Compounds and reagents

The native toxins TcdA and TcdB from C. difficile VPI 10,463 were generously provided by Klaus Aktories (University of Freiburg, Germany) and purified as described earlier [29]. The peptide α-defensin-6 was purchased from PeptaNova (Sandhausen, Germany).

Peptide design

The antimicrobial segment of angiogenin was predicted as previously shown [7] by AMPA, http://tcoffee.crg.cat/apps/ampa/do [49, 50], CAMPR3, http://www.camp.bicnirrh.res.in/prediction.php [54] using the Rational Design of Antimicrobial Peptides tool, and then evaluated by CAMPR3-Predict Antimicrobial Peptide tool, Antibp2, http://crdd.osdd.net/raghava/antibp2/ [32], ClassAMP, http://www.bicnirrh.res.in/classamp/predict.php [27], Peptide AMP Scanner, https://www.dveltri.com/ascan/v2/ascan.html [51], and iAMPpred, http://cabgrid.res.in:8080/amppred/server.php [37].

Synthesis of Angie peptides

The Angie peptides Angie 1, 3, 5, 6, and 7 were obtained from PSL Heidelberg (PSL, Heidelberg, Germany) using F-moc chemistry [38].

The reference Angie was synthesized on site (CFP, Ulm, Germany) as previously described by Harms et al. [22]. Briefly, the peptides were synthesized via standard Fmoc solid-phase peptide synthesis using a Liberty Blue microwave synthesizer (CEM Corporation, Matthews, NC, USA) and then purified using reversed-phase high-performance liquid chromatography (Waters, Milford, MA, USA), employing an acetonitrile/water gradient under acidic conditions on a Phenomenex C18 Luna column (particle size 5 μm, pore size 100 Å). The pure peptide was lyophilized on a freeze-dryer (Labconco, Kansas City, MI, USA), and the molecular mass was verified by liquid chromatography–mass spectrometry (LC-MS; Waters, Milford, MA, USA). For the experiments the peptides were dissolved in water.

Half-life determination of Angie 5 in human plasma

The half-life of Angie 5 in plasma was calculated according to a work by Freisem et al. [15], with minor modifications. A sample (0.5 ml) of human plasma was spiked with 20 µM Angie 5 and incubated at 37 °C. Aliquots (50 µl) were separated at 0, 15, 30, 60, and 120 min, respectively, and mixed with 250 µl 0.1% TFA in acetonitrile at − 20 °C. The mixture was centrifuged at 13,000 rpm for 30 s, and 150 µl of the supernatant was mixed with 150 µl 20% acetic acid in ice. The samples were analyzed with an Axima Confidence MALDI-TOF MS (Shimadzu) in linear mode using exactly the same measurement conditions for all samples spotted on a 384-well plate. Wells were coated with 0.5 µl of 10 mg/ml CHCA previously dissolved in TFA/water/acetonitrile/2-propanol (2.5/47.5/25/25, v/v), and the solvent was allowed to evaporate. Then, each sample (0.5 µl), previously mixed with matrix (0.5 µl), was applied onto the dry pre-coated well, and the solvent was allowed to evaporate. Laser shots were automatically done following a regular circular raster of a diameter of 2000 μm and spacing of 200 μm on each well; 100 profiles were acquired per sample, and 20 shots were accumulated per profile. An accelerating voltage of 20 kV was applied to the ion source. Measurements of each sample were done in triplicate. The measurement and MS data processing (peak area calculation) were controlled with MALDI-MS Application Shimadzu Biotech Launchpad 2.9.8.1 (Shimadzu). The half-life was calculated using GraphPad Prism version 10.3.1 for Windows, GraphPad Software, Boston, Massachusetts USA, www.graphpad.com. Data (signal area vs. time) was fitted to a one-phase decay curve.

