Vancomycin-Loaded Isogenous Membrane Vesicles for Macrophage Activation and Intracellular Methicillin-Resistant Staphylococcus aureus Elimination

Introduction

Infectious diseases caused by bacteria, particularly multidrug-resistant (MDR) strains, pose a serious threat to human health.¹ Methicillin-resistant Staphylococcus aureus (MRSA) is a prominent MDR pathogen known for its rapid spread and ability to cause various infections, ranging from common wound infections to lethal pneumonia, osteomyelitis, and sepsis.² The resistance of MRSA to multiple clinically relevant drugs, such as methicillin, erythromycin, and clindamycin, complicates treatment and control efforts.³ Structure-based virtual screening has identified potential inhibitors targeting key S. aureus determinants, including the glycopeptide-resistance regulator GraR,⁴ methicillin-resistance-associated FmtA,⁵ GTPase YsxC,⁶ and lipophilic membrane,⁷ which may serve as leads for novel anti-MRSA agents. Notably, MRSA can subvert the innate immune system of macrophages and proliferate intracellularly, promoting chronic and recurrent infections.^8,9 Vancomycin (VAN) is considered as a last-resort drug for severe MRSA infections.¹⁰ However, the poor cellular penetration and rapid macrophage clearance of VAN limit its efficacy against intracellular pathogens.^9,11,12 Therefore, enhancing drug delivery to achieve high intracellular accumulation represents a promising strategy for eradicating persistent MRSA infections.

Nanodrug carriers substantially improve transmembrane permeability, intracellular accumulation, and antimicrobial activity of encapsulated drugs against intracellular infections caused by MDR pathogens.^13,14 A variety of nanodrug carriers, such as liposomes,¹⁵ poly(lactic-co-glycolic acid) particles,¹⁶ exosomes,¹⁷ and inorganic nanoparticles,¹⁸ have been developed to enhance drug delivery efficiency to macrophages. Among these, bacterial membrane vesicles (MVs), lipid-bound nanoparticles (20–400 nm in diameter) spontaneously released during bacterial growth, have emerged as attractive nanodrug carriers.^19,20 These MVs are enriched with proteins, peptidoglycans, DNA, and RNA, most of which are pathogen-associated molecular patterns (PAMPs) capable of targeting and activating immune cells, including macrophages, to kill intracellular pathogens.¹⁹ While MVs derived from Gram-negative bacteria such as Escherichia coli have been widely explored for antibiotic delivery,¹⁸ they frequently contain lipopolysaccharide (LPS), a membrane–bound endotoxin that contributes to bacterial pathogenicity.²⁰ In contrast, Gram-positive bacteria such as S. aureus produce LPS-free MVs with superior biocompatibility.²⁰ Additionally, S. aureus MVs demonstrate high macrophage uptake efficiency.²¹ Notably, multiple studies have revealed that MVs are favorably internalized by their homologous bacteria,^22–24 implying an intrinsic targeting capability of bacterial MVs.

In our previous study, we constructed an attenuated S. aureus strain, RN4220ΔagrA, and demonstrated the hypotoxicity of ^ΔagrAMVs to macrophages.²⁵ Herein, we hypothesized that isogenic ^ΔagrAMVs encapsulated with VAN would be favorably targeted and taken up by MRSA-infected macrophages, enhancing intracellular bacterial killing. To evaluate the nanodrug carrier potential of S. aureus MVs, we prepared VAN-loaded ^ΔagrAMVs (^ΔagrAMV-VAN). These nanoparticles exhibited a sustained-release profile and were sufficiently internalized by macrophages. Moreover, ^ΔagrAMV treatment promoted macrophage polarization toward the M1 phenotype, which facilitated the clearance of intracellular MRSA. ^ΔagrAMV-VAN demonstrated excellent safety, and treatment with ^ΔagrAMV-VAN effectively killed MRSA in macrophages in both in vitro and in vivo models. Collectively, this study provides a new design for isogenic nanomedicines that combines efficient drug delivery with potent antimicrobial activity against intracellular infections.

Materials and Methods

Bacterial Strains and Culture

The S. aureus strain RN4220ΔagrA was generated previously,²⁵ and the strain USA300 (FPR3757) was kindly provided by Dr. Min Li (Shanghai Jiao Tong University, China). S. aureus strains were cultured in brain heart infusion (BHI) medium (Oxoid, UK) or on BHI agar (BHIA) at 37 °C.

Preparation of ^ΔagrAMVs and ^ΔagrAMV-VAN

^ΔagrAMVs were prepared from the culture supernatant of S. aureus strain RN4220ΔagrA as previously described.^23,25 Briefly, a single colony of RN4220ΔagrA grown on BHIA was picked, inoculated into fresh BHI broth, and cultivated at 37 °C for 12 h with agitation. Subsequently, the culture was diluted 1:100 in 300 mL of fresh BHI broth and incubated at 37 °C for another 12 h. The supernatant containing ^ΔagrAMVs was collected by centrifugation at 6000 ×g for 10 min, followed by centrifugation at 10,000 ×g for 10 min, and filtered through a 0.22 μm membrane filter (Merck Millipore, USA) to remove dead cells and cellular debris. The resulting solution was concentrated by ultrafiltration using a 100 kDa hollow fiber membrane column (GE Healthcare, USA) to enrich ^ΔagrAMVs. After ultracentrifugation at 200,000 ×g with a rotor (HITACHI, Japan) for 3 h, the ^ΔagrAMV pellets were resuspended in 4 mL of 50% (v/v) Optiprep density gradient solution (Alere Technologies AS, Norway), followed by sequential layering with 2 mL of 40% Optiprep, 2 mL of 20% Optiprep, and 1.5 mL of 10% Optiprep. After centrifugation at 200,000 ×g for 3 h at 4°C, the solution was divided into five fractions from top to bottom, transferred into sterile tubes, and analyzed by SDS-PAGE (Servicebio, China). The ^ΔagrAMVs located between the 20% and 40% Optiprep layers were carefully extracted, concentrated using an ultrafiltration tube (Millipore), resuspended in phosphate-buffered saline (PBS), and stored at −80 °C until use.

