Enhancing the Antimicrobial Potency of Self‐Assembling Antibiotics by Co‐Assembly‐Based Aggregation Modulation

ABSTRACT

The activity of many antimicrobial peptides (AMPs), a promising alternative to classical antibiotics, depends strongly on their aggregation‐propensity and weak peptide association favor antimicrobial function, whereas stable assembly can reduce activity. Thus, modulation of intermolecular interactions offers a feasible way to enhance the potency of existing AMPs. However, a general strategy to achieve this objective has not been established. Here, we demonstrate that the spectrum of activity of Gram‐positive‐specific self‐assembling AMP Fmoc‐phenylalanine can be broadened by co‐assembly with non‐antibiotic Fmoc‐glutamic acid. Biophysical assays confirmed that the co‐assembled system disrupted bacterial membrane integrity, leading to cell death. The enhanced potency correlated with the reduced mechanical rigidity of the co‐assembled hydrogel, as determined by rheological measurements, and molecular dynamics simulations further revealed that heterogeneous non‐covalent interactions were detrimental to the fibril stability. These findings suggest that rationally designed co‐assembling partners that weaken stabilizing non‐covalent interactions can serve as a common strategy to enhance the antibacterial efficacy of existing AMPs.

Co‐assembly of self‐assembling antimicrobial peptides (AMPs) with non‐AMP molecules enhances antimicrobial efficacy by weakening stabilizing non‐covalent interactions within native AMP assemblies. The entry describes how heterotypic co‐assembly modulates peptide packing, increases structural dynamics, improves membrane interaction, and ultimately amplifies antimicrobial activity while offering a tunable strategy for designing more potent peptide‐based therapeutics.

1. Introduction

Antibiotic resistance is evolving as a significant global health concern. A systematic analysis of the global burden of antimicrobial resistance predicted more than 10 million worldwide fatalities associated with antimicrobial resistance by 2050, of which nearly two million can be directly attributed to resistant pathogens [1]. The paucity of new antibiotic discoveries and approvals in recent times has further compounded the problem [2]. An effective strategy to circumvent the stagnant antibiotic discovery pipeline is to potentiate the activity of existing antibiotics to overcome resistance [3, 4]. The identification of clavulanic acid, a β‐lactamase inhibitor, and its excellent effectiveness in potentiating the activity of penicillin‐based antibiotics against resistant strains represents a successful demonstration of the above approach [5]. Several antibiotic adjuvants have subsequently been identified, showing excellent synergistic action with clinically approved antibiotics and, in many cases, broadening the spectrum of Gram‐positive‐specific antibiotics, making them cytotoxic against Gram‐negative pathogens [6, 7, 8, 9, 10, 11, 12]. Antimicrobial peptides (AMPs), another diverse class of antibiotics, have emerged as attractive alternatives to conventional antibiotics [13]. This is largely attributed to their unique mechanisms of action, associated with a lower propensity for resistance development [14, 15]. AMPs target bacterial membranes in a non‐specific manner to disrupt membrane integrity, in contrast to conventional antibiotics, which act by binding to specific proteins essential for bacterial survival. However, strategies to augment the activity of existing AMPs have been much less explored.

Recent reports have indicated the critical role of self‐assembly in the functioning of natural and synthetic AMPs and the dependence of antimicrobial activity on the nature of self‐assembly architectures [16, 17, 18, 19, 20, 21]. Chemical modifications of AMPs to induce self‐assembly, or conjugation of non‐assembling AMPs to self‐aggregating scaffolds, have been demonstrated to improve antimicrobial efficacy, physiological stability, and biocompatibility [22, 23, 24, 25, 26]. The dependence of antimicrobial activity on self‐assembly was clearly demonstrated by a study involving the dipeptide diphenylalanine, which exhibits strong antimicrobial activity due to its ability to form fibrillar assemblies, whereas glycylglycine, which lacks self‐assembly propensity, is inactive [27].

Although self‐assembly is critical for the antimicrobial activities of AMPs, the aggregation propensity and stability of assembled AMP nanostructures have been indicated to show a reciprocal relationship. In a notable study, Conticello and co‐workers demonstrated that end‐capped AMPs, which form stable filamentous aggregates, exhibit minimal antibacterial activity, while uncapped analogs forming less stable fibrils retain strong potency [28]. The enhanced activity correlates with the fact that the membrane disruption ability of AMPs requires reorganization of the peptide aggregates observed in solution to favorably interact with the lipid environment, and weaker peptide–peptide interactions could facilitate such processes. This is further validated by the chameleon‐like peptide system, which is capable of assembling into either cross‐α or cross‐β architectures, but only the cross‐α state is cytotoxic [29, 30, 31]. The higher thermodynamic stability of cross‐β fibrils, a common structural motif of amyloid plaques, compared to cross‐α fibrils corroborates this inverse relationship [4, 32]. Although these studies primarily focused on exploring the relationship between the physical properties of self‐assembled AMPs and antimicrobial potency, they nevertheless pointed to the fact that modulation of the activity of existing AMPs is feasible, analogous to the role played by adjuvants with respect to classical antibiotics, by interfering with their aggregation propensity. However, a general mechanistic approach allowing the reorganization of the self‐assembly features of existing AMPs and affording enhancement of the antimicrobial activity is yet to be established.

In this study, we demonstrate that the antibacterial spectrum of the hydrogel‐forming AMP Fmoc‐phenylalanine (Fmoc‐F), which is naturally active only against Gram‐positive bacteria, can be broadened through a co‐assembly strategy without any chemical modification of the original sequence. Co‐assembly of Fmoc‐F with Fmoc‐glutamic acid (Fmoc‐E), which by itself is non‐aggregating and biologically inactive, resulted in significant antibacterial activity against Gram‐negative Escherichia coli. Microscopy, spectroscopy, and bacterial growth assays revealed that the co‐assembled hydrogel disrupts bacterial membrane integrity, leading to reduced cell viability. Rheological analysis revealed that the enhanced potency is correlated with a reduction in the mechanical rigidity of the composite hydrogel. Furthermore, molecular dynamics simulations indicated that intermolecular hydrogen bonding and hydrophobic interactions between Fmoc‐F and Fmoc‐E destabilize the assembly, leading to a more dynamic and biologically active structure.

2. Experimental

2.1. Materials and Methods

Fmoc‐L‐phenylalanine and Fmoc‐L‐glutamic acid were acquired from SRL, India. For antibacterial assays, S. aureus (MTCC strain accession no. 96) and E. coli (MTCC strain accession no. 1302) have been employed. SRL, India supplied citric acid (SRL, India), Luria‐Bertani broth (LB) procured from Himedia, India. Polymyxin B, N‐Phenyl‐1‐Naphthylamine, pyrene were sourced from Sigma–Aldrich.