Bacterial culturing

All bacterial strains used for the susceptibility testing are listed in Supplementary Table 1. The C. difficile strain used is the strain VPI 11186 and PCR negative for cdtB, tcdA, and tcdB genes. All bacteria were cultivated on Tryptone Soya Agar with Sheep Blood (Thermo Scientific, Waltham, MA, USA) at 37 °C and 5% CO2. C. difficile was cultivated under anaerobic conditions, created by an GENbag anaer bag (bioMérieux, Marcy-l’Étoile, France). Liquid cultivation of C. difficile was performed in Brain-Heart infusion medium (Oxoid, Dardilly, France) supplemented with 0.5% yeast (Gibco) and 0.4 g/l L-cysteine (Fluka-Honeywell Research Chemicals, Morris Plains, NJ, USA) (BHI). After inoculation, the liquid culture was overlayed with sterile liquid Vaseline (VWR, Radnor, PA, USA) to achieve anaerobic conditions. For liquid cultivation of P. aeruginosa, A. baumannii, and E. coli, bacteria were inoculated in lysogeny broth (LB-Miller) and incubated at 37 °C with shaking at 160 rpm. E. faecium, S. aureus, and K. pneumoniae were grown in Todd-Hewitt Broth (Oxoid, Dardilly, France) supplemented with 5% yeast at 37 °C and 5% CO2.

Radial diffusion assay

To investigate antimicrobial activity of the Angie peptides an overlay-assay was performed, as previously described [52]. In short, bacteria were inoculated in liquid agarose with a density of 2 × 107 cells per plate. Wells were put into the solidified 1% agarose and filled with the different Angie peptides (100 µM). After 3 h of incubation at 37 °C an overlay with nutrient agar was performed. Plates were incubated at 37 °C and 5% CO2 and inhibition zones were measured after overnight incubation. For C. difficile, some modifications were made to account for the anaerobic conditions. In detail, 1 ml of a C. difficile overnight culture was directly added the agarose, mixed and a plate was poured. After drying for 5 min at 4 °C, wells were put in the agarose and filled with 10 µl of the Angie peptides 1, 3, and 5 in concentrations ranging from 100 µM to 1 mM or the various Angie peptides (5 mM and 100 µM for the reference Angie). The plate was incubated at 37 °C in a GENbag anaer bag for 3 h, then an overlay with 10 ml BHI-Agar was conducted. After overnight incubation, inhibition zones were measured. As a positive control LL-37 (Anaspec, Fremont, CA, USA) was used at a concentration of 1 mg/ml for C. difficile and 100 µg/ml for all other bacteria.

Transmission Electron microscopy of C. difficile

To investigate effects of Angie 5 on C. difficile, transmission electron microscopy was performed, as previously described [20]. Shortly, C. difficile was grown for 3 h and cells were harvested by centrifugation (2 min, 8800 xg). The pellet was reconstituted in 10 mM phosphate solution and either Angie 5 or water was added. For the incubation period (37 °C, 1 h), an overlay with sterile liquid Vaseline was performed to ensure anaerobic conditions. The samples were subsequently fixed using 3.5% glutaraldehyde, 1% saccharose in phosphate buffer. Bacterial cells were postfixed in osmium tetroxide and dehydrated in a graded series of propanol. Finally, cells were stained with uranyl acetate, embedded in Epon and ultrathin sections were prepared using standard procedures. A Jeol 1400 Transmission Electron Microscope was used to analyze the samples and at least 25 pictures per sample were taken. The experiment was conducted once.

