Liposomal Encapsulation Reduces the Cytotoxic Effects of Gramicidin S in Monolayer and Spheroid Fibroblast Cultures

TL;DR

Liposomal encapsulation reduces the harmful effects of gramicidin S on fibroblast cells in both 2D and 3D cultures.

Contribution

This study demonstrates that liposomal encapsulation significantly mitigates gramicidin S cytotoxicity and validates 3D spheroids as a better toxicity model.

Findings

Liposomal GS preserved 80.3–82.2% cell viability at 75 µg/mL, compared to 0.5% with free GS.

Spheroid cultures showed higher resistance to GS toxicity, with liposomal GS preserving viability above 84.8%.

Liposomal encapsulation likely reduces toxicity by limiting direct membrane disruption.

Abstract

Background/Objectives: Gramicidin S (GS) is a cyclic antimicrobial peptide with strong antibacterial activity but significant cytotoxicity toward mammalian cells. This study evaluated GS-induced cytotoxicity in L929 fibroblast cells using both traditional 2D monolayer cultures and more physiologically relevant 3D spheroid models, and assessed whether liposomal encapsulation could mitigate toxicity and improve biocompatibility. Methods: L929 cells were cultured in monolayers and spheroids and treated with free GS or GS encapsulated in liposomes of varying lipid compositions. Cell viability and morphology were evaluated after 24 h of exposure using standard cytotoxicity assays. Results: Control liposomes, regardless of tested lipid type or concentration, showed no adverse effects on cell morphology or viability. Free GS caused pronounced, dose-dependent cytotoxicity in monolayers, decreasing viability to 11.0 ± 1.9% and 0.5 ± 1.1% at 50 and 75 µg/mL, respectively. By contrast, encapsulation in liposomes significantly reduced toxicity (p < 0.05), preserving 80.3–82.2% viability at 75 µg/mL depending on formulation, corresponding to protection factors exceeding 160-fold (80.3% vs. 0.5%). Spheroid cultures showed slightly higher resistance to GS; free GS reduced viability to 2.9%, while liposomal GS preserved it above 84.8%, depending on lipid composition. Conclusions: Liposomal encapsulation effectively reduces GS-induced cytotoxicity, likely by limiting direct membrane disruption. Moreover, spheroid models provide a more physiologically relevant and predictive platform for toxicity testing, while the results support nanoliposomes as a practical delivery strategy to enhance the safety of antimicrobial peptides during preclinical development.

Figures

Full text

1. Introduction

Drug redeployment could be a fast-track strategy for addressing growing challenges such as microbial resistance, malaria, cancer treatment, and protein-folding diseases [1,2,3,4,5]. By improving drug delivery through appropriate vehicles or encapsulation strategies, it may be possible to broaden the therapeutic applicability of already approved drugs or active pharmaceutical ingredients, thereby reducing the need for prolonged and costly clinical trials focused on long-term safety and severe adverse effects [6,7,8].

One class of evolutionarily proven self-defense substances, antimicrobial peptides (AMPs), due to their unique structure and nonspecific mechanism of action, exhibit a range of additional beneficial effects with potential anti-inflammatory, anticancer, radioprotective, and cryoprotective properties [9,10].

Gramicidin S (GS), an antimicrobial cyclic peptide composed of 10 L- and D-amino acid residues and derived from Bacillus brevis, exhibits potent antibiotic activity [11,12]. The molecular mechanism by which GS disrupts cell membranes has been extensively studied. GS disrupts cell membranes through direct binding to the lipid bilayers of bacterial and eukaryotic cells. This interaction is primarily entropy-driven, with GS localized beneath the polar headgroup region of the membrane and its polar groups oriented toward the aqueous phase. The membranotropic activity of GS is determined by its amphiphilic structure: positively charged L-ornithine residues are located on one side of the molecule, while hydrophobic D-phenylalanine residues occupy the opposite side [13]. Electrostatic interactions with negatively charged bacterial lipids facilitate peptide insertion into the membrane, resulting in increased lipid cross-sectional area, reduced acyl-chain ordering, and bilayer thinning, which promotes pore formation and membrane lysis, similarly to other gramicidins [14]. Nuclear magnetic resonance 31P-NMR and X-ray diffraction studies demonstrate that GS induces phospholipid bilayer thinning at low concentrations and promotes the formation of non-lamellar phases at higher concentrations [15,16]. At micromolar concentrations, GS alters the transmembrane potential and impairs eukaryotic cell viability [17]. In addition to pore formation, GS can induce lipid phase segregation, leading to delocalization of essential membrane proteins and inhibition of cellular processes [18].