VAN-loaded ^ΔagrAMVs (^ΔagrAMV-VAN) were prepared as previously described.²⁶ In brief, VAN (Sigma-Aldrich, USA) and ^ΔagrAMVs were mixed at mass ratios of 1:1, 1:2, or 2:1 and subjected to sonication using a UP-50H Ultrasonic cell disruptor (Hielscher, Germany) at 30% power for six cycles (4 s pulse, 2 s pause). The mixture was then incubated at 37 °C for 1 h to allow ^ΔagrAMVs reformation. Free VAN was removed by ultrafiltration using an Amicon^® Ultra-centrifugal filter (MWCO = 100 kDa, GE Healthcare, UK). The purified ^ΔagrAMV-VAN was aseptically filtered through a 0.22 μm syringe filter (Millipore) and stored at −80°C.

Characterization of ^ΔagrAMVs and ^ΔagrAMV-VAN

^ΔagrAMV samples were dropped onto copper grids, allowed to sediment naturally for15 min,negatively stained with 2% (v/v) phosphotungstic acid for 15 s, and air-dried for 1 h.²⁰ The morphology of ^ΔagrAMVs and ^ΔagrAMV-VAN was observed using transmission electron microscopy (TEM; HT7700, Hitachi, Japan). The particle size distribution and zeta potential of ^ΔagrAMVs and ^ΔagrAMV-VAN were measured by dynamic light scattering (DLS) using a Nanoparticle Size Analyzer (Zatasizer Nano ZSP, Malvern, USA).

VAN Release

High-performance liquid chromatography (HPLC) was performed to quantify ^ΔagrAMV-VAN as previously reported.²⁶ Briefly, freshly prepared ^ΔagrAMV-VAN was enclosed in a dialysis bag (MWCO:100 kDa; Ruiswbio, USA) and incubated at 37 °C with shaking. At each time point, 1 mL aliquots were collected and analyzed by HPLC. Each aliquot was mixed with an appropriate volume of acetonitrile in a centrifuge tube. After sonication, the supernatant was collected by centrifugation at 18,000 ×g for 10 min, filtered through a 0.22 μm syringe filter, and transferred to an HPLC auto-sampler. Approximately 1 mL of the sample was injected into the HPLC system (Agilent 1260, Agilent Technologies, CA) equipped with a C18 column (extended-C18, 250 mm × 4.6 mm, 5 μm, 100 Å, Agilent). The mobile phase (KH₂PO₄ :acetonitrile, 90.5: 9.5, v/v, pH = 3.2) was used at 30 °C with a flow rate of 1 mL/min. VAN elution was monitored at 236 nm, and a standard curve was established using VAN concentrations ranging from 3.125 to 50 µg/mL with OpenLAB CDS ChemStation Edition software. The release profile of VAN from ^ΔagrAMV-VAN was evaluated in PBS (pH 7.4).

Direct Killing of MRSA by ^ΔagrAMV-VAN

The bactericidal activity of ^ΔagrAMV-VAN against planktonic MRSA was determined as previously described.²⁷ Briefly, overnight cultured MRSA USA300 was diluted with PBS to 1×10⁶ colony-forming units per milliliter (CFU/mL), and 100 μL of the bacterial suspension was added to each well of a 96-well plate. Next, 100 μL of PBS, VAN (5 μg/mL), ^ΔagrAMVs (25 μg/mL), ^ΔagrAMVs (25 μg/mL) plus VAN (5 μg/mL), or ^ΔagrAMV-VAN (25 μg/mL) was added to each well. Afterward, the plates were incubated at 37 °C for 24 h. At 0, 2, 6, 12, and 24 h post-incubation, 10 μL aliquots were collected from each well for CFU enumeration. After 24 h of treatment, bacterial cells were collected and stained using a LIVE/DEAD BacLight Bacterial Viability kit (APExBIO, USA). Cell viability was assessed using confocal laser scanning microscopy (LSM880, Zeiss, Germany).