Cell lines

All materials for the cultivation of cell lines were purchased from Gibco (Thermo Fisher Scientific, Waltham, MA, USA), unless indicated differently. The experimentally used cell lines included Vero cells (African green monkey kidney cells; DSMZ, Braunschweig, Germany), HeLa cells (human cervical carcinoma cells; DSMZ, Braunschweig, Germany), and CaCo-2 cells (human epithelial colorectal adenocarcinoma cells, ATCC HTB-37, Manassas, VA, USA) which were cultivated under humidified conditions at 37 °C with 5% CO2. The cells were trypsinized and reseeded every two to three days with a maximum of 25 times, while Vero cells and HeLa cells were cultivated in minimum essential medium (MEM), supplemented with 10% FCS, 1 mM sodium pyruvate, 0.1 mM non-essential amino acids and 100 U/ml penicillin and 100 g/ml streptomycin. CaCo-2 cells were cultivated in Dulbecco’s Modified Eagle Medium (DMEM), supplemented with 10% FCS, 1 mM sodium pyruvate, 0.1 mM non-essential amino acids and 100 U/ml penicillin and 100 g/ml streptomycin. For the performed intoxication experiments, cells were seeded in respective culture dishes one or two days before and treated in FCS-free media with toxins and the respective compounds.

Cytopathic cell rounding assay

For the analysis of cell rounding due to intoxication with TcdA and or TcdB cell morphology assays were conducted. Therefore, the respective cell line was seeded in 96-well plates one or two days prior to treatment with toxins, Angie peptides or water (solvent control) in FCS-free medium. The treated cells were incubated using humidified conditions at 37 °C with 5% CO2. Afterwards, the cell morphology was monitored using light microscopy every hour for at least 6 h using a Leica DMi1 microscope connected to a Leica MC170 HD camera (both Leica Microsystems GmbH, Wetzlar, Germany). Rounded and non-rounded cells were counted using the online software Neuralab (https://neuralab.de).

Glucosylation status of intracellular Rac1

For the analysis of the glucosylation status of intracellular Rac1, Vero cells were seeded one day before the treatment with toxin, Angie peptides or water (solvent control) in 24 well plates in FCS-free medium. After treatment, the cells were incubated with the approaches for 2 h or according to incubation times of the time series. At the end of the intoxication time, whole-cell lysates were harvested in 2.5x Laemmli (0.3 M Tris-HCl, 10% SDS, 37.5% glycerol, 0.4 mM bromophenol blue, 100 mM DTT) and samples were heat-denatured at 95 °C for 10 min. Afterwards, samples were subjected to SDS-PAGE and immunoblotting.

Gel electrophoresis and Immunoblotting

After sample preparation SDS-PAGE and immunoblotting was performed. For separation of proteins gel electrophoresis was conducted, using 8 or 12.5% acrylamide gels, depending on the size of the protein of interest. Subsequently, semi-dry Western blotting was performed for the transfer of proteins from the gels onto nitrocellulose membranes, which was controlled by staining the membranes with Ponceau-S (AppliChem GmbH, Darmstadt, Germany). Next, the membranes were blocked at room temperature, using 5% skim milk powder in PBS-T (PBS containing 0.1% Tween 20) for at least 30 min. After blocking, washing steps were performed in PBS-T, followed by incubation with primary and secondary antibodies, again separated by washing steps with PBS-T. For detection of non-glucosylated Rac1 a respective primary antibody was used, primary mouse anti-Rac1 antibody (1:1000, #610651; clone 102; BD Biosciences, Heidelberg, Germany), while Hsp90 was detected as loading control, using a primary mouse anti-Hsp90 antibody (1:1000, Santa Cruz Biotechnology, Heidelberg, Germany). Total Rac1 was detected using the primary mouse anti-Rac1 antibody (Clone 23A8, Sigma-Aldrich Chemie GmbH, Germany). As secondary antibodies Horseradish peroxidase (HRP)-coupled goat anti-mouse IgG (H + L) secondary antibody (#31430; Thermo Fisher Scientific, Waltham, USA) and HRP-coupled mouse IgG kappa-binding protein (m-IgGκ BP-HRP; Santa Cruz Biotechnology, Dallas, USA; sc −516,102) were used respectively. After incubation with secondary antibodies, washing steps were performed and signals were detected using Pierce ECL Western blotting substrate (Thermo Fisher Scientific, Waltham, MA, USA), using the iBright 1500 system (Thermo Fisher Scientific). The signal quantification was performed using the ImageJ software v.1.52.a (NIH).