The available values for the minimum inhibitory concentration (MIC) of GS range from 3 to 10 µg/mL [12]. Similar values have been reported for the minimum bactericidal concentration (MBC) [19,20]. GS also exhibits activity against certain pathogenic fungi, such as Candida spp., with a minimum fungicidal concentration (MFC) of approximately 200 µg/mL [12,21]. The promising antibacterial activity of GS is, unfortunately, associated with notable cytotoxicity toward eukaryotic cells [19]. GS exerts significant hemolytic toxicity and nephrotoxicity at concentrations of approximately 10 µg/mL and more [19,22]. The hemolytic effect of GS has been widely recognized since early studies [23]. In addition, GS has been shown to induce platelet disaggregation [24]; kinetic light-scattering analyses suggest that this process involves the detachment and dissociation of individual inactivated platelets from the surface of platelet aggregates. Furthermore, GS reduces the adhesive behavior of specific cell types and inhibits proliferation in multiple malignant cell lines [25]. While these effects are indicative of toxicity, they also point to a potential utility of GS as an antitumor agent; however, its pronounced adverse toxicity necessitates careful and cautious evaluation.

Liposomal nanocontainers are widely employed in targeted therapeutic applications, including cancer, antimicrobial, anti-inflammatory, and gene-based treatments, owing to their high biocompatibility and natural biodegradability [1,3,8,26,27]. Their well-defined physicochemical properties and predictable membrane behavior enable the design of controlled lipid-based systems for the efficient transport of bioactive compounds. Dipalmitoylphosphatidylcholine (DPPC) is a widely distributed phospholipid and a major structural component of biological membranes. Owing to its bilayer-forming properties, DPPC is commonly used in the preparation of liposomal systems. The incorporation of cholesterol (CHOL) into DPPC-based bilayers is known to reduce membrane fluidity and permeability while enhancing overall membrane stability [28,29]. Cardiolipin (CL), a phospholipid characteristic of bacterial and mitochondrial membranes, is frequently employed to introduce negative surface charge into lipid assemblies. Modulating the cardiolipin content during liposome formulation enables controlled adjustment of surface electrostatic properties, such as the zeta potential [30,31]. These lipid characteristics can be strategically exploited to design lipid-based delivery systems optimized for specific biomedical applications. In particular, we tested different combinations and ratios of DPPC, CL, and CHOL for liposome preparation and characterization. Liposomes were characterized in terms of stability, surface charge, size, and other physicochemical properties using biochemical and biophysical methods, and the results have been previously reported [29]. The interaction between the liposomes and the carried drug, GS, was also investigated and reported [29,31]. In the present study, we performed our analysis of four types of liposomes, ranging from simple DPPC liposomes to more complex formulations composed of DPPC, CL, and CHOL in the most promising ratios identified in previous studies [31].

In this study, the antimicrobial peptide GS was investigated in its bound form to liposomal nanocarriers, designed to enhance GS delivery and, importantly, reduce its cytotoxicity. Various liposomal formulations, including those containing cardiolipin or cholesterol, have previously been evaluated for their interaction with GS [29,31]. In model membranes, GS exhibits a higher affinity for negatively charged lipids than for zwitterionic or neutral phospho- and glycolipids, with reduced interaction in the presence of cholesterol [29]. After insertion into the bilayer, GS preferentially associates with interfacial regions, perturbing lipid packing and compromising membrane integrity. As a result, GS disrupts liquid-crystalline bilayers and enhances nonspecific permeability in both model and biological membranes [15,16]. At low concentrations, GS induces bilayer thinning, whereas at higher concentrations it promotes the formation of inverted nonlamellar cubic phases in phospholipid dispersions [29,31]. In this study, in addition to our previous work that mostly focused on the biochemical and biophysical properties of liposomal formulations, we investigate the biological effects of liposome-encapsulated GS.

The L929 fibroblast cell line was used in this study to compare the toxicity of liposome-carried GS with that of free GS. This cell line is widely employed in compound toxicity and biocompatibility assays [32] in accordance with International Organization for Standardization, ISO 10993-5 and ISO 10993-12 guidelines. L929 cells can grow as confluent monolayers or be cultured into 3D spheroid structures, which more accurately mimic the complex, three-dimensional architecture of tissues and tumors [33,34].

The aim of this study was to assess the extent of interaction between GS-loaded liposomes and both monolayer cultures and three-dimensional spheroids, based on the hypothesis that liposomal encapsulation and lipid composition critically modulate GS–cell interactions and thereby reduce GS-induced cytotoxicity. We investigated how variations in liposomal composition influence cellular adhesion and viability. These analyses provide insight into liposome–cell interactions across different cellular architectures, supporting the rational optimization of GS-liposomal formulations to improve biocompatibility and therapeutic efficacy in biomedical applications.

