Bacterial strains

The bacterial strains used in the study included laboratory strains and clinical isolates of Gram-negative and Gram-positive pathogens from the collection of the Gamaleya National Center for Epidemiology and Microbiology (Russia), the State collection of pathogenic microorganisms and cell cultures “SCPM-O-B” and the collection of Gabrichevsky Moscow Research Institute of Epidemiology and Microbiology.

For animal studies K. pneumoniae M9 and P. aeruginosa 38–16 isolates were used. Strains were cultivated in GRM-1 media (pancreatic hydrolysates of fish meal (15.0 g/L) and casein (10.0 g/L), yeast extract (2.0 g/L), NaCl (3.5 g/L), glucose (1.0 g/L), SRCAMB) at 37 °C, at 240 rpm overnight. Antibiotic susceptibility testing of these strains was performed by the broth microdilution method using Mueller–Hinton broth (Thermo Fisher Scientific), according to ISO recommendations [17]. Results were interpreted according to The European Committee on Antimicrobial Susceptibility Testing. Breakpoint tables for interpretation of MICs and zone diameters. Version 12.0, 2022 (https://www.eucast.org).

Protein expression and purification

LysECD7 was obtained as previously described [16]. To express LysECD7-SMAP its coding sequence (Additional file 1) was artificially synthesized in a pAL-TA commercial vector (Evrogen Ltd.). Thereafter endolysin ORF was amplified from a pALTA-LysECD7-SMAP clone and integrated into the expression vector pET-42b (+) (Evrogen Ltd.), resulting in a pET42b-LysECD7-SMAP plasmid (kanamycin resistance). All constructs were checked for errors via Sanger sequencing.

The expression vector was introduced into the competent Escherichia coli strain BL21 (DE3) pLysS (chloramphenicol resistance) using a heat shock transformation protocol. The E. coli were grown in LB broth (37 °C, 240 rpm) to OD600 = 0.55–0.65 and then induced with 1 mM β-D-1-thiogalactopyranoside (IPTG) at 37 °C for 4 h. The cells were harvested by centrifugation (6000 × g for 20 min at 4 °C) and resuspended in a lysis buffer (20 mM Tris HCl, 250 mM NaCl, 0.1 mM EDTA, 1 mM PMSF pH 8.0). Then, the suspension was disrupted by APV Laborhomogenisierer 2000 (SPX FLOW, Inc.) and the cell debris was removed by centrifugation (10,000 × g for 30 min at 4 °C). The supernatant was filtered through a 0.2 μm filter and diluted fivefold with 0.36 M NaCl, 50 mM Tris–HCl pH 7.5.

The protein was purified using ÄKTA avant 150 system with XK 50/30 Column (GE Healthcare), prepacked with Capto SP ImpRes resin (GE Healthcare). The cell lysate was applied to a column pre-equilibrated with binding buffer (0.36 M NaCl, 50 mM Tris–HCl pH 7.5). Then, the column was washed with the binding buffer. The protein fractions were eluted using a linear gradient to a 100% elution buffer (0.55 M NaCl, 50 mM Tris–HCl pH 7.5). The collected protein fractions were concentrated with Labscale TFF System (Merck-Millipore) supplied with Pellicon® XL50 Ultracel® 10 kDa membrane (Merck-Millipore).

During the next step, protein solution was purified with XK 50/30 Column, prepacked with Superdex 75 resin (GE Healthcare) and preequilibrated with 0.55 M NaCl, phosphate-buffered saline (PBS) pH 7.4, and, finally, desalted on HiPrep™ 26/10 Desalting column (GE Healthcare) to PBS pH 7.4 buffer. LysECD7-SMAP fractions were collected and concentrated with Pellicon® XL50 Ultracel® 10 kDa membrane to protein concentration 20 mg ml−1 and sterilized by 0.2 μm filtration.

The purity of the protein was determined by 16% SDS-PAGE (molecular mass 17.2 kDa). The protein concentration was measured using a spectrophotometer (Implen NanoPhotometer, IMPLEN) at 280 nm and calculated using a predicted extinction coefficient (1.44 (mg/ml)−1 cm−1).