In vitro glucosylation status of Rac1 from Whole-Cell lysates

For the analysis of the in vitro glucosylation status of Rac1, cell lysate was generated from CaCo-2 cells. Therefore, CaCo-2 cells were seeded in 10 cm culture dishes and grown for two to three days. Afterwards, cells were washed and frozen for cell lysis. Cell lysates were collected in glucosylation buffer (50 mM HEPES, 100 mM KCl, 2 mM MgCl2, 1 mM MnCl2, 100 mg/L BSA, pH 7.5) or in glucosylation buffer without BSA (50 mM HEPES, 100 mM KCl, 2 mM MgCl2, 1 mM MnCl2, pH 7.5), centrifuged at 10,000 x g for 1 min, the supernatant was transferred into a new tube, and the protein concentration was determined at the Nanodrop. The Angie peptides or water (solvent control) and 10 nM TcdB were mixed in glucosylation buffer with or without BSA and directly added to 40 µg CaCo-2 cell lysate. Subsequently, the samples were incubated for 2 h at 37 °C. After that, the samples were subjected to gel electrophoresis and immunoblotting. Rac1, total Rac1, and Hsp90 as a loading control were detected as described above and signals were quantified using the ImageJ software v.1.52.a (NIH).

Actin-Staining and fluorescence microscopy

For analysis of the cytoskeletal modification upon toxin treatment staining and fluorescence microscopy experiments were performed where Vero cells were seeded and grown for one day in 18-well µ-slides (ibidi GmbH, Gräfelfing, Germany). Therefore, cells were treated with Angie peptides or water (solvent control) and intoxicated with TcdB in FCS-free medium for 2 h at 37 °C. After treatment, the cells were washed with PBS, fixed with 4% paraformaldehyde for 20 min, permeabilized using 0.4% (v/v) Triton X-100 in PBS for 5 min if required, and quenching was performed for 2 min in glycine (100 mM in PBS). This was followed by a blocking step for 1 h at 37 °C in PBS-T (PBS containing 0.1% Tween 20) containing 10% normal goat serum (Jackson ImmunoResearch, West Grove, PA, USA) and 1% BSA. For visualization of the cytoskeleton, F-actin was stained for 1 h at 37 °C, using the membrane-permeant SiR-actin (SiR-actin kit, Spirochrome, Stein am Rhein, Switzerland). Finally, cell nuclei were stained for 5 min using Hoechst 33,342 (1:10,000, Thermo Fisher Scientific, Waltham, MA, USA). After completing the staining procedure, the slides were examined via fluorescence microscopy using the BZ-X810 Keyence fluorescence microscope with a Plan Apochromat 40X objective and BZ-X filters (Keyence Deutschland GmbH, Neu-Isenburg, Germany) and BZ-X800Viewer v1.3.0.

Cellular binding assay

For the analysis of TcdB binding to cells, Vero cells were seeded one day before treatment in 24 well plates. Then, cells were put on ice for 30 min and subsequently, Angie peptides or water (solvent control), and TcdB were added to cells and incubated for 1 h on ice in FCS-free medium. Cell lysates were collected in 2.5x Laemmli and samples were heat-denatured at 95 °C for 10 min. Afterwards, samples were subjected to gel electrophoresis and immunoblotting, while TcdB was detected, using an anti-TcdB antibody (1:1000, Anti-Clostridium difficile Toxin B antibody, Abcam, Cambridge, UK) and Hsp90 served as loading control. Signals were quantified using the ImageJ software v.1.52.a (NIH).