2.2. Preparation of Co‐Assembled Hydrogel

Fmoc–F (fluorenylmethyloxycarbonyl‐phenylalanine) was prepared at a concentration of 30 mm in phosphate buffer (100 mm, pH 7.4). Heating at 90°C for 10 min dissolved Fmoc‐F, forming a clear solution. 20 mm of Fmoc‐E (fluorenylmethyloxycarbonyl glutamic acid) was dissolved in the phosphate buffer and the pH was adjusted with 1(M) NaOH to achieve a final pH of 7.4. After forming these two clear solutions of Fmoc‐F and Fmoc‐E, they were mixed in fixed molar ratios to produce working samples. Samples formed stable hydrogels following heating at 90°C and gradual cooling at room temperature.

2.3. Structural Assessment of the Hydrogel

To investigate the nanoscale morphology of the fibrillar networks in our co‐assembled hydrogel, we used transmission electron microscopy (TEM). Approximately 10–15 µL of hydrogel was diluted tenfold with Milli‐Q water to reduce sample viscosity and improve dispersion, while preserving the integrity of the fibrillar structures. For TEM grid preparation, 5 µL of the diluted sample was deposited onto a carbon‐coated copper grid (300 mesh) and left undisturbed for 1 min to allow the fibrils to adsorb. Filter paper were used to blot the excess liquid. The grids were then negatively stained for 60 s with a 2% (w/v) aqueous solution of uranyl acetate, followed by blotting and air drying at room temperature for at least 15 min. TEM imaging was performed using a JEOL JEM‐2100 microscope operated at 120 kV. At least three independently prepared hydrogel samples were analyzed to confirm the reproducibility of the observed morphological features.

The critical aggregation concentration (CAC) of the self‐assembled hydrogel was determined using a pyrene fluorescence assay with minor modifications from a well‐established method [33]. Briefly, a concentrated pyrene stock solution (800 µm) was prepared in DMSO and subsequently diluted with 100 mm phosphate buffer (PB) to obtain a working concentration of 16 µm. A fixed volume of this pyrene solution was added to peptide samples with concentrations ranging from 1 to 30 mm, maintaining a constant total volume of 1 mL. The mixtures were gently vortexed and equilibrated at room temperature for 1 hour to allow pyrene molecules to partition into the hydrophobic domains generated during peptide self‐assembly. Fluorescence emission spectra were collected using a spectrofluorometer with an excitation wavelength of 334 nm, emission range from 360 to 420 nm, and a slit width of 5 nm. The ratio of the first (P₁, 373 nm) to the third (P₃, 395 nm) vibronic peaks (P₁/P₃) was plotted as a function of peptide concentration. The CAC value was determined from the inflection point of the sigmoidal decrease in the P₁/P₃ ratio, obtained by nonlinear least‐squares fitting using a four‐parameter Boltzmann function, signifying the onset of peptide self‐assembly and formation of hydrophobic domains.

The rheological behavior of the gel was determined using an MCR 702e Rheometer (Anton Paar, Austria) with a stainless‐steel parallel plate geometry (diameter of the upper plate = 25.00 mm and the lower plate = 50.00 mm) and a measuring gap size of 1.00 mm for all tests. The temperature was adjusted to 25 ± 0.10°C for each measurement, and the gels were carefully placed on the lower plate that was fastened to the linear drive motor. The entire system was closed to minimize the solvent evaporation. For the frequency sweep, the strain was adjusted to 1%, and for the amplitude sweep, the frequency was adjusted to 1 Hz.

Circular dichroism (CD) spectroscopy was used to examine the supramolecular organization and secondary structural features of Fmoc‐amino acids and their co‐assembly. Stock solutions (10 mm) of Fmoc‐F, Fmoc‐E, and the H1 were prepared in 0.1 M PB, then diluted with deionized water to a final concentration of 200 µm to reduce baseline distortions caused by scattering effects. The samples were equilibrated before CD spectra were recorded using a Jasco J‐1500 spectropolarimeter with a 1‐mm quartz cuvette at 25°C. Spectra were collected across 190–260 nm wavelength range at a scan speed of 50 nm/min, a bandwidth of 1 nm, and a data pitch of 0.5 nm. Each spectrum averaged three scans to enhance the signal‐to‐noise ratio. Baseline correction was performed using the solvent spectrum, and ellipticity was reported in millidegrees (mdeg).

2.4. Antibacterial Activity

To assess the antibacterial efficacy of the co‐assembled hydrogels, they were exposed to bacterial strains (E. coli and S. aureus), representing model Gram‐negative and Gram‐positive bacteria, respectively. Briefly, overnight cultures in LB of both bacterial strains were used to obtain secondary cultures of 0.5 McFarland turbidity standard. They were then diluted 1000‐fold and seeded onto the already formed hydrogels corresponding to a final count of 10⁵ CFUmL⁻¹. At different time intervals, turbidity measurements of bacterial growth co‐incubated with hydrogel at 37°C were performed at a wavelength of 600 nm using a BioTek Epoch microplate spectrophotometer until the negative control (bacteria without any treatment) reached the stationary phase. Kanamycin mono‐sulfate was used as a positive control throughout the experiments. In addition, the inhibitory effects of hydrogels on bacteria were assessed using a higher bacterial load (OD_600nm = 0.5, 10⁸ CFUmL⁻¹) by counting the bacterial population before and after treatment. Aliquots were collected at different time points, followed by serial dilution on agar plates. Finally, the percentage decrease in viable cells compared to the negative control (bacteria without any treatment) was calculated, representing the inhibitory effect of the prepared hydrogel and validating the antibacterial efficacy obtained from the time‐dependent kinetic data.

2.5. Outer Membrane Perturbation Assay

To assess whether hydrogel disrupts the E. coli outer membrane, a 1‐N‐phenylnaphthylamine (NPN) intake assay was performed. At a mid‐logarithmic phase (OD_600nm ≈ 0.5), E. coli cultures were harvested by centrifugation at 3500 × g for 6 min at 4°C, washed once with sterile 1X phosphate‐buffered saline (PBS), and diluted in 1X PBS to an OD_600nm of 0.05. For the assay, 100 µL samples were prepared in black 96‐well fluorescence microplates by combining equal volumes (50 µL each) of bacterial suspension and the respective test solutions, resulting in a final optical density (OD) of the bacterial suspension of 0.025. The experimental conditions included Fmoc‐F (10 mm), Fmoc‐E (10 mm), co‐assembled, citric acid (positive control), PBS (negative control), and PB (solvent control). A final concentration of 20 µm NPN (prepared in 0.25% DMSO in PBS) was present in all test combinations. The bacteria‐containing samples were incubated in the dark for 2 h at 37°C, allowing them to interact with the outer membranes of E. coli. The fluorescence emission spectra were recorded using a Cary Eclipse fluorescence spectrophotometer with excitation at 356 nm and emission scanned from 375 to 600 nm. An increase in fluorescence intensity was interpreted as an indication of enhanced membrane permeability, reflecting increased NPN incorporation into the lipid bilayers. Statistical analyses were performed using one‐way ANOVA, and p values were determined using Dunnett's multiple comparisons test.

