Reprogramming of the Caseinolytic Protease by ADEP Antibiotics: Molecular Mechanism, Cellular Consequences, Therapeutic Potential

Abstract

Rising antibiotic resistance urgently calls for the discovery and evaluation of novel antibiotic classes and unique antibiotic targets. The caseinolytic protease Clp emerged as an unprecedented target for antibiotic therapy 15 years ago when it was observed that natural product-derived acyldepsipeptide antibiotics (ADEP) dysregulated its proteolytic core ClpP towards destructive proteolysis in bacterial cells. A substantial database has accumulated since on the interaction of ADEP with ClpP, which is comprehensively compiled in this review. On the molecular level, we describe the conformational control that ADEP exerts over ClpP, the nature of the protein substrates degraded, and the emerging structure-activity-relationship of the ADEP compound class. On the physiological level, we review the multi-faceted antibacterial mechanism, species-dependent killing modes, the activity against carcinogenic cells, and the therapeutic potential of the compound class.

Keywords: ClpP; acyldepsipeptide; antibiotic; conformational control; drug discovery; mode of action; protease.

FIGURE 1

Impact of ADEP on ClpP at the molecular level. Conformational dynamics of ClpP and conformational control exerted by ADEP. Upper row, side view on the tetradecameric barrel of ClpP, switching between the “compressed” and “extended” state as the two endpoints of the conformational transition. The term “compact” is used for an intermediate state (not shown). ADEP stabilizes the extended, active conformation. A single ClpP protomer within the barrel is highlighted in orange. Single protomer showing the dynamic α5 helix (right). Middle, top view on the apical surface of ClpP, depicting the closed pore in apoClpP and the widened pore upon ADEP4 binding. Magnified H-pocket formed by two neighboring protomers and occupancy by ADEP4. Bottom, magnified catalytic triads of the compressed, inactive vs. extended, active conformation. In the nucleophilic attack of the catalytic serine on the peptide bond carbonyl, the serine hydroxyl proton is abstracted by the histidine imidazole, and the positive charge at the histidine is stabilized by the carboxy group of the aspartate. The hydrogen network between the catalytic triad is essential for the interaction and the optimal distance for a hydrogen bond ranges from 2.7 to 3.3 Å. S. aureus ClpP structures are shown: Compressed barrel (PDB code: 4EMM) (Ye et al., 2013); compressed monomer (PDB code: 3QWD) (Geiger et al., 2011), extended barrel with closed (PDB code: 6TTY) and open pore (PDB code: 6TTZ) (Malik et al., 2020).

FIGURE 2

Impact of ADEP on ClpP on the physiological level. ADEP acts by a dual mechanism. The binding of ADEP to the hydrophobic pockets at the apical surface of ClpP causes rapid and efficient displacement of all cooperating Clp-ATPases (red box). Consequently, all the natural functions of the Clp protease in protein homeostasis and regulatory proteolysis are inhibited, of which examples are given. In M. tuberculosis, which depends on a functioning Clp protease for survival, the blocked Clp-ATPase/ClpP interaction is the cause of death. Conformational control of ClpP by ADEP bestows independent proteolytic capabilities to the ClpP core (green box). A variety of non-native substrates are untimely degraded in a concentration-dependent manner, of which examples are given. The indicated members of the Firmicutes and other bacterial species die by self-digestion.

FIGURE 3

Prominent ADEP congeners discussed in this review and structure-activity-relationship. Exemplary ADEP congeners are depicted that were prominently featured in the scientific literature on the compound class over the last 15 years. Compounds only described in patents are not shown. Natural products are shown in green boxes, and all other depicted structures were obtained by total synthesis. The natural product ADEP1 (“factor A”) represents the progenitor of the compound class, while “factor B” lacks the MePro and is 4-fold less active against S. aureus. Compounds synthesized originally to improve the activities against staphylococci, streptococci, and enterococci are shown in red boxes. ADEP2, ADEP4, and ADEP5 originate from an initial optimization campaign directed at improving the activity against S. aureus and increasing chemical and metabolic stability. Bis-fluorination of Phe led to strongly enhanced antibacterial activity. Rigidification of the macrolactone core by exchanging N-MeAla for pipecolic acid increased activity further. Reduction of the number of double bonds in the side-chain increased stability. Removal of the two terminal double bonds was sufficient to prevent sensitivity to light and temperature. This modification also increased metabolic stability, although ADEP4 was still a high-clearance drug. Removal of the α,β-double bond in ADEP2 enhanced metabolic stability further but led to a loss in antibacterial activity. ADEP2 was a medium clearance drug and still highly active but less active than ADEP4. ADEP5 illustrates that N-MeAla allows the attachment of bulkier moieties. ADEP5 has a substantially higher solubility. Further rigidification of the macrocycle by replacing Ser for allo-Thr brought an additional increase in antibacterial activity. Introduction of a urea moiety into the side-chain allowed to omit the α,β-double bond without loss of antibacterial activity. Within the ureadepsipeptide series (blue box), metabolic stability is markedly improved. ADEP26 (ADEP-14) showed very good activity against Neisseria (anti-Neisseria activity indicated by a yellow box). Against Neisseria and E. coli, the multiple-unsaturated side-chain is superior to the mono-unsaturated one. The same was observed for Streptomyces, against which ADEP1 proved superior to ADEP4. ADEP-28 (ADEP 1g, ADEP B315) and ADEP-41 (ADEP 1f) were featured as particularly active against human mitochondrial ClpP (grey box). Fragment 14 represents the minimal structural element required for ClpP activation and deregulation towards unregulated proteolysis, although removing the macrocycle reduces potency greatly. Fragment 2a binds to a yet unknown binding site at mycobacterial ClpP and does not interfere with ClpX binding.

FIGURE 4

ClpP activators from other structural classes.