Macrophage-mimicking nanodiscs for treating systemic infection caused by methicillin-resistant Staphylococcus aureus

Abstract

Antibiotic-resistant bacteria represent a critical public health challenge, underscoring the urgent need for innovative therapeutic strategies. Inspired by the unique properties of cell membrane–derived nanodiscs, particularly their ultrasmall size and intrinsic membrane functions, we develop macrophage membrane–derived nanodiscs (denoted “MФ-NDs”) as a nanomedicine for the treatment of systemic bacterial infections caused by methicillin-resistant Staphylococcus aureus (MRSA). Our findings demonstrate that MФ-NDs interact directly with bacteria, disrupting their membranes, inducing leakage of intracellular contents, and ultimately causing bacterial death. Meanwhile, MФ-NDs reduce intracellular bacterial count. In a mouse model of systemic MRSA infection, treatment with MФ-NDs significantly improves survival rates in both therapeutic and preventative contexts. Moreover, MФ-NDs show a lower propensity to induce bacterial resistance compared to conventional small-molecule antibiotics. No acute toxicity is observed in mice treated with MФ-NDs. Overall, this study underscores the straightforward fabrication and promising potential of MФ-NDs for addressing antibiotic-resistant bacterial infections.

Macrophage membrane–derived nanodiscs can directly disrupt bacterial membranes and lower the number of bacteria inside cells.

INTRODUCTION

Bacterial infections remain a significant global health challenge, causing millions of deaths annually (1). Since the discovery of penicillin in 1928, antibiotics have been the cornerstone of bacterial infection management (2, 3). However, the growing prevalence of antibiotic resistance is increasingly undermining their efficacy (4, 5). Among the most notable antibiotic-resistant pathogens is methicillin-resistant Staphylococcus aureus (MRSA), a bacterium responsible for a wide range of infections, from mild skin conditions to life-threatening diseases affecting the lungs, heart, and bloodstream (6, 7). MRSA is resistant to an entire class of beta-lactams. Although vancomycin remains the primary antibiotic for the treatment, resistant strains are emerging, complicating treatment strategies and threatening future antibiotic strategies (8). Nonantibiotic agents such as antimicrobial peptides or polymers have a low likelihood of inducing resistance (9, 10). However, they are typically associated with cellular toxicity, limiting their use in systemic applications. Consequently, the development of alternative therapeutic strategies beyond traditional antibiotics has become imperative (11, 12).

Macrophages are essential components of the innate immune system, serving as the first line of defense against pathogens including MRSA. Boosting macrophage activity offers a promising approach to combating MRSA infections (13, 14). However, MRSA bacteria have evolved sophisticated mechanisms to evade and subvert MФ-mediated defenses (15). For example, MRSA suppresses the production of reactive oxygen species (ROS), a key bactericidal mechanism in macrophages (16). During phagocytosis, MRSA interacts with Toll-like receptor 2 on the macrophage surface, activating c-Jun N-terminal kinase, which inhibits superoxide production and thus allows MRSA to persist within macrophages (17, 18). Similarly, MRSA mitigates nitric oxide (NO)–mediated bactericidal activity by producing flavohemoglobin, which detoxifies NO into nitrate, thereby enhancing bacterial survival within the host (19, 20). Furthermore, MRSA uses virulence factors such as alpha toxins, which bind to macrophage membrane receptors like a disintegrin and metalloproteinase domain-containing protein 10, disrupting cellular integrity and inducing cell death (21, 22). These immune evasion strategies highlight the urgent need for innovative therapeutic approaches capable of counteracting MRSA’s defenses and restoring macrophage functionality.

In the past few decades, the field of nanomedicine has advanced significantly to combat drug-resistant bacteria (23–25). Among the various nanoparticle platforms, recently emerged cellular nanodiscs (CNDs) have expanded the potential of biomimetic antimicrobial strategies. Derived from cell membranes, CNDs preserve the functional properties of their source membranes while offering a smaller size and increased surface area, which enhance their therapeutic effectiveness. For example, nanodiscs engineered from red blood cell (RBC) membranes effectively neutralize pore-forming toxins by leveraging the RBC membrane’s natural toxin-binding properties (26). Similarly, nanodiscs derived from bacterial outer membranes maintain immunogenic antigens and adjuvant properties, making them suitable for antibacterial vaccination (27). Moreover, neuronal and platelet membrane–derived nanodiscs use their native membrane proteins to neutralize neurotoxins and pathological antibodies, respectively (28, 29). These advancements underscore the versatility of biomimetic nanodiscs and their potential in diverse therapeutic applications.