2.6. Assessment of Bacterial Morphology by Scanning Electron Microscopy

To assess the morphological effects of the hydrogels on both S. aureus and E. coli, bacterial cultures were exposed to hydrogel sample at 37°C for four hours. Following treatment, cells were collected by centrifugation at 3000 rpm at 4°C for five min, then washed three times with PBS (1X) to remove residual reagents. The cell pellets were subsequently fixed in 2.5% glutaraldehyde (prepared in PBS) at room temperature for three hours. The samples were then subjected to a graded ethanol dehydration sequence (50%, 70%, 90%, and 95% v/v), culminating in incubation in absolute ethanol for 30 min. Dehydrated cells were mounted onto sterile glass coverslips and air‐dried under ambient conditions. Dried specimens were then prepared for morphological analysis by sputter‐coating with a thin layer of gold using a high‐vacuum coater to enhance conductivity. Scanning electron microscopy (SEM) was performed using a Zeiss instrument operated at an accelerating voltage of 10 kV. The presented images are representative of three independent experiments and consistently demonstrate the morphological changes observed in E. coli cells following exposure to hydrogel.

2.7. Bacterial Live/Dead Staining Fluorescence Studies

Bacterial viability following exposure to the hydrogel surface was checked using a Live/Dead fluorescence assay, based on established protocols [34]. Briefly, mid‐log phase E. coli cultures (OD600 ≈ 0.5) were treated with the hydrogel and incubated for four hours at 37°C. The cells were then harvested by centrifugation at 3000 rpm at 4°C for 6 min and washed thrice with sterile 1× PBS to remove any residual compounds. The resulting bacterial pellet was resuspended in 1 mL of PBS, and an equal volume of dye solution was added (containing 3 µgmL⁻¹ DAPI and 10 µgmL⁻¹ propidium iodide). The samples were incubated at room temperature for 15 min in the dark. Following staining, excess dye was removed via centrifugation at 3000 rpm for 8 min, and the cells were gently resuspended in fresh PBS. Fluorescence imaging was done using a Zeiss fluorescence microscope. DAPI‐positive cells (blue fluorescence) represented the total bacterial population, whereas PI‐positive cells (red fluorescence) indicated membrane‐compromised, non‐viable bacteria. All imaging experiments were conducted in triplicate using biologically independent sample sets.

2.8. Cell Cytotoxicity Assay

HEK293 human embryonic kidney cells were seeded into 96‐well plates at a density of cells per well in 100 µL of complete DMEM (DMEM,10% FBS and 1% penicillin‐ streptomycin). The cells were then subjected to overnight incubation at 37°C in a 5% CO₂ atmosphere to facilitate adherence. After cell attachment, the medium was replaced with either Fmoc‐F, co‐assembled hydrogel, or Fmoc‐E in serum‐free DMEM. The control groups included untreated HEK293 cells (negative control) and cells treated with 1% Triton X‐100 (positive control). All treatments were conducted in triplicate and incubated at 37°C for 24 hours. Following the treatment period, 100 µL of MTT solution (0.5 mgmL⁻¹, in serum‐free DMEM) was added to each well, and plates were incubated for 4 h at 37°C to allow the formation of formazan crystals. After incubation, the medium was removed, and 100 µL of DMSO was added to each well to solubilize the formazan crystals, resulting in a purple solution. An additional incubation at 37°C in the dark for 30 min was performed to dissolve any residual formazan crystals. The absorbance at 570 nm for these purple solutions was measured using a BioTek Epoch microplate spectrophotometer. The percentage of living cells after treatment with the testing agent relative to the untreated control group was considered for cell viability calculations. Data are represented as mean ± standard deviation from three independent biological replicates. Statistical analysis was performed by One Way ANOVA, whereas the ‘p’ values were evaluated by ‘Dunnett’ hypothesis testing.

2.9. Quantification of the MIC of Antibiotics and Their Synergy Potential Studies

The broth microdilution method has been used to measure the minimum inhibitory concentration (MIC) of the antibiotics following the protocol of Clinical and Laboratory Standards Institute (CLSI) guidelines [35]. Serial two‐fold dilutions of each antibiotic were prepared in sterile M9 minimal medium and incubated with E. coli. The MIC was defined as the lowest antibiotic concentration at which visible bacterial growth was inhibited. To assess potential synergistic interactions, sub‐inhibitory concentrations of antibiotics were combined with Fmoc‐F or Fmoc‐E. E. coli cultures were grown to mid‐logarithmic phase (OD_600nm ≈ 0.5), harvested via centrifugation at 1100 × g for 12 min at 4°C, and resuspended in M9 minimal media. The resulting suspension was diluted to a final concentration of 5 × 10⁵ CFU mL⁻¹. A final volume of 100 µL per well was used for the synergy‐test experiments, which were conducted in 96‐well tissue culture plates. 50 µL of bacterial suspension and 50 µL of test solution, which included 40 µL of M9 medium and 10 µL of each compound stock (antibiotic and/or Fmoc‐F or Fmoc‐E), were placed in each well. Bacterial growth was measured by optical density at 600 nm after incubation for 16 h at 37°C. All experiments were conducted in triplicate with independent biological replicates for each experiment. Growth inhibition curves are constructed to evaluate the efficacy of the combination treatments. Statistical analysis was performed by One Way ANOVA, whereas the ‘p’ values were evaluated by ‘Dunett’ hypothesis testing.

