Antibacterial activity of a dual peptide targeting the Escherichia coli sliding clamp and the ribosome

Abstract

The bacterial processivity factor, or sliding clamp (SC), is a target of choice for new antibacterial drugs development. We have previously developed peptides that target Escherichia coli SC and block its interaction with DNA polymerases in vitro. Here, one such SC binding peptide was fused to a Proline-rich AntiMicrobial Peptide (PrAMP) to allow its internalization into E. coli cells. Co-immunoprecipitation assays with a N-terminally modified bifunctional peptide that still enters the bacteria but fails to interact with the bacterial ribosome, the major target of PrAMPs, demonstrate that it actually interacts with the bacterial SC. Moreover, when compared to SC non-binding controls, this peptide induces a ten-fold higher antibacterial activity against E. coli, showing that the observed antimicrobial activity is linked to SC binding. Finally, an unmodified bifunctional compound significantly increases the survival of Drosophila melanogaster flies challenged by an E. coli infection. Our study demonstrates the potential of PrAMPs to transport antibiotics into the bacterial cytoplasm and validates the development of drugs targeting the bacterial processivity factor of Gram-negative bacteria as a promising new class of antibiotics.

Fig. 1. Expression of SC binding peptide in E. coli is toxic. (A) E. coli BL21 (DE3) pLysS cells were transformed with either pET-M30, pET-M30-8F or pET-M30-8A vectors expressing His-GST, and respectively the 8F SC binding peptide (MGPRQLDLF) or the SC non-binding mutant 8A (MGPRALDAA) in fusion with His-GST. After 1 hour incubation at 37 °C, one tenth of the transformation mixtures were plated on LB agar plates containing kanamycin, chloramphenicol and either 0 or 0.1 mM IPTG. The picture was taken after 18 hours incubation at 37 °C. (B) The growth of transformed E. coli BL21 (DE3) pLys cells in Luria broth complemented or not with IPTG was followed over a period of 5 h 30 min by measurement of OD₆₀₀ every 15 minutes in a TECAN M200 plate reader. Open symbols: no IPTG in the medium; closed symbols: 0.5 mM IPTG; circles: empty vector (pETM30); squares: expression of SC non binding peptide (pETM30-8A); triangles: expression of SC binding peptide (pETM30-8F).

Fig. 2. Analysis of peptide penetration into E. coli cells. (A) Confocal LSM of D22 E. coli cells exposed to the three different FITC labeled peptides (9 μM) and stained with Nile Red. Composite image were obtained by merging the FITC and Nile Red channels and clearly show the cytoplasmic localization of FITC. (B) Three-dimensional analysis of the FITC-Py (9 μM) peptide penetration in the bacterial cytoplasm. D22 cells were treated as in (A), except that they were additionally stained with DAPI (blue). It clearly appears that the two fluorescent signals from DAPI (blue) and FITC (green) are located within the cell, while Nile Red stains the lipidic phases of membranes. This unambiguously indicates the cytosolic localization of the peptide. (C) Representative fluorescent microscopy images of D22 E. coli cells exposed to the three different FITC labeled peptides (9 μM) and stained with DAPI. The uptake of the bifunctional peptides is clearly less efficient than the one of FITC-Py. (D) Relative FITC fluorescence quantification (average ± SD) normalized to that measured with FITC-Py peptide. Methods and quantification protocol are detailed in the Experimental section.

Fig. 3. In cellulo targeting of the ^EcSC by the P₇ peptide. E. coli D22 cells were grown in absence (no peptide) or presence of the indicated FITC-labelled peptides (16 μM) before extensive wash and lysis. Pull-down of the FITC-labelled peptides on the different pre-cleared lysates was carried out using anti-FITC antibody bound to protein G sepharose beads. Finally, the presence of the ^EcSC in the lysates (input) and in the pull-down fractions (ImmunoPrecipitated, IP) was assessed by western blotting using a polyclonal anti-DnaN antibody (Ab246, kindly provided by Charles McHenry and Kenneth Marians). The specificity and sensitivity of the anti-DnaN antibody are shown using whole cell extracts (input) and 5, 10 and 20 ng of purified ^EcSC (DnaN), respectively. Full gel image and total protein stain of this representative assay are shown in SI.5 (ESI†). The specific FITC-Py-P₇ mediated co-immunoprecipitation of the SC was confirmed in 3 independent biological replicates.

Fig. 4. In vivo survival assay using Drosophila melanogaster infection model. Seven-day-old female kenny^C02831 flies were subjected to sterile injury (A) or septic injury (B) with a thin tungsten needle previously dipped in sterile PBS alone or in an E. coli suspension diluted in PBS and kept at 29 °C. 1 and 24 hours after injury, 18.4 nL of peptides (1 mM) diluted in PBS or PBS alone were injected into the flies body cavity. Data represents means ± standard errors of 3 independent experiments. Individual experiments and their respective log-rank analysis are presented in SI.5 (ESI†). The survival rate at 144 hours (C) and the Lethal Time 50% (LT50) (D) represent the mean ± standard errors of the value determined for each of the 3 independent experiments and data were analyzed by the One-Way ANOVA test: *P < 0.05, **P < 0.01, ***P < 0.001. ns: not significant.