Andreas Schumacher

Medicinal chemistry optimization of acyldepsipeptides of the enopeptin class antibiotics.

The therapy of life-threatening infections is significantly weakened by the global spread of antibiotic resistance. Among Gram-positive bacteria, the development of resistance of staphylococci, streptococci, and enterococci is of particular concern. Methicillin-resistant Staphylococcus aureus (MRSA) is a major cause of complicated nosocomial infections, and its prevalence in hospitals has increased during the last decade. Furthermore, the frequent reliance on vancomycin as a last line of defense has led to a high level of vancomycin-resistant nosocomial isolates. Streptococcus pneumoniae is an important community-acquired respiratory tract pathogen, and a leading cause of morbidity and mortality. More than 30% of US isolates are penicillin-resistant (penicillin-resistant S. pneumoniae, PRSP) and many of these strains are also resistant towards a large number of other antibiotics. Novel antibacterial agents with unprecedented mechanisms of action, which are devoid of pre-existing cross-resistances, are therefore very necessary. Most of the currently marketed antibiotic classes originate from the secondary metabolism of bacteria or fungi, which emphasizes that natural products are a valuable source of novel antibacterial agents. Thus, depsipeptides of the enopeptin family are of interest. In 1982, the isolation of depsipeptide antibiotics A54556A and B (Scheme 1, 1 and 2, respectively), from a mixture of eight individual depsipeptidic factors (A– H), produced by aerobic fermentation of Streptomyces hawaiiensis (NRRL 15010) was described. Enopeptin A (3) and B (4) (Scheme 1) were isolated in 1991 from a culture broth of Streptomyces sp. RK-1051, found in a soil sample from Tsuruoka City, Japan. The enopeptin structure consists of a lactone core composed of five S-configured amino acids and a lipophilic acylated phenylalanine side chain attached to a serine nitrogen. This macrocyclic, peptidic structure made these acyldepsipeptides interesting candidates for total synthesis. The level of interest was further increased by mode of action studies with B. subtilis, demonstrating impaired bacterial cell division and induction of filamentation as the underlying causes of antibacterial activity (Figure 1). Applying reversed genomics technologies, Brçtz-Oesterhelt and co-workers were able to demonstrate that the lead structure 1 acts by binding to caseine lytic protease (ClpP), the core unit of a major bacterial–protease complex. In order to protect the bacterial cell from the destructive power of this universal protease, ClpP is tightly regulated and requires a ClpATPase, and often other accessory proteins for activation. Binding of 1 to ClpP eliminates the requirement of Clp-ATPases and other regulatory factors for proteolytic degradation. Therefore, uncontrolled proteolysis leads to inhibition of bacterial cell division and eventually cell death. With respect to drug discovery, natural acyldepsipeptides 1 and 2 displayed only limited in vitro activity against the Grampositive pathogen S. aureus (Table 1) and Gram-negative bacteria were not susceptible. Furthermore, 1 and 2 were not effective in standard mouse models of lethal bacterial infection, and their physicochemical and pharmacokinetic profile was dominated by poor aqueous solubility and high clearance. Thus, 1 and 2 were not favorable candidates for drug development. Finally, natural enopeptin acyldepsipeptide antibiotics 1 and 2 are challenging lead structures from a chemical and synthetic viewpoint. Several functional groups limited their stability: (a) the lactone core was readily hydrolyzed in basic and acidic aqueous media; (b) the acylated serine hydroxy group eliminated readily under non-aqueous basic conditions; (c) the conjugated triene was sensitive to temperature and light (cyclization and aromatization reactions) ; and (d) solubility was not sufficient for parenteral application. However, the novel target and the absence of cross-resistance to established antibiotics strongly encouraged the initiation of a medicinal chemistry program, the objective of which was to improve these deficiencies using a thorough understanding of the lead conformation, based on x-ray structure analysis of the synthetic congener 5 (Figure 2). Crystallization of 5 from toluene gave a solvate with two toluene molecules, whereas solvent free crystals were obtained from aqueous acetonitrile. In the case of the solvate crystal, a Scheme 1. Natural enopeptin depsipeptide antibiotics.