Antimicrobial activity of a short guanidine mimic immobilised on contact lenses

Abstract

The advancement of antimicrobial contact lenses presents a promising strategy for mitigating microbial keratitis. This study investigated the antimicrobial activity of four guanidine-substituted anthranilic amide peptidomimetics (GAMPs), identifying RK1083 as the most potent candidate. The minimum inhibitory concentrations ranged from 20 to 86 µM, with therapeutic indices between 2 and 22. All tested GAMPs exhibited resistance to proteolytic degradation. RK1083 was covalently immobilized onto contact lenses using carbodiimide chemistry, oxazoline plasma deposition, and plasma immersion ion implantation (PIII). The modified lenses demonstrated increased nitrogen content (≥3%), changes in surface charge, and improved hydrophilicity. Adhesion of Staphylococcus aureus was reduced by 5 log₁₀, while Pseudomonas aeruginosa adhesion decreased by ≥5 log₁₀ on oxazoline and PIII-treated lenses, and by ≥3 log₁₀ on carbodiimide-treated lenses. RK1083-coated surfaces exhibited no cytotoxicity toward corneal epithelial cells, and carbodiimide-treated lenses maintained antimicrobial activity post-sterilization. These results underscore RK1083’s potential for enhancing antimicrobial contact lens surfaces with improved bacterial resistance.

Introduction

Microbial keratitis, infection of the cornea, can occur during contact lens wear, and in severe cases can lead to scarring and vision loss [1]. Other microbially-driven adverse events during contact lens wear include Contact Lens-induced Acute Red Eye, Contact Lens Peripheral Ulcers and Infiltrative Keratitis, grouped together as non-infectious corneal infiltrative events (NiCIEs) [2]. MK and NiCIEs are believed to be initiated by microbial adhesion to contact lenses [3, 4]. P. aeruginosa is one of the most common causes of contact lens-related MK [5]. Several other bacteria including Staphylococcus aureus are associated with NiCIEs [3]. One way of reducing these adverse events during lens wear is to produce antimicrobial contact lenses.

Several antimicrobial contact lenses have been investigated, including those containing silver, selenium and antimicrobial peptides (AMPs) [6–8]. AMPs have potent activity, even against multidrug-resistant strains of bacteria, as well as fungi and Acanthamoeba that can also cause MK [9]. AMPs exert their antimicrobial effects through non-receptor interactions, making it challenging for microbes to develop resistance against them [10, 11]. A Phase III clinical trial showed that NiCIEs were reduced by over 65% when the AMP Mel4 was immobilized on contact lenses [12]. However, its activity progressively decreased during wear, with loss of activity after six nights of wear, most likely due to the action of tear proteases [13]. This suggests that an AMP coating with improved protease stability might result in greater reductions in NiCIEs.

One way of improving the stability of AMPs is to produce synthetic mimics (AMPMs) such as α-peptides, β-peptides, and peptoids [14]. Short AMPMs composed of amphiphilic backbones of 3’-amino-[1,1’-biphenyl]-3-carboxylic acid with arginine and tryptophan amino acids had good antibacterial activity, with MICs of 1-9 µM for S. aureus and E. coli [15]. Aryl-alkyl-lysine-based AMPMs had an MIC of 10 µg mL⁻¹ against S. aureus [16, 17]. AMPMs with a biphenyl backbone had MICs of 15.6 µM against S. aureus and 7.8 µM against E. coli [17, 18]. AMPMs with an anthranilamide backbone containing naphthoyl side chains and various hydrophilic cationic groups (amino, quaternary ammonium, and guanidino) had MICs ranging from 3-25 µM [19]. These guanidine substituted anthranilic amide peptidomimetic AMPMs (GAMPs) also had a good therapeutic index (i.e. antimicrobial concentration greater than toxic concentration), and also disrupted preformed S. aureus biofilms [17]. GAMPs can kill S. aureus, disrupt the membranes of E. coli, and have anti-biofilm activity [20]. A GAMP coated onto hydroxyapatite by non-covalent and covalent methods reduced the adhesion of E. coli by 3 log₁₀ colony forming units [19]. One of these GAMPs, RK758, has been found to have synergistic activity with fluoroquinolones against S. aureus and P. aeruginosa [21].