2. Results

2.1. Effects of GS-Free Control Liposomes on Cells

Liposomes, previously investigated by our team as potential carriers for GS, must first be evaluated for their inherent effects on cellular function, independent of any pharmaceutical loading, under the assumption of a neutral baseline activity. Given their ability to fuse with the cell surface membrane, liposomes may induce localized alterations in membrane composition and structure, which could influence key cellular processes such as viability and adhesion. These effects could be particularly dependent on the lipid composition and concentration of the liposomes.

Therefore, in the initial phase of this study, we assessed the impact of various lipid formulations and concentrations of unloaded (control “pure”) liposomes on fundamental cellular behaviors—specifically viability and monolayer confluence.

Morphologically, following 24-h incubation with pure liposomes, no significant changes were observed in the visual structure of L929 cells cultivated as a monolayer, regardless of liposome concentration or lipid composition. A representative image of L929 cells incubated with DPPC/CL/CHOL liposomes (85/5/10 mol%) is shown in Figure 1a. The cellular morphology was predominantly oval or polygonal (Figure 1a,b). The cells contained a clearly visible nucleus with one or several nucleoli. Nuclei appeared round, with a homogeneous structure and no signs of fragmentation. Cells undergoing mitosis were also visible in the field of view (Figure 1a,b).

To assess the effect of liposomal lipid composition on cellular viability, cells were incubated for 24 h with liposomes of distinct lipid profiles, and viability percentages were subsequently quantified (Table 1). Cell condition was assessed through fluorescein diacetate/ethidium bromide (FDA/EB) staining for viability. Liposome treatment for 24 h, across all tested lipid compositions and liposome quantity, did not reduce the viability of L929 cells nor cause their detachment from the culture surface. Monolayer cell viability remained consistently 98.5–99.8% (Table 1), closely matching the control untreated cells, which were previously established and referenced as having 100% viability. These results indicate no significant deviation from the control benchmark. Based on these results, a liposomal suspension volume of 400 µL was selected for subsequent experiments, and in control conditions where liposomes were not added, the total volume was normalized to ensure equivalence across all experimental conditions.

2.2. Effects of GS on Cells in Monolayer Culture

In the subsequent phase, cell viability in the L929 cell line was assessed following incubation with free GS at concentrations of 10, 25, 50, and 75 µg/mL. While treatment with 10 and 25 µg/mL resulted in a slight reduction in cell viability during 1-h exposition, exposure to 50 and 75 µg/mL induced strong statistically significant decline, with viability dropping to 18.6 ± 2.9% and 1.9 ± 1.6%, respectively (Table 2). For 24 incubation time, cell viability decreased to 11.0 ± 1.9% and 0.5 ± 1.1% at GS concentrations of 50 and 75 µg/mL, respectively.

2.3. Effects of GS-Loaded Liposomes on Cells in Monolayer Culture

In the next phase, the viability of L929 cells was analyzed following incubation with liposomes of various lipid compositions loaded with GS at concentrations of 10, 25, 50, and 75 µg/mL, and viability was assessed using FDA/EB staining. The results indicated that liposomes loaded with the GS at 10 and 25 µg/mL did not significantly affect cell viability. However, treatment with liposomes containing GS at concentrations of 50 and 75 µg/mL led to a slight decrease in cell viability, depending on the composition of the liposomes (Table 3). For example, for more complex liposomes, contained DPPC/CL/CHOL, the 50 and 75 µg/mL of GS decreased vitality to 79.0 ± 15.1% and 80.3 ± 9.2% correspondingly (Table 3), representing a reduction of approximately 20% compared to the intact control liposomes (Table 1).

The partial impact on L929 cells following treatment with liposome-encapsulated GS was significantly lower compared to the pronounced cytotoxic effect observed with free GS (Table 2). At a concentration of 50 µg/mL, cell viability was 7.2-fold higher with liposomal GS (79.0%) than with free GS (11.0%), while at 75 µg/mL, viability was improved by 160.6-fold (80.3% vs. 0.5%, Table 2 and Table 3), highlighting the protective role of liposomal delivery in mitigating GS-induced toxicity.

Following the viability assays, the proliferative capacity of L929 monolayer cells was evaluated by assessing monolayer confluence. At GS concentrations of 10 µg/mL and 25 µg/mL, a slight reduction in monolayer confluence was observed across all liposome formulations (Table 4). However, monolayer confluence was notably affected at higher GS concentrations (Table 4). In liposomes composed of DPPC/CL/CHOL with 50 and 75 µg/mL GS, confluence significantly declined to 71.6 ± 5.0% and 69.0 ± 11.2%, respectively (Table 4), indicating an approximate 30% reduction compared to the control sample without GS (Table 4, first column).