Crystallization

Crystallization screening was performed on an Oryx4 crystallization robot (Douglas Instruments Ltd) in MRC 2-well crystallization plates (Hampton Research (HR)) and Combi Clover Junior crystallization plates (Rigaku Reagents), respectively, employing the sitting drop vapour diffusion method. Proteins solutions of LysECD7 in 100 mM Tris–HCl buffer pH 7.5 and LysECD7-SMAP in 100 mM MES buffer pH 9.5 were used for the initial crystallization trials. Commercial crystallization screens were used: JCSG-plus, Structure Screen 1&2, Morpheus, the PGA screen, MIDAS (Molecular Dimensions Ltd) and Index (HR). The proteins with concentrations of 5 mg ml−1 and 10 mg ml−1 were mixed with precipitant solutions in ratios 1:1 and 2:1, equilibrated against 50–70 µl precipitant solution in the reservoirs at 293 K. Optimization of successful crystallization conditions was performed by variation of salt, PEG and protein concentration [18]. Later on, various additives were used as well as the microseeding technique, aimed to adjust crystal size and quality. 3D crystals of LysECD7 with a period of three weeks were grown under condition containing 0.2 M Ammonium sulfate, 20% w/v Polyethylene glycol 3350 pH 6.0 with concentrations of 5 mg ml−1. The LysECD7-SMAP crystals were obtained with the precipitant composed of 0.2 M Potassium sodium tartrate tetrahydrate, 2.0 M Ammonium sulfate and 0.1 M Sodium citrate 5.6 during one month.

Data collection and processing

X-ray diffraction data for LysECD7 and LysECD7-SMAP at 100 K were collected at the BESSY II electron-storage ring on the MX14.1 beamline operated by the Helmholtz-Zentrum Berlin (Berlin-Adlershof, Germany) to the 2.5 and 1.7 Å resolution, respectively. Data were recorded on the Pilatus detector S3 2 M (Dectris). The dataset was collected on the crystals from different crystallization conditions. 1800 diffraction images each were collected with 0.1 oscillation angle. Collected data were processed using XDS program [19] with the XDSAPP graphical user interface [20].

The search model for LysECD7 protein was generated by ARCIMBOLDO light [21, 22], based on the combination of locating small model fragments with density modification with the program SHELXE [23]. LysECD7 and LysECD7-SMAP crystal structures were determined by automated model-building pipeline using models from BUCCANEER [24]. LysECD7 solved structure was used as a template for the molecular replacement phasing using MOLREP [25]. The structures of LysECD7 and LysECD7-SMAP were further refined by REFMAC5 from the CCP4 package [26] and manually modelled using COOT program package [27]. MolProbity server was used for final model geometry validation. For determination of the protein assembly, PDBePISA [28] was applied. The structures of both proteins were deposited to the Protein Data Bank. For LysECD7 accession code is 8Q2G (LysECD7-SMAP structure is undergoing validation in PDB). Figures with the 3D structure were prepared using the program PyMOL. The complete information about the data collection and refinement statistics are provided in Additional file 1.

For intrinsic disordered regions prediction PONDR® software with VL-XT, VSL2 and VL3-BA algorithms was used (http://www.pondr.com), the calculation and presentation of various sequence parameters associated with disordered protein sequences were provided with CIDER (Classification of Intrinsically Disordered Ensemble Regions (http://pappulab.wustl.edu/CIDER/analysis/). For AMP prediction AxPEP Server with AmPEP and Deep-AmPEP30 instruments (https://cbbio.online/AxPEP/?), AMPfun web server (http://fdblab.csie.ncu.edu.tw/AMPfun/index.html) and Antimicrobial Peptide Scanner vr.2 (https://www.dveltri.com/ascan/v2/ascan.html) were applied.

Molecular dynamics simulations

PyMOL was used for initial structures editing including amino acids substitution (virtual mutagenesis) and concatenation of structures as well as results visualization. As no SMAP-peptide could be resolved using the obtained protein crystal structure, and LysECD7-SMAP was artificially fused to homologous SMAP structure from PDB ID 5Z2O (RALRRLARKIAHAVKKYG) to produce LysECD7-SMAP model. The modified SMAP used in this study differs in several positions including 3 substitutions and absence of trailing glycine (RKLRRLKRKIAHKVKKY-) thus, the 5Z2O structure was additionally edited to match the target sequence and modified SMAP structure was linked with resulting in initial LysECD7-SMAP model.