Inspired by the recent developments in CND technology, this study explores the therapeutic potential of macrophage membrane–derived nanodiscs (MФ-NDs) for treating MRSA infections. As shown in Fig. 1A, we fabricate MФ-NDs by mixing styrene–maleic acid (SMA) copolymers with purified macrophage plasma membranes, enabling spontaneous self-assembly into nanodiscs. MФ-NDs directly interact with MRSA, disrupting bacterial membrane integrity, causing cellular content leakage, and ultimately leading to bacterial death. Meanwhile, MФ-NDs can also act as decoys for host macrophages, sequestering MRSA virulence factors and reducing bacterial pathogenicity. This decoy effect may strengthen macrophage-mediated bactericidal activity. Our results show that MФ-NDs can reduce intracellular bacterial survival and enhance ROS and NO production by macrophages. In addition, MФ-NDs exhibit a lower likelihood of inducing drug resistance compared to traditional small-molecule antibiotics. Systemic administration of MФ-NDs shows negligible acute toxicity, underscoring their safety profile. Collectively, these results highlight the potential of MФ-NDs as a safe and effective alternative to conventional antibiotics for treating MRSA infections.

Fig. 1. Fabrication and characterization of MΦ-NDs.

(A) Schematic illustration of the fabrication of MΦ-NDs and their effects against methicillin-resistant Staphylococcus aureus (MRSA), including bacterial membrane disruption and reducing intracellular bacterium. (B) Size (diameter, nm) and surface zeta potential (ζ, mV) of macrophage vesicles (MΦ-Vs) and MΦ-NDs measured by dynamic light scattering (DLS). (C) A representative transmission electron microscopy (TEM) image of MΦ-NDs. (D) Stability of MΦ-NDs assessed by measuring their size (diameter, nm) with DLS in 1× PBS at 4°C for 14 days and in 50% fetal bovine serum (FBS) at 37°C for 72 hours. h, hours. (E) Pharmacokinetic (PK) profile of MΦ-NDs following intravenous injection of Alexa Fluor 647–labeled MΦ-NDs (AF647-MΦ-NDs). (F) Biodistribution of MΦ-NDs in mice at 24, 48, and 72 hours post–intravenous injection of AF647-MΦ-NDs. (G to I) Neutralization of proinflammatory cytokines, including (G) IL-6, (H) TNF-α, and (I) IFN-γ. In all studies, n = 3 independent experiments using the same batch of cell membranes. In (E) and (F), the MΦ-ND dosage was 20 mg/kg (membrane protein). Data in (B) and (F) are shown as mean + SD, while data in (D), (E), and (G) to (I) are presented as mean ± SD. Statistical analysis was performed using a two-tailed Student’s t test, *P < 0.05.

RESULTS

The macrophage membrane was harvested from THP-1 cells using a centrifugation method previously established (30, 31). The membrane was then sonicated to form macrophage membrane–derived vesicles (MΦ-Vs), which were subsequently mixed with SMA copolymers. Apolipoproteins or their peptide mimics have also been used for nanodisc formation (32, 33). In the study, we chose SMA copolymers because of their consistent performance in CND development. As a result of this mixing, the size of the MΦ-Vs reduced from 154.2 ± 2.5 to 21.2 ± 1.9 nm with a final polydispersity index value of 0.19 ± 0.03 (Fig. 1B and fig. S1). The surface zeta potential values shifted from −22.5 ± 0.3 to −25.2 ± 1.1 mV. Transmission electron microscopy (TEM) revealed disc-shaped nanostructures with a uniform size distribution, a clear change from the morphology of cell membrane vesicles (Fig. 1C and fig. S2). These results are consistent with previous characterization of CNDs, confirming the successful formation of MΦ-NDs (26, 27). When stored in 1× PBS at 4°C for 14 days or 50% fetal bovine serum (FBS) at 37°C for 72 hours, the MΦ-ND samples showed negligible changes in size, demonstrating excellent buffer storage stability and biostability (Fig. 1D).