2.4. Effects of GS on Cells in Spheroid Culture

At next step, we developed and tested spheroid forms of L9292 cells to establish a more complex, organism-like cellular model and to evaluate the three-dimensional mode of interaction between carrier liposomes and cells. Liposomes of varying compositions were added to the spheroids and incubated for 1–24 h. As shown in Figure 2, no significant changes in the structural integrity of the spheroids were observed after treatment with control GS-free liposomes. No signs of spheroid destruction were detected; they retained their rounded shape. The morphology of cells migrating from the spheroids was typical of the L9292 cell line, displaying a single well-defined nucleus with one or more nucleoli. The nuclei were round, homogeneous in structure, and showed no signs of fragmentation.

Structural (morphological) observations of spheroids treated with free GS revealed only minor, non-significant changes after 1 h of exposure at concentrations of 10 and 25 µg/mL. In contrast, treatment with GS at higher concentrations elicited pronounced toxic responses in L929 cells already after 1 h: cell morphology was changed and cell viability dropped to 34.8 ± 18.7% for 50 µg/mL GS (Table 5). A 24-h treatment with GS further significantly reduced cell viability in spheroids to 34.5 ± 15.8%, 18.1 ± 12.0%, 5.3 ± 5.6%, and 2.9 ± 3.8% following exposure to 10, 12, 50, and 75 µg/mL GS, respectively (Table 5).

2.5. Effects of GS-Loaded Liposomes on Cells in Spheroid Culture

Finally, we assessed the viability of L929 cells cultured as spheroids following treatment with liposome-encapsulated GS to evaluate whether liposomal delivery could mitigate GS-induced toxicity. Across all tested lipid compositions and GS concentrations, liposome-carried GS caused only minimal, non-significant effect on the cell viability (Table 6). For instance, liposomes composed solely of DPPC and loaded with 75 µg/mL GS reduced viability from 92.9 ± 6.1% to 84.8 ± 7.8%, without statistical significance. Liposomes of more complex composition (DPPC/CL/CHOL) at the same GS concentration showed also no viability reduction, with values of 89.2 ± 6.9% and 89.9 ± 0.8%, respectively (Table 6).

The study of GS effects at various concentrations on the viability of L929 cell lines revealed strong differences in cellular responses to the GS depending on culture conditions. Cells grown in monolayers exhibited high sensitivity to GS and cells forming spheroids showed greater resistance to GS. These results highlight the potential advantages of three-dimensional (3D) cultures over two-dimensional (2D) cultures in supporting cell survival under antimicrobial exposure, and highlight the relevance of culture format in future GS-related research across varying cellular microenvironments.

3. Discussion

Two-dimensional cell cultures are widely used in cytotoxicity studies due to their simplicity, but they have notable limitations, including a monolayer structure, restricted cell–cell interactions, and greater exposure to nutrients and oxygen than in vivo conditions [35,36]. The 3D models help address these limitations by better mimicking intercellular connections and communication. Consequently, spheroid-based 3D cultures provide more accurate predictions of pharmacological treatment outcomes than 2D models [37,38]. We applied both cellular models to evaluate the cytotoxic effects of the free and encapsulated GS.

In this study, we reported a significant reduction in the cytotoxicity of GS toward L929 fibroblast cells when GS was delivered in liposomal form. To support cell viability, we demonstrated that at 50 µg/mL, GS-liposomes with a more complex DPPC/CL/CHOL composition preserved 2D-cultured L929 cell viability at 79.0%, compared to just 11.0% with free GS. This pronounced difference suggests that liposomal encapsulation reduces direct cellular exposure to GS by altering its interaction with the cell membrane. The liposomal carrier likely promotes a diffuse kinetic of GS release, resulting in a slower, more homogeneous distribution of the active compound over time rather than an immediate high local concentration. Such a release profile may act as a protective barrier, mitigating acute cytotoxic effects while maintaining bioavailability. Zeta potential plays a crucial role in the activity of free antimicrobial peptides by governing their electrostatic interactions with target cells and membranes, whereas liposomal formulation markedly alters surface charge and, consequently, modulates AMP–target interactions, particularly during the initial contact phase. In the present study, the precise mechanisms underlying GS interactions with cellular membranes were not directly investigated; elucidating these processes will constitute an important focus of our future research.