Amber16 was used for molecular dynamics simulations and results analysis. Molecular dynamics simulations were performed with Amber ff14SB [29] force field in 3 stages: (1) Energy minimization, 2500 Steepest Descend algorithm steps, 2500 Conjugate Gradient algorithm steps. (2) Heating: 0.002 ps time step, 9000 steps from 0 to 310 K, then 1000 steps at 310 K, Langevin thermostat, no pressure control. (3) Production run: 200 ns total, 2 × 100 ns (50000000 steps) sequental runs, 0.002 ps time step, 310 K, Langevin thermostat, Berendsen barostat, 1 bar pressure control, 8 Å cut-off for non-bonded interactions. The SHAKE algorithm was applied to consider constrain bonds involving hydrogen atoms.

Antibacterial in vitro activity test

Overnight bacterial cultures of A. baumannii (n = 6), P. aeruginosa (n = 12), K. pneumoniae (n = 9), B. cepacia complex (n = 5), Achromobacter spp (n = 4), H. influenzae (n = 4), Staphylococcus aureus (n = 8), and S. haemolyticus (n = 3) strains were diluted 30-fold in Mueller–Hinton broth and grown to the exponential phase (OD600 = 0.6). Subsequently, the cells were harvested by centrifugation (6000 × g, 10 min) and resuspended in the same volume of PBS pH 7.4. Each suspension was diluted 100-fold in the 20 mM Tris–HCl pH 7.5 to a final density of approximately 106 cells/ml. Afterwards, 100 μl of the bacterial suspensions and 100 μl of the protein solutions with 200 µg ml−1 concentration (final protein concentration is 100 µg ml−1) were mixed in 96-well plates, with a buffer lack of endolysin used as the negative control. The mixtures were incubated at 37 °C for 5, 10, 30, 60 or 120 min to study the kinetics of antibacterial action and for 30 min to study spectrum of activity, with shaking at 200 rpm and then were diluted tenfold in PBS pH 7.4.

Subsequently, 100 μl of each dilution was plated onto Mueller–Hinton agar, and the number of bacterial colonies was counted after an overnight incubation at 37 °C. All experiments were performed in triplicate, and the antibacterial activity was expressed as follows: Antibacterial activity (%) = 100%—(CFUexp/CFUcont) × 100%, where CFUexp is the number of bacterial colonies in the experimental culture plates, and CFUcont is the number of bacterial colonies in the control culture plates. Antibacterial activity was arbitrarily regarded as meaningful when it was higher than 33%. For all strains, three independent replicates were done.

Antibiofilm activity

A. baumannii 50–16 and K. pneumoniae biofilm-forming strains were cultured overnight in TSB medium (BD Tryptic Soy Broth). Then culture was diluted 1:50 in fresh TSB, added 100 µl to wells of 96-well sterile polystyrene cell culture plates and incubated for 24 h at 37 °C, 250 rpm to allow biofilm formation. After incubation, the wells’ contents with planktonic cells were shaken off, the plate was washed with 200 µl of PBS pH 7.4 three times and air-dried for approximately 10 min. Then 100 µl of endolysin solutions in final concentration of 100, 1000 µg ml−1 or 20 mM Tris–HCl pH 7.5 as a negative control were added to the wells and incubated at 37 °C, 250 rpm for 2 h. After incubation, wells were twice washed with 200 µl of PBS pH 7.4 and air-dried. Washed biofilms were stained with 0.1% aqueous solution of crystal violet (Applichem) for 20 min at RT followed by triple rinsing with PBS. The remaining stain was re-solubilized in 200 µl of 33% acetic acid and OD590 of obtained solution was measured using SPECTROstar NANO Absorbance Reader (BMG LABTECH). Alcian blue staining was performed with 1% solution in 3% acetic acid for 30 min and measurement of OD615. All experiments were performed in six replicates.

The biofilms formation interpretation was done accordingly to [30]. Briefly, weak biofilm was defined at ODc < ODAb ≤ 2 × ODc, moderate biofilm at 2 × ODc < ODAb ≤ 4 × ODc, and strong biofilm at 4 × ODc < ODAb, where ODc is the cut-off value calculated as three standard deviations (SD) above the mean OD of the negative control, ODAb = the optical density of A. baumannii 50–16 well stained with crystal violet.