Given the excellent biological performance of GAMPs in solution and when attached to non-contact lens surfaces, this study aimed to evaluate the antimicrobial efficacy, toxicity, and structural functionality of GAMPs bound to contact lenses.

Materials and methods

Peptidomimetics synthesis

The synthesis of the GAMPs has been described in Australian Provisional Patent Application No. 2021902457 and was developed according to a procedure outlined previously [22]. Fig. 1 shows the structure of the GAMPs used in the current study.

Fig. 1.

Structure of the GAMPs tested in this study

Bacteria

In our study, we utilized ten strains each of Pseudomonas aeruginosa and Staphylococcus aureus. Several of these strains have been previously reported as resistant to traditional antibiotics and were primarily isolated from eye infections. The susceptibility profiles and molecular sequencing details of these strains have been analyzed and published in prior studies [10, 23–29].

Antimicrobial activity of the compounds

The minimum inhibitory and bactericidal concentrations (MIC and MBC) of the GAMPs RK758, RK1078, RK1079, and RK1083 were determined using a modified microbroth dilution method in accordance with Clinical Laboratory Standard Institute (CLSI) guidelines. Briefly, 5×10⁶ colony forming units (CFU) mL⁻¹ of each strain was added to serial dilutions (250 to 0.24 µg mL⁻¹) of GAMPs in Muller Hinton broth. After incubating for 24 h at 37°C with shaking, samples were plated on tryptic soya agar (TSA-Oxoid, Basingstoke, UK) to assess bacterial growth. The MIC was defined as the lowest concentration of GAMPs that achieved ≥90% reduction in CFU, while the MBC was defined as the lowest concentration achieving ≥99.99% reduction in CFU, compared to untreated controls. All experiments were performed independently in triplicate.

Hemolysis

Hemolysis of the GAMPs was conducted using a standard assay [30] with both horse (HRBCs; Sigma, Aldrich, St Louis, MO, USA) and human blood (ethics approval number HC2-10550). Hemolysis was assessed as the release of haemoglobin from RBCs by measuring the absorbance at 540 nm of the supernatants after centrifugation at 20 g for 10 mins, with MilliQ water serving as the positive control and phosphate buffered saline (PBS; NaCl 8 g l⁻¹, KCl 0.2 g l⁻¹, Na₂HPO₄ 1.15 g l⁻¹, KH₂PO₄ 0.2 g l⁻¹; pH 7.4) as the negative control. The percentage of hemolysis at each dilution was determined and the concentrations causing 10% and 50% lysis are reported. Three independent experiments were analyzed in triplicate.

Therapeutic index and the selective ratio

The safety of the compounds was assessed using the therapeutic index, calculated as the ratio of the peptoid concentration causing 50% hemolysis to the geometric mean of the MIC values, which represents the central tendency of bacterial susceptibility. The geometric mean is the nth root of the product of all values. The selectivity ratio was similarly determined using the concentration causing 10% hemolysis [31].

The effect of proteases on the MICs of the GAMPs

GAMP susceptibility to proteases was assessed following a previously described protocol [32]. GAMPs were incubated with the proteases trypsin or proteinase K for 24 h at 37°C, followed by determination of theirs MICs (see above) against P. aeruginosa strain 216. GAMPs incubated in PBS without proteases served as controls. Three independent experiments were repeated in triplicates. (n = 3)

Contact lenses

Etafilcon A contact lenses (Johnson & Johnson Vision Care Inc., Jacksonville, FL, USA) were removed from their packs and washed three times in 2 mL sterile PBS. The lenses were covalently immobilized with the GAMP 1083 (see below) or left untreated as controls.