In vitro Cellular Uptake of ^ΔagrAMVs and ^ΔagrAMV-FITC-VAN

Mouse RAW264.7 macrophages (HaiStar Biotech, Beijing, China; Cat. No. TCM-C766) were cultured in Dulbecco’s Modified Eagle Medium (DMEM; Zeta Life, USA) supplemented with 10% fetal bovine serum (FBS; Thermo Fisher Scientific, USA) and 1% (v/v) penicillin-streptomycin solution (Zeta Life, USA) at 37 °C with 5% (v/v) CO₂. For cell uptake experiments, RAW264.7 cells (1 × 10⁵ cells/well) were seeded in confocal dishes (In Vitro Scientific, China) and incubated overnight at 37 °C. ^ΔagrAMVs, liposomes (Abnova, China), and ^ΔagrAMVs pretreated with proteinase K (2 μg/mL; Beyotime, China) were labeled with 30 μM DiO dye (Dalian Meilun Biotech, China) at 37 °C for 90 min. Cells were then treated with DiO-labeled ^ΔagrAMVs, liposomes, or proteinase K-treated ^ΔagrAMVs (PK; 5 μg per dish). After 90 min of incubation at 37 °C with 5% (v/v) CO₂, the culture medium was removed, and cells were washed twice with PBS. Cells were fixed with 500 μL of 4% paraformaldehyde (Biosharp, China) at room temperature for 15 min, followed by nuclear staining with 200 μL of 4’,6-diamidino-2-phenylindole (DAPI; Dowobio, China) for 10 min. After washing twice with PBS (4 min each), cells were observed under a super-resolution laser scanning confocal microscope (LSM880, Zeiss, Germany). DiO (green fluorescence) and DAPI exhibited maximum excitation wavelengths of 484 nm and 358 nm, respectively.

FITC-labeled VAN (FITC-VAN; Ruixibio, Xi’an, China) was used to prepare ^ΔagrAMV-FITC-VAN formulation at a 1:2 mass ratio. For uptake experiment, RAW264.7 macrophages were treated with either FITC-VAN (3 μg/dish) or ^ΔagrAMV-FITC-VAN (15 μg/dish). Fluorescence intensity was detected by confocal laser scanning microscopy (CLSM).²⁰

Inflammatory Factor Detection

RAW264.7 macrophages were seeded in 24-well plates at a density of 1×10⁵ cells/well and incubated overnight at 37 °C with 5% (v/v) CO₂. The following day, cells were treated with VAN (5 μg/mL), liposomes (5 μg/mL), or ^ΔagrAMVs (5 μg/mL). An equal volume of PBS was used as a negative control. After 12 h of treatment, culture supernatants were collected, and the levels of inflammatory factors TNF-α and IL-6 were detected using an enzyme-linked immunosorbent assay (ELISA) kit (MEIMIAN, China).²⁰

Macrophage Polarization Analysis

RAW264.7 macrophages cultured in confocal dishes were treated with ^ΔagrAMVs (5 μg/mL) for 12 or 24 h. Untreated cells were used as blank controls, while PBS-treated cells served as negative controls. After fixation with 4% (v/v) paraformaldehyde, macrophage polarization was assessed using anti-mouse CD86 antibodies (Multisciences, China). CD86⁺ cells, representing M1-polarized macrophages, were analyzed by flow cytometry (Challenbio, China) and CLSM (LSM880) as described.^28,29

MRSA Killing by ^ΔagrAMV-Activated Macrophages

Methicillin-resistant S. aureus (MRSA) infection was performed as previously described.³⁰ For assessing ^ΔagrAMV-mediated macrophage killing of MRSA, RAW264.7 cells were seeded in 24-well plates (1 × 10⁵ cells/well) and cultured overnight at 37 °C with 5% CO₂. Cells were then infected with MRSA USA300 at a multiplicity of infection (MOI) of 15 (approximately 5×10⁶ CFU/well). After 2 h of infection, extracellular bacteria were eliminated by treatment with 20 μg/mL lysostaphin (Sigma-Aldrich, USA) for 20 min. Afterward, fresh DMEM medium supplemented with VAN (2.5 μg/mL), liposomes (2.5 μg/mL), or ^ΔagrAMVs (2.5 μg/mL) was added, followed by incubation for an additional 12 h. PBS was used as a negative control. Intracellular bacterial counts were determined by plate assay.¹⁸ In brief, infected cells were collected by centrifugation at 6000 ×g for 10 min at 4 °C. Cell pellets were treated with 15 µg/mL lysostaphin for 20 min at room temperature, washed thrice with PBS, and lysed with Hank’s Balanced Salt Solution (Servicebio, China) containing 0.2% (v/v) Triton X-100 for 10 min. Subsequently, intracellular CFUs were quantified using BHI agar plate dilution assays.

For evaluating ^ΔagrAMV-activated macrophage killing of MRSA, RAW264.7 cells were pretreated for 2 h with VAN, liposomes, or ^ΔagrAMVs at various concentrations (0, 2.5, 5.0, and 10.0 μg/mL). Following treatment, MRSA infection was conducted as above. After lysostaphin treatment (20 μg/mL) to eliminate extracellular bacteria and replacement with fresh DMEM, intracellular MRSA counts were detected 12 and 24 h post-infection.

Intracellular MRSA Killing by ^ΔagrAMV-VAN

RAW264.7 cells were cultured in 24-well plates and infected with MRSA USA300 at an MOI of 20 as described above. The infected cells were then cultured in fresh DMEM supplemented with ^ΔagrAMV-VAN (25 μg/mL). Control groups included MRSA-infected cells treated with PBS, VAN alone (5 μg/mL), ^ΔagrAMVs alone (25 μg/mL), and ^ΔagrAMV (25 μg/mL) + VAN (5 μg/mL). Intracellular bacterial loads were counted at 12 and 24 h post-infection.

In vivo Anti-MRSA Effect of ^ΔagrAMV-VAN

Female BALB/c mice (6–8 weeks old) were obtained from Chongqing Byrness Weil Biotechnology Co., Ltd. (Chongqing, China). All animal experiments were approved by the Laboratory Animal Welfare and Ethics Committee of Army Medical University (Approval No. AMUWEC2020735).