2.10. MD Simulations

The molecular dynamics simulation was performed in GROMACS 2023.4 using the OPLS‐AA force field [36, 37, 38]. The models of Fmoc‐F and Fmoc‐E were prepared by the builder program in PyMOL, and the initial structure was depicted in Figure S6A [39, 40]. The C‐terminus of Fmoc‐F bears a carboxylate anion, resulting in a net charge of −1 per molecule. Both carboxylic groups present in Fmoc‐E exist in the deprotonated state; accordingly, each Fmoc‐E molecule carries a net charge of −2. The selected protonation states are consistent with the expected charge profiles of the molecules at pH 7.4, under which the antibacterial assays were performed. As the parameters for the Fmoc group are not available in the default OPLS‐AA force field, the Fmoc parameters optimized by Ren and co‐workers were incorporated into the force‐field parameter files [41]. The complete set of Fmoc parameters is available in the Supporting Information of the article published by Ren and co‐workers [41]. To construct the simulation system, 100 Fmoc‐F and 50 Fmoc‐E molecules were randomly distributed within a cubic simulation box measuring 9 nm × 9 nm × 9 nm. To assess the reproducibility and robustness of the self‐assembly process, three independent MD simulations were performed using distinct initial molecular configurations. The initial arrangements were generated by varying the rotational orientation of the PyMOL‐generated molecular structures during insertion into the simulation box. Specifically, molecules were randomly rotated about the x‐, y‐, and z‐axes, only about the z‐axis, or inserted without rotational perturbation for Replicas 1, 2, and 3, respectively (Figure S6B,D). Following system construction, each Replica was subjected to equilibration and production protocol. In all simulations, the systems were solvated with TIP3P water. The charge was neutralized by the addition of 200 Na⁺ ions, which were randomly inserted into the simulation box by replacing solvent water molecules [42]. The system was then energy‐minimized to remove high‐energy contacts using the steepest descent method, a first‐order optimization algorithm employed to locate the local minimum of a differentiable function. Table 1 depicts the different types of molecules and ions present in the simulation box. Following this, the system was equilibrated under NVT and NPT conditions at 310K and 1 bar for 100 ps and 2000 ps, respectively, with the heavy atoms position restrained. For Replica 1, the temperature and pressure were controlled by V‐rescale thermostat and Berendsen barostat during equilibration phase whereas V‐rescale thermostat and Parrinello−Rahman barostat were employed during production run [43, 44]. In replica run 2 and 3, V‐rescale thermostat and C‐rescale barostat were used for temperature and pressure coupling, respectively for both equilibration and production run. A summary of the thermostat and barostat used during simulation was reported in Table 1. The thermodynamic equilibrium of the system was assessed by monitoring parameters such as temperature, pressure, and density. As shown in Figure S7, during the NPT equilibration, the system maintained the target temperature of 310 K throughout the simulation. The target pressure was reached within the first few picoseconds and fluctuated symmetrically around 1 bar, confirming the convergence of both pressure and temperature. The density of the system also converged within a few picoseconds. These features were consistently observed across all replicas, indicating the attainment of thermodynamic equilibrium. The production MD run was performed for a total of 1000 ns, in which the LINCS constraint algorithm was used to constrain the bonds, and PME was employed for long‐range electrostatics with a non‐bonded interactions cut‐off of 1 nm. MD run was performed with a leap‐frog integrator with a step size of 2 fs [45]. Atomic restrained was not applied during the production run. The trajectory was saved every 10 ps for data analysis. The inbuilt GROMACS tools have been utilized for the post‐MD analysis. The molecular topology files, force‐field parameter files, and coordinate files corresponding to the final frames of the simulations have been included in the Supporting Information document (MD_parameter). The per‐molecule SASA was calculated by dividing the total SASA value by the corresponding number of molecules. For example, for the co‐assembled system at 500 ns, the total SASA was ∼ 250.37 nm² and the number of molecules was 150, yielding a per‐molecule SASA of 1.669 nm², or ∼1.7 nm²/molecule. Hydrogen‐bonding (H‐Bond) and radial distribution function (RDF) analyses were also performed using the standard built‐in GROMACS utilities. It is important to note that both the time‐dependent cluster number analysis and RDF calculations were carried out over the final 400 ns of the simulation, during which stable cluster formation was observed.

TABLE 1.

System details and MD parameters. Number of molecules and ions present in each simulation box and MD parameters used during the simulation.

	Replica 1	Replica 2	Replica 2
Number of components
Fmoc‐F	100	100	100
Fmoc‐E	50	50	50
Na+	200	200	200
Water	20900	20916	20890
Equilibration
Thermostat	V‐rescale	V‐rescale	V‐rescale
Barostat	Berendsen	C‐rescale	C‐rescale
Production run
Thermostat	V‐rescale	V‐rescale	V‐rescale
Barostat	Parrinello−Rahman	C‐rescale	C‐rescale
Production time	1000 ns	1000 ns	1000 ns
PBC box size (nm)	9 × 9 × 9	9 × 9 × 9	9 × 9 × 9

3. Results and Discussions

3.1. Co‐Assembly of Fmoc‐F and Fmoc‐E into Functional Hydrogels

We recently reported the broad‐spectrum antibacterial effect of hydrogels prepared from a co‐assembled system of Fmoc‐Phenylalanine (Fmoc‐F) and Fmoc‐Lysine (Fmoc‐K) [46]. Fmoc‐F exhibits activity only against Gram‐positive pathogens, while Fmoc‐K is ineffective against any class of bacteria. The enhanced bactericidal potency of the composite system was attributed to the positively charged Fmoc‐K, as molecular dynamics simulations indicated a novel fibrillar architecture in which Fmoc‐F forms the core, surrounded by Fmoc‐K. We conclude that the peripheral placement of positively charged lysine residues in the co‐assembled fibers permits better interactions with the negatively charged bacterial membrane. This observation is consistent with the current understanding of sequence–function relationships of AMPs, whose sequences mandatorily contain several positively charged residues [47]. The enhanced activity was also facilitated by the reduction in the mechanical rigidity of the co‐assembled hydrogels. In this study, we sought to further verify whether the stability of the co‐assembled fibers or the presence of positively charged lysine residues are the determining factors for the observed broad‐spectrum effect. Accordingly, we replaced Fmoc‐K with Fmoc‐glutamic acid, which carries two negative charges and therefore does not aid the interaction between the fibers and the bacterial membrane, but at the same time may influence the physicochemical properties of the Fmoc‐F‐based hydrogels. Thus, this study aims to provide important insights into developing a general strategy for modulating the efficacy of existing self‐assembling AMPs through a co‐assembly approach. Fmoc‐F formed self‐supporting hydrogels in phosphate buffer (PB) at physiological pH. To prepare the hydrogel, Fmoc‐F and Fmoc‐E were separately dissolved in 100 mm PB (pH 7.4) and heated to 90°C to obtain clear solutions. Subsequently, the individual solutions were mixed and gradually cooled to room temperature to form co‐assembled hydrogels composed of 10 mm Fmoc‐F and 10 mm Fmoc‐E, denoted as H1 (Figure S1A). Notably, under the same conditions, Fmoc‐E alone was unable to undergo hydrogelation and remained in the solution state (Figure S1A). The formation of the co‐assembled hydrogels provides an opportunity to compare the antibacterial activity of the individual components and the composite system.

3.2. Investigation of the Antibacterial Activity of Co‐Assembled Hydrogel

The antibacterial activity of the hydrogels was studied against the Gram‐positive bacterium S. aureus and the Gram‐negative bacterium E. coli, which served as primary model organisms for evaluating antibiotic efficacy and represent the most critical pathogenic strains requiring urgent intervention to mitigate their adverse effects on human health [1]. The Fmoc‐F hydrogel has been established to demonstrate excellent potency against Gram‐positive pathogens. To validate the inherent activity of Fmoc‐F in the co‐assembled system, we tested the antibacterial properties of H1 against S. aureus. As shown in Figure 1A, the Fmoc‐F and H1 hydrogels exhibited comparable activity, validating the retention of the original characteristics of Fmoc‐F in the co‐assembled system. Subsequently, the antibacterial potency of the hydrogels against the Gram‐negative E. coli was examined (Figure 1B). The time‐dependent growth kinetic assay revealed that H1 demonstrated excellent growth inhibition compared to the Fmoc‐F hydrogel, which afforded only a 50% reduction in viability. In contrast, Fmoc‐E had negligible adverse effects on the viability of S. aureus and E. coli at concentrations up to 10 mm, confirming the non‐antibiotic nature of this compound (Figure S1B,C). The broadening of the activity spectrum of Fmoc‐F in the presence of negatively charged Fmoc‐E confirmed that the presence of a positively charged residue is not critical for enhancing the efficacy of self‐assembling antibiotics [46].