At a GS concentration of 75 µg/mL, cell viability was 80.3% with liposomal GS, but drastically dropped to just 0.5% with free GS. This remarkable protection factor (160.6-fold) highlights the ability of liposomal encapsulation to mitigate the threshold-dependent cytotoxicity of free GS by attenuating its rapid membrane-disruptive effects [39]. Proliferative properties were assessed by measuring monolayer confluence (Table 4). At GS concentrations of 10–15 µg/mL, only a slight effect on cell adhesion and proliferation was observed across all liposome formulations, whereas higher GS concentrations led to a more pronounced reduction in cell confluency. These changes may indicate the onset of early cellular stress responses, potentially involving focal adhesion–associated proteins such as integrins, focal adhesion kinase, and paxillin, which are critical for cell–substrate attachment and mechanotransduction [40,41]. In parallel, alterations in actin cytoskeletal dynamics, including remodeling of F-actin, and modulation of actin-regulatory proteins such as cofilin or Rho GTPases, may contribute to the observed effects [42,43]. Even when GS is delivered in liposomal form, these responses could reflect transient changes in cell surface receptor expression, membrane lipid organization, or downstream signaling pathways (e.g., FAK–PI3K–Akt or MAPK signaling), ultimately affecting cell spreading and proliferation without inducing overt cytotoxicity [9,11,30,44]. These mechanisms represent promising avenues for future investigation and will be explored in subsequent studies.

The effects of GS on fibroblast spheroid models showed a similar trend. The pronounced toxicity of free GS was markedly attenuated when GS was delivered via liposomal carriers. Free GS resulted in cell viabilities of 34.5 and 18.1% at concentrations of 10 and 25 µg/mL, respectively, with near-complete loss of viability at 50 and 75 µg/mL. In contrast, liposomal GS provided strong cytoprotection across all tested concentrations. Notably, at 75 µg/mL, liposomal formulations restored cell viability to levels ranging from 84.8% to 94.7%, depending on lipid composition. The results obtained with liposomal formulations of GS do not increase its intrinsic antimicrobial activity but significantly enhance its practical therapeutic potential by expanding the safety margin. Liposomal encapsulation markedly reduces GS-induced cytotoxicity toward mammalian cells, the main limitation to its clinical application, thereby enabling the use of high antimicrobial concentrations that would otherwise be restricted by toxicity and improving the overall therapeutic index.

The exact mechanism of action of GS and its liposomal formulation remains to be fully elucidated. It is evident that the initial interaction of GS with the cell membrane triggers a cascade of structural changes [14,39], ranging from subtle regulatory alterations to rapid membrane disruption at higher GS concentrations as described in Introduction section. Liposomal encapsulation effectively reduces GS-induced cytotoxicity, likely by limiting direct membrane disruption. In addition, the effect of GS on Na^+^/K^+^-ATPase has been reported [45]. Although this effect is less pronounced than that observed for gramicidin A, it may nonetheless contribute to GS-induced cytotoxicity. GS can also affect mitochondrial membranes by increasing their permeability to alkali metal cations, leading to uncoupling of oxidative phosphorylation, altered respiration, and mitochondrial swelling, indicative of disrupted membrane integrity and ion gradients [46]. Such mitochondrial damage is plausibly accompanied by oxidative stress and lipid peroxidation, resulting in oxidative modification and functional impairment of proteins, as demonstrated for other gramicidins and membrane-active toxic compounds [47,48]. Importantly, such effects are likely attenuated in eukaryotic cells when GS is associated with liposomes and therefore represent a plausible mechanistic component that warrants further in-depth investigation. Note that GS release kinetics data are not currently available, as additional studies are required to assess different experimental conditions and liposome interactions with target membranes and cells.

Another critical aspect is the potential of GS to induce oxidative stress, or conversely its susceptibility to free radical mediated damage, which remains unclear and warrants further investigation. GS exposure may influence the intracellular balance of reactive oxygen species, potentially through effects on mitochondrial electron transport chain complexes or activation of NADPH oxidases (NOX1/NOX4). GS has been shown to interact with mitochondrial membranes [46], which may contribute to ROS production. Elevated ROS levels could promote lipid peroxidation of membrane, leading to the formation of reactive aldehydes such as malondialdehyde (MDA) and 4-hydroxynonenal (4-HNE), which are known to disrupt membrane integrity and protein function [49,50,51]. Moreover, some publications fail to specify the exact type of gramicidin used [52], which complicates the interpretation of results.

An emerging perspective in drug delivery emphasizes the repurposing of clinically approved compounds via innovative delivery platforms, optimizing their therapeutic index while circumventing conventional pharmacokinetic limitations. We investigated the effects of the known but only partially characterized antimicrobial peptide GS, administered in both free and liposome-encapsulated formats, on the morphofunctional properties of L929 fibroblast cells cultured under two-dimensional monolayer and three-dimensional spheroid conditions. Unlike two-dimensional cultures, spheroids known to exhibit greater resistance to antitumor agents [53]. This is due to the unique structure of spheroids, where the gradient of substances in the inner region differs from the peripheral one. In addition, the cells of the inner layers of the spheroid are in a dormant state, demonstrating a decrease in oxygen and nutrient consumption. As a result, a population of cells remains in the spheroid center, showing reduced sensitivity to toxic agents compared to peripheral cells. This phenomenon should be taken into account when interpreting research results, as central cells may remain intact, affecting survival outcomes in response to active pharmacological compounds. Since spheroids recapitulate key cellular states associated with tumor metastasis and dissemination, enhancing the sensitivity of spheroid-based cultures to antitumor agents represents a significant experimental and therapeutic challenge. This biological complexity must be carefully accounted for when designing, conducting, and interpreting studies aimed at evaluating treatment efficacy in three-dimensional cellular models.