Microscopy

To provide planktonic cells microscopy an overnight A. baumannii 50–16 and S. aureus 73–14 bacterial cultures were diluted in an LB broth, grown to an exponential phase (OD600 = 0.6) and harvested by centrifugation. Subsequently, cell pellets were washed twice with distilled water and resuspended in water to a final density of approximately 108 cells/ml (McFarland 0.5). Afterwards, 150 µl of the bacterial suspensions were mixed with 150 µl of the protein water solutions in to the 10 µg ml−1 final concentration and incubated at 37 °C for 30 min with shaking at 200 rpm. Distilled water was used as the negative control. Mixtures were fixed by adding 600 µl of 10% formaldehyde solution for 24 h. Next, 10 µl of mixtures were placed on a slide and dried at room temperature conditions. The samples were mounted to stubs and sputter-coated with a gold layer (5 nm) in an SPI-Module Sputter/Carbon Coater System (SPI Inc.). The sputtering samples were analyzed by means of a scanning dualbeam electron microscope Quanta 200 3D (FEI Company) using the high vacuum mode (10 kV).

Bacterial biofilms microscopy

Sterile glass circle cover slides (Hampton Research) were plunged into overnight A. baumannii 50–16 culture in TSB medium and incubated at 37 °C for 24 h without shaking for biofilm formation. Then slides were washed with distilled water three times and air-dried. Two slides were submerged into 20 mM Tris–HCl pH 7.5 control buffer, another two into 100 µg ml−1 LysECD7-SMAP solution for 2 h at room temperature. Afterwards, all slides were again washed with distilled water three times and air-dried. One slide from each group was fixed in 10% formaldehyde vapors for 24 h and analyzed using scanning electron microscopy after coating with a gold layer as described above.

Another two slides were stained with 0.1% aqueous solution of crystal violet for 20 min at room temperature, washed three times with water and dried at room temperature. These slides were analyzed using an Axiostar plus Transmitted Light Microscope (Zeiss AG) at 400 × magnification.

Peptidoglycan binding assay

Bacterial PG was isolated from E. coli M15 strain as described by [31] with modifications. Briefly, E. coli cells were grown in LB to OD600 = 0.7, centrifuged at 6000 × g at 4 °C for 10 min and resuspended in 20 ml of ice-cold water per liter of bacterial culture. Cell’s suspension was poured into an equal volume of 8% SDS solution and stirred in a boiling water bath for 1 h. Sample was cooled to RT and PG was pelleted by ultracentrifugation at 130,000 × g at room temperature for 1 h. Resuspended in 20 ml water PG was poured into boiling 8% SDS solution and incubated for 15 min. Then peptidoglycan was washed four times with room temperature water, centrifuged 1 h at 130,000 × g for each wash to remove residual SDS. PG was resuspended in 10 ml water, supplied with equal volume of 10 mM Tris–HCl pH 7.0 and 0.12 mg ml−1 trypsin and incubated for 45 min to hydrolyze proteins trapped in PG sacculi. This mixture was added to an equal volume of boiling 8% SDS solution and incubated for 15 min in a boiling-water bath. PG was washed four times with water at room temperature as described earlier. Purified PG was resuspended in 1 ml water and stored at 4 °C. PG concentration was determined by OD206 measurement according to the standard curve made with Micrococcus lysodeikticus PG (Sigma-Aldrich).

The binding of endolysin to PG was assessed after the 30 min incubation of 5 μg protein with peptidoglycan in 10:1, 1:1, 1:10 and 1:20 LysECD7-SMAP:PG ratios at RT for. Mixtures were separated by centrifugation (16,000 × g, 15 min, 4 °C) and unbound fractions (supernatant) were analyzed with 16% SDS-PAGE gels following Coomassie brilliant blue (CBB) staining.

Antibacterial activity was assessed after incubation of endolysin (10 μg) and peptidoglycan in 10:1, 1:1, 1:10 and 1:100 LysECD7-SMAP:PG ratios mixtures for 30 min as described above using in vitro activity test with A. baumannii 50–16 strain.