FIGURE 1.

(A) Bacterial growth kinetics assay showing the effect of Fmoc‐F and H1 on S. aureus. (B) Growth inhibition effect of Fmoc‐F and H1 on E. coli. (C) Colony forming unit (CFU) counting assay demonstrating the quantitative reduction of S. aureus in the presence of Fmoc‐F and H1 over time. (D) Quantitative reduction in E. coli CFU count in the presence of Fmoc‐F and H1 over time. (E, F) Bacterial growth kinetics assay showing the effect of varying Fmoc‐F concentrations in combination with a fixed concentration of Fmoc‐E against S. aureus (E) and E. coli (F).

Counting the colony‐forming units (CFU) is a standardized method that allows the determination of viable cells and is suitable for examining the effect of antibiotics in the presence of a higher initial bacterial load, thereby simulating more clinically relevant infection conditions [35].

Accordingly, we investigated the antibacterial efficacy of H1 against both E. coli and S. aureus using the colony‐forming unit (CFU) assay, starting with a relatively high initial bacterial density (OD_600nm = 0.5). Representative colony‐forming plates of untreated and treated bacterial cultures (with Fmoc‐F and H1) were also analyzed to visually assess the bactericidal activity (Figure 1C,D, Figures S2 and S3). H1 demonstrated a rapid and sustained antibacterial effect against E. coli, with the viable bacterial count decreasing by approximately 95% within 4 h of treatment and reaching >99% reduction after 12 h. The corresponding agar plates showed a drastic reduction in colony formation in treated samples compared to the untreated control, confirming the potent killing efficiency of the co‐assembled hydrogel. Importantly, no detectable bacterial regrowth was observed even after 24 h of incubation, further establishing the strong bactericidal activity of H1 against Gram‐negative pathogens (Figure 1D, Figure S3). As expected, the Fmoc‐F hydrogel alone displayed a negligible reduction in the E. coli CFU count, while a remarkable inhibitory effect against S. aureus was observed throughout the incubation period, both in the individual Fmoc‐F hydrogel and in the co‐assembled state, as evidenced by the substantial reduction in colony formation on the corresponding treated plates compared to the untreated control. These findings further support the selective Gram‐positive antibacterial activity of Fmoc‐F (Figure 1C,D, Figures S2 and S3). The concentration dependence of the active agent Fmoc‐F on the bactericidal effect was also tested by varying the concentration of Fmoc‐F in the co‐assembled hydrogels while maintaining a constant 10 mM concentration of Fmoc‐E (Figure 1E,F). As shown in Figure 1F, E. coli cell viability exhibited a strong dose dependence with respect to Fmoc‐F, with the inhibitory effect starting to diminish below 5 mm Fmoc‐F. A similar experiment with Gram‐positive S. aureus displayed potent activity above 0.75 mm Fmoc‐F in the presence of 10 mm Fmoc‐E (Figure 1E). This observation is consistent with the fact that the monomeric state of Fmoc‐F is sufficient to exert antibacterial effects against Gram‐positive bacteria. However, potency toward Gram‐negative bacteria critically depended on the fibrillar assembly of Fmoc‐F, but in a co‐assembled state rather than as a single component.

3.3. Mechanistic Characterization of the Antibacterial Activity of the Co‐Assembled Hydrogels

A major difference between Gram‐negative and Gram‐positive bacteria is the presence of an outer membrane in the former, which restricts the entry of many antibiotics, making them ineffective [48]. Accordingly, we studied the interaction of the co‐assembled hydrogels and Fmoc‐F with the bacterial membrane to probe their membrane‐disrupting ability. This is also consistent with the fact that AMPs exert their effect by perturbing the integrity of the outer membrane in Gram‐negative pathogens. The 1‐N‐phenylnaphthylamine (NPN) uptake assay is a sensitive method for evaluating changes in the outer membrane permeability [49].

An enhancement in NPN fluorescence intensity was observed upon treatment of E. coli with the Fmoc‐F hydrogel, and the effect became more pronounced in the presence of H1 (Figure 2A). This clearly demonstrated that co‐assembly facilitated stronger interactions with, and greater disruption of, the bacterial envelope, in agreement with the enhanced efficacy of the composite hydrogels against Gram‐negative strains. To further support hydrogel‐induced membrane disruption, scanning electron microscopy (SEM) was employed to observe the morphological changes in E. coli and S. aureus following exposure to H1 (Figure 2B–E) [50]. Under SEM, E. coli appeared as cylindrical‐shaped organisms with regular topology and uniform dimensions, whereas S. aureus exhibited a spherical morphology with smooth surfaces and largely independent geometry (Figure 2B,D). In contrast, E. coli treated with H1 revealed substantial membrane deformation and increased surface roughness, indicating compromised membrane integrity (Figure 2C). Similarly, extensive alteration of the membrane architecture, manifested as changes in shape, size, and surface smoothness, as well as the fusion of individual bacterial cells, was commonly observed in S. aureus following incubation with H1 (Figure 2E). The membrane integrity of cells under physiologically relevant conditions, in contrast to the extensive post‐processing required for SEM, was further examined by staining E. coli with two different dyes: DAPI and propidium iodide (PI). DAPI can permeate the cell membrane and label intracellular DNA, whereas PI intercalates and stains DNA only when the membrane is compromised. The cells treated with H1 exhibited extensive staining with both DAPI and PI, indicating that antibiotic treatment resulted in significant membrane disruption, which allowed the entry of PI. In comparison, untreated E. coli exhibited minimal PI staining, confirming normal biochemical profiles (Figure 2F, Figure S4). These observations support the broad‐spectrum antibacterial nature of the co‐assembled hydrogels, suggesting that interactions between the self‐assembled fibrils and the bacterial membrane may play a critical role in the observed bactericidal properties.

FIGURE 2.

(A) NPN fluorescence assay showing the percentage of E. coli membrane disruption induced by H1, Fmoc‐F, and Fmoc‐E. ‘p’ value, < 0.0001, resulting statistical significance with R² value of 0.9989. (B) Scanning electron microscopy (SEM) image illustrating the intact surface morphology of untreated E. coli. (C) SEM image revealing pronounced morphological deformation and membrane rupture in E. coli following H1 treatment. (D) SEM image showing the smooth and intact surface of untreated S. aureus. (E) SEM image depicting severe surface damage and structural collapse of S. aureus upon H1 treatment. The scale bar represents 1 µm.(B‐E) (F) Fluorescence microscopy images of control and H1‐treated E. coli stained with DAPI (blue, viable cells) and PI (red, membrane‐compromised cells). Scale bar is 20 µm.