In future studies, a more comprehensive investigation of DPPC/CL/CHOL formulations for GS delivery is warranted, as this lipid composition is attracting increasing interest for the delivery of various therapeutic agents, including antitumoral anthracyclines [44,54]. At the same time, the expanding scope of antimicrobial peptide applications underscores the growing need for effective delivery systems capable of mitigating cytotoxicity and improving therapeutic performance [7,8,55,56]. The reported results are particularly relevant for the development of safer antimicrobial formulations, including advanced topical applications, localized delivery systems, and, following further validation, potentially systemic administration. In addition, the improved safety profile may broaden the applicability of GS in biomedical contexts where membrane activity is desirable, such as antibiofilm strategies or combination therapies.

Additionally, incorporating specific agents—such as proteolytic enzymes or modulators of gap junctional intercellular communication—into the liposomal formulation appears promising for facilitating drug penetration and extracellular matrix degradation, which is actively synthesized by spheroid cells [57]. By improving GS diffusion through the matrix, this strategy may enhance intratumoral delivery and therapeutic efficacy. Moreover, such an approach could advance our understanding of liposome–spheroid interactions and support the development of more effective drug delivery strategies in three-dimensional cellular models.

4. Materials and Methods

All chemicals were sourced from Merck–Sigma-Aldrich (Darmstadt, Germany) and were of analytical grade or higher purity.

4.1. Preparation and Characterization of Gramicidin S–Containing Liposomes

Liposomes were prepared following the protocol described previously [58], with minor adjustments. Briefly, stock solutions were prepared by dissolving L-α-dipalmitoylphosphatidylcholine (DPPC), cardiolipin (CL), and cholesterol (CHOL) in ethanol at concentrations of 100 mg/mL, 5 mg/mL, and 10 mg/mL respectively. These lipids were then blended in ratios: 100% DPPC for pure DPPC liposomes; 95 mol% DPPC and 5 mol% CL for DPPC/CL; 90 mol% DPPC and 10 mol% CHOL for DPPC/CHOL liposomes, and 85 mol% DPPC, 5 mol% CL and 10 mol% CHOL for DPPC/CL/CHOL. To remove any residual organic solvents, the lipid mixture was first subjected to nitrogen drying and then placed under reduced pressure using a rotary vacuum evaporator (Eppendorf, Hamburg, Germany) for at least 3 h at 40 °C. This process ensured the formation of a consistent and well-dehydrated lipid film. The film was hydrated with Hanks’ Balanced Salt Solution (pH 7.4) to yield a liposomal suspension. Lipid membrane extrusion was performed at 45 °C, above the main phase transition temperature of the lipid (T_m_ approx. 41 °C), to ensure that the bilayer remained in the fluid (liquid-crystalline) phase during hydration and vesicle formation. Extrusion through polycarbonate membranes with a nominal pore size of 100 nm produced liposomes with mean diameters of 80–100 nm and a polydispersity index (PDI) of 0.24, as determined by dynamic light scattering, at a final lipid concentration of 10 mM in aqueous medium. Gramicidin S (GS, PubChem CID: 73357) was dissolved in ethanol and loaded into liposomes. The encapsulation efficiency was 86 ± 5% for the more complex DPPC/CL/CHOL lipid formulation and was comparable for the other lipid formulations under the temperature conditions used in this study. Encapsulation efficiency was determined as the percentage of GS loaded into the liposomes. GS concentration was quantified using a biological activity assay (see Section 4.2). Specifically, the concentration of GS in aliquots collected immediately before addition to the liposome suspension (prior to association) and in the final unbound GS fraction after completion of the association incubation was determined. Note, a systematic investigation of encapsulation efficiency across a broader range of formulation parameters and experimental conditions is currently underway in separated study to establish and further optimize GS loading.

For liposome-carried GS, drug release profiles were studied using the dialysis method in separate study. GS, when spontaneously released from the liposomes, diffused through a dialysis membrane into the release medium (L929 cell culture medium) at 37 °C under gentle stirring. Released GS was quantified using a biological activity assay. For the more complex DPPC/CL/CHOL (85/5/10 mol%) lipid composition, the spontaneous drug release into the medium after 1 h, 24 h, and 48 h was 8.7 ± 3.2%, 15.3 ± 4.2%, and 18.1 ± 6.1%, respectively.