To study the pharmacokinetic (PK) profile of MΦ-NDs, fluorescence-labeled MΦ-NDs were administered via the tail vein of mice after confirming negligible dye release from the formulation (fig. S3). Blood samples were collected at various time points, and fluorescence intensity was measured. Initially, the MΦ-ND signal declined rapidly after injection, but the decay slowed significantly after approximately 4 hours. Using a two-compartment model from previous CND circulation studies, the elimination half-life of MΦ-NDs was calculated to be 9.4 ± 0.3 hours (Fig. 1E) (34, 35). In the biodistribution study, fluorescence-labeled MΦ-NDs were also injected into mice through the tail vein at a dose of 20 mg/kg (protein concentration). Fluorescence intensities were measured from the heart, liver, spleen, lung, kidney, and blood (Fig. 1F). When normalized per gram of tissue, MΦ-NDs were predominantly localized in the liver and spleen, the primary organs of the reticuloendothelial system (RES). The fluorescence in the blood decreased over time, while the signal in the liver increased, suggesting that the RES is responsible for MΦ-ND clearance.

We expect MΦ-NDs to inherit the protein function of the source membrane. To test this hypothesis, we measured the capacity of MΦ-NDs to bind with representative inflammatory cytokines including interleukin-6 (IL-6), tumor necrosis factor–α (TNF-α), and interferon-γ (IFN-γ). The binding profile shows a sigmoidal dose-response curve when the MΦ-ND was plotted against the IL-6 concentration (Fig. 1G and table S1). Approximately 50.2 ± 8.5 μg/ml of MΦ-NDs neutralized half of the IL-6 [median effective concentration (EC₅₀)]. The binding between the MΦ-NDs and TNF-α (Fig. 1H and table S2) or IFN-γ (Fig. 1I and table S3) also shows a similar sigmoidal profile with EC₅₀ values of 60.7 ± 4.2 and 43.7 ± 5.8 μg/ml, respectively. Notably, MΦ-NDs demonstrated a superior capacity to neutralize cytokines compared to CNDs derived from platelet, neuronal, or RBC membranes, highlighting their unique effectiveness in cytokine neutralization (fig. S4).

After the fabrication of MΦ-NDs, we evaluated their inhibitory and bactericidal efficacy against MRSA in vitro. As shown in Fig. 2A, increasing the concentration of MΦ-NDs in the bacterial culture resulted in decreased bacterial growth, evidenced by the increased clarity of the culture media. In contrast, cultures treated with RBC membrane–derived nanodiscs (RBC-NDs) remained turbid, indicating continued bacterial growth. To quantify the inhibition, we plotted bacterial growth against nanodisc concentrations. As shown in Fig. 2B and table S4, MΦ-NDs exhibited apparent bacterial inhibition, with a minimum inhibitory concentration (MIC) of approximately 40 μg/ml. In contrast, RBC-NDs showed negligible inhibition of MRSA growth under the same conditions. To assess bactericidal efficacy, we inoculated bacterial cultures onto agar plates for colony counting. As shown in Fig. 2C, MΦ-NDs significantly reduced the number of MRSA colonies in a concentration-dependent manner. On the basis of bacterial enumeration, MΦ-NDs achieved a minimum bactericidal concentration (MBC) of approximately 100 μg/ml (Fig. 2D and table S5). In contrast, RBC-NDs resulted in only a 22.5% reduction in bacterial colonies at the same concentration. Notably, compared with CNDs made from human platelets, neurons, or RBCs, MΦ-NDs showed the highest efficacy (fig. S5). These findings indicate that MΦ-NDs are highly effective in inhibiting MRSA.

Fig. 2. Inhibitory and bactericidal effects of MΦ-NDs against MRSA bacteria in vitro.

(A) Images of MRSA cultures after 24-hour treatment with various concentrations of MΦ-NDs or RBC-NDs. (B) OD₆₀₀ values of MRSA cultures from (A). (C) MRSA colony formation after 24-hour treatment with various concentrations of MΦ-NDs or RBC-NDs followed by inoculation on agar plates. (D) Colony counts from the agar plates shown in (C). (E) Dose-dependent bacterial DNA leakage from MRSA treated with MΦ-NDs or RBC-NDs. (F) Time-dependent DNA leakage from MRSA following treatment with MΦ-NDs or RBC-NDs. In (E) and (F), DNA concentrations were quantified by measuring the absorbance at 260 nm. (G) Propidium iodide (PI, Ex/Em = 544/620 nm) uptake by MRSA bacteria after incubation with MΦ-NDs or RBC-NDs. Data in (D), (E), and (G) are presented as mean + SD. Data in (B) and (F) are presented as mean ± SD. Statistical analysis was performed using a two-tailed Student’s t test, *P < 0.05, **P < 0.01, ****P < 0.0001. In all studies, n = 3 independent experiments using the same batch of THP-1 or RBC membranes.

To explore the mechanism of bacterial inhibition, we evaluated the integrity of the MRSA membrane following exposure to MΦ-NDs. MRSA bacterial morphology showed signs of membrane disruption (fig. S6). In a DNA leakage assay, treatment with MΦ-NDs led to dose-dependent DNA leakage (Fig. 2E and table S6). DNA leakage was significantly higher in bacteria treated with MΦ-NDs than those treated with RBC-NDs. Time-dependent measurements revealed that DNA leakage from bacteria treated with MΦ-NDs occurred much faster than in the group treated with RBC-NDs, suggesting that MΦ-NDs inhibit MRSA growth by compromising the bacterial membrane (Fig. 2F and table S7). We further assessed bacterial membrane permeability using propidium iodide (PI) uptake assay to confirm such membrane damage. As shown in Fig. 2G and table S8, increasing the concentration of MΦ-NDs resulted in higher PI uptake by the bacteria, indicating concentration-dependent membrane rupture. In contrast, bacteria treated with RBC-NDs exhibited much lower PI uptake. These results demonstrate that MΦ-NDs effectively disrupt the bacterial membrane, contributing to their antibacterial activity. Notably, several receptors on the macrophage membrane—such as scavenger, complement, Fc, and TLR-2 receptors—are involved in MRSA recognition and facilitate phagocytosis (6). MΦ-NDs inherit these receptors from their source macrophages and exhibit preferential binding to MRSA, a feature that may contribute to their bactericidal efficacy (fig. S7).

Macrophages recognize and bind to S. aureus, initiating phagocytosis to engulf the bacteria. However, MRSA can survive within macrophages, leading to intracellular infection. In the study, we also investigated the capacity of MΦ-NDs to reduce intracellular bacterial count after MRSA bacteria are taken up by macrophages. THP-1 cells were chosen as a representative macrophage-like cell line. First, we mixed MRSA bacteria with various concentrations of MΦ-NDs (5, 10, 25, 50, and 100 μg/ml) and the mixtures were added to THP-1 cells at a multiplicity of infection (MOI) of 20. As shown in Fig. 3A and table S9, a higher dose of MΦ-NDs resulted in greater bacterial killing, while RBC-NDs showed negligible bacterial inhibition. We also studied the effect of MOI on antibacterial efficacy. In the study, MRSA bacteria [1 × 10⁶, 5 × 10⁶, 1.0 × 10⁷, 2.0 × 10⁷, and 5.0 × 10⁷ colony-forming unit (CFU)/ml] were mixed with MΦ-NDs (50 μg/ml), and the mixture was added to THP-1 cells (MOIs of 1, 5, 10, 20, and 50, respectively, Fig. 3B and table S10). In the control groups, MRSA bacteria were mixed with RBC-NDs at the same conditions. The intracellular MRSA was then counted. In both groups, a higher MOI value led to more intracellular MRSA. However, with the presence of MΦ-NDs, the bacterial counts were significantly lower: At MOIs of 1, 5, 10, 20, and 50, the intracellular MRSA counts were reduced to 49.8 ± 21.6%, 42.5 ± 15.6%, 28.8 ± 9.43%, 21.0 ± 9.11%, and 35.7 ± 6.11% of the corresponding group mixed with RBC-NDs, respectively. These results demonstrate that MΦ-NDs effectively reduce MRSA survival within macrophages.

Fig. 3. MΦ-NDs enhanced macrophage bactericidal activity against MRSA in vitro.

(A) Concentration-dependent macrophage bactericidal activity against MRSA bacteria. MRSA was preincubated with MΦ-NDs at varying concentrations (5, 10, 25, 50, and 100 μg/ml) and then incubated with THP-1 macrophages at a multiplicity of infection (MOI) of 20. Intracellular bacterial burdens were determined by colony counting. (B) MOI-dependent macrophage bactericidal activity. MRSA was preincubated with MΦ-NDs at 100 μg/ml and then incubated with THP-1 macrophages at different MOIs (1, 5, 10, 20, and 50). Intracellular bacterial burdens were determined by colony counting. (C) Time-dependent reactive oxygen species (ROS) induction. MRSA was preincubated with MΦ-NDs or RBC-NDs at 100 μg/ml and then incubated with THP-1 macrophages for different durations (15, 30, 45, 60, 90, and 120 min). ROS levels were measured and normalized. MRSA in the control group received no pretreatment. (D) Concentration-dependent ROS induction. MRSA was preincubated with MΦ-NDs or RBC-NDs at various concentrations (5, 10, 25, 50, and 100 μg/ml), then incubated with THP-1 macrophages for 30 min. ROS levels were measured and normalized. (E) Time-dependent nitric oxide (NO) induction. MRSA was preincubated with MΦ-NDs or RBC-NDs at 100 μg/ml and then incubated with THP-1 macrophages for different durations (1, 2, 4, 8, 12, and 24 hours). NO levels were measured and normalized. (F) Concentration-dependent NO induction. MRSA was preincubated with MΦ-NDs or RBC-NDs at various concentrations (5, 10, 25, 50, and 100 μg/ml) and then incubated with THP-1 macrophages for 8 hours. NO levels were measured and normalized. MRSA bacteria in the control group were not mixed with MΦ-NDs. Data are presented as mean + SD. Statistical analysis was performed using a two-tailed Student’s t test, *P < 0.05, **P < 0.01, ***P < 0.001. In all studies, n = 3 independent experiments using the same batch of THP-1 or RBC membranes.

Previous studies have demonstrated that MRSA bacteria suppress macrophage function by reducing the production of key bactericidal molecules such as ROS and NO, which are hallmarks of macrophage defense against bacteria (36, 37). Given the inhibitory effect of MΦ-NDs against MRSA, we hypothesized that MΦ-NDs facilitate host cells to produce ROS and NO, likely through the neutralization of bacterial virulence factors, and therefore boost macrophages’ capability to defend against MRSA invasion. To test this hypothesis, we mixed MRSA with MΦ-NDs or RBC-NDs and then added the mixture to THP-1 cells. MRSA-induced ROS production was much higher in THP-1 cells added with MΦ-NDs than those with RBC-NDs (Fig. 3C and table S11). A higher concentration of MΦ-NDs resulted in a more prominent enhancement of ROS production, and such enhancement was negligible in RBC-ND control groups (Fig. 3D and table S12). We observed similar effects on NO production. MRSA-induced NO production was much higher in THP-1 cells added with MΦ-NDs than those with RBC-NDs (Fig. 3E and table S13). A higher concentration of MΦ-NDs resulted in a more prominent enhancement of NO production, and such enhancement of NO production from THP-1 cells was negligible in RBC-ND control groups (Fig. 3F and table S14). These results support the idea that MΦ-NDs enhance macrophage ROS and NO production during MRSA infection, thereby strengthening macrophage immune response against MRSA intracellular invasion.

After confirming the antibacterial efficacy of MΦ-NDs against MRSA in vitro, we proceeded to evaluate their therapeutic potential in vivo using a mouse model of MRSA infection. To establish lethal MRSA infection, we injected varying doses of MRSA bacteria (0.8 × 10⁹, 1.2 × 10⁹, 1.4 × 10⁹, and 1.6 × 10⁹ CFU/kg, respectively) to the mice through their tail veins. Mouse survival was monitored for 72 hours. As shown in Fig. 4A, the survival rate decreased as the MRSA dosages increased. A dose of 1.2 × 10⁹ CFU/kg MRSA bacteria resulted in 50% mortality [median lethal dose (LD₅₀)]. This dose was selected for subsequent experiments. Next, we challenged mice with MRSA (1.2 × 10⁹ CFU/kg) and then MΦ-NDs (5, 10, and 15 mg/kg) 1 min later, both intravenously through the tail vein. As shown in Fig. 4B and table S15, MRSA injection alone resulted in 50% mortality. However, MΦ-ND injection at 5 mg/kg delayed the mouse death. MΦ-ND dosage of 10 and 15 mg/kg increased the survival rates to 83 and 100%, respectively, demonstrating a dose-dependent improvement in survival. To further test the therapeutic effectiveness, mice were infected at the LD₅₀ dose and treated with MΦ-NDs (15 mg/kg) at 5 or 15 min postinfection. As shown in Fig. 4C and table S16, the survival rate of mice in both treatment regimens increased to 83%. To evaluate prophylactic efficacy, we injected MΦ-NDs (15 mg/kg) into mice through the tail vein first. After 1, 5, and 15 min, we injected MRSA, also through the tail vein. As shown in Fig. 4D and table S17, an interval of 1 min resulted in a survival rate of 100%, and intervals of 5 and 15 min resulted in survival rates of 83 and 67%, respectively. Besides survival rates, injection of MΦ-NDs also lowered circulating bacteria in mice (fig. S8). Together, these findings demonstrate that MΦ-NDs enhance survival both as a therapeutic intervention and as a prophylactic measure.

Fig. 4. In vivo efficacy of MΦ-NDs against MRSA.

(A) Establishment of a lethal mouse model for MRSA infection. Mice were challenged with MRSA at various concentrations (0.8 × 10⁹, 1.2 × 10⁹, 1.4 × 10⁹, and 1.6 × 10⁹ CFU/kg) intravenously via tail veins. Following the inoculation, the mice were monitored for 72 hours, during which their survival rates were recorded. (B) Dose-dependent treatment efficacy of MΦ-NDs against MRSA. Mice were first injected with MRSA (1.2 × 10⁹ CFU/kg, LD₅₀) via the tail vein. After 1 min, the mice received MΦ-NDs (5, 10, and 15 mg/kg). The survival rates were monitored for 72 hours. (C) Time-dependent efficacy in a therapeutic regimen. Mice were challenged with MRSA (1.2 × 10⁹ CFU/kg, LD₅₀). The mice were then administered with MΦ-NDs (15 mg/kg) at three different time intervals: 1, 5, and 15 min following the MRSA injection. The survival rates from each group were recorded for 72 hours. (D) Time-dependent efficacy in a preventative regimen. Mice first received MΦ-NDs (15 mg/kg) at intervals of 1, 5, and 15 min before the MRSA injection. Their survival rates were monitored over 72 hours. All mice were humanely euthanized following the 72-hour observation period. In all datasets, n = 6 mice. Statistical analysis was performed using Log-rank and Mantel-Cox test.

We have shown that MΦ-NDs inhibited MRSA by disrupting bacterial membrane and boosting macrophage activity, likely through the neutralization of the bacterial virulence factors. These effects do not directly interfere with bacterial resistance pathways. Therefore, we hypothesize that MΦ-NDs are less likely to elicit drug resistance that is common to small-molecule antibiotics. To test this hypothesis, we compared MΦ-NDs with linezolid and vancomycin commonly used against MRSA. In this study, MRSA was incubated with MΦ-NDs or antibiotics at three sub-bactericidal concentrations (¹/₂ MIC, ²/₃ MIC, and MIC) for 14 days. As shown in Table 1, MRSA treated with linezolid at ¹/₂ MIC (1.0 μg/ml), ²/₃ MIC (1.3 μg/ml), and MIC (2.0 μg/ml) developed resistance on days 9, 8, and 5, respectively. Similarly, MRSA exposed to vancomycin at ¹/₂ MIC (0.5 μg/ml), ²/₃ MIC (0.7 μg/ml), and MIC (1.0 μg/ml) acquired resistance on days 8, 6, and 4, respectively. In contrast, MRSA treated with MΦ-NDs at all tested concentrations—¹/₂ MIC (20.0 μg/ml), ²/₃ MIC (26.7 μg/ml), and MIC (40.0 μg/ml)—did not develop resistance throughout the 14-day study.

Table 1. Resistance development of MRSA.

MRSA bacteria were incubated with different levels of MΦ-NDs, linezolid, or vancomycin over a span of 14 days. The concentrations tested include the minimal inhibitory concentration (MIC), two-thirds of the MIC (²/₃ MIC), and half of the MIC (¹/₂ MIC). In the table, “+” denotes “resistant” and “−” denotes “not resistant.”

Formulation	Concentration	Time (days)
		1	2	3	4	5	6	7	8	9	10	11	12	13	14
MΦ-ND	1/2 MIC	−	−	−	−	−	−	−	−	−	−	−	−	−	−
	2/3 MIC	−	−	−	−	−	−	−	−	−	−	−	−	−	−
	MIC	−	−	−	−	−	−	−	−	−	−	−	−	−	−
Linezolid	1/2 MIC	−	−	−	−	−	−	−	−	+	+	+	+	+	+
	2/3 MIC	−	−	−	−	−	−	−	+	+	+	+	+	+	+
	MIC	−	−	−	−	+	+	+	+	+	+	+	+	+	+
Vancomycin	1/2 MIC	−	−	−	−	−	−	−	+	+	+	+	+	+	+
	2/3 MIC	−	−	−	−	−	+	+	+	+	+	+	+	+	+
	MIC	−	−	−	+	+	+	+	+	+	+	+	+	+	+

Last, we assessed the acute toxicity of MΦ-NDs in mice. In the study, MΦ-NDs (20 mg/kg, protein weight) were injected into mice intravenously through the tail vein. After 24 hours of administration, we counted the RBCs, platelets, and white blood cells (WBCs), and their values were found to be comparable to the baseline levels (Fig. 5A and table S18). Populations of WBCs including monocyte, neutrophil, lymphocyte, eosinophil, and basophil also remained at their baseline levels (Fig. 5B and table S19). Meanwhile, the levels of representative cytokines including IL-6, TNF-α, and IL-1β remained unchanged (Fig. 5C). In addition, a comprehensive serum chemistry panel was performed, and the results remained consistent with the baseline levels (Fig. 5D). These findings from blood analysis showed the short-term safety of MΦ-NDs in vivo. Moreover, the tissues collected from main organs (heart, liver, spleen, lung, and kidney) were analyzed histologically after sectioning and staining with hematoxylin and eosin (H&E). The histological analysis revealed similar appearances between the control and treated mice, without any apparent irregularities or signs of acute toxicity caused by MΦ-NDs (Fig. 5E). These results collectively indicate that MΦ-NDs exhibit good biosafety.

Fig. 5. In vivo biosafety of MΦ-NDs.

Measurements of (A) blood cell counts and (B) white cell subpopulation counts 24 hours post–intravenous administration of PBS or MΦ-NDs at a dosage of 20 mg/kg. WBC, white blood cell; RBC, red blood cell; and PLT, platelet. (C) Measurements of representative inflammatory cytokine levels in blood 24 hours after the administration of PBS or MΦ-NDs. (D) Analysis of a blood biochemical panel 24 hours after the administration of PBS or MΦ-NDs. ALB, albumin; ALP, alkaline phosphatase; ALT, alanine transaminase; AMY, amylase; TBIL, total bilirubin; BUN, blood urea nitrogen; CA, calcium; PHOS, phosphorus; GLU, glucose; NA⁺, sodium; K⁺, potassium; TP, total protein; and GLB, globulin. (E) Histological sections of major organs (heart, liver, spleen, lung, and kidney) were stained with hematoxylin and eosin (H&E) 24 hours after administering PBS or MΦ-NDs (20 mg/kg). Scale bar, 100 μm. Data are presented as mean + SD. Statistical analysis was performed using a two-tailed Student’s t test (P > 0.05).

DISCUSSION

In summary, we successfully developed MФ-NDs using natural macrophage membrane and demonstrated their antibacterial capabilities against MRSA. Comprising small discoidal cell membrane fragments stabilized by synthetic polymer chains, these MФ-NDs were significantly smaller than MΦ-Vs and showed high stability in biological solutions. We validated their ability to inhibit and eradicate bacteria by directly disrupting bacterial membrane, with a lower likelihood of inducing resistance compared to traditional small-molecule antibiotics. We also demonstrated their ability to enhance macrophage-mediated bactericidal activity, likely due to the neutralization of bacterial virulence factors. In a mouse model of MRSA systemic infection, treatment with MФ-NDs effectively improved survival rates in both therapeutic and preventative settings. Biosafety studies further revealed that MФ-NDs had negligible acute toxicity. Notably, MФ-NDs also inhibited other bacteria such as Escherichia coli (fig. S9). CNDs derived from different macrophage-like cells also inhibited MRSA (fig. S10). Collectively, these findings demonstrate MФ-NDs as a promising formulation against MRSA or other bacterial infections.

Despite their promise, the development of CNDs from cell membranes faces various challenges, including sourcing high-quality membranes and ensuring the functional retention of membrane proteins and lipids. Issues such as scalability, batch variability, and long-term stability remain underexplored, while additional obstacles include high production costs and regulatory complexities. Moreover, immune responses to nonautologous macrophage membrane materials upon repeated dosing need to be carefully evaluated for future translational development. Advances in engineering, nanomaterial manufacturing, and cell membrane biology are gradually addressing these hurdles. Given the broad-spectrum interactions between macrophages and various microbes, MФ-NDs hold great potential as a versatile antimicrobial strategy effective against a wide range of microbial pathogens.

MATERIALS AND METHODS

Cell culture and membrane derivation

THP-1 cells (TIB-202, American Type Culture Collection), a human monocytic cell line, were grown in RPMI 1640 medium (Gibco) enhanced with 10% FBS (Hyclone) and 1% penicillin-streptomycin (Gibco). These cells were kept at 37°C in a 5% CO₂ humidified chamber and were regularly checked for mycoplasma every 2 weeks. Cells were allowed to grow to 2 × 10⁶ cells/ml before the membrane derivation. For the membrane derivation, the cells were washed three times with 1× PBS using centrifugation (800g, 5 min) and then suspended in a hypotonic lysing solution containing tris-HCl (30 mM, pH 7.5), d-mannitol (225 mM), sucrose (75 mM), EGTA (0.2 mM), and a mix of protease and phosphatase inhibitors (all from Millipore-Sigma). Cells were homogenized using a Kinematica Polytron PT-2000 and then centrifuged at 7600g for 25 min at 4°C. The supernatant was collected and centrifuged again at 29,600g for 35 min at 4°C to pellet the membrane. The supernatant was removed. The protocol for obtaining RBC membrane involved suspending washed human RBCs (BioIVT) in 0.25× PBS in an ice bath for 20 min for hypotonic treatment, followed by centrifugation at 20,000g for 10 min to pellet the membrane. The supernatant was discarded, and the pellet was resuspended in 0.25× PBS. This process was repeated four times. Both membranes were stored in 0.2 mM EDTA at −80°C for future use. The membrane quantity was determined using a bicinchoninic acid assay (BCA) protein assay kit (Thermo Fisher Scientific).

Fabrication of cell membrane–derived nanodiscs

To fabricate MΦ-NDs, MΦ membrane was first sonicated using a batch sonicator (Fisherbrand Model 120, at 100 W for 30 s) to generate MΦ membrane vesicles. These vesicles were then mixed with styrene maleic acid copolymer (SMALP 200, Cube Biotech) at a protein-to-polymer weight ratio of 1:5. The mixture was sonicated in an ice water bath (Fisherbrand Model 120, at 100 W for 30 s) and then stirred overnight at 4°C. The sample was then centrifuged at 150,000g for 30 min. The supernatant containing MΦ-NDs was collected. These MΦ-NDs were then concentrated using ultrafiltration [Amicon, 100 kDa molecular weight cutoff (MWCO)] and washed with 1× PBS to eliminate residual SMALP 200. For the fabrication of fluorescence-labeled MΦ-NDs, the previously fabricated MΦ-NDs were mixed with either Alexa Fluor 647–hydroxysuccinimide (Alexa Fluor 647–NHS) or fluorescein isothiocyanate–N-hydroxysuccinimide (FITC-NHS, both from Thermo Fisher Scientific) at a dye-to-nanodisc weight ratio of 1:50 in sodium bicarbonate buffer (0.8 M, pH 8) and allowed to react for 2 hours. After the reaction, the sample was washed with 1× PBS using ultrafiltration (Amicon, 100 kDa MWCO) to remove residual dyes. Membrane concentration was measured using the BCA assay (Thermo Fisher Scientific). The same procedure was used to fabricate RBC-NDs.

Physicochemical characterization of MΦ-NDs

The size and surface zeta potential of MΦ-NDs were determined using dynamic light scattering (DLS, Malvern ZEN 3600 Zeta-sizer) at a membrane concentration of 0.5 mg/ml. Nanodisc morphology was visualized using a transmission electron microscope (TEM, JEOL 1200 EX II) after negative staining with 1% uranyl acetate. For stability assessment, MΦ-NDs (1 mg/ml) were incubated in 1 ml of 50% FBS at 4°C for 3 days. The size was measured every 12 hours for 24 hours using DLS. For storage stability, MΦ-NDs were stored in 1× PBS at 4°C (1 mg/ml). The size was measured every 2 days for 14 days using DLS.

Animal care and injections

All animal studies were approved under the guidelines of the University of California San Diego (UCSD) Institutional Animal Care and Use Committee. In all animal studies, 4-week-old ICR mice (Harlan Laboratories, male) were used. Mice were housed in an animal facility at UCSD under federal, state, local, and NIH guidelines for animal care. In the study, no inflammation or any abnormal appearance was observed at the sites of injection.

Assessment of PK and biodistribution profiles

To characterize MΦ-ND PK profile, Alexa Fluor 647–labeled MΦ-NDs (20 mg/kg, protein weight) were intravenously injected via the tail vein (n = 3), and blood samples were collected at predetermined time points (1, 5, 15, and 30 min, and 1, 4, 8, 24, and 48 hours) via submandibular puncture. After the blood collection, fluorescence intensity was measured (Tecan M200). PK parameters were calculated using a two-compartment model. To assess MΦ-ND biodistribution, Alexa Fluor 647–labeled MΦ-NDs (20 mg/kg, protein weight) were injected intravenously through the tail vein. At 24, 48, and 72 hours after the injection, major organs (heart, liver, spleen, lung, and kidney) and blood were collected from the mice (n = 3). The collected organs were then weighed and homogenized in 1× PBS for fluorescence measurement (Tecan M200). The measured intensity was multiplied by the corresponding organ weight to obtain the total organ fluorescence. Then, the relative distribution of the MΦ-NDs in each organ was calculated.

Cytokine neutralization by MΦ-NDