Anti-bacterial properties of hyperbranched poly (epsilon-lysine) peptides dendrons for wound dressing applications: molecular specie-specific effects

Abstract

Background

Wound dressings, as tulles or hydrogels, are the devices of choice for the management of wounds and they are also made available to clinicians in formulations integrating silver nanoparticles as antibacterial agents. Nevertheless, the management of wounds is significantly affected by bacterial infections that limit healing to only 45% of the cases.

Results

Poly(epsilon-Lysine) branched peptides, called dendrons, were synthesised with different levels (also known as generations) of branching by an automated solid-phase peptide method. Production at hundreds of miligram scale and over 90% of purity were obtained. A study of the antimicrobial properties of these branched peptides, when in soluble form or used to impregnate commercial wound dressings, was performed on two bacterial species relevant to wound infections, Staphylococcus aureus and Pseudomonas aeruginosa. The data showed the antibacterial activity of these molecules, when presenting three generations of molecular branching, on Staphylococcus aureus both in planktonic and biofilm conditions.

Conclusions

The present study shows that hyperbranched poly(epsilon-Lysine) dendrons can be used for wound S. aureus infection prevention, as an alternative to silver, either in soluble form or when used to impregnate the fibres of clinically-available, tulle- or hydrogel-based wound dressings.

Background

Chronic wounds are a significant social and economic burden. A 2017/2018 cohort study estimated that the UK NHS had spent £8.3 billion on acute/chronic wound care (3.8 million patients) [1]. The analysis also showed 59% of chronic wounds healed when there was no infection compared with 45% if infection was ascertained or suspected [1].

Indeed, if not effectively cleared by macrophages, contaminating bacteria promote the continued production of pro-inflammatory cytokines, subsequently prolonging the inflammatory phase [2–4]. Resultant tissue necrosis and debris create a favorable environment for biofilm formation by encouraging bacterial adherence to the wound surface [3–5]. Multiple studies report S. aureus and P. aeruginosa as the most common microbes found in infected wounds and responsible for biofilm formation in chronic wounds, with S aureus contributing to the majority (43% to 63%) of the cases [3, 4].

As a result, researchers have become increasingly interested in the study and development of novel antimicrobials acting through varying mechanisms of action such as natural and synthetic antimicrobial peptides [6, 7]. In particular, cationic peptides have been shown to exert an antibacterial mechanism of action whereby direct interaction with the negatively charged bacterial membrane causes membrane permeability and disruption and, as a consequence, the lysis of the bacterial cell [6, 7]. This mechanism of action is advantageous because of both its relative rapidity and its lack of interference with metabolic pathways linked to drug resistance [6].

In 2019, it was reported that over 3000 different wound dressings existed within the wound care market [8], allowing healthcare professionals the flexibility to provide more efficient wound care strategies depending on the wound conditions (e.g. varying exudate volumes, occurrence of infection) [9–11].

In the last two decades, wound dressings, both tulle- and hydrogel-based, have been impregnated with silver nanoparticles because of the ascertained antibacterial properties of ionic silver that can be gradually liberated at relatively high concentrations from the large surface area of the nanoparticles and lead, as in the case of the cationic peptides, to the disruption of the bacterial cell membrane and DNA damage [12, 13]. At the same time, some evidences of bacterial resistance to silver have been reported [14] and silver nanoparticles on infected wounds appear to provide benefits only when used in the first few days of treatment with no evidence of improve percentage of healing of non-infected wounds [15, 16].

In this study, hyperbranched poly(epsilon-Lysine) peptides (i.e. dendrons) with increasing molecular branching were synthesised by a solid-phase peptide synthesis method and assessed for their:

• degree of purity after synthesis.

• antibacterial properties on clinical isolates of two relevant bacterial species; S. aureus and P. aeruginosa.

• Suitability for integration in a range of clinically-available wound dressings.

• Ability to confer antibacterial properties to a range of wound dressings.

Methods

Dendron synthesis

Three formulations of poly(epsilon-lysine) dendrons exposing an arginine (R) residue at their carboxy terminal and with varying degrees of branching generations (i.e. RGen₁K, RGen₂K and RGen₃K) were synthesized using the Liberty Blue microwave peptide synthesizer (CEM, US). The automated solid-phase peptide synthesis was performed using with Fmoc (9-fluorenylmethyloxycarbonyl) protected amino acids using a Fmoc Rink Amide resin scaffold (CEM, 100–200 mesh, 0.58mmol/g and the Fmoc-protected amino acids Fmoc-Arg(Pbf)-OH (Novabiochem^®, Merck) and Fmoc-Lys(Fmoc)-OH (Iris Biotech, GmbH). Synthesis scales of 0.1mmol and 0.25mmol per batch were used.

All reagents were prepared in anhydrous solvent synthesis grade N-dimethylformamide (DMF) with a concentration of 0.2 M. All amino acids were coupled by N-Diisopropylethylamine (DIPEA)/O-(1H-Benzotriazol-1-yl)-N, N,N’, N’ -tetramethyluronium hexafluorophosphate agents in DMF. Fmoc-deprotection was performed by 20% v/v piperidine in DMF. HBTU, DIPEA, Piperidine and DMF were all of analytical grade sourced from Fisher Scientific, UK.

Following final deprotection of the uppermost branching generation from Fmoc groups, the newly synthesized dendrons, still attached to the resin, were washed eight times with 5 ml of HPLC grade dichloromethane, methanol and diethyl ether (Fisher Scientific, UK), respectively.

