3. Conclusions

Our trials to obtain a general building block derived from levulinic acid which contains one additional carbon at C3 were severely hampered by the instability of such compounds. The major problem is the retro-Michael reaction. For the synthesis of precursors of this sort it proved to be crucial to have the adequate protecting groups for the three functional groups which have to be present in the final product. The use of carboxylates enhances the retro-Michael reactivity. Using furans as substitute for a carboxylate avoids this problem. On the other hand the furan ring has to be stabilised by conjugation. The moment the substituent on the furan ring is an alkyl substituent or worth a hydroxy alkyl substituent the furan ring is too labile to be a valid protecting group for our purposes. The approaches reported in this paper clearly demonstrate the difficulties to set the desired functionalities in place. In the end the best way to obtain compounds with the right arrangement of functional groups proved to be the Mukaiyama aldol reaction. The ease of elimination giving the a,b-unsaturated compound 18 illustrates the general difficulty to obtain building blocks of the structure envisaged. As the major goal of our synthetic effort is to obtain one or several building blocks which can be easily modified and used for the creation of biased libraries, the approach reported here has been unsatisfactory despite the fact that a series of interesting compounds have been obtained. Some of these compounds have already been tested as inhibitors of the enzyme porphobilinogen synthase and the interesting results of these inhibition studies will be reported elsewhere.[84]


Continue to Experimental Part
Go to References
Back to main menu