A murine abdominal infection model was established as previously described.¹⁸ Briefly, 20 BALB/c mice were intraperitoneally inoculated with 5×10⁷ CFU of MRSA USA300 and randomly allocated into five groups: (1) PBS (negative control), (2) VAN (10 mg/kg; positive control), (3) ^ΔagrAMVs (6 mg/kg), (4) ^ΔagrAMV+VAN (equivalent individual doses), and (5) ^ΔagrAMV-VAN (6 mg/kg). After 24 h of treatment, mice were anesthetized with 1% (w/v) pentobarbital sodium, followed by peritoneal lavage with 3 mL sterile PBS. All animals were euthanized on day 2 post-infection. Peritoneal lavage fluid was centrifuged at 6000 ×g for 10 min at 4 °C to pellet cells. Bacterial loads in both the supernatant (extracellular) and cell lysates (intracellular) were counted separately.

Safety Assessment

RAW264.7 cells were used to assess the cytotoxicity of ^ΔagrAMV-VAN using the 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) assay as previously described.¹⁸ In brief, RAW264.7 macrophages were seeded in 96-well plates (2 × 10³ cells/well) and cultured for 24 h at 37 °C with 5% (v/v) CO₂. The culture medium was replaced with fresh medium supplemented with different concentrations of ^ΔagrAMVs (0, 25, 50, 100, and 200 μg/mL). Cell viability was detected 2, 6, 12, and 24 h after treatment using a Cell Counting Kit (Zeta Life).

A mouse abdominal infection model was generated (n = 3 per group) and treated with PBS, VAN (10 mg/kg), ^ΔagrAMVs (6 mg/kg), or ^ΔagrAMV-VAN (6 mg/kg). After 24 h of treatment, MRSA-infected mice were sacrificed, and organs (including the heart, liver, spleen, lungs, and kidneys) were harvested and fixed with 4% (v/v) paraformaldehyde. Histopathological examination was then performed on organ sections using hematoxylin and eosin (H&E) staining. Serum IL-6 and TNF-α levels at 6 h post-treatment were determined using an ELISA kit (MEIMIAN).

Statistical Analysis

All experiments were repeated at least three times, and the data are expressed as the mean ± standard deviation (SD) or standard error of the mean (SEM). Statistical comparisons were performed using GraphPad Prism 9.5 program (GraphPad Software, USA). The significance of differences between groups was assessed using Student’s t-test, one-way analysis of variance (ANOVA), or two-way ANOVA. A P-value of < 0.05 was considered statistically significant.

Results and Discussion

Preparation and Characterization of ^ΔagrAMVs and ^ΔagrAMV-VAN Particles

The quorum-sensing accessory gene regulator (agr) is a major factor that controls S. aureus virulence.^31,32 We previously showed that agr deletion remarkably decreased the virulence of MVs in S. aureus strain RN4220.²⁰ Therefore, we used the engineered RN4220ΔagrA as the starting strain.^{25 ΔagrA}MVs were prepared from the culture supernatant of RN4220ΔagrA. After density gradient centrifugation, the ^ΔagrAMVs was verified by SDS-PAGE, and were found to be mainly located between 20% and 40% Optiprep (Figure S1). TEM analysis exhibited the nanoscale vesicles of ^ΔagrAMVs (Figure 1a). The protein concentration of ^ΔagrAMVs was determined (Figure S2), and VAN-loaded ^ΔagrAMVs were prepared using an ultrasonic method as previously described.¹⁸ VAN-loaded liposomal formulations often have low encapsulation efficiencies, ranging from 0.1% to 9%.¹¹ To achieve high drug-loading efficiency, ^ΔagrAMVs:VAN mass ratios of 1:1, 1:2, and 2:1 were used. The results indicated that the ^ΔagrAMVs:VAN ratio of 1:2 achieved the highest VAN loading ratio of 21.6%, which was significantly higher than those at 1:1 (6.3%) and 2:1 (2.0%; P < 0.001; Figure 1b). TEM observations demonstrated that ^ΔagrAMV-VAN maintained its spherical shape (Figure 1c). DLS measurements indicated that the mean size of ^ΔagrAMVs was 71 nm, whereas the particle size of ^ΔagrAMV-VAN increased to 101 nm (Figure 1d). Xie et al¹⁸ found that the size of VAN-loaded and polydopamine-encapsulated mesoporous SiO₂ nanoparticles increased from 106 to 122 nm after coating with E. coli MVs. This phenomenon indicated the versatile potential of bacterial MVs as drug carriers.

Figure 1 Characterization of ΔagrAMVs and ΔagrAMV-VAN. (a) Transmission electron microscopy (TEM) image showing the spherical morphology of ΔagrAMVs (scale bar: 200 nm). (b) VAN loading efficiency with various ΔagrAMVs:VAN mass ratios. (c) TEM image of ΔagrAMV-VAN nanoparticles. (d) Particle size distribution of ΔagrAMVs and ΔagrAMV-VAN detected by dynamic light scattering (DLS). (e) PDI values of ΔagrAMVs and ΔagrAMV-VAN detected by DLS. (f) Zeta potential of ΔagrAMVs and ΔagrAMV-VAN measured by DLS. (g) In vitro VAN release profile from ΔagrAMV-VAN in neutral PBS, quantified by HPLC. Data are shown as mean ± SD (n = 3). Statistical significance was calculated by Student’s t-test, ** P < 0.01, and *** P < 0.001.

