give \nthe compounds 104 and 105 respectively. These compounds, as well as the glycosylated \nfuranones 92, 96a, 96b, 96c, 96d, 102, 104, and 105, were tested but none of these synthetic \nfuranones were active as chitinase inhibitors.