Immobilisation onto antimicrobial surfaces for contact lenses

Three methods were used to bind RK1083 to etafilcon A contact lenses: carbodiimide chemistry, oxazoline plasma, and plasma immersion ion implantation, and these have been described in detail previously [33]. The carbodiimide chemistry forms covalent bounds between carboxylic acid groups on the lenses and amide groups of the GAMPs. Oxazoline was deposited by plasma treatment, depositing oxazoline rings, isocyanates, alkynes, and nitriles on the lens surface. These then can interact with amide groups of the GAMPs to produce intermediate imidazole derivatives [34–36]. This subsequently results in the formation of amide or ester bonds. Amide bonds are thermally stable, but ester bonds are susceptible to hydrolysis at high temperatures [37]. The third binding approach utilized plasma immersion ion implantation (PIII) with nitrogen gas that activates surfaces, producing radicals that subsequently enable covalent attachment of molecules via a relatively non-specific process [38]. For all attachment strategies, process control lenses were manufactured by following the techniques but omitting the addition of the GAMP RK1083. These were then tested, in addition to the 1083-coated lenses and uncoated lenses, in the experiments described below.

Following activation, lenses were washed with sterile PBS three times, then the GAMP RK1083 was introduced at a concentration of 1 mg mL⁻¹. All reactions were left to progress for 24 h at 4°C. Subsequently, lenses were immersed in 2 ml of 10% sodium chloride solution (Sigma Aldrich) for 2 h, followed by washing three times in sterile PBS.

X-ray photoelectron spectroscopy

The chemical elemental composition (atomic percentage) of the surface-coated species on the contact lenses was examined using X-ray photoelectron spectroscopy (XPS; ESCALAB220-iXL, VG Scientific, West Sussex, UK). The lenses were dried in air at room temperature. XPS utilized monochromatic Al Kα radiation with a photon energy of 1486.6 eV and a source power of 120 W, operating under a vacuum pressure of 10⁻⁸mbar or lower. By analysing the kinetic energy of electrons emitted from the surface, XPS identified the elemental composition and provided detailed information on the chemical and electronic states of atoms up to a depth of 10 nm. Variations in atomic concentration were interpreted as evidence of peptidomimetic coatings on the contact lenses. Two replicates were included (n = 2).

Surface charge

The charge of the lenses was assessed using a Zetasizer (SurPASSTM3, Anton Paar, Graz, Austria) with 10 mM KCl as the test solution. The instrument was calibrated across acidic and alkaline pH ranges. Antimicrobial lenses were washed in PBS to eliminate debris, ensuring the sample cell was free of air bubbles. The measurement involved applying an electric field to the sample and determining the electrophoretic mobility over 10–20 zeta cycles. Three independent experiments performed in triplicate. Three independent experiments performed in triplicate (n = 3).

Wettability

The surface wettability of peptidomimetic-bonded contact lenses was measured using a contact angle goniometer (serial number L2004A1, Ossila BV, Biopartner, 2333 BD Leiden, Netherlands) with the sessile drop technique [39]. Air-dried lenses were placed on a slide on the goniometer table, and deionized water was applied. The contact angle was measured by increasing the water volume without changing the solid-liquid interface area. High-speed imaging and video recording tracked changes in the contact angle, analyzed with software using manual or automated methods. (Contact angle goniometer Serial number L2004A1; Ossila BV, biopartner, 2333 BD Leiden, Netherlands). All surface measurements repeated three times (n = 3).

Antibacterial activity of lenses

Established protocols were used, with minor adjustments [12, 40]. S. aureus and P. aeruginosa strains were grown in tryptone soya broth (Oxoid, Basingstoke, UK) overnight at 37°C. After washing cells in PBS, they were resuspended in PBS to an optical density at 660 nm and further diluted to obtain 10⁶ CFU mL⁻¹. An aliquot (1 mL) was dispensed into wells of 24 well polystyrene plates (Greiner Bio-One; Monroe 28110, North Carolina USA). Lenses were then placed into the wells and incubated at 37°C for 18 h with shaking (120 revolutions per minute). After incubation, lenses were washed with PBS and transferred to container containing 2 mL PBS and stirred for one minute. The resulting slurry was diluted (1/10) and aliquots plated on TSA(TSA; Oxoid, Basingstoke, UK). After incubation at 37°C for 18 h, the number of CFU were calculated. Three independent experiments performed in triplicate. Three independent experiments performed in triplicate (n = 3).