The peptides were cleaved from the rink-amide resin by its resuspension in a solution of 95% trifluoroacetic acid (TFA), 0.25% H₂O and 0.25% triisopropylsilane (TIIPS), (95:2.5:2.5 v/v/v) at room temperature for 3 h. TFA and TIIPS were sourced from Thermo Scientific, UK. For every 50 mg of resin-peptide 1 ml of cleavage solution was used. For each cleavage, 200 mg of peptide-resin and 4mls of cleavage solution were used. After the 3 h incubation period, the solution of cleaved dendrons was filtered through glass Pasteur pipettes filled with glass wool (0.5 cm thick) into a 50mL Falcon tube containing chilled (−20 °C) diethyl ether. This step led to the precipitation of the peptide forming a visible white pellet. The precipitated pellets were made firmer by centrifugation for 5 min at 3500 rpm and collected by discarding the diethyl ether solution supernatant. The precipitated peptide pellet was washed in 20mL diethyl ether three times before being allowed to dry overnight under the fume hood and finally freeze dried.

Dendron characterisation

Dendron yields (mg) were calculated by a gravimetric method where the weight of the freeze-dried samples was measured by subtracting the weight of the empty falcon tube (prior to cleavage) from the final weight of the peptide containing falcon tube after lyophilization.

To identify the characteristic spectra of each dendron Fourier-Transformed Infrared spectroscopy (FTIR) measurements were performed using a PerkinElmer FTIR Spectrophotometer Spectrum Two. Peptide samples were measured within a spectral range of 4000–450 cm⁻¹ at a resolution of 4 cm⁻¹. Each spectrum was collected from 16 scans.

A Bradford Assay was conducted to estimate protein concentration of peptide samples following lyophilisation. Stock solutions (2 mg/ml) of RGen₁K, RGen₂K and RGen₃K samples as well as the that of the protein standard Bovine Serum Albumin (BSA) were prepared in a synthetic body exudate (SBE). Two-fold serial dilutions were prepared from stock solutions. Samples (10 µl) were added to a 96-well-plate (ThermoScientific™, Denmark) to which 290 µl of ready to use Bradford reagent (Pierce™ Bradford Protein Assay Kit, Thermo Scientific) were added. A blank of just SBE was also prepared. After 5 min at room temperature the absorbance of the samples and standards were measured at 595 nm using an ELX600 spectrophotometer (Agilent Biotek, USA). The absorbance values obtained for the standard solutions were plotted against their corresponding protein concentrations to generate a standard curve. Protein concentrations of each type of dendron samples were determined by comparing their absorbance values to the standard curve obtained from the BSA standard solutions.

Purity of peptides was determined by a high pressure liquid chromatography (HPLC, Agilent Technologies,1100 Series) equipped with a 5 μm Fortis C18 150 × 4.6 mm column (Phenomenex, UK) using a timed gradient of 100-0% 0.1% v/v TFA in water (solvent A) and 0–100% 0.1% v/v TFA in acetonitrile (solvent B) with a flow rate of 1mL/min for 15 min, and detection at 223 nm. 2 mg/mL samples of each peptide were prepared in methanol and an injection volume of 20 µl was used. Column was heated to 25 °C. All chemicals including water used were of HPLC grade purity from Fisher Scientific and solvents A and B were degassed for 10 min prior to use.

Mass spectra of dendrons were obtained with a hybrid quadrupole orbitrap mass spectrometer (Q Exactive, ThermoFisher) by direct injection using an in-house platinum coated fused silica emitter (Nanospray Flex™ Ion Source, ThermoFisher). The mass spectrometer was operated in the positive mode, spectra recorded at a resolution of 140,000@200 m/z, with a spray voltage of 1.6 kV, capillary temperature at 250 °C, and S-lens Rf level at 50. Peptide samples (2 mg/mL) were prepared in 50% v/v acetonitrile in deionized water (LC/MS grade) sourced from Fisher Scientific, UK.

The average size and size distribution of the dendrons were measured using a dynamic light scattering device (DLS, Zetasizer Pro, Malvern, UK). Samples (2 mg/ml) of each peptide were prepared in deionised water and measured in a disposable cuvette at 25 °C. Each scan was performed in triplicate. Each dendron sample reading was proceeded with a blank of deionized water measured under the same conditions so to establish baseline.

For this study, commercially available dressings specifically used in the management and treatment of moderate to heavily exudating chronic wounds were functionalised: N-A^® (Systagenix Wound Management Limited, North Yorkshire, UK), Kaltostat^® (ConvaTec Limited, Flintshire, UK), KerraCel™ (Crawford Healthcare Limited, Cheshire, UK) and Aquacel™ Extra™(ConvaTec Limited, Flintshire, UK).

Wound dressing coating and characterization

Wound dressings were functionalized with the different types of dendron by harnessing their swelling properties. Using an autoclaved cork borer 1 × 1 cm³, cuttings of each dressing were prepared and soaked in 5 to 7 ml of a 1 mg/mL solution of RGen₃K in sterile 75% v/v ethanol in water. Dressings were soaked for an hour after which residual peptide solution was discarded and dressings were allowed to dry overnight.