Polydispersity index (PDI) determination revealed that ^ΔagrAMV-VAN had a decreased PDI from 0.6 (^ΔagrAMVs) to 0.38 (Figure 1e), along with an increase in the absolute value of the membrane surface potential (Figure 1f). These data indicate that the uniformity of ^ΔagrAMV particles improved after loading VAN under our experimental conditions. HPLC was performed to determine the VAN content in ^ΔagrAMV-VAN according to a standard curve (Figure S3). The results showed that approximately 80% of the VAN was released from ^ΔagrAMV-VAN within 8 h under neutral conditions (Figure 1g). Overall, the prepared ^ΔagrAMV-VAN maintained a spherical, nanosized shape with relative homogeneity compared to ^ΔagrAMVs, and the VAN-loaded ^ΔagrAMV particles exhibited a sustained release effect.

MRSA Killing Effect of ^ΔagrAMV-VAN In Vitro

To address the resistance of S. aureus, many structure-based virtual screening methods have been developed to target antibiotic-resistance-associated determinants such as GraR and FmtA.^4,5 The screened compounds or inhibitors are potential antimicrobial agents against MRSA after molecular dynamics, pharmacophore modeling, and binding free energy analyses.^4–7 However, the application of these identified molecules against MRSA infections can only be realized after further experimental evaluation. To assess the anti-staphylococcal effect of ^ΔagrAMV-VAN, agar plate counting was performed as previously described.¹⁰ As shown in Figure 2a, the minimal inhibitory concentration (MIC) of ^ΔagrAMV-VAN against MRSA USA300 was 12.5 μg/mL. Approximately 1×10⁶ CFUs of S. aureus USA300 cells were treated with 10 μg/mL VAN, 50 μg/mL ^ΔagrAMVs, ^ΔagrAMV+VAN (10 μg/mL VAN + 50 μg/mL ^ΔagrAMVs), or 50 μg/mL ^ΔagrAMV-VAN. Plate counting was conducted at 0, 2, 6, 12, and 24 h post-culture, and the results indicated that ^ΔagrAMVs exhibited no bacteriostatic activity, showing bacterial counts similar to those of the PBS control (Figure 2b). However, the VAN-loaded ^ΔagrAMV formulation (^ΔagrAMV-VAN) showed antimicrobial activity comparable to that of the free VAN and the ^ΔagrAMV+VAN combination, except at the 6-h time point, where the ^ΔagrAMV-VAN treatment group showed a slightly higher survival rate than the free VAN and ^ΔagrAMV+VAN groups (Figure 2c). We hypothesized that the slow-release effect of ^ΔagrAMV-VAN might account for this phenomenon. Notably, complete elimination of MRSA USA300 at 12 h post-incubation was achieved in the VAN, ^ΔagrAMV+VAN, and ^ΔagrAMV-VAN groups (Figure 2b and c).

Figure 2 In vitro antibacterial activity of ΔagrAMV-VAN. (a) Minimum inhibitory concentration (MIC) of ΔagrAMV-VAN against MRSA USA300. (b) Growth inhibition of MRSA USA300 by ΔagrAMV-VAN. Approximately 1×106 CFU of bacteria were treated with 50 μg/mL ΔagrAMV-VAN for diverse times as indicated, and the grown bacteria on agar plates were shown. The treatment of PBS, free VAN (10 μg/mL), ΔagrAMVs (50 μg/mL), or ΔagrAMV+VAN combination (10 μg/mL VAN + 50 μg/mL ΔagrAMVs) served as controls. (c) The time-killing kinetics of ΔagrAMV-VAN against MRSA USA300. (d) CLSM images of MRSA USA300 after 24 h treatment with ΔagrAMV-VAN, stained with LIVE/DEAD BacLight Bacterial Viability Kit. Green fluorescence represents live bacteria, while red one indicates dead cells. Scale bars: 20 μm.

CLSM, equipped with a LIVE/DEAD cell staining kit, allowed direct visualization of bacterial inactivation, where live bacteria (labeled with calcein-AM) appeared green, and dead cells (stained with propidium iodide, PI) appeared red.³³ As shown in Figure 2d, treatment with VAN, ^ΔagrAMV+VAN, and ^ΔagrAMV-VAN considerably increased red fluorescence (dead cells) compared to the PBS and ^ΔagrAMV groups. This finding was consistent with the plate counting results (Figure 2b), indicating that the encapsulation of VAN in S. aureus ^ΔagrAMVs did not compromise the antibacterial activity of VAN.

Cellular Uptake of ^ΔagrAMV Nanoparticle by Macrophages in vitro

Studies have shown that MVs derived from S. aureus can be efficiently taken up by macrophages, a process mediated by motor proteins on the MV surface.³³ While primary macrophages derived from human peripheral blood mononuclear cells may offer greater physiological relevance for modeling intracellular infections,³⁴ stable macrophage cell lines such as human THP-1 and murine RAW264.7 have been widely utilized in developing intracellular infection models.^35–37 To evaluate the potential of ^ΔagrAMVs as a vehicle for intracellular drug delivery, we used the RAW264.7 cell line to assess the uptake efficiency of ^ΔagrAMVs by macrophages. Commercial liposomes, which are established drug delivery carriers,¹⁵ served as controls. DiO-labeled ^ΔagrAMVs or liposomes were co-incubated with RAW264.7 cells, and the green fluorescence intensity within macrophages was visualized using CLSM (Figure 3a). The fluorescence intensity was significantly higher in the ^ΔagrAMV group than in the liposome group (P < 0.01; Figure 3b). However, uptake was inhibited when ^ΔagrAMVs were pretreated with proteinase K (PK) for 1 h (^ΔagrAMVs (PK)), suggesting that this phenomenon may be attributed to the degradation of surface-associated proteins on ^ΔagrAMVs.³³