In vitro precipitation assay

For the precipitation analysis, TcdB (50 ng equals 2 ng/µl) and inhibitors, α-defensin-6 (6 µM) and Angie peptides or water (solvent control) were incubated for 30 min in a total volume of 25 µl PBS. Next, the samples were centrifuged for 20 min, 14,000 rpm at 4 °C for separation of supernatant and pellet fraction. Therefore, the supernatant was transferred into a new tube and the pellet was resuspended in PBS. The samples were subjected to gel electrophoresis and immunoblotting. TcdB was detected using an anti-TcdB antibody (1:1000, Anti-Clostridium difficile Toxin B antibody, Abcam, Cambridge, UK).

In Silico prediction of the complex TcdB-Angie 5

The structural prediction of the TcdB-Angie 5 complex was carried out using three web-based tools: AlphaFold3 (https://alphafoldserver.com/) [1], HPEPDOCK (http://huanglab.phys.hust.edu.cn/hpepdock/) [57], and PEP-SiteFinder (https://bioserv.rpbs.univ-paris-diderot.fr/services/PEP-SiteFinder/) [44]. In all cases, the TcdB protein was designated as the receptor, while the Angie 5 peptide served as the ligand. The amino acid sequence and three-dimensional (3D) structure of TcdB were retrieved from the Protein Data Bank [8] under the accession code 6oq5 [12].

For AlphaFold3, the receptor and peptide sequences were provided as input, whereas for HPEPDOCK and PEP-SiteFinder the 3D structure of TcdB and the peptide sequence were used. All docking simulations were performed in a blind manner using the default parameters of each server. Since AlphaFold3 generates only five docking models per run, the docking procedure was repeated three times with this tool. The resulting models from each server were clustered based on a root-mean-square deviation (RMSD) threshold of 20 Å, yielding the ten most representative clusters per server. The best structure from each cluster was selected based on the scoring function of the respective server.

To facilitate comparison and identify the most stable complex, the binding energy of the selected structures was re-evaluated using the Prodigy server (https://rascar.science.uu.nl/prodigy/) [56]. The TcdB-Angie 5 complex with the lowest predicted binding energy was considered the most probable conformation. Interaction analysis of the selected complex and the identification of key binding residues (hot spots) in the peptide were performed using PPCheck (https://caps.ncbs.res.in/ppcheck/) [47].

To validate these findings, in silico mutagenesis was conducted by manually modifying the Angie 5 peptide within the complex to generate the other Angie peptide variants (Table 1). The binding energy of these TcdB-peptide complexes was subsequently calculated using Prodigy and compared with experimental data from inhibitory assays.

In vitro autoprocessing assay of TcdB

To investigate the effects of the Angie peptides on the intrinsic cysteine protease activity of TcdB an in vitro autoprocessing assay of TcdB was conducted. TcdB with or without the different Angie peptides was incubated for 1 h at 37 °C in a 20 mM Tris-HCl buffer containing 150 mM NaCl at pH 7.4. The autoprocessing activity was induced by the addition of 1 mM inositol hexakisphosphate (Santa Cruz Biotechnology). To inhibit the autoprocessing of TcdB, a positive control containing 1 mM N-ethylmaleimide (NEM) (Sigma Aldrich by Merck) was added. The reaction was stopped by the addition of Laemmli buffer. The samples were incubated for 10 min at 95 °C and subjected to SDS-PAGE and immunoblotting, while TcdB was detected, using an anti-TcdB antibody (1:1000, Anti-Clostridium difficile Toxin B antibody, Abcam, Cambridge, UK).

Reproducibility of experiments and statistics

All performed experiments were conducted independently from each other at least three times. The number of replicates (n) for experiments or tested conditions is given in the figure legends, while representative results are shown in the figures. If not stated otherwise in the figure legends, the statistical analysis performed was a one-way ANOVA in combination with Dunnett’s multiple comparison test using GraphPad Prism Version 9 (GraphPad Software Inc., San Diego, CA, USA). The obtained p values are depicted as follows: ns = not significant p > 0.05, * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.