Non-functionalised wound dressings used as negative controls were also soaked in sterile 75% v/v ethanol in water and dried overnight. The functionalisation of wound dressings was performed under sterile conditions using a laminar flow hood and confirmed by assesing samples of each dressing from every batch with 200 µL of Bradford reagent (Pierce™ Bradford Protein Assay Kit, Thermo Scientific).

Assessment of anti-bacterial properties of soluble dendrons RGen₁K, RGen₂K and RGen₃K dendrons and of RGen₃K-impregnated dressings

Typed/referenced strains of S. Aureus NCTC 10,788 (SA) and P. Aeruginosa NCIMB 10,548 (PA) were obtained from the National Collection of Type Cultures, UK and The National Collection of Industrial, Food and Marine Bacteria, UK respectively.

For growth, strains were cultured in Müeller-Hinton broth (Oxoid, UK) for 18 h at 37 °C. The inoculum turbidity was adjusted to 0.5 McFarland standard (1 × 10⁶ CFU/ml) as recommended by the Clinical and Laboratory Standards Institute (CLSI) guidelines (Susceptibility Testing Subcommittees (clsi.org)).

Broth microdilutions set-up in 96-well plates (Nunclon™ Delta Surface Treatment ThermoScientific™, Denmark) were performed to determine the minimum inhibitory concentrations (MIC) of each dendron formulation against each of the two bacterial species under investigation. Stock solutions of each peptide prepared in Müeller-Hinton broth (MHB) underwent a two-fold serial dilution to obtain final concentrations ranging from 0.0625 to 4 mg/ml. Vancomycin hydrochloride and Gentamycin sulfate (Thermo Fisher Scientific, UK) were included as positive controls whilst broth only and normal uninhibited bacterial growth were used as negative controls. Plates were incubated for 48 h at 37 °C with optical density measurements taken at 600 nm using an ELX600 spectrophotometer (Agilent Biotek, USA) at 0, 1, 2, 3, 4, 5, 6, 24 and 48 h post inoculation. For each concentration of peptide, the optical density was plotted versus time. The MIC was the lowest concentration of dendron that inhibited the growth of the tested bacterial strain.

The minimum bactericidal concentrations (MBC) of each type of dendron were determined by the inoculation of Mueller Hinton agar (MHA) plates with 50 µl of MIC broth samples at different concentrations following the final spectrophotometric readings at 600 nm (48 h). Plates were then incubated for 24 h at 37 °C after which growth was evaluated. MBC was defined as the dendron concentrations that did not show any bacterial growth on freshly infected agar plates. Both MIC and MBC determinations were performed under static conditions, without agitation.

Microbial suspensions of S. aureus and P. aeriginosa in sterile phosphate buffered saline pH = 7.4 (PBS) were swabbed on plates of iso-sensitest agar (Oxoid, UK). Nutrient, Mueller Hinton and cation-adjusted Mueller Hinton agars (Oxoid, UK) were also tried and tested however were found to yield significantly reduced or no zones of inhibition.

A series of peptide solutions with concentrations of 0.125, 0.25, 0.5, 1, 2, 4, 8, 16 mg/ml were prepared in sterile PBS. An aliquot (20 µl) of each peptide concentration was loaded onto 6 mm diameter diffusion discs before being directly placed on the readily swabbed agar plates. RGen₃K functionalised wound dressings were wetted with PBS before being directly placed on the recently swabbed agar plates. Vancomycin and gentamycin (Oxoid, UK) diffusion discs were used as positive controls.

Plates were incubated for 24 h at 37 °C. Zones of inhibition (ZOI) were defined as the area surrounding the point of disc/wound dressing application where bacterial growth was impeded and therefore were measured by their widest diameter using ImageJ software.

Results obtained from the MIC and MBC determination experiments showed Rgen₃K to have a bactericidal effect on S. Aureus at lower concentrations than RGen₂K and RGen₁K. RGen₃K was therefore deduced the most effective and commercially preferred antimicrobial formulation of dendron. As a result, antibiofilm activity was only evaluated for the RGen₃K dendrimer.

Antibiofilm activity was investigated in two stages, the first observed effects of the dendron when in solution whilst for the second stage of the experiment RGen₃K functionalised wound dressings were used (prepared as shown in Sect. 2.3). This allowed us to observe and compare the antimicrobial and antibiofilm activities of RGen₃K and commercial wound dressings both separately and when used in conjunction with one another.

No dendrimer exhibited an inhibitory effect against P. aeruginosa, so biofilm impact experiments only focused on S. aureus, with all experiments conducted under static conditions.

When evaluating the antibiofilm activity of planktonic peptides, final working concentrations, 0.5 mg/ml and 1.0 mg/ml were used. S. aureus biofilms were set up in cell-culture treated 12-well plates (Nunclon™ Delta Surface Treatment ThermoScientific™, Denmark). Each well was made up to 1 ml with inoculum in MHB. In order to better observe and evaluate the impact of the dendron on the various stages of biofilm growth and development each well was treated with 0.5 ml of RGen₃K solution (in MHB) at various timepoints, 0, 4, 8 and 24 h with 0 h being the point of inoculation. For the control, 0.5 ml of sterile MHB was added instead. Throughout this experiment plates were incubated at 37 °C, under static conditions. Biofilms were then quantified 4, 8, 24, 48 h post inoculation. This allowed for the impact to biofilm growth and development to be observed as well as demonstrate how the point of application could impact effect efficacy. Also, the above time range was meant to consider the typical time of application of wound dressings that are changed every 3 or 7 days.