Figure 3 Cellular uptake of ΔagrAMV nanoparticles by macrophages. (a) Protease K (PK) pretreatment decreased macrophage internalization of ΔagrAMVs. Liposomes served as negative controls. Green signals within cells represent internalized ΔagrAMVs, and blue ones indicate cell nuclei (40× magnification), Scale bars: 10 μm. (b) Quantitative analysis of ΔagrA MV uptake. Fluorescence intensity was measured in 149 liposome-treated cells, 142 ΔagrAMVs-treated cells, or 122 ΔagrAMVs (PK)-treated cells. (c) Enhanced cellular uptake of VAN through ΔagrAMV encapsulation. Scale bars: 10 μm. (d) Quantitative analysis of FITC-VAN uptake. Fluorescence intensity was measured in 173 FITC-VAN-treated cells or 190 ΔagrAMV-FITC-VAN-treated cells. Data are shown as mean ± SEM. Statistical significance was measured by Student’s t-test, *P < 0.05, **P < 0.01, and *** P< 0.001.

To determine whether VAN loading affected ^ΔagrAMV uptake efficiency, we examined the internalization of ^ΔagrAMV-VAN nanoparticles by RAW264.7 cells. Fluorescein isothiocyanate (FITC)-labeled VAN (FITC-VAN) was used to generate the nanomaterial ^ΔagrAMV-FITC-VAN. CLSM revealed that the uptake of ^ΔagrAMV-FITC-VAN by RAW264.7 macrophages was remarkably enhanced compared to free FITC-VAN (Figure 3c). Quantitative fluorescence analysis confirmed that the ^ΔagrAMV-FITC-VAN formulation significantly improved VAN uptake efficiency (P < 0.001; Figure 3d). Overall, these results indicate that ^ΔagrAMV-VAN can efficiently deliver VAN to macrophages. Given that high intracellular drug concentrations are crucial for killing intracellular pathogens,¹² the enhanced internalization of VAN mediated by ^ΔagrAMVs may facilitate the clearance of intracellular MRSA within macrophages.

Effect of ^ΔagrAMVs on Macrophage Polarization

Macrophages exhibit two activation phenotypes in response to external stimuli: M1 and M2, which are closely associated with inflammatory responses.^28,38 M2 macrophages characterized by an altered metabolic state or impaired bactericidal activity, can provide a niche for the long-term survival of intracellular bacteria, contributing to persistent and recurrent infections.^3,39 The MVs secreted by bacteria contain diverse biomolecules, such as proteins, lipids, nucleic acids, and metabolites, which play crucial roles in bacterial physiology like resistance transfer.^19,40 Studies have shown that bacterial MVs containing various PAMPs can modulate macrophage activation.^41–43 To determine the effect of S. aureus RN4220ΔagrA-derived ^ΔagrAMVs on macrophage activation, we treated RAW264.7 cells with ^ΔagrAMVs for 12 h. As shown in Figure 4a and b, ^ΔagrAMV treatment considerably increased TNF-α and IL-6 levels compared to the PBS, VAN, and liposome groups, which displayed similar cytokine profiles. The increased TNF-α and IL-6 levels suggest M1 polarization of macrophages, a phenotype associated with enhanced bacterial clearance.^38,44

Figure 4 Effect of ΔagrAMVs on macrophages polarization. (a and b) Cytokine levels of (a) TNF-α and (b) IL-6 secreted from RAW264.7 macrophages after 12 h treatment with PBS, VAN, Liposome, and ΔagrAMVs. (c and d) Flow cytometric analysis of M1 polarization: (c) Representative flow cytometry plots of CD86+ macrophages, (d) Quantification of CD86+ cell populations post-treatment. (e and f) Immunofluorescence analysis: (e) CLSM images of CD86+ macrophages (red: CD86, blue: DAPI; Scale bars: 10 μm), and (f) Quantitative fluorescence intensity of CD86 staining. All experiments were performed in triplicate. Data are shown as mean ± SD (cytokine/flow data) or SEM (fluorescence intensity). The difference was analyzed by one-way or two-way ANOVA. *P < 0.05, **P < 0.01, *** P< 0.001, and ns represents no significance.

CD86, a hallmark of M1 macrophages,² was analyzed by flow cytometry. Treatment with S. aureus ^ΔagrAMVs for 12 or 24 h substantially increased the proportion of CD86⁺ macrophages compared to the untreated and PBS control groups (Figure 4c and d), further supporting M1 polarization of RAW264.7 cells. This finding was further confirmed by immunofluorescence staining (Figure 4e and f). Collectively, these results demonstrate that S. aureus RN4220ΔagrA-derived ^ΔagrAMVs induce M1 macrophage polarization, thereby facilitating intracellular bacterial clearance.