Lipopolysaccharides binding assays

Purified bacterial Salmonella typhimurium LPS (#437650, Calbiochem®, Merck KGaA) was used in the experiment. LysECD7-SMAP (5 μg) was preincubated with LPS in 1:1 LysECD7-SMAP:LPS ratio at RT for 3 h. Mixtures were separated by centrifugation (16,000 × g, 15 min, 4 °C) and supernatants were analyzed on 16% SDS-PAGE gels with CBB staining. LPS-LysECD7-SMAP ELISA binding analysis was done using LPS from S. typhimurium, murine monoclonal antibodies IgG1 towards LPS from S. typhimurium (#1E6cc, HyTest Ltd.) and HRP-conjugated polyclonal goat-anti mouse IgG (#L20/01, HyTest Ltd.). For this, the wells of polystyrene 96-well plate (Xema-Medica) were coated with 1 µg of LysECD7-SMAP per well in PBS buffer pH 7.4 (VWR Chemicals) at 4 °C for 24 h and blocked with 200 µl of 0.025% BSA for 1 h at 37 °C, 600 rpm. Then, 0.3 µg of S. typhimurium LPS was added to each well and inoculated for 1 h at 37 °C, 600 rpm. Incubation was performed in PBS buffers containing 0.05% Triton X100, 1 mM EDTA, and supplemented with 0, 100 or 500 mM NaCl, these buffers were also used as negative controls. After incubation the wells were thrice washed with PBS supplemented with 0.05% Triton X100, and 0, 100 or 500 mM NaCl. Then, murine IgG1 monoclonal antibodies were diluted 1:4000, and 100 µl were added per well and incubated for 1 h at 37 °C, 600 rpm. Wells were washed three times with PBS with 0.05% Triton × 100 and incubated with 100 µl of 1:50000 diluted HRP-conjugated goat anti-mouse IgG polyclonal antibodies (#L20/01, HyTest Ltd.) for 1 h at 37 °C, 600 rpm. Afterwards, wells were washed with PBS with 0.05% Triton × 100, and 0, 100 or 500 mM NaCl, and incubated with TMB substrate (Xema–Medica) for 10 min in dark and stopped with 100 µl of 10% HCl solution. The optical density was measured with Multiscan FC (Thermo Fisher Scientific) at 450 nm. Normalization of data was performed using subtraction of negative control optical density.

Antibacterial activity was assayed as mentioned above after the 3 h incubation of mixtures of endolysin (5 μg) and peptidoglycan in 5:1, 1:1 and 1:5 LysECD7-SMAP:LPS ratios.

DNA binding assay

Bacterial DNA was isolated from an overnight culture of A. baumannii 50–16 strain with DNeasy UltraClean Microbial Kit (QIAGEN). DNA concentration was measured with Qubit DNA HS Assay Kit (Thermo Fisher Scientific) on Qubit 3.0 fluorimeter.

Endolysin (280 ng) was incubated with DNA in 10:1 and 2:1 LysECD7-SMAP:DNA ratio at room temperature for 30 min. Then mixtures were analyzed by electrophoresis in 1% agarose gel. Electrostatic interactions effect was assessed by adding 0, 50, and 150 mM NaCl to the mixtures during the incubation. Also, antibacterial activity of LysECD7-SMAP (1 μg ml−1) in the presence of DNA was assessed after incubation for 30 min as described above using in vitro activity test with A. baumannii 50–16 strain.

Safety assessment

The hemolytic activity of endolysins was determined against human red blood cells (RBC). Fresh human blood was obtained in a heparin-containing tube and was centrifuged at 500 × g for 10 min, 4 °C. The pellet was washed three times with PBS buffer, and a solution of 10% RBC in PBS was prepared. RBC solution was mixed in a ratio of 1:1 with endolysins solutions in concentration of 100 µg ml−1 and 1 mg ml−1 in 20 mM Tris–HCl buffer (pH 7.5). PBS buffer was used as a negative control and 0.1% Triton X-100 was used as a positive control. The reaction mixtures were incubated for 1 h at 37 °C, harvested by centrifugation (500 × g, 10 min, 4 °C) and the absorbance at the wavelength of 405 nm of the supernatants was measured. All experiments were performed in triplicate, and the hemolysis rate was estimated as follows: Hemolysis (%) = (ODexp—ODPBS)/(ODTriton—ODPBS) × 100%, where ODexp is OD405 in the experimental mixtures, ODPBS is OD405 in the negative control (PBS) mixtures, and ODTriton is OD405 in the positive control (0.1% Triton × − 100) mixtures.