This experiment was then repeated using RGen₃K dendrimer functionalized wound dressings which were each added to different wells (already made up to 1.5 ml with inoculum in MHB)at various timepoints 0, 4, 8, 24 h, 0 h being the point of inoculation. Biofilms were then quantified 4, 8, 24 and 48 h post inoculation. Non-functionalised wound dressings were used as a negative control with exception to non-functionalised Kaltostat^® wound-dressings which due to the established antimicrobial properties of alginate were used as a positive control.

To quantify biofilms the applied bacterial suspensions and functionalized wound dressings were carefully removed. Wells were then washed with sterile deionised water three times in order to remove non-adhered cells and fragmented dressing before adding 1mL 0.5% w/v crystal violet (PL.7002 Prolab Diagnostics, UK) for 20–30 min at room temperature. Crystal violet was washed away with sterile deionised water and the plate was allowed to air dry. Dry 30% v/v acetic acid was then added to solubilise residual crystal violet. The crystal violet and acetic acid solution were transferred to 96-well plates and measured at 604 nm using an ELX600 spectrophotometer (Agilent Biotek, USA). The amount of residual dye present after washing directly reflected the amount of cells in the biofilm. These OD₆₀₄ nm measurements were plotted into bar charts/graphs.

Inhibition % could be calculated using the following equation:

These biofilm quantification timepoints (4, 8, 24 and 48 h post inoculation) correspond with key sequential stages of biofilm formation; initial attachment, microcolony formation/early maturation and mature biofilm development which allows for a more informed evaluation of antibiofilm activity.

Statistical analysis

Experiments were performed in triplicate and findings were presented as mean ± standard deviation. Biofilm quantification data was evaluated by Analysis of Variance (ANOVA) and post ad hoc Tukey’s test using excel software. Values were considered significant at p values ≤ 0.05.

Results

Yields of dendron synthesis

Table 1 shows the yield of synthesis for all the dendron formulations (i.e. RGen₁K, RGen₂K, RGen₃K). The percentage yield for each dendron differed substantially depending on both the peptide and the chosen synthesis scale (mmol). Higher % yields for all three peptides were achieved when using a high synthesis scale. Furthermore, higher % yield were more consistently achieved with the larger dendrons RGen₃K and RGen₂K where the yield percentage at a synthesis scale of 0.25 mmol reached 82% and 75%, respectively. In particular, it was noted that during the cleavage of RGen₁K the resin showed a tendency towards aggregation even following incubation in the cleavage solution. This aggregation made collection using a Pasteur pipette challenging and resulted in a significant amount of resin-peptide being left in the flask.

Table 1.

Theoretical, actual and percentage yields of each dendron formulation according to their individual MWs and the set synthesis scale. Actual values were obtained by gravimetric analysis of the precipitated and freeze-dried product powder

Dendron Formulation	Theoretical MW (g/mol)	Synthesis Scale (mmol)	Maximum Theoretical Yield (mg)	Expected Theoretical Yield (mg)	Actual Yield (mg SD)	%
RGen 1 K	558.42	0.1	56	19	10 2.5	54
		0.25	140	47	29  5.7	62
RGen 2 K	1070.80	0.1	107	36	23  4.6	64
		0.25	267.5	89	67 4.2	75
RGen 3 K	2095.56	0.1	209.5	70	44 5.6	63
		0.25	524	175	144  29.7	82

It was anticipated that the mass of peptide obtained after cleavage would be a third of the total mass of peptide obtained from the synthesizer. The maximum theoretical yield of each peptide was calculated as 0.1 and 0.25 mmol of the peptide’s theoretical MW (g/mol) as shown in Table 1.

The yield of reaction of each type of dendron was also evaluated by a protein Bradford assay using bovine serum albumin as standard curve (R² = 0.9975). When the dendron freeze-dried powders were dissolved at a concentration of 4 mg/ml and tested by the Bradford assay they showed concentrations of 10.94 µg/ml for RGen₁K, 149.22 µg/ml for RGen₂K and 278.35 µg/ml for RGen₃K.

FTIR

The FTIR spectra for RGen3K, RGen2K and RGen1K consistently exhibited the same characteristic peaks to varying %T intensities. The characteristic peaks identified in Fig. 1 corresponded to the expected functional groups .

Fig. 1.

FTIR spectra of dendrons with increasing branching generation. A RGen₁K, (B) RGen₂K, (C) RGen₃K. While showing reproducibility of the main peptide peaks, the increasing generations show differences in peaks ratios in the regions 1178 cm⁻¹ and 1125 cm⁻¹ (arrows)

Figures 1 A-C show the FTIR analysis of the three generations of dendrons. For each dendron formulation, the analysis showed the typical primary amide peaks carbonyl stretching vibrations at 1659–1660 cm^− 1 and of the R-NH₂ at 720, 798 and 835 cm^− 1. Secondary amine peaks in the region 1529–1532 cm^− 1 were also found. Changes between peak ratios were observed in the shoulders of the peaks at 1179 cm^− 1 and at 1125 cm⁻¹ (Figs. 1 A-C, arrows). The former showing the shoulder emerging as a separate peak, the latter showing a shoulder gradually disappearing as the number of branching generations increased. These changes indicated the emergence of conformational changes in the dendron structures as the molecular branching increases.

Peaks typical of the stretching of the N-H band in the 3000 cm⁻¹ region were also observed.