Sterilization effect on RK1083-bound antimicrobial activity

Since contact lenses are typically sterilized by autoclaving during production, the effect of this process on the antimicrobial efficacy of the lenses was examined. After being coated, the lenses were sterilized in PBS at 121°C for 15 min. Following sterilization, the capacity of the lenses to reduce the adhesion of P. aeruginosa 6294 was evaluated as previously described [41]. Two independent experiments were conducted (n = 2).

Cytotoxicity

The cytotoxicity of peptidomimetic-coated lenses was assessed using a direct contact method as per ISO 10993-5:2009. Immortalized human corneal epithelial cells (HCE-T) were cultured in DMEM/F-12 supplemented with 10% FBS, 10 ng/mL EGF, and 1% ITS (Life Technologies, Sydney, Australia). HCE-T were seeded at 1×10⁴ cells/well in 24-well plates and incubated for 24-48 h to achieve approximately 80% confluence [42]. Lenses were incubated with the cells for 24 hours, followed by MTT assay following previous method. Formazan crystals were solubilized in DMSO, and absorbance was measured at OD_570nm, comparing results to cells grown without lenses. For validation three experiments were run and the results averaged (n = 3).

Data analysis

Statistical analyses were conducted using GraphPad Prism version 10.2.0 (392) software (GraphPad Software, La Jolla, CA, USA). Comparisons were made using one-way ANOVA and two-way ANOVA (analyses of variance) with a Tukey’s test of multiple comparisons. The statistical significance threshold was set at p < 0.05.

Results

MIC and MBC of the GAMPs

The MICs and MBCs are given in Table 1. Overall, the MBCs were approximately double the MICs for strains of P. aeruginosa but usually less than double for strains of S. aureus. GAMP RK1083 gave the lowest average MIC (27.6 μM) across the P. aeruginosa strains, but its average MIC (11.3 μM) was slightly higher that of other GAMPs (MIC ≤ 11.0 μM) across the S. aureus strains.

Table 1.

The minimum inhibitory and bactericidal concentrations of the GAMPs against strains of P. aeruginosa and S. aureus