Cytotoxicity assays

Cytotoxicity was measured for HEK293 (ATCC-CRL-1573) cells by MTT assay. Cells were seeded at 4.5 × 104 cells per well in 96-well plates and cultivated in 100 µl of DMEM medium supplemented with 10% fetal bovine serum, 2 mM glutamine, 50 U ml−1 penicillin and 50 µg ml−1 streptomycin for 24 h in a humidified atmosphere of 5% CO2 and 95% air at 37 °C. Afterwards, the medium was removed and 50 µl of endolysins serial dilutions in DMEM (from 1000 µg ml−1 to 16 µg ml−1) were added into each well. DMEM was used as a negative control and 0.1% Triton × − 100 was used as a positive control. Mixtures were incubated for 1 h (37 °C, 5% CO2) and the cells were then stained with 10% MTT solution in DMEM (final stain concentration is 0.5 µg ml−1) for 4 h (37 °C, 5% CO2). After incubation, the wells content was replaced by 100 µl of DMSO and the absorbance of the solutions at the wavelength of 570 nm was measured using SPECTROstar NANO (BMG LABTECH). All experiments were performed in triplicate, and the survival of HEK293 cells was estimated as follows: Survival (%) = (ODexp—ODmin)/(ODmax—ODmin) × 100%, where ODexp is OD570 in the experimental culture wells, ODmin is OD570 in the positive control (0.1% Triton ×− 100) culture wells, and ODmax is OD570 in the negative control (DMEM) culture wells.

Alternatively, HEK293 cells were seeded into 24-wells plates in completed DMEM (Gibco), supplemented with 1% Antibiotic/Antimycotic solution (Capricorn) and 10% fetal bovine serum (HyClone), the day before the experiment. Then, the cells were treated with different concentrations of LysECD7-SMAP and incubated at 37 °C and 5% CO2. After 48 h of incubation, the cells monolayer was fixed with 5% formaldehyde solutions and stained with 0.5% crystal violet (Sigma) and cells disaggregation was assessed.

Determination of bacterial resistance

Resistance development was tested using the repeated exposures of endolysins on bacterial cultures in plate lysis assay and antibacterial assay on initial and passed strains of A. baumannii 50–16 and K. pneumoniae Ts 104–14. Overnight bacterial cultures were diluted in LB broth and grown at 37 °C, 250 rpm to an exponential phase (OD600 = 0.6). Subsequently, 200 µl of the bacterial suspension was spread over a Petri dish with LB agar and air-dried at room temperature. Then, 10 µl of 1 mg ml−1 endolysins solutions or control (20 mM Tris–HCl buffer, pH 7.5) were dropped onto the bacterial lawn, and the dish was treated for 16–18 h at 37 °C. Next day, bacteria from the sub-lethal lysis zone with a not fully cleared lawn were scraped and incubated in LB broth to exponential phase at 37 °C with constant shaking to generate a new bacterial lawn for the next passage. Ten passages were done for A. baumannii and K. pneumoniae. Cultures from initial and passed strains were used to assess antibacterial activity of endolysins as described before in 50 µg ml−1 concentration. All experiments were performed in triplicate.

Endolysin dosage forms formulation

To obtain dosage forms for parenteral administration, 20 mg ml−1 sterile LysECD7-SMAP solution was diluted with sterile PBS pH 7.4 to final concentrations of 2.5 mg ml−1 or 5 mg ml−1, supplied with 0.01% Poloxamer 188 and freeze-dried using FreeZone 2.5 Liter Benchtop Freeze Dryer (Labconco) resulting in lyophilized powder for injections. Lyophilized powders were dissolved in deionized water immediately before the experiment.

Gel preparations for topical administration were formulated with LysECD7-SMAP, containing 5 mg ml−1 or 10 mg ml−1 of protein, 1% Natrosol 250 HHX (w/w) and 1.5% PEG 1500 (w/w). The gel base was mixed separately and sterilized at 120 °C for 15 min. Further sterilized endolysin solution in appropriate concentration was added to the gel and mixed thoroughly. Placebo gel was prepared the same way using PBS solution instead of active compound.

Animal models

All animals were housed in separate cages with controlled temperature (20–24 °C) and humidity (45–65%), fed with a balanced rodent diet and water ad libitum.

Murine sepsis model

Four-week-old female BALB/c mice (weight 16–18 g) were intraperitoneally inoculated with 200 µl of K. pneumoniae M9 suspension in PBS. Based on toxicological study results the sepsis infection in mice was established by inoculating 1 × 103 CFU per animal of K. pneumoniae M9. At 2 h after infection mice were intravenously treated with 100 µl of LysECD7-SMAP solutions at protein concentrations of 2.5 mg ml−1 (n = 33) or 5 mg/ml (n = 33). For 15 animals the survival analysis was conducted and 6 animals were euthanized to assess the bacterial loads at each time point (1st before the beginning of treatment, 3rd and 7th days post infection, PI). Control groups included mice treated with 100 µl of PBS pH 7.4 (placebo group, n = 33), 100 µl of gentamycin 40 mg ml−1 (positive control, n = 24) or 100 µl of ampicillin 100 mg ml−1 (negative control, n = 24), in antibiotic groups 3 mice were euthanized for bacterial assessment at each time point. Also 33 mice were left untreated. The animals were treated twice a day with 10 h interval for 5 days.