Effect of ^ΔagrAMVs on MRSA Killing by Macrophages

Since M1 macrophages promote pathogen clearance,^2,38 we investigated the effect of ^ΔagrAMVs on macrophage antibacterial activity. RAW264.7 macrophages were infected with MRSA USA300 at an MOI of 15. After 2 h of infection, cells were treated with PBS, VAN, liposomes, or ^ΔagrAMVs for 12 h. Plate counting revealed a remarkable reduction in intracellular MRSA load in macrophages treated with ^ΔagrAMVs compared to the PBS control (Figure 5a). Although VAN treatment also considerably reduced MRSA burden in macrophages, liposomes did not contribute to MRSA killing by macrophages. These data indicate that ^ΔagrAMVs promote macrophage-mediated clearance of intracellular MRSA.

Figure 5 Enhancement of MRSA clearance in macrophages by ΔagrAMVs. (a) Promotion of macrophage intracellular bacterial clearance by ΔagrAMVs. RAW264.7 cells were infected with MRSA USA300 at an MOI of 15, followed by treatment with PBS, VAN, liposomes, or ΔagrAMVs after 2 h post-infection. Representative images of surviving bacterial colonies on BHI agar plates after 12 h of treatment (up panel), and quantitative analysis of bacterial counts (bottom panel). (b) Protective effect against MRSA infection. RAW264.7 cells were pretreated for 2 h with PBS, VAN, liposomes, or ΔagrAMVs, and then infected with USA300. Intracellular MRSA load was counted by plate assay after 12 h infection. (c) Dose-dependent protection by ΔagrAMVs. RAW264.7 cells were pretreated for 2 h with various concentrations of ΔagrAMVs before MRSA infection. Bacterial counts measured after 12 h infection. All experiments were repeated three times. Data are shown as mean ± SD. Statistical significance was calculated by one-ANOVA, **P < 0.01, *** P< 0.001, and ns represents no significance.

Next, we determined whether ^ΔagrAMVs could prevent MRSA infection in macrophages. RAW264.7 cells were pretreated with PBS, VAN, liposomes, or ^ΔagrAMVs for 2 h prior to MRSA infection. After 12 h of infection, intracellular bacteria were counted using a plate assay.^{45 ΔagrA}MV pretreatment significantly decreased intracellular MRSA burden compared to PBS (P < 0.001; Figure 5b), with dose-dependent protection observed (Figure 5c). In contrast, neither VAN nor liposomes conferred prophylactic effects against MRSA infection (Figure 5b). These results indicate that S. aureus ^ΔagrAMVs protect macrophages from MRSA infection and that ^ΔagrAMV-activated macrophages efficiently eliminate intracellular MRSA.

Effect of ^ΔagrAMV-VAN on Intracellular MRSA Killing

Given the macrophage-activating effect of S. aureus ^ΔagrAMVs, we determined whether ^ΔagrAMV encapsulation enhances the efficacy of VAN against intracellular MRSA in macrophages. RAW264.7 cells were seeded at 1×10⁵ cells/well and cultured overnight in 5% CO₂. Cells were then infected with S. aureus USA300 at an MOI of 20 for 2 h, followed by treatment with PBS (control), VAN alone, ^ΔagrAMVs alone, ^ΔagrAMV+VAN, or ^ΔagrAMV-VAN for 12 or 24 h. As shown in Figure 6a–c, ^ΔagrAMV-VAN treatment considerably reduced intracellular MRSA number compared to VAN alone at both time points (P < 0.01 or 0.001). While ^ΔagrAMVs and VAN combination also remarkably decreased intracellular MRSA burden, the ^ΔagrAMV-VAN group showed the lowest bacterial survival rate (Figure 6b and c), indicating that VAN encapsulation in ^ΔagrAMVs synergistically promotes bacterial clearance. This effect may be attributed to either ^ΔagrAMV-induced macrophage activation or improved intracellular VAN delivery via ^ΔagrAMVs, both of which could augment bactericidal activity against intracellular MRSA.^18,38 Notably, the specific mechanism of MRSA neutralization was not elucidated. A recent study revealed that E. coli-derived MVs are preferentially internalized by E. coli and other Gram-negative bacteria like Klebsiella pneumoniae, but not by Gram-positive bacteria like S. aureus,³² highlighting the homotypic targeting capability of bacterial MVs.³¹ We hypothesize that VAN-loaded isogenous MVs may selectively target MRSA-infected macrophages, thereby improving intracellular bacterial clearance. Overall, the ^ΔagrAMV-VAN formulation showed potent intracellular bactericidal activity against MRSA, indicating significant potential as a therapeutic strategy for MRSA-refractory infections.

Figure 6 ΔagrAMV-VAN eliminates intracellular MRSA in vitro and in vivo. (a) Survival of USA300 on BHI plates. RAW264.7 macrophages were infected with MRSA USA300 (MOI = 20) for 2 h, then treated with PBS, VAN, ΔagrAMVs, ΔagrAMV+VAN, or ΔagrAMV-VAN for 12 or 24 h. Intracellular MRSA was quantified by plate counting at: (b) 12 h and (c) 24 h post-treatment. (d) In vivo bacterial clearance. BALB/c mice (n = 4 per group) were intraperitoneally challenged with 5×107 CFU of MRSA USA300, then treated intraperitoneally with PBS, VAN, ΔagrAMVs, ΔagrAMV+VAN, or ΔagrAMV-VAN. Peritoneal fluids were collected after 24 h for quantification of: (e) Extracellular and (f) Intracellular bacterial loads. Data are shown as mean ± SD. Statistical significance was measured by one-way ANOVA, *P < 0.05, **P < 0.01, *** P< 0.001, and ns represents no significance.