3.4. Insights Into the Physicochemical Features of Co‐Assembled Hydrogel

The enhanced activity of co‐assembled H1, compared to its constituent building blocks, was assessed by examining the physicochemical properties of the hydrogel. It can be presumed that the co‐assembly may impose geometric restrictions on the optimal self‐association of Fmoc‐F, which is negatively charged at pH 7.4, in the presence of negatively charged Fmoc‐E, leading to an increase in the critical aggregation concentration (CAC). The higher CAC would increase the monomer concentration of Fmoc‐F in equilibrium with the hydrogel phase and may explain the observed broad‐spectrum effect, as the monomeric form of Fmoc‐F is bactericidal against Gram‐positive bacteria. The CAC of Fmoc‐E and H1 was measured using a pyrene fluorescence assay, a well‐established technique for determining the CAC of self‐assembling biomolecules [51]. The assay revealed an equilibrium monomeric concentration of Fmoc‐F in the single‐component assembly and H1 hydrogel of 4.9 and 5.3 mm, respectively (Figure 3A,B, Figure S5D,E). The similar value of Fmoc‐F concentration indicated that the potency of H1 may stem from the altered physicochemical features of the co‐assembled hydrogel, rather than from differences in equilibrium monomer concentration. Accordingly, the physical properties of the hydrogels were evaluated by rheological measurements (Figure 3C–F). The strain sweep experiment was performed at a constant frequency of 1 Hz and revealed differential behavior for Fmoc‐F and H1 hydrogels (Figure 3C,E). The storage moduli (Gʹ) and loss moduli (Gʹʹ) of both gels decreased with increasing strain, but Gʹ of H1 showed much higher sensitivity to the applied strain, confirming an early onset of structural reorganization. Furthermore, H1 reached the Gʹ and Gʹʹ crossover point at much lower strain, indicating that the hydrogel network in the co‐assembled system can be more easily disrupted compared to the Fmoc‐F hydrogel (Figure 3E). Frequency sweep experiments demonstrated a consistently higher Gʹ than Gʹʹ, a crucial attribute for maintaining a stable hydrogel state. However, the mechanical rigidity of the co‐assembled H1 was substantially lower than that of Fmoc‐F, as evidenced by more than a 50% reduction in the storage modulus (Figure 3D,F). Thus, rheological measurements confirmed the formation of a physicochemically weaker Fmoc‐F hydrogel in the presence of Fmoc‐E. The direct association between the individual components in the self‐assembled state, stabilized by various noncovalent forces such as π‐π stacking, Hydrogen Bonding, and van der Waals interactions, may account for the observed properties. The morphological features of H1, Fmoc‐F, and Fmoc‐E were examined using transmission electron microscopy (TEM). As expected, Fmoc‐F exhibited a fibrillar morphology with diameters around 20±5 nm (Figure S5B) [46]. In contrast, Fmoc‐E formed dispersed, poorly defined amorphous aggregates, consistent with its low self‐assembly propensity (Figure S5C). Interestingly, H1 also displayed a fibrillar morphology, with fiber diameters comparable to those observed for the parent Fmoc‐F assemblies (Figure S5A). These results indicate that while co‐assembly significantly altered the physicochemical properties of the system, it preserved the characteristic fibrillar morphology of Fmoc‐F.

FIGURE 3.

(A) Determination of the critical aggregation concentration (CAC) of Fmoc‐F. (B) CAC determination of the co‐assembled hydrogel with respect to Fmoc‐F concentration. (C–F) Rheological characterization of the resulting hydrogels. Oscillatory strain sweep profiles of Fmoc‐F (C) and H1 (E) recorded at a constant frequency of 1 Hz, and frequency sweep profiles of Fmoc‐F (D) and H1 (F), demonstrating the differential viscoelastic behavior of Fmoc‐F and H1.

Circular dichroism (CD) spectroscopy was performed on Fmoc‐F, Fmoc‐E, and H1 to investigate their supramolecular organization. As shown in Figure S5F, Fmoc‐F exhibited a strong positive maximum at ∼195 nm, a negative maximum around 206 nm, and well‐resolved positive bands in the 218–230 nm region. The high intensity of these CD signals indicated the formation of a highly ordered self‐assembled supramolecular architecture, resulting in strong excitonic coupling between chromophores. The positive band at ∼195 nm can be attributed to electronic transitions associated with the C─N bond connecting the Fmoc group to the amino acid. The negative band at ∼206 nm may have originated from excitonic interactions between phenylalanine residues within the assembled state, with overlapping contributions from multiple π–π ^* and n–π^* electronic transitions involving the C─N bond [52]. The positive Cotton effects observed between 218 and 230 nm are indicative of the higher‐order organization of the Fmoc moieties within the supramolecular ensemble [53]. H1 retained the characteristic spectral features of Fmoc‐F, but with reduced ellipticity, suggesting preservation of the overall supramolecular architecture accompanied by partial perturbation of aromatic packing upon incorporation of Fmoc‐E. These altered structural features of H1 may be responsible for the observed enhanced antibacterial activity. In contrast, Fmoc‐E displayed substantially weaker CD signals, indicating limited aromatic organization and a lower degree of ordered self‐assembly. Notably, a weak blue‐shifted negative band at ∼203 nm was also observed, which can be primarily attributed to electronic transitions associated with the C─N linkage [52].