To access the endolysin therapeutic effect general examination of animals, body weight and survival rate were observed daily for 8 days after the infection. Also, microbial contamination of spleen and lungs homogenates (on the 1st, 3rd and 7th days PI) and blood samples (on the 1st and 3rd days) were assessed by inoculation of blood or organ homogenates dilutions on GRM-1 agar medium. CFU count of K. pneumoniae was estimated after plates cultivation for 18 h at 37 ºC.

Murine pneumonia model

Four-week-old female BALB/c mice (weight 16–18 g) were intranasally inoculated with 20 µl of K. pneumoniae M9 suspension in PBS (2 × 106 CFU per animal). At 2 h after infection mice were intravenously treated with 100 µL of LysECD7-SMAP solutions at protein concentrations of 2.5 mg ml−1 (n = 24) or 5 mg ml−1 (n = 24). Control groups included mice treated with 100 µl of PBS (placebo group, n = 24) or 100 µl of gentamycin 40 mg ml−1 (positive control, n = 24). Also 24 mice were left untreated. In this model we did not used ampicillin control group as it showed low effectiveness in the sepsis in combating the infectious agent comparable to the level of untreated animals. In each group, for 15 animals the survival analysis was conducted and 3 animals were euthanized to assess the bacterial loads at each time point (1st before the beginning of treatment, 3rd and 7th days, PI). The animals were treated twice a day with 10 h interval for 5 days.

To access the endolysin therapeutic effect general examination of animals, body weight and survival rate were observed daily for 8 days after infection. Also, microbial contamination of spleen and lungs homogenates (on the 1st, 3rd and 7th days PI) and blood samples (on the 1st and 3rd days) were assessed by inoculation of blood or organ homogenates dilutions on GRM-1 agar medium. CFU count of K. pneumoniae was estimated after plates cultivation for 18 h at 37 ºC.

Diabetic wound infection

In the diabetic wound model 19 male leptin-deficient db/db mice (BKS.Cg-Dock7m + / + Leprdb/J, Wildtype for Dock7m, Homozygous for Leprdb, The Jackson Laboratory) were maintained until their blood glucose (non-fasted) concentration reached at least 15 mmol l−1 (≥ 270 mg dl−1) or higher [32] for two days. The blood was collected from a warmed tail vein and analyzed by glucose analyzer (OneTouch Select® Plus Flex, LifeScan Johnson&Johnson). The age of mice at start of experimental time frame was 9 weeks.

When type 2 diabetes mellitus developed, as indicated with the blood test, the mice backs were shaved with an electric razor and four full-thickness skin wounds (d = 4 mm) were made with biopsy gun Dermo-Punch (Sterylab, Italy). Each wound was used to study either wound swabs contamination, either dermal graft homogenates contamination, either wound-closure course or histological examination.

Five minutes later each wound was infected with 1.5 × 108 CFU per animal of P. aeruginosa 38–16 clinical isolate (25 µl in PBS). Mice were randomly allocated into experimental and control groups and at 24 h PI, the infected wounds were epicutaneously treated with 50 µl of LysECD7-SMAP gel at protein concentration 10 mg ml−1 (n = 8), placebo gel (n = 8) or left untreated (n = 3). The animals were treated twice a day with 10 h interval for 5 days.

To access the endolysins therapeutic effect the dynamics of wound healing, wound size and microbial contamination were observed on 1, 3, 7, 10 and 14 days after infection. CFU count of P. aeruginosa was estimated after the swabs of the wounds with sterile cotton balls wetted with saline were performed, serially diluted and plated on Agar B Medium. After 14th day of experiment, the mice (n = 4 in experimental groups and n = 3 in untreated group) were euthanized by CO2 inhalation, heart blood samples were taken and the infected dermal grafts were excised and homogenized in 1 ml of PBS. The solutions were serially diluted, plated on Pseudomonas Agar B Medium (for Fluorescein), containing peptone (20.0 g l−1), agar (15.0 g l−1), glycerol (10.0 g l−1), MgSO4·7H2O (1.5 g l−1), K2HPO4 (1.5 g l−1) and bacterial CFU were counted after an overnight incubation at 37 °C. Wound-closure course (digital planimetry) was analyzed by tracing the wound margin. For this, the pixel area of wound image was calculated using Adobe Photoshop CS6 software.

Histology

One of four wounds in LysECD7-SMAP, placebo and untreated group was used for histological examination in the diabetic wound model. For mice euthanized on the 14th day dermal grafts were excised with biopsy gun (d = 8 mm, Sterylab) and fixed with 10% formaldehyde. Then biopsy samples were processed and embedded in paraffin (Leica TP 1020), and sectioned at 4 μm (MICROM HM340). These sections were deparaffinized, dehydrated, applied on glass slides and stained with hematoxylin/eosin according to the manufacturer’s instructions to assess wound healing processes. Tissue sections were analyzed with Leica DM LB2 microscope equipped with a Leica D300 digital camera (Leica).

Bacteria genotyping

DNA was extracted from pure cultures using the QIAamp DNA Mini Kit (Qiagen) or RIBO-prep (Amplisens) as recommended by the manufacturers. To control for DNA contamination an additional sample containing MQ water was processed in parallel. DNA quantification were evaluated with Qubit 4 and Qubit dsDNA HS (Thermo Fisher Scientific, USA). 16S rRNA genes were amplified using Phusion High Fidelity DNA Polymerase (NEB, USA) and primer pair 27F and 1492R [33] according to the manufacturer’s instructions. PCR product purification was conducted with Cleanup S-Cap (Evrogen). The DNA sequencing was performed with the BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, USA) following manufacturer instruction. The reaction with BigDye v3.1 was purification using xTerminator® BigDye Purification kit (Applied Biosystems, USA) and the sequencing product was by the 3500 Genetic Analyzer (Thermo Fisher Scientific, USA). Sequencing analyses were performed using DNA Baser Sequence Assembler v.5.15 (Heracle BioSoft S.R.L.). Taxonomic identification was performed using BLAST.

Burn wound infection

In the burn wound model induced with P. aeruginosa 38–16 clinical isolate (sputum of ICU patient), backs of 8–10 weeks old male Wistar rats (180–210 g) were shaved with an electric razor. To induce skin burn, a copper plate (surface area 150 mm2) was heated to 300 °C and then applied on the shaved backs of the rats for 30 s. To restore the water-electrolyte balance 1 ml of PBS buffer was injected subcutaneously to the burn areas. Five minutes later the burn was infected with 2 × 109 CFU per animal of P. aeruginosa (100 µl in PBS). P. aeruginosa 38–16 shows resistance towards ampicillin; chloramphenicol; ceftazidime; cefotaxime; gentamicin; meropenem; polymyxin and tetracycline, moderate resistance to polymyxin B, as estimated by MIC assay microdilution method in Mueller–Hinton broth (Difco) [34]. The strain was not lethal in animal models when applied epicutaneously, however, in the absence of treatment, a stable infection develops with the preservation of wounds contamination up to 3 weeks.

At 24 h PI, the infected burns were epicutaneously treated with 500 µl of LysECD7-SMAP gel at protein concentrations of 5 mg ml−1 (n = 7) or 10 mg ml−1 (n = 7). Control groups included rats treated with 500 µl of vehicle-treated animals (placebo group, n = 7). Also 7 animals were left untreated. The animals were treated twice a day with 10 h interval for 5 days.

To assess the endolysins therapeutic effect the dynamics of burn wound healing, wound size and microbial contamination were observed on 1, 3 and 7 days after infection. CFU count of P. aeruginosa was estimated after the swabs of the wounds with sterile cotton balls wetted with saline were performed, serially diluted and plated on Pseudomonas Agar B Medium. After 7 days of experiment, the rats were sacrificed by CO2 inhalation, and the infected dermal grafts and animal spleens were excised and homogenized in 1 ml of PBS. The solutions were serially diluted, plated on Agar B Medium and bacterial CFU were counted after an overnight incubation at 37 °C. The wound-closure rate during the treatment course was monitored with the planimetry.

Statistical analysis

Data were analysed using GraphPad Prism 9.0 software. A value of p < 0.05 was considered statistically significant. The methods used for the comparative tests are indicated in the captions of the figures and tables.