The antibacterial efficacy of liposome-loaded GS was not the primary objective of this study; however, to verify that GS retained antibacterial activity, the minimum inhibitory concentration (MIC) was determined. The MIC values were 2 µg/mL for free GS and 8 µg/mL for DPPC/CL/CHOL liposome-associated GS against E. coli. In present study, the amount of GS encapsulated in liposomes was calculated to yield final concentrations of 10, 25, 50, and 75 µg/mL in the culture medium. All L929 cell treatments were normalized based on the encapsulated GS content, accounting for encapsulation efficiency, to ensure equivalent GS exposure between free and liposomal formulations. The dose range was selected based on the reported hemolytic concentrations of GS, which is approximately 10 µg/mL or higher (19, 22).

Precision during drying, rehydration, and incubation was maintained by weighing samples with a Mettler XP26 microbalance (Mettler-Toledo, Columbus, OH, USA). The zeta potential of liposomes with various compositions under different conditions was examined in our previous work [31]. In particular, the DPPC/CL/CHOL formulation used in this study and identified as the most promising exhibited a zeta potential of −28 ± 3 mV. Liposome stability during storage was investigated in a separate study [31], which demonstrated that the lipid structures remained stable for at least 10 days at 4 °C, while GS was retained within the liposomes in the corresponding cell culture medium for up to 5 days, with leakage below 5%. Nevertheless, in the present study, liposomes were used immediately following preparation.

4.2. Determination of GS Concentration by Biological Activity Assay

The concentration of GS was determined using a biological activity assay. A calibration curve was generated using GS standard solutions ranging from 0 to 200 µg/mL. Escherichia coli (ATCC 25922) was used as the indicator strain. Bacterial cultures were incubated with GS standards in Luria–Bertani (LB) medium at 37 °C with shaking at 180 rpm for defined time intervals (1, 2, 3, 4, 8, 16, and 24 h). After incubation, bacterial growth was evaluated using three independent readout methods: (i) optical density measurement by spectrophotometry at 600 nm (OD_600_); (ii) colony-forming unit (CFU) enumeration on agar plates; and (iii) imaging analysis, performed using a proprietary device (SmartAst, developed by Eltek S.p.A., Casale Monferrato, Alessandria, Italy). This system consists of a centrifuge equipped with integrated optics combined with a microfluidic device, enabling direct analysis of bacterial proliferation within a short time using very small sample volumes (a few microliters). As described in [59], bacterial growth becomes detectable by this device from the earliest duplication cycles, whereas the kinetics of growth inhibition are determined by the biological characteristics and mode of action of the antibacterial agent. The concordance and reproducibility of all three methods were verified prior to the start of the experiments.

4.3. Cultivation of L929 Cell Line in Mololayer and Spheroid Forms

L929 fibroblast cells were cultivated in DMEM/F12 medium (Dulbecco’s Modified Eagle Medium/Nutrient Mixture F-12; Biowest, Nuaillé, France), supplemented with 10% fetal bovine serum, along with 50 μg/mL each of penicillin and streptomycin (Biowest, France). Cultures were maintained at 37 °C in a humidified incubator under a 5% CO_2_ atmosphere. For monolayer formation, cells were seeded into plastic culture flasks (SPL Life Sciences, Republic of Korea) at an initial density of 2 × 10^5^ cells/mL and confluence was reached. To compare confluence values, the confluence of untreated control cell cultures was set as 100%.

Low attachment plates (Corning, CA, USA) were applied for spheroid formation. L929 fibroblast cells were seeded into 6-well plates at an initial density of 2–5 × 10^5^ cells/mL. Initial cell aggregation was observed in the first 3 days, followed by compaction and spheroid formation on the 5th day. Spheroids were cultured for 10–12 days. Half of the nutrient medium was changed every 3 days. Spheroids were collected by centrifugation, filtered through a mesh with a pore diameter of 200 μm to exclude large spheroids then filtered through a mesh with a pore diameter of 30 μm to remove single cells and small cell aggregates. The diameter of the resulting spheroids was 100–200 μm. The representative spheroids are shown in Figure 3. Note, oversized spheroids were excluded from the experimental analysis to ensure experimental consistency and reliability. It is described that when the spheroid size increases beyond a critical diameter, diffusion limitations for oxygen, nutrients, and test compounds become significant. Indeed, large spheroids often develop hypoxic or necrotic cores, which can markedly alter cellular metabolism, viability, and drug sensitivity. Moreover, limited drug penetration into the inner layers of oversized spheroids can lead to heterogeneous exposure, resulting in variable or underestimated treatment effects. This size-dependent heterogeneity increases experimental variability and compromises the comparability of results across samples. Therefore, filtering out large spheroids ensured a more uniform population with predictable diffusion properties, improved reproducibility, and more accurate interpretation of treatment-induced effects.