In vivo Anti-MRSA Efficacy of ^ΔagrAMV-VAN

The persistent intracellular presence of S. aureus is a major contributor to recurrent infections,⁵ primarily due to the limited ability of antibiotics to penetrate host cells.¹² Our results demonstrated that encapsulation of VAN in ^ΔagrAMVs increased its cellular internalization (Figure 3c and d) and promoted intracellular MRSA clearance by RAW264.7 macrophages (Figure 6a–c). To evaluate the in vivo antibacterial efficacy of ^ΔagrAMV- VAN, we established a mouse model of abdominal infection as previously described.¹⁷ Peritoneal fluid samples were collected 24 h post-treatment with PBS, VAN, ^ΔagrAMVs, ^ΔagrAMV+VAN, or ^ΔagrAMV-VAN (n = 4 per group). As shown in Figure 6d, both VAN and ^ΔagrAMV+VAN treatments substantially reduced extracellular bacterial loads compared to PBS (P < 0.001). While ^ΔagrAMV-VAN also decreased extracellular bacteria (P < 0.01), its efficacy was lower than VAN alone or the physical mixture (Figure 6e). Notably, the ^ΔagrAMV+VAN combination displayed superior intracellular bactericidal effects compared to VAN alone, consistent with in vitro results (Figure 6a–c). Although statistical difference was not reached relative to the combination group, ^ΔagrAMV-VAN treatment exhibited the lowest intracellular MRSA burden among all groups (Figure 6f). Overall, these data indicate that the ^ΔagrAMV-VAN formulation efficiently eradicates intracellular MRSA in vivo, highlighting its therapeutic potential for persistent infections.

Safety Assessment

Compared to MVs secreted by the wild-type RN4220 strain, ^ΔagrAMVs derived from S. aureus RN4220ΔagrA exhibit lower toxicity.²⁵ To assess the biocompatibility of ^ΔagrAMV-VAN, RAW264.7 cells were treated with various concentrations (0, 25, 50,100, and 200 µg/mL) of ^ΔagrAMV-VAN for different durations (2, 6, 12, and 24 h). Cell viability remained unaffected within the first 12 h of treatment (Figure 7a). Interestingly, a 24 h treatment with ^ΔagrAMV-VAN remarkably increased cell proliferation compared to untreated controls. This effect may be attributed to uncertain proliferative molecules involved in ^ΔagrAMV nanoparticles, though the precise mechanism warrants further investigation.

Figure 7 Safety assessment of ΔagrAMV-VAN. (a) Toxic effects of ΔagrAMV-VAN on RAW264.7 macrophages. RAW264.7 cells were treated with diverse concentrations (0, 25, 50, 100, and 200 µg/mL) of ΔagrAMV-VAN for the indicated time periods. Cell viability was detected using a Cell Counting Kit. The viability of untreated cells (0 µg/mL) was set as 100%, and the relative viability of other groups was calculated and shown. (b) H&E staining of mouse organ sections. Organs from BALB/c mice challenged with PBS, VAN, ΔagrAMVs, and ΔagrAMV-VAN were collected, sectioned, and subjected to H&E staining. Scale bars: 50 μm. (c) Changes in mouse body weight. (d) Blood IL-6 levels in mice 6 h post-challenge. (e) Blood TNF-α levels. Data are shown as mean ± SD. Statistical significance was measured by one-way or two-way ANOVA; *** P<0.001, and ns indicates no significance.

We next investigated the in vivo toxicity of ^ΔagrAMV-VAN in a murine infection model. BALB/c mice were intraperitoneally injected with PBS, VAN, ^ΔagrAMVs, or ^ΔagrAMV-VAN. After 24 h post-injection, mouse organs (heart, liver, spleen, lungs, and kidneys) were harvested and sectioned for analysis. H&E staining revealed normal histomorphology in all all treatment groups, with no pathological alterations observed (Figure 7b). Additionally, ^ΔagrAMV-VAN treatment did not affect body weight or systemic levels of IL-6 and TNF-α compared to controls (Figure 7c–e). Collectively, these findings demonstrate that the ^ΔagrAMV-VAN formulation is biocompatible in vitro and non-toxic in vivo, supporting its potential as a safe and effective therapeutic for persistent MRSA infections.

Conclusion

We developed ^ΔagrAMV-VAN to achieve intracellular uptake and accumulation of therapeutic agents by MRSA-infected macrophages, enabling sustained elimination of intracellular MRSA. However, this study has several limitations. First, although multiple mass ratios (1:1, 1:2, and 2:1) were determined, the VAN loading efficiency in ^ΔagrAMV-VAN remained limited (21.6%). Novel preparation methods, such as using hybrid cellular MVs,^27,46 may improve VAN loading capacity and warrant further exploration. Second, while MRSA can infect virtually all human organs, the tissue-targeting capability of ^ΔagrAMV-VAN was not evaluated. Additional infectious models, such as endocarditis and meningitis, are needed to conduct the therapeutic efficacy of ^ΔagrAMV-VAN in diverse pathological contexts. Third, although bacterial MVs contain certain PAMPs that may modulate macrophage activation,^41–43 the specific PAMPs in ^ΔagrAMVs responsible for M1 macrophage polarization remain unidentified. Future studies employing proteomic analysis and genetic screening could elucidate the mechanistic basis of ^ΔagrAMVs-mediated macrophage activation. In summary, this study provides a promising strategy against intracellular MDR infections and offers a rational framework for designing exogenous antibacterial nanomaterials.