Bacterial type	Strain number	Isolation site	Resistance to traditional antibiotics	Minimum inhibitory and bactericidal concentrations (MIC and MBC) of the peptidomimetics (µM)
				758		1078		1079		1083
				MIC	MBC	MIC	MBC	MIC	MBC	MIC	MBC
P. aeruginosa	PA01	Wound	CHL, TET (R); CEFT, CIP, TOB (S)36	20	41	39	82	37	74	21	43
	31	Cornea	CIP, LEV, GEN, TOB, IMI, CEF (R); PIP, PMB (S)	41	ND	19	31	18	37	43	86
	37	Cornea	CIP, LEV, GEN, TOB, IMI, CEF (R); PIP, PMB (S)	20	ND	19	39	18	37	21	43
	123	Cornea	IMI, PMB (R); CIP, LEV, GEN, TOB, PIP, CEF (S)	41	ND	19	39	18	18	43	43
	216	Cornea	CIP, PIP, IMI, CEF, PMB (R), LEV, GEN, TOB (S)	20	41	39	82	37	74	21	21
	219	Cornea	CIP, LEV, GEN, TOB, PIP, IMI, CEF (R); PMB (S)	20	41	39	82	74	ND	21	43
	224	Cornea	CIP, IMI, (R); LEV, GEN, TOB, PIP, CEF, PMB (S)	41	ND	39	39	18	37	21	43
	233	Cornea	CIP, IMI, CEF, (R); LEV, GEN, TOB, PIP, PMB (S)	41	41	39	39	74	ND	21	43
	235	Cornea	CIP, IMI, PIP, CEF, (R); LEV, GEN, TOB, PMB (S)	81	41	39	39	37	74	43	43
	6294	Cornea	ND	20	ND	39	39	18	37	21	43
Mean MIC				34.5	41	33	51.1	34.9	48.5	27.6	45.1
S. aureus	34	Cornea	CEF, AZI, PMB (R); CIP, OXA, GEN, VAN, CHL (S)37	10	10	9.8	9.8	9	18	10	10
	38	Cornea	ND	5	10	4.9	9.8	4.6	9.3	5.4	10
	41	Contact lens	CIP, CEF, AZI (R); OXA, GEN, VAN, CHL, PMB (S)37	10	20	4.9	9.8	4.6	9.3	10	10
	65	Cornea	ND	5.2	10	9.9	9.9	4.6	9.3	10	10
	110	Cornea	CIP, CEF, OXA, CHL, AZI, PMB (R); GEN, VAN, CHL (S)37	10	20	9.8	9.8	4.6	9.3	10	10
	111	Cornea	CIP, CEF, OXA, AZI, PMB (R); GEN, VAN (S)37	20	20	4.9	9.8	9.3	9.3	10	10
	113	Cornea	CIP, CEF, OXA, AZI (R); GEN, VAN, PMB (S)37	10	10	4.9	9.9	4.6	9.3	21	21
	114	Cornea	CIP, CEF, AZI (R); OXA, GEN, VAN, PMB (S)37	10	10	9.9	19.7	9.3	9.3	10	10
	117	Contact lens	CIP, CEF, AZI (R); OXA, GEN, VAN, PMB (S)37	10	10	9.9	9.9	4.6	4.6	5.4	5.4
	ATCC 6538	“lesion”	ND	20	20	9.8	19	9.3	9.3	21	21
Mean MIC				1	14	7.9	11.7	6.5	9.87	11.3	11.7

MRSA methicillin resistant S. aureus, MSSA methicillin susceptible S. aureus, AMP amphotericin, AZI azithromycin, CEF cefepime, CEFTA ceftazidime, CIP ciprofloxacin, CHL chloramphenicol, GEN gentamicin, IMI imipenem, LEVO levofloxacin, MERO meropenem, MOX moxifloxacin, OXA oxacillin, PIP piperacillin, TOB tobramycin, VAN vancomycin

Hemolytic profiles of the GAMPs

The concentration that produced 50% hemolysis of horse or human red blood cells (RBCs) (HC₅₀) are given in Table 2. GAMP RK1078 had the lowest HC₅₀ of 37 µM with horse RBCs. RK758 and RK1083 were tested with human RBCs, and these had HC₅₀s of 164 µM and 344 µM, respectively.

Table 2.

The hemolytic concentration of the GAMPs and their therapeutic indices (TI)

Peptoid	Horse HC50 (µM)	Human HC50 (µM)	Geometric mean of MIC (μM)	TI horse blood (ratio HC50 to geometric mean MIC)	TI human blood (ratio HC50 to geometric mean MIC)
RK758	41 ± 0.5	164 ± 0.3	20	2	8
RK1078	37 ± 0.9	ND	13	2	ND
RK1079	39 ± 1.0	ND	18	2	ND
RK1083	86 ± 0.7	344 ± 0.5	15	6	22

ND-not determined

The safety of a drug can be assessed using its therapeutic index (TI) [31]. This compares the concentration that produces toxicity (HC₅₀) to the concentration that inhibits microbial growth (MIC). The TIs of the GAMPs are given in Table 2. With horse RBC, RK1083 gave the highest TI of 6, while all other compounds had TIs of 2. For RK758 and RK1083, their TIs were higher with human RBCs, at 8 and 22, respectively.

The effect of proteases on the MICS of the GAMPs RK758 and RK1083

Compounds RK758 and RK1083 maintained their antimicrobial efficacy against P. aeruginosa strain 216 even after incubation with trypsin or proteinase K, as their MICs did not change after incubation with either protease.