3.5. Molecular Dynamics (MD) Simulation of the co‐assembly Process

To decipher the underlying interaction between Fmoc‐F and Fmoc‐E in the co‐assembled state at the molecular level, all‐atom molecular dynamics (MD) simulations were performed [54, 55, 56]. In addition to the negatively charged amino acid conjugates, the system also comprised sodium ions to neutralize the charges. A total of three independent MD runs were carried out, termed as Replica 1, Replica 2, and Replica 3. The assembly of individual monomers was observable at the early stages of simulation, and over a period of 1 µs, cluster formation incorporating all the monomers became visible (Figure 4A). To assess the stability of the aggregated structure, we evaluated the time‐dependent solvent‐accessible surface area (SASA) and H‐bonding pattern over the course of the simulation. SASA provides a direct assessment of the association of hydrophobic molecules, and a reduction in SASA confirms the burial of the hydrophobic parts of the molecule [57]. The SASA was calculated by rolling a spherical probe of radius 0.14 nm, which is a standard radius of a water molecule, over the Van der Waals surface of the Fmoc‐F and Fmoc‐E molecules at different time points of simulation, following the algorithm developed by Eisenhaber and coworkers [58]. The SASA of the molecules decreased sharply at the beginning and reached an equilibrium value of ∼1.7 nm²/molecule around 500 ns in Replica 1, indicating the assembly of monomers into a stable aggregate (Figure 4B). In contrast, the SASA of Fmoc‐F reduced much faster, reaching the equilibrium value of ∼1.7 nm²/molecule by 300 ns, confirming its faster association and high aggregation propensity. The assembly of Fmoc‐E was the slowest, and its SASA exhibited a higher level of fluctuation, with an average value of 1.68 nm²/molecule over the last 400 ns of simulation (Figure 4B). Importantly, the similar average SASA of individual residues over the simulation period (Figure S8B), along with the per‐molecule values described above, indicated efficient burial of Fmoc‐E residues along with Fmoc‐F in the core of the fibril, as supported by the analysis of the co‐assembled architecture (vide infra). Replica 2 and 3 demonstrated very similar SASA values of ∼1.7 nm²/molecule for both the co‐assembled system as well as Fmoc‐F in the range of 300 to 400 ns (Figure S8A,B). In contrast, Fmoc‐E exhibited higher fluctuation. H‐bond formation between the individual building blocks is an important indicator of stable self‐assembly. The number of H‐bonds between Fmoc‐F and Fmoc‐E increased substantially at the expense of H‐bonding with the solvent (water and ions) and attained a constant value around 500 ns, coinciding with the timeframe observed in SASA analysis, confirming the formation of a stable co‐assembled system (Figure S8C). In Replica 2 and 3, the number of H‐bonding between Fmoc‐F and Fmoc‐E reached maxima around 250 ns, but maintained a similar trend of increase in the number of H‐bonding between the constituent molecules and corresponding decrease in the number of H‐bonds with solvent molecules (Figure S8C).

FIGURE 4.

Molecular dynamics (MD) simulations illustrating the co‐assembly behavior of Fmoc‐F and Fmoc‐E. (A) Representative snapshots from the MD trajectory at different time intervals, with Fmoc‐F and Fmoc‐E shown in green and red, respectively. (B) Time‐dependent variation in the solvent‐accessible surface area (SASA) of Fmoc‐F (green), Fmoc‐E (red), and their co‐assembled system (black). (C) Radial distribution function (RDF) plots corresponding to different molecular pair arrangement within the co‐assembled system: Fmoc‐F–Fmoc‐F (green), Fmoc‐E–Fmoc‐E (red), and Fmoc‐F–Fmoc‐E (black). r denotes the distance between CG atom of the Fmoc residues. (D) Hierarchical organization of the largest cluster observed during the simulation, with a cross‐sectional view highlighting the arrangement of Fmoc groups within the core. (E) Formation of continuous π–π stacking interactions involving both Fmoc‐F and Fmoc‐E molecules, with the relevant atoms depicted as spheres.

A closer analysis of the evolution of the aggregation process revealed that homogeneously distributed monomers co‐assembled into two distinct aggregates over the course of the simulation. A plot of the number of clusters with respect to time, as well as the presence of molecular ensembles of two different sizes with higher frequency, corroborated the visual trajectory analysis (Figure 4A,C, Figure S9A,B). An important aspect of the assembly process is the presence of positively charged Na⁺ ions in the vicinity of the co‐assembled geometry, with almost 76% of the Na⁺ ions located within van der Waals distances of the atoms belonging to the amino acid moieties. The cations played a crucial role in screening the charges of glutamic acid, enabling it to co‐assemble with Fmoc‐F. The abundance of metal ions near glutamic acid further supported this notion. Considering the negative potential of the bacterial membrane, charge screening may also be critical in facilitating the interaction between the co‐assembled fibers and the membrane.

We further analysed the largest cluster to decipher the organization of individual building blocks. As shown in Figure 4D and S10, the aggregate adopted an elongated shape, and it can be surmised that further association of such nucleating modules may afford mature fibrils required for hydrogelation. The hydrophobic Fmoc groups are mostly located in the core of the aggregate and take part in extensive and contiguous π‐stacking interactions with another Fmoc‐F or Fmoc‐E (Figure 4E). In addition, π‐π interactions between the side chains of Phe and Fmoc groups are also abundant (Figure S11). The main chains and side chains of Fmoc‐F and Fmoc‐E have several H‐bond acceptors and a single H‐bond donor, and these sites are involved in extensive H‐bonding with solute or solvent molecules, conferring necessary stability to the co‐assembled system (Figure S12). The analysis of cluster formation during the production run of replicas 2 and 3 revealed the formation of two predominate cluster, and a close‐up inspection of the biggest cluster showed the presence of the Fmoc group at the core, similar to that of Replica 1 (Figure S9A, Figure S10). The pair distribution function (RDF) provides critical information about the spatial arrangement of molecules by describing how the local density of neighboring molecules varies relative to a reference molecule. The RDF profile was determined with respect to the CG atom of the Fmoc group for each pair of molecules under consideration (Figure S6A,C) [59]. As shown in Figure 4C, the RDF profile of Fmoc‐F: Fmoc‐F, Fmoc‐E: Fmoc‐E, and Fmoc‐F: Fmoc‐E pairs displayed dominant peaks at approximately 4.5–5.3 Å. These peaks can be ascribed to the π–π stacking interactions among Fmoc molecules within the aggregate [59]. In addition, secondary peaks near 8.2 Å and 10.1 Å indicate the higher‐order structural organization. Interestingly, the presence of two peaks in the case of Fmoc‐F: Fmoc‐E pair of Replica 1 corresponded well with the presence of two aggregates with different numbers of atoms (2000 and 5000, respectively) in the simulation box, where the larger aggregate showed a higher frequency of occurrence (Figure S9B). In comparison, Replica 2 produced two aggregates with comparable sizes (3000 and 4000 atoms, respectively) and frequency of occurrence (Figure S9B). These properties have been reflected in the RDF feature of Replica 2, exhibiting a broad peak in the range of 7.9 to 10.1 Å (Figure S9C). In Replica 3, a large aggregate of 6000 atoms (the second aggregate contains only 1000 atoms or around 20 molecules) predominantly contributes to the observed singlet peak around 9.2 Å (Figure S9B,C). Thus, the RDF signals between the Fmoc‐F:Fmoc‐E pair indicated close molecular proximity and heterogeneous higher‐order organization within the co‐assembled state, primarily driven by intermolecular π–π stacking between Fmoc‐F and Fmoc‐E. In summary, the MD simulations revealed that the co‐assembly architecture comprises both Fmoc‐F and Fmoc‐E in the fibrillar core, forming various non‐covalent interactions with each other. This heterogeneous association may be responsible for the weaker assembly and reduced mechanical rigidity compared to the pristine Fmoc‐F assembly.

Based on this observation, a schematic depiction relating the enhanced potency of Fmoc‐F in the presence of Fmoc‐E is presented in Figure 5. Fmoc‐F, which shows excellent aggregation propensity and antibacterial efficacy against Gram‐positive bacteria, has minimal activity against Gram‐negative pathogens. The presence of the adjuvant, Fmoc‐E, which has negligible antibacterial activity but is capable of interfering with the self‐assembling attributes of Fmoc‐F, leads to the formation of weakly associated aggregates. The reduced stability of the co‐assembled hydrogel enables the facile reorganization of the building blocks, facilitating interaction with the bacterial membrane and resulting in effective membrane disruption and the killing of Gram‐negative pathogens. The presence of Na⁺ ions surrounding the fibers composed of negatively charged building blocks further aids this process [57]. Thus, our current study, in combination with previous work³⁴, indicates that modulating the self‐assembly properties of AMPs by incorporating

FIGURE 5.

Schematic illustration of the interaction between co‐assembled fibrils and the E. coli membrane. The fibrils of Fmoc‐F have limited ability to permeate and disrupt the bacterial membrane owing to the stronger intermolecular interactions between Fmoc‐F molecules in the self‐assembled fibrils (left). In contrast, the co‐assembled fibrils have weaker intermolecular interactions between the constituent building blocks, Fmoc‐F and Fmoc‐E, allowing better membrane permeabilization and leading to the loss of membrane integrity (right).

non‐antibiotic, biocompatible adjuvants capable of forming non‐covalent interactions with the targeted AMP can provide a generalized strategy to enhance the antibacterial potency of self‐assembling AMPs, especially those developed for wound healing and topical applications.

3.6. Assessment of Cellular Cytotoxicity of Co‐Assembled Hydrogel

To compare the relative toxicity of the co‐assembled systems and their constituent building blocks, we evaluated in vitro cell viability using human embryonic kidney (HEK293) cells, a commonly used human cell line for early‐stage biocompatibility testing. The cells were treated with the co‐assembled systems, as well as with Fmoc‐F and Fmoc‐E separately, at a fixed concentration of 1.5 mM, since higher concentrations produced hydrogels that were not amenable to cell culture assays [46]. The number of viable cells was estimated using the 3‐(4,5‐dimethylthiazol‐2‐yl)‐2,5‐diphenyltetrazolium bromide (MTT) assay, which provides an indirect assessment of cellular metabolic activity [60]. The co‐assembled systems exhibited cell viability of approximately 90%, comparable to that of Fmoc‐F and Fmoc‐E alone, indicating a high level of biocompatibility (Figure 6). The comparable viability of the components and their combination suggests that co‐ assembly can serve as an attractive strategy for formulating potent antibacterial composites without compromising the biocompatibility of the original building blocks.

FIGURE 6.

Assessment of the cytotoxicity of H1, Fmoc‐F, and Fmoc‐E using the MTT assay in HEK293 human embryonic kidney cells. A 2% Triton X‐100 solution served as the positive control. ‘p’ value is 0.0001 (<.05), resulting in statistical significance with R² value of 0.9916.

3.7. Exploration of the Adjuvant Potential of the Co‐Assembled System with Conventional Antibiotics

This work has established the adjuvant features of Fmoc‐E in potentiating the activity of the Gram‐positive‐specific antibiotic Fmoc‐F, extending its efficacy to a broad‐spectrum one. The NPN assay described above also indicated that the co‐assembled system facilitated the penetration of the hydrophobic dye NPN through the outer membrane, a known barrier limiting the potency of many clinically approved antibiotics against Gram‐negative bacteria [48]. Membrane‐active AMPs, at concentrations far below their MIC, and related peptidomimetics or derivatives with minimal or no bactericidal activity, have been established as potent adjuvants that act synergistically with many classical antibiotics, especially those with hydrophobic characteristics, leading to substantial reductions in their MIC values [6, 61]. Following the reported adjuvant capabilities of AMPs and the cues from the NPN assay, we explored the ability of the co‐assembled hydrogels to enhance the efficacy of two different antibiotics, erythromycin and tetracycline, against E. coli. Erythromycin targets the 50S ribosomal subunit but is unable to penetrate the outer membrane of Gram‐negative pathogens, whereas tetracycline has broad‐spectrum activity but is a major substrate of efflux pumps. As shown in the Figure 7A,B, the MICs of both erythromycin and tetracycline were reduced fourfold, from 500 to 125 µgmL^{− 1}, when co‐administered with sub‐inhibitory concentrations of the co‐assembled hydrogel (3 mm Fmoc‐F and 10 mm Fmoc‐E), demonstrating excellent synergistic functionality.

FIGURE 7.

(A) Endpoint growth profiles of E. coli illustrating the minimum inhibitory concentration (MIC) of erythromycin and its modulation in the presence of a sub‐inhibitory concentration of Fmoc‐F (3 mm) and Fmoc‐E (10 mm), as determined from bacterial growth kinetics assays. (B) Endpoint E. coli growth analysis depicting the MIC of tetracycline and the synergistic effect observed upon co‐treatment with a sub‐inhibitory concentration of Fmoc‐F (3 mm) and Fmoc‐E (10 mm). Statistical significance was determined by comparing the values of testing agents with that of the negative control, resulting in an overall ‘p’ value < 0.0001 with an R² value of 0.9957.

The observed synergy may result from the composite system's ability to disrupt bacterial membranes, thereby facilitating the better accumulation of the respective drugs at their target sites at significantly lower doses. In summary, the co‐assembly of Fmoc‐F and Fmoc‐E not only functions as an antimicrobial agent but also acts as a biocompatible adjuvant platform that substantially enhances the potency of conventional antibiotics against resistant bacterial strains.

4. Conclusions

In conclusion, we have developed co‐assembled hydrogel formulation in which the combined system demonstrated higher potency in reducing bacterial cell viability compared to the treatment with individual components. The enhanced activity resulted from the weaker association between native self‐assembling antibiotics in the co‐assembled fibers, aided by different non‐covalent interactions as probed by MD simulation and rheological measurement. Thus, this study provided a generalized strategy to augment the activity of self‐assembling AMPs, especially involving formulation comprising hydrogels relevant to wound healing and topical application, by incorporating biocompatible non‐antibiotic agents having the potential to interfere with the aggregation of target AMP [62]. Moreover, the membrane‐disrupting capability of the co‐assembled fibers reported in this study has great potential as an adjuvant to conventional antibiotics, as evidenced by the reduction in the MIC of erythromycin and tetracycline.

Author Contributions

Sudipta Mondal and Bodhisattwa Das Gupta conceived and designed the experiments. Bodhisattwa Das Gupta, Rahul Kuiri, Thangavel Vijayakanth, Nandita Ghosh, Ranik Konar and Sudha Shankar performed the experiments. Sudipta Mondal and Bodhisattwa Das Gupta wrote the manuscript, and all authors commented on the draft and proofread the manuscript.

Funding

The authors gratefully acknowledge financial support from DST‐SERB, Government of India for core research grant (CRG/2021/002797).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File 1: mabi70221‐sup‐0001‐SuppMat.pdf.

Supporting File 2: mabi70221‐sup‐0002‐DataFile.zip.

Data Availability Statement

The data that supports the findings of this study are available in the supplementary material of this article.