Novel analogs of the cyclodepsipeptide didemnin B

The didemnins are cyclodepsipeptides isolated from a marine tunicate. Most contain the same twenty-three membered macrocycle and differ only in the side chain attached to the threonine amide of the macrocycle. They exhibit a wide variety of biological properties, including antitumor, antiviral and immunosuppressive activities. Didemnin B was the first marine natural compound to enter clinical trials. Unfortunately, it showed toxic side effects and did not demonstrate significant efficacy. However, determination of its mechanism(s) of action and structural-activity relationships might eventually lead to an effective drug. Previous work (chemical manipulations and structural analysis via NMR and X-ray crystal work) demonstrated that the N,O -diMetyrosine might be important for biological activity. Our goal was to prepare analogs of didemnin B in which the N,O-diMetyrosine was replaced both by coded N-methylated amino acids and by constrained tyrosine and phenylalanine analogs in order to determine the importance of the tyrosine. The Joullié group has developed the only stereocontrolled synthesis of the macrocycle and it is this route that made these analogs synthetically possible. Chapter One deals with the replacement of N,O-diMetyrosine with N-Meleucine and N-Mephenylalanine. The first substitution was designed to probe the importance of aromaticity in the tyrosine\u27s side chain and the second was to probe the role of the methoxy group. Chapter Two discusses the replacement of N,O-diMetyrosine with two constrained analogs: L-1,2,3,4-tetrahydroisoquinoline-3-carboxylic acid (Tic) and L-1,2,3,4-tetrahydro-7-methoxyisoquinoline-3-carboxylic acid (MeOTic). Both analogs restrict rotation about the χ1 and χ2 bonds of the tyrosine moiety. These restrictions will help to tell us whether or not the conformation of the tyrosine side chain is important. Each chapter discusses the biological testing results available to date. Preliminary results show that the three analogs prepared at this time have activity comparable to the natural compound. Thus the tyrosine may not be involved in receptor binding as once thought. Alternatively, it still may be involved in binding to a receptor, but these particular variations do not have deleterious effects