The above data indicate that the GAMPs were active against all strains of P. aeruginosa and S. aureus. As RK1083 had a relatively low MIC against P. aeruginosa and good TIs, this was chosen to be bound to contact lenses.

XPS analysis

Table 3 shows that binding of RK1083 resulted in 70.9, 22.9, and 3.5 to %C, %O and %N after EDC/NHS coupling and 54.9, 24.6, and 7.9 after oxazoline, and 59.5, 29.2, 3.8 after PIII coating. EDC/NHS coating of RK1083 resulted in increased carbon-carbon single bond (C-C; C-H) species from 31.5% in process control to 60.2%, but a decrease in N-C = O from 6.1% to 0.06% (Table 3). Similar changes in the carbon-carbon single bond species were seen after coupling with oxazoline and PIII. However, the N-C = O species increased after coupling of RK1083 with oxazoline or PIII compared to their respective process control surfaces (Table 3).

Table 3.

Surface elemental analysis (XPS) of GAMP RK1083 covalently attached to contact lenses; mean ± SD (n = 2)

Major elements (Atom%)	Untreated lens	Process control	RK1083
Carbodiimide surfaces
Carbon	70.0 ± 4.8	66.0 ± 0.5	70.9 ± 3.7
Oxygen	25 ± 5.4	30.1 ± 0.6	22.9 ± 3.0
Nitrogen	0.9 ± 0.8	1.2 ± 0.9	3.5 ± 0.07
C/O ratio	2.8	2.2	1.6
N/C ratio	0.01	0.01	0.06
Oxazoline-coated surfaces
Carbon	71.2 ± 0.01	63.0 ± 4.1	54.9 ± 1.9
Oxygen	23.0 ± 0.4	23.0 ± 2.3	26.6 ± 6.8
Nitrogen	1.5 ± 0.1	7.9 ± 0.1	3.9 ± 1.0
C/O ratio	3.1	2.7	2.1
N/C ratio	0.02	0.13	0.07
PIII-treated surfaces
Carbon	71.2 ± 0.01	60.2 ± 0.02	59.5 ± 2.7
Oxygen	23.0 ± 0.4	24.1 ± 0.4	29.2 ± 0.7
Nitrogen	1.5 ± 0.1	9.35 ± 0.1	3.8 ± 0.3
C/O ratio	3.1	2.5	2.0
N/C ratio	0.02	0.16	0.06

Contact lens surface wettability

The contact angles of the contact lens surfaces after coating with the three different techniques are presented in Table 4. The process control lenses showed changes in contact angles compared to the untreated lenses, which further changed after coupling of RK1083. After the addition of RK1083 via carbodiimide, the surface was more hydrophobic, as it was after coupling via PIII, but coupling via oxazoline resulted in a similar low contact angle compared to the untreated control lenses (Table 4).

Table 4.

Contact angles of the etafilcon A lenses before and after coupling of RK1083 (mean ± SD, n = 3)

Contact lens	Coupling strategy
	Carbodiimide	Oxazoline	PIII
	Contact angle (o); Mean ± SD
Untreated lens	29 ± 0.2	29 ± 0.2	29 ± 0.2
Process control	33 ± 0.0	74 ± 1.4	55 ± 0.5
RK1083	83 ± 1.0	28 ± 3.8	71 ± 1.0

Surface charge of contact lenses

Figure 2 shows the changes to the zeta potential (charge) of the lenses after coupling RK1083 via the three different strategies. Activation of the lenses with carbodiimide did not affect the surface charge, whereas activation by either oxazoline or PIII produced lower surface charges. Specifically, upon coupling with 1083, EDC/NHS surfaces had a surface charge of -3.5 mV, whereas oxazoline and PIII lenses had lower surface charges of -19 mV and 5 mV, respectively, compared to their process controls (Fig. 2).

Fig. 2.

Surface charge (zeta potential) of lenses after activation and coating with RK1083, Values shown are mean and standard deviation (n = 3). *p < 0.05; ***p < 0.001

Ability of RK1083 bound to contact lenses to reduce adhesion ofS. aureus and P. aeruginosastrains

When RK1083 was bound via any method, the adhesion of all strains of S. aureus was lowered significantly (p < 0.01), for most strains by ≥ 5.5 ± 0.4 log₁₀ CFU/lens. The exception was for strain ATCC 6538, which had a reduction of 3.1 ± 1.2 log₁₀ CFU/lens with PIII coating. For P. aeruginosa strains, activation of the surfaces by oxazoline and PIII and coupling of RK1083 resulted in ≥ 5.4 ± 1.0 log₁₀ CFU/lens for all three strains. Activation of the lens by carbodiimide coupling of RK1083 resulted in significantly less reduction compared to the other activation strategies (p < 0.01) but still produced significant (p < 0.01) reductions in adhesion of all three strains compared to process control lenses (Figs. 3,4).

Fig. 3.

The effect of RK1083 when bound via three different attachment strategies on the adhesion of strains of S. aureus (A) and P. aeruginosa (B) strains, Reductions of RK1083 coated lenses were all significantly different (p < 0.001) compared to control lenses. Values shown are mean and standard deviations (n = 3)

Fig. 4.

The autoclave effect on the antimicrobial activity of surface-bound RK1083, ∗Significant difference (p < 0.02). Values shown are mean and standard deviations (n = 2)

The effect of sterilization on the antimicrobial activity of bound RK1083

Control lenses showed no significant differences among the untreated controls, irrespective of autoclaving (p > 0.1). All RK1083-coated lenses exhibited significantly lower bacterial adhesion than their corresponding controls (p < 0.0001), with the largest reductions in the unautoclaved oxazoline- and PIII-bound groups. Autoclaving did not affect the activity of carbodiimide-bound RK1083 (p > 0.99) but significantly reduced the activity of the oxazoline- and PIII-bound coatings (p ≈ 0.02), which nonetheless remained more effective than untreated lenses.

Cytotoxicity

The cytotoxicity was measured for carbodiimide and oxazoline-bound RK1083 lens surfaces. There was no effect of the amount of MTT produced by corneal epithelial cells after addition of the lenses coated with RK1083 via EDC/NHS or oxazoline, indicating no cytotoxic effect of these lenses (Fig. 5).

Fig. 5.

Cytotoxicity of RK1083-bound surfaces on corneal epithelial cells

Discussion

This study has demonstrated that GAMPs are active against antibiotic resistant strains of S. aureus and P. aeruginosa isolated for ocular conditions. The GAMPs needed higher concentrations to cause the hemolysis of 50% of RBCs from horses or humans than their MICs, giving therapeutic indices of ≥ 2 with horse RBCs and ≥ 8 when tested with human RBCs. Furthermore, the GAMPs were resistant to proteolytic digestion, maintaining their antimicrobial activity even in the presence of these enzymes. This study also demonstrated that the GAMP RK1083 could be bound to the surface of contact lenses by three different strategies. This resulted in the production of lenses that could reduce the adhesion of strains of S. aureus and P. aeruginosa. These lenses have the potential to progress further in preclinical animal models for safety and efficacy during lens wear, and then possibly onto clinical trials in humans, as was done previously for the melimine and Mel4 AMP-coated lenses [12, 43–47].

The same attachment strategies as used in the current study were used previously to bind the AMP melimine, Mel4, and the AMPM peptoids [40]. Very similar results were obtained when comparing the antibacterial activity of RK1083, melimine and peptoids. All three attachment strategies produced contact lenses that could reduce the adhesion of S. aureus and P. aeruginosa to, usually, undetectable levels when RK1083 or the peptoids TM5 or TM18 were covalently attached. This may indicate that RK1083 and the peptoids have similar antimicrobial activities when bound to surfaces. GAMP RK1083 and the peptoids have similar MICs of 18-37 µM against various bacteria with [32].

Binding of RK1083 to the etafilcon A lenses through carbodiimide (EDC/NHS) chemistry resulted in increased nitrogen levels, consistent with findings from melimine-bound surfaces [40], and for TM5 and TM18. PIII-treated surfaces showed a significantly different atomic percentage when compared to carbodiimide-treated surfaces. This difference is likely due to the presence of abundant free radicals on PIII-treated surfaces, which reacted with the peptidomimetics and contribute to the increased nitrogen atomic percentage. This observation is consistent with previous studies where melimine bound to PIII-treated PVC surfaces demonstrated a similar increase in atomic percentage. Moreover, Mel4 bound to five different FDA-classified non-ionic contact lens materials also showed a significant increase in nitrogen levels compared to control lenses, with the exception of lotrafilcon materials [48]. These findings suggest that the type of surface treatment and the presence of reactive species significantly influence the atomic composition and the binding efficiency of peptidomimetics on different surfaces [49].

The effect on lens wettability after coating of RK1083 via EDC/NHS or PIII was different compared to melimine [40] and the peptoids whereby the binding of RK1083 resulted in more hydrophobic surfaces. This is most likely because RK1083 contains aromatic rings and bromine that confer hydrophobicity, whereas the peptoids only contain aromatic rings [32]. On the other hand, binding of RK1083 via oxazoline produced a hydrophilic surface. This may indicate different orientations of RK1083 on the surfaces. Future experiments using, for example, time-of-flight secondary ion mass spectrometry (ToF-SIMS) with principal component analysis [50], may be able to shed light on any structural differences. The effect of covalently bound RK1083 on the overall surface charge was similar to that for the peptoids [41], which indicates that the surface-bound cationic peptidomimetics in general exert their antimicrobial activity through interaction with the outer surfaces of bacteria regardless of the overall surface charge and hydrophobicity [51, 52].

This study provides significant insights into the development of antimicrobial contact lenses using peptidomimetic compounds, in this case the GAMP RK1083. By demonstrating the successful covalent attachment of peptidomimetics to contact lens surfaces through the three strategies, this study advances the understanding of how to maintain the physical integrity of lenses while enhancing their antibacterial properties. The findings underscore the versatility and effectiveness of carbodiimide-treated surfaces, which exhibited stable antibacterial activity relatively resistant to autoclaving and producing surfaces non-toxic to corneal epithelial cells.

Further research is needed to fully exploit the antibacterial potential of bound RK1083 and develop effective antimicrobial lenses that can be used in humans. The study highlights the importance of anthranilic acid-based peptides as promising candidates for future antibacterial drug development efforts.

Despite the promising results, this study has several limitations. The focus was on a limited number of bacterial strains, which may not fully represent the broad spectrum of potential pathogens [3, 53–55]. Therefore, broader testing on diverse bacterial strains is necessary to generalize the findings. Additionally, the long-term stability and effectiveness of the RK1083 antimicrobial coatings were not assessed. While the in vitro results are encouraging, further in vivo studies are required to confirm the efficacy and safety of the antimicrobial contact lenses. Moreover, the exact mechanisms underlying the observed antimicrobial activity were not elucidated, necessitating further investigation. Although RK1083 partly retained its activity post-sterilization, the impact of repeated sterilization cycles on activity effectiveness and lens properties was not studied, which could be useful for their use in clinical settings. Furthermore, a more in-depth analysis of why the coatings lose activity after autoclaving may help design more robust coatings in the future.

Conclusions

This study demonstrated that the guanidine mimetic RK1083 can be effectively attached to contact lens surfaces using three different methods, leading to lenses that significantly reduce the adhesion of S. aureus and P. aeruginosa. These lenses were found to be non-toxic to mammalian HCE-T cells, indicating their suitability for further testing in preclinical models and potentially in clinical trials. The research showed that RK1083, melimine, and peptoids exhibit similar antibacterial activities when bound to surfaces, and they share comparable mechanisms of action in solution, primarily involving bacterial membrane disruption. Future studies should aim to further explore these mechanisms and assess the long-term stability and effectiveness of the antimicrobial coatings.

Funding

Open Access funding enabled and organized by CAUL and its Member Institutions.

Compliance with ethical standards

Conflict of interest

The authors declare no competing interests.