Short proline-rich antimicrobial peptides (PrAMPs) are a promising class of antibiotics that use novel mechanisms, thus offering the potential to overcome the health threat of multiresistant pathogens. The peptides bind to the bacterial 70S ribosome and can inhibit protein translation. We report that PrAMPs can be divided into two classes, with each class binding to a different site, and thus use different lethal mechanisms. Oncocin-type peptides inhibit protein translation in Escherichia coli by binding to the exit tunnel of the 70S ribosome with half maximal inhibitory concentrations (IC50 values) of around 2 to 6 μmol  L(-1), whereas apidaecin-type peptides block the assembly of the large (50S) subunit of the ribosome, resulting in similar IC50 values. The revealed mechanisms should allow the design of new antibiotics to overcome current bacterial resistance mechanisms.

A novel concept to release peptidic drugs systemically by serum proteases from a PEGylated precursor makes it possible to tune release kinetics to fit the medical needs. Drug release depends on the size of the PEG polymer and the sequence and length of the peptide linker. The antimicrobial activities of the prodrugs were even better than those of the free peptides, whereas direct PEGylation abolished the peptide activity.

The bacterial protein DnaK promotes folding of newly synthesized polypeptide chains, refolding of misfolded proteins, and protein trafficking. Assisted refolding is especially important under stress conditions induced by antibiotic therapies reducing the desired bactericidal effects. DnaK is supposedly targeted by proline-rich antimicrobial peptides (PrAMPs), but Escherichia coli ΔdnaK mutants and wild type strains are equally susceptible indicating further intracellular targets, such as the 70S ribosome. Crystal structures of PrAMPDnaK- complexes revealed forward and reverse binding modes at the substrate binding domain. Here, we used these ligand-target structures for the first time to rationally optimize peptides using molecular modeling and docking leading to the prediction of four-residue long sequences for improved binding to DnaK. When these sequences were used to replace the original sequence stretch in Onc72, most peptides showed significantly reduced dissociation constants (Kd) determined by fluorescence polarization. In a second approach, the X-ray structures of Api88 and Onc72 bound to DnaK were examined to predict substitutions prone to stronger interactions. Among the 36 peptides obtained from both approaches, six derivatives bound to DnaK with more than 10-fold higher affinities (Kd values in the low micromolar to nanomolar range). Peptides binding stronger to DnaK showed the same minimal inhibitory concentrations against wild type E. coli as the original peptide, but were slightly less active for ΔdnaK mutants. However, one peptide was able to overcome the resistance in an E. coli mutant lacking the SbmA transporter obligatory for the uptake of PrAMPs including Api88 and Onc72. Thus, it´s tempting to speculate that DnaK might be involved in the translocation of PrAMPs into E. coli.

The Gram-negative bacterium Pseudomonas aeruginosa is a life-threatening nosocomial pathogen due to its generally low susceptibility toward antibiotics. Furthermore, many strains have acquired resistance mechanisms requiring new antimicrobials with novel mechanisms to enhance treatment options. Proline-rich antimicrobial peptides, such as the apidaecin analog Api137, are highly efficient against various Enterobacteriaceae infections in mice, but less active against P. aeruginosa in vitro. Here, we extended our recent work by optimizing lead peptides Api755 (gu-OIORPVYOPRPRPPHPRL-OH; gu = N,N,N′,N′-tetramethylguanidino, O = L-ornithine) and Api760 (gu-OWORPVYOPRPRPPHPRL-OH) by incorporation of Ile-Orn- and Trp-Orn-motifs, respectively. Api795 (gu-O(IO)2RPVYOPRPRPPHPRL-OH) and Api794 (gu-O(WO)3RPVYOPRPRPPHPRL-OH) were highly active against P. aeruginosa with minimal inhibitory concentrations of 8–16 and 8–32 μg/mL against Escherichia coli and Klebsiella pneumoniae. Assessed using a quartz crystal microbalance, these peptides inserted into a membrane layer and the surface activity increased gradually from Api137, over Api795, to Api794. This mode of action was confirmed by transmission electron microscopy indicating some membrane damage only at the high peptide concentrations. Api794 and Api795 were highly stable against serum proteases (half-life times >5 h) and non-hemolytic to human erythrocytes at peptide concentrations of 0.6 g/L. At this concentration, Api795 reduced the cell viability of HeLa cells only slightly, whereas the IC50 of Api794 was 0.23 ± 0.09 g/L. Confocal fluorescence microscopy revealed no colocalization of 5(6)-carboxyfluorescein-labeled Api794 or Api795 with the mitochondria, excluding interactions with the mitochondrial membrane. Interestingly, Api795 was localized in endosomes, whereas Api794 was present in endosomes and the cytosol. This was verified using flow cytometry showing a 50% higher uptake of Api794 in HeLa cells compared with Api795. The uptake was reduced for both peptides by 50 and 80%, respectively, after inhibiting endocytotic uptake with dynasore. In summary, Api794 and Api795 were highly active against P. aeruginosa in vitro. Both peptides passed across the bacterial membrane efficiently, most likely then disturbing the ribosome assembly, and resulting in further intracellular damage. Api795 with its IOIO-motif, which was particularly active and only slightly toxic in vitro, appears to represent a promising third generation lead compound for the development of novel antibiotics against P. aeruginosa.

Recent surveillance data on antimicrobial resistance predict the beginning of the post-antibiotic era with pan-resistant bacteria even overcoming polymyxin as the last available treatment option. Thus, new substances using novel modes of antimicrobial action are urgently needed to reduce this health threat. Antimicrobial peptides are part of the innate immune system of most vertebrates and invertebrates and accepted as valid substances for antibiotic drug development efforts. Especially, short proline-rich antimicrobial peptides (PrAMP) of insect origin have been optimized for activity against Gram-negative strains. They inhibit protein expression in bacteria by blocking the 70S ribosome exit tunnel (oncocin-type) or the assembly of the 50S subunit (apidaecin-type binding). Thus, apidaecin analog Api137 and oncocin analog Onc112 supposedly bind to different nearby or possibly partially overlapping binding sites. Here, we synthesized Api137/Onc112-conjugates bridged by ethylene glycol spacers of different length to probe synergistic activities and binding modes. Indeed, the antimicrobial activities against Escherichia coli and Pseudomonas aeruginosa improved for some constructs, although the conjugates did not bind better to the 70S ribosome of E. coli than Api137 and Onc112 using 5(6)-carboxyfluorescein-labelled Api137 and Onc112 in a competitive fluorescence polarization assay. In conclusion, Api137/Onc112-conjugates showed increased antimicrobial activities against P. aeruginosa and PrAMP-susceptible and -resistant E. coli most likely because of improved membrane interactions, whereas the interaction to the 70S ribosome was most likely not improved relying still on the independent apidaecin- and oncocin-type binding modes. Copyright © 2016 European Peptide Society and John Wiley & Sons, Ltd.