HPLC

Figures 2A-C show the HPLC chromatograms of the three dendron formulations. In all cases, the typical peak of dendrons appears at an elution time of approximately 3 min. A peak eluting approximately at 1.9 min also appeared that was attributed to the solvent peak. The higher mAU value (> 1200 mAu) of the RGen₃K dendron clearly show the higher amounts obtained from the synthesis of this type of dendron with a higher number of molecular branching. Smaller peaks attributed either to contaminant or dendron aggregates were observed in the case of the RGen₁K and RGen₂K formulations.

Fig. 2.

HPLC of dendrons of varying molecular branching. A RGen₃K, (B) RGen₂K, (C) RGen₁K. The elution peak of each formulation was observed at approximately 3 min

Mass spectrometry

Mass spectrometry data confirmed the relatively higher degree of purity obtained from the synthesis of RGen₃K (Figs. 3A-C).

Fig. 3.

Mass spectrometry of dendrons of varying molecular branching. A RGen₁K, (B) RGen₂K, (C) i. RGen₃K, ii. RGen₃K expanded. Each spectrum also reports the theoretical molecular weight of each dendron. [M + H] + x annotations indicate dendron fragments and their charge as identified by the Protein Prospector software

As expected, the spectrum of RGen₁K dendron showed a limited number of fragments all clearly assigned to charged fragments of the whole branched peptide.

Spectra for RGen₂K (Fig. 3B) exhibited peaks for an [M + H]⁺ ion at 1070.7958 m/z (an almost exact match to the theoretical MW), an [M + H]⁺² ion at 535.9019 m/z, an [M + H]⁺³ ion at 357.6047 m/z and an [M + H]⁺⁴ ion at 268.4559. The largest peaks within the spectra were those of [M + H]⁺³ and [M + H]⁺⁴. Results confirmed the RGen₂K peptide (Theoretical MW 1070.7997 g/mol) to be dominant species within the sample.

Spectra for RGen₃K (Fig. 3 Ci and ii) exhibited high volumes of peaks from multiple species. The largest peaks within the spectra were of [M + H]⁺⁴ and [M + H]⁺⁵ at 524.89 m/z and 420.1182 respectively.

Various molecular ions of higher charged states were identified within each spectrum. As expected, higher charged states, ≥[M + H]⁺⁴ were observed in the RGen₂K and RGen₃K spectra. No single charged ions < [M + H]⁺⁴ could be identified within the spectra for RGen₃K. Indeed, as the branching generations increased so did the presence of multiply charged species. Additional peaks of variants of dendrons suggesting the presence of truncated (i.e. incomplete synthesis) or greater lysine branches were also observed.

DLS

As expected, RGen₃K exhibited the largest Z-average (4110 ± 454.19 nm) when compared to RGen₂K (586.3 ± 166.90 nm) and RGen₁K (496.5 ± 106.95 nm).

The polydispersity Index (PDI) results given showed that the RGen₃K had a relatively broad particle size distribution suggesting potential aggregation of the dendrons. PDI values decreased with the number of branching generations showing that both RGen₂K and RGen₁K were more dispersed than RGen₃K.

Wound dressing coating

The study of the poly(epsilon-Lysine) dendron as surface functionalization moiety of wound dressings focused on the RGen₃K formulation as it showed better yield of synthesis and purity. The presence of the cast film was assessed by staining of the uncoated and RGen₃K-coated wound dressings, NA, Kaltostat, KerraCel and Aquacel. Among these dressings, NA, Kerracel and Aquacel were based on cellulose derivatives (i.e. NA: rayon, Kerracel and Aquacel: carboxymethylcellulose), while Kaltostat is an alginate-based dressing.

Figure 4 shows the different Bradford staining results as obtained from the uncoated and RGen3K-coated dressings.

Fig. 4.

Assessment of RGen₃K functionalization of clinically-available wound dressings

The staining procedure showed the uniform coating of the tulle-based and hydrogel based carboxymethyl cellulose dressings, the presence of the dendrons resulting in a blue staining. However, in the case of Aquacel and Kaltostat a background staining did not enabled a clear difference between the control and the coated biomaterials. This was particularly the case for the alginate-based dressing Kaltostat, while a more intense blue/purple staining appeared in the coated Aquacel.

Antibacterial properties of soluble dendrons

The antimicrobial activity of soluble RGen₁K, RGen₂K and RGen₃K were measured against bacterium species in a concentration dependent manner and their growth patterns (lag, log and stationary phases) were compared to those of untreated microbial culture (Fig. 5).

Fig. 5.

Bacterial growth curves for S. aureus (SA) and P. aeruginosa (PA). Plots show concentrations at which dendron formulations (i.e. RGen₁K, RGen₂K, RGen₃K) inhibit log phase (blue arrows) and stimulate log and lag phases (red arrows). Lower concentrations showed no significant effect on bacterial growth

When RGen₁K was used to treated S. aureus in the range of concentrations (0.125 to 16 mg/ml no inhibition of the growth pattern was observed when compared to that of the untreated control. RGen₂K showed to inhibit the S.aureus growth pattern in the range of concentrations 0.5 to 16 mg/ml, this dendron MIC being determined to be 0.5 mg/ml. RGen₃K showed an inhibitory effect on S. aureus across a wider range of concentrations; i.e. 0.25 to 16 mg/ml, with its MIC determined to be 0.25 mg/ml. RGen₃K also expressed an inhibitory activity characterised by the extension of the lag phase and subsequent delay to the log phase over a slightly wider range of concentration compared to RGen₂K and RGen₁K.

These results demonstrated that the tested dendron formulations had both a concentration and molecular branching-dependent inhibitory impact on the growth of S. aureus.

In the case of P. aeruginosa, RGen₂K and RGen₃K showed an inhibitory effect during the bacterium log phase starting from 0.5 mg/ml. However, all dendron formulations led to a higher growth at all the tested concentrations.

Noticeably, at concentrations of RGen₂K and RGen₃K lower than their MIC, the log phase consistently started earlier and its transition into the stationary phase occurred sooner than the untreated control by leading to a higher bacterial growth. The final absorbance measurements at 48 h were also considerably higher compared to those of the untreated bacterial strains. This growth pattern was also observed at all concentrations in the case of RGen₁K.

The results of S. aureus growth inhibition obtained for RGen2K and RGen3K prompted a study of the dendron MBA (Fig. 6). When evaluating the MBC plates, RGen₂K treated S. aureus, no growth was observed for concentrations 4-1 mg/ml consistently. Growth was observed at a concentration of 0.5 mg/ml. The MBC was therefore determined to be 1 mg/ml. For RGen₃K treated S.aureus, no growth was observed for all tested concentrations 4-0.5 mg/ml consistently. Growth was observed at a concentration of 0.25 mg/ml. MBC was therefore determined to be 0.5 mg/ml.

Fig. 6.

Photographs of MBC plates for dendron samples microdilutions at concentrations ranging from 0.5 to 4 mg/ml

Agar diffusion tests were also performed to determine the ZOI of the different dendron formulations on both bacterial species under investigation (Table 2).

Table 2.

ZOI induced by dendron formulations on S.aureus agar plates

Dendron Concentrations (mg/ml)
Dendron Formulation		0.125	0.25	0.5	1.0	2.0	4.0	8.0	16.0
RGen1K	ZOI Diameter (mm)	-	-	-	-	-	-	-	-
RGen2K	ZOI Diameter (mm)	-	-	-	-	-	-	7.30 ± 0.30	8.64 ± 0.52
RGen3K	ZOI Diameter (mm)	-	-	-	-	-	7.57 ± 0.37	8.33 ± 0.52	10.20 ± 0.56
Gentamycin	ZOI Diameter (mm)	31.34 ± 0.50

ZOI only developing for S. aureus treated with concentrations 4–16 mg/ml for RGen₃K and 8–16 mg/ml for RGen₂K. No ZOI were observed at any concentration of RGen₁K. This general pattern of concentration- and molecular branching-dependent antibacterial activity was however still observed. Like in the case of the microdilution tests, no ZOI were observed for P. aeruginosa at any concentration or dendrimer generation. The lack of cohesion between the two susceptibility tests was attributed to the inhibition of the diffusion of the highly cationic peptides in the negatively-charged agarose gels.

In the case of S.aureus tests, RGen₃K at the highest concentration of 16 mg/ml showed a mean zone of growth inhibition of 10.20 ± 0.563 mm, while the ZOI induced by RGen₂K was at the highest tested concentration of 16 mg/ml.

The prevention and eradication of S. aureus biofilm by soluble RGen₃K were also investigated (Figs. 7A-D). When RGen₃K was tested at 0.5 mg/ml and 1.0 mg/ml an antibiofilm activity could be detected. This preventative antibiofilm activity was considered both concentration and time dependent.

Fig. 7.

RGen₃K inhibitory activity on S.aureus biofilm formation and growth. A RGen₃K spiking at 0 h inoculation time, (B) RGen₃K spiking at 4 h post-inoculation, (C) RGen₃K spiking at 8 h post-inoculation, (D) RGen₃K spiking at 24 h post-inoculation. * indicates data significantly different at p ≤ 0.05

When spiked with RGen₃K solutions at time 0 h of inoculation an initial stimulation of bacterial growth was observed when compared to the control. This was followed by a significant inhibition at 4, 8, 24 and 48 h of incubation (Fig. 7A). A similar pattern observed when bacterial cultures were spiked with RGen₃K solutions at time 4 h post inoculation (Fig. 7B).

However, when bacterial samples were spiked with RGen₃K solutions at 8 h (Fig. 7C) and 24 h (Fig. 7D) post inoculation no significant antibiofilm activities were observed. The initial stimulation rather than an inhibitory activity was observed in the case of the spiking at 8 h.

No ZOI was observed when the experiments were performed on P. aeruginosa cultures.

Antibacterial properties of RGen₃K-functionalized wound dressings

Wound dressings functionalised with RGen₃K at 1.0 mg/ml also showed to biofilm prevention activity. This preventative antibiofilm activity was considered both concentration and time dependent.

When RGen₃K wound dressings were added to bacterial solutions at time 0 h of inoculation (Fig. 8A) a significant difference in biofilm biomass compared to the untreated biofilm growth control and the untreated wound dressings was observed at 4, 8, 24 and 48 h incubation. Over the 48 h period, although biofilm biomass increased, it was significantly inhibited compared to the untreated dressings; this inhibitory effect was 19%, 40% 21% and 43% in RGen₃K-functionalized NA, Kaltostat, KerraCel and AquaCel, respectively.

Fig. 8.