4.4. Cell Treatment by GS or Liposome-Carried GS

To evaluate the cytotoxic effects of free GS or GS delivered via liposomes, L929 cells were incubated for 1 or 24 h at 37 °C in a CO_2_-enriched atmosphere. Treatments included either free GS, within a concentration range of 10 to 75 µg/mL, or liposomal formulations containing GS at matching concentrations, prepared using various lipid compositions (see above). For each experimental condition, either 100 or 400 µL of liposome suspension in DMEM/F12 medium was added to each well. To ensure uniform distribution and enhance interaction between the liposomes and the cells, the plates were gently agitated at regular intervals.

4.5. Determination of Cell Viability

To assess cell viability, two fluorescent probes, fluorescein diacetate (FDA) and ethidium bromide (EB), commonly used in molecular biology were employed [60,61,62,63]. FDA is a nonpolar, nonfluorescent molecule that readily crosses the plasma membrane and is hydrolyzed by intracellular esterases in viable cells, releasing fluorescent, polar fluorescein. Fluorescein accumulates within living cells, interacts with intracellular components, and emits green fluorescence at 520 nm upon excitation at 488–490 nm. In contrast, EB penetrates only cells with compromised plasma membranes and binds to double-stranded nucleic acids. Upon DNA binding, ethidium bromide undergoes a shift in its excitation and emission spectra from ultraviolet absorption peaks at approximately 210 nm and 285 nm in solution to longer wavelengths, typically 300–360 nm for excitation and 485–526 nm for emission. Free ethidium bromide fluoresces at around 605 nm, whereas DNA-bound ethidium bromide exhibits a slight blue shift, emitting at approximately 590 nm, and can be efficiently excited at 543 nm. Stained samples were analyzed using a confocal laser scanning microscope (LSM 510 META, Carl Zeiss, Germany), with excitation at 488 nm for FDA and 543 nm for EB, and emission detection at 520 nm and 590 nm, respectively. Cell viability was expressed as a percentage and defined as the ratio of cells exhibiting only green fluorescence to the total number of cells, multiplied by 100%, based on the evaluation of at least 100 high-power fields (HPF). In selected experiments, the MTT assay was additionally performed to confirm the viability data. The viability percentages obtained by the MTT assay were statistically consistent with those determined using the HPF-based method.

For spheroids, to ensure high-quality quantitative analysis of cells within three-dimensional structures, Z-stacks (15 nm intervals) were acquired using a confocal microscope. A representative image is shown in Figure 4 for methodological illustration. The quantification was conducted with LSM Image Examiner (version 3.2, Carl Zeiss, Germany) software, where images in stacks were analyzed in the depth up to 100 µm. The percentage of dead cells was defined as the ratio of cells showing EB positivity, while the percentage of live cells was defined as the ratio of cells showing fluorescein diacetate (FDA) positivity, multiplied by 100%. Cell viability was calculated and presented in the Results as the percentage of live cells.

4.6. Determination of Cell Monolayer Confluence

The confluence of the hematoxylin–eosin-stained monolayer was assessed using a previously described method [25] by scanning the bottom of the well plate with an Epson Perfection V10 scanner (Epson, Nagano, Japan). The relative area occupied by the adherent cell monolayer was quantified using AxioVision Rel. 4.8 software (Carl Zeiss, Oberkochen, Germany) and expressed as a percentage.

4.7. Statistical Analysis

Data from five independent cellular and liposome preparations (biological replicates) are presented in the tables as mean ± standard deviation (SD). Statistical significance was assessed using the Mann–Whitney U test for pairwise comparisons or the Kruskal–Wallis test for overall group comparisons, performed with IBM SPSS Statistics for Windows (version 29.0; IBM Corp., Armonk, NY, USA). An asterisk (*) indicates statistically significant differences at p < 0.05.

4.8. Data Sharing Statement

The data supporting the findings of this study are available within the article.

5. Conclusions

In this study, we investigated the effects of the antimicrobial peptide GS, administered in both free and liposome-encapsulated formats, on the morphofunctional properties of L929 fibroblast cells cultured under two-dimensional monolayer and three-dimensional spheroid conditions. Our findings demonstrated a significant reduction in GS-induced toxicity when delivered via liposomal carriers, particularly in spheroid cultures. These results underscore the potential of liposomal delivery systems to enhance biocompatibility and modulate cellular responses across dimensional culture models that better mimic native tissue architecture.