Total synthesis and initial structure-activity relationships of longicatenamycin A.

Natural products have provided the majority of lead structures for marketed antibacterials. In addition, they are biological guide principles to new therapies. Nevertheless, numerous "old" classes of antibiotics such as the longicatenamycins have never been explored by chemical postevolution. Longicatenamycin A is the first defined longicatenamycin congener that has been totally synthesized and tested in pure form. This venture required the de novo syntheses of the non-proteinogenic amino acids (2S,3R)-beta-hydroxyglutamic acid (HyGlu), 5-chloro-D-tryptophan (D-ClTrp), and (S)-2-amino-6-methylheptanoic acid (hhLeu). In the key step, the sensitive HyGlu building block was coupled as a pentafluorophenyl active ester to the unprotected H-D-ClTrp-Glu-hhLeu-D-Val-D-(Cbz)Orn-OH fragment. This first total synthesis of longicatenamycin A provided new congeners of the natural product (deacetyllongicatenamycin, dechlorolongicatenamycin, and longicatenamycin-A-amide).