S.aureus biofilm inhibition activity by RGen₃K-functionalized wound dressings. A Wound dressing application at 0 h inoculation time, (B) Wound dressing application at 4 h post-inoculation, (C) Wound dressing application at 8 h post-inoculation, (D) Wound dressing application at 24 h post-inoculation. * indicates data significantly different at p ≤ 0.05

This biofilm growth pattern was also observed when RGen₃K functionalised wound dressings were added to bacterial samples at 4 h post inoculation (Fig. 8B). After 48 h of incubation 19%, 34% 47% and 75% inhibition of biofilm by treated NA, Kaltostat, KerraCel and AquaCel respectively were observed.

However, when bacterial biofilms were spiked with RGen₃K-functionalized wound dressings at 24 h and 48 h post inoculation no significant difference between the treated and untreated groups were observed (Figs. 8C and D).

Discussion

Dendrimers are hyperbranched polymeric nanomaterials that can be obtained from a range of monomers and synthesized by either liquid or solid phase methods [17]. When synthesized in liquid phase, dendrimers acquire a spherical structure with their molecular branching expanding in all directions. However, this type of synthesis requires rather cumbersome purification steps to separate the final products from unwanted solvents and byproducts [17]. On the contrary, the solid-phase synthesis provides tree-like structures, known as dendrons, rather than spherical ones offering the advantage of varying the molecular design according to the technological application under development and making the elimination of solvents and contaminants easier by the simple and rapid washing of the solid support resin on which the synthesis is carried out prior to product collection [17]. In particular, the synthesis of peptide-based dendrons has become much easier to perform, reproducible and scalable because of the availability of automated peptide synthesizers [18]. Among others, poly(epsilon-Lysine) dendrons have been synthesized for a range of biomedical applications, mainly as nanocarriers for therapeutics (e.g. drugs, siRNA) [17].

In this work, the fine-tuning of the dendron molecular design has been harnessed to control the size of the dendron as well as the positive charges that have generally been linked to the antimicrobial properties of a range of polycationic polymers [19].

It has been hypothesized that, unlike the linear and more flexible chains of polycationic polymers, the more rigid dendron structure would enable the exposure of positive charges at a relatively high density in proximity of the bacterial cell surface and present an overall molecular size able to penetrate within the bacterial cell. Also, when compared to other natural peptides with known antibacterial properties, the hyperbranched structure of dendrons is likely to prevent the digestion by proteolytic enzymes ubiquitously present in biological fluids and quickly digesting linear peptides [20].

In this study, the design of the poly(epsilon-Lysine) dendrons also included an Arginine residue at the carboxy terminal of the peptides to counteract the negative charge of the carboxylic group with an additional positive charge. The synthesis of these peptides was then performed to obtain dendrons of increasing branching levels or generations; 4, 8 and 16 molecular branches. The increase in branching generations led to an increase of the exposed positively-charged amino groups as well as to both their increasing special density and size of the macromolecules. The combination of characterization methods confirmed the reproducibility of the solid-phase synthesis, particularly in the case of the larger dendrons made of three generations of branching (RGen₃K) where the yield of reaction as well as the purity of the final reaction product were higher than for RGen₁K and RGen₂K.

The Bradford’s assay and the DLS measurements also confirmed the increasing of amino groups and the increasing size of the dendrons.

When tested for their antibacterial properties as soluble molecules on S. aureus and P. aeruginosa referenced strains, patterns of inhibition of S. aureus growth and biofilm formation were observed. Noticeably, in planktonic S. aureus cultures, all types of dendrons, tested at different concentrations, showed a very short time frame of the log phase where these macromolecules appear to slightly favor rather than inhibit growth. It can be speculated that in this phase, dendrons interact with the surface of more than one bacterial cells offering an anchoring substrate capable of promoting growth. This short phase is clearly followed by an inhibitory effect clearly dependent on the number and density of positive charges. Indeed, RGen₁K (1 branching generation, 4 amino groups per molecule) did not show any significant antibacterial effect, while RGen₂K (2 branching generations, 8 amino groups per molecule) and RGen₃K (3 branching generations,16 amino groups per molecule) showed increasing antibacterial properties. The data also demonstrated that the increased hydrodynamic size of the three molecules did not play a role in determining the dendron antibacterial properties suggesting that either the penetration of the molecules in the bacterial cells is not necessary to obtain an antibacterial effect or that they were all capable of penetrating into it. These properties could be defined as bactericidal for both RGen₂K and RGen₃K as their MBC/MIC ratio was less than 4 [21]. It can therefore be argued that, albeit with a varying efficacy, both these types of dendrons are able to disrupt the S. aureus membrane. Indeed, previous work has demonstrated the antibacterial effect of positively-charged poly(amino amide) dendrimers (PAMAM) on this bacterium species as compared to the lack of it in the case of polyanion dendrimers [22]. However, PAMAM are obtained by liquid phase synthesis that requires more laborious purification processes and their cytotoxicity towards tissue cells has also been shown [22]. On the contrary, the lack of significant cytotoxicity in peptide dendrimers has widely been ascertained [23, 24].

In the case of P. aeruginosa, none of the dendrons tested appeared to have an antibacterial effect. However, two distinct observations were made when the log and lag phase of the bacterial growth curves were examined. At dendron concentrations of 0.5 mg/ml and above, both RGen₂K and RGen₃K led to an inhibition of the log growth phase (Fig. 2, PA, blue arrows) followed by level of both log and lag phases higher than the control (Fig. 2, PA, red arrows). It can be reasonably argued that dendrons can initially penetrate the cell wall of this gram-negative species to cause an early antibacterial activity. Indeed, it has been demonstrated that the spiking of P. aeruginosa cultures with polyamines increases the susceptibility of this bacterial species to a range of antibacterial agents. On the contrary, polyamines have also been shown to induce resistance to cationic peptides, aminoglycosides, and quinolone antibiotics in the same bacterial pathogen [25] as well as to stimulate the gene expression of the extracellular virulence factors such as type III secretion system (T3SS) effectors [26]. Indeed, the biosynthesis of polyamines in bacteria involves a series of decarboxylases, which utilize the amino acids Arginine, Lysine (monomers of the RGen_xK dendrons) and ornithine as substrates [25, 26]. The spiking of P. aeruginosa cultures with spermidine, another polycationic peptide, also induces the expression of the efflux pump oprH-phoPQ operon [25, 26]. These findings might explain why, after an initial slowing of the log phase, the spiking with RGen₂K and RGen₃K seems to accelerate P. aeruginosa planktonic growth (Fig. 2, PA, red arrows).

The study of biofilm prevention provided consistent results, whereby S. aureus biofilm formation could be prevented only when the addition of RGen₃K was performed within 48 h post inoculation (Figs. 7A and B). No effect was observed on P. aeruginosa biofilms.

When compared to silver nanoparticles, an antibacterial agent widely used in a range of clinically-available wound dressings, it appears that the poly(epsilon-lysine) dendrons here investigated may not provide the same wide spectrum of bactericidal activities of silver [27]. However, silver-impregnated wound dressings have been shown to be effective only when applied to the wound in the early phases of wound treatment and the optimal conditions of silver controlled release are difficult to define because of the dual release of nanoparticles first and the gradual oxidation of the element that results in its antibacterial effect [27]. Also, in vitro studies have shown Ps aeruginosa resistance to silver [28]. In addition, adverse reaction to silver have been reported with evidences of cytotoxicity likely to be caused by the accumulation of silver in tissues [29]. In this respect, the flexible design of dendrons gives the opportunity to tune the physical interactions (e.g. electrostatic, hydrophobic, hydrogen bonds) of these nanobiomaterials with the surface of the dressings and, as a consequence, to tune their controlled release. Furthermore, previous papers have clearly demonstrated the absence of cytotoxicity of this type of dendrons and highlighted their tissue regeneration potential [30, 31].

The proof of synthesis scalability and level of product purity of the RGen₃K dendrons encouraged its use as an antibacterial additive for clinically available dressings both in their tulle and hydrogel formulations [32]. A simple staining procedure clearly demonstrated the efficient entrapment of these molecules in most of the dressings examined and clearly demonstrate the improved S. aureus biofilm prevention effect of these novel formulations and the eradication of forming biofilms in their early phases (up to 8 h) of insurgence.

Conclusions

Wound dressings, engineered as either tulle or hydrogels for the management of both acute and chronic wounds, have been developed to protect wounds from external contamination and accidental traumas as well as to keep the wound bed moist while removing excessive volumes of exudates [8, 10]. However, these devices, are not able to prevent and eradicate infections such as those caused by S. aureus and P.aeruginosa [27–29] unless antibacterial agents are added to their mesh [14, 33]. While the impregnation of dressings with silver nanoparticles has been widely used, their limited efficacy requires alternative solutions [15]. Poly(epsilon-Lysine) dendrons emerge as an alternative wound dressing additives to prevent for S. aureus wound infections. These antibacterial properties, combined with their demonstrated potential as drug nanocarriers [17, 22–24], may lead to a novel generation of antibacterial wound dressings with enhanced healing properties.

Abbreviations

RGen₁K

Poly(epsilon-lysine) dendron with an arginine at the carboxyl terminal and 1 generation of molecular branching (4 molecular arms)

RGen₂K

Poly(epsilon-lysine) dendron with an arginine at the carboxyl terminal and 2 generations of molecular branching (8 molecular arms)

RGen₃K

Poly(epsilon-lysine) dendron with an arginine route and 3 generations of molecular branching (16 molecular arms)

RGen_xK

Poly(epsilon-lysine) dendron with an arginine at the carboxyl terminal and x generations of molecular branching

P.A (Figures only)

Pseudomonas aeruginosa (P. aeruginosa)

S.A (Figures only

Staphylococcus aureus (S. aureus)

Authors’ contributions

Conceptualization: M.S.; Methodology: M.S., G.N.; Validation: M.S., S.S., G.N.; Formal Analysis: G.N., M.S., S.S.; Data Curation: G.N., M.S., S.S.; Writing Original Draft Preparation: M.S., Writing Review and Editing: M.S.; S.S., G.N.; Supervision: M.S., S.S.; Project Administration: M.S., Funding Acquisition: M.S.; All authors have read and agreed to the published version of the manuscript.

Funding

This work has been supported by the UKRI EPSRC grant n. EP/W023164/1, by the EC Horizon Europe HORIZON-CL4-2024-RESILIENCE-01-TWO-STAGE Project number: 101177924 and by Ms Georgina Nicolaou’s master’s in research.

Data availability

The datasets supporting the conclusions of this article are available in the University of Brighton repository, https://researchdata.brighton.ac.uk/id/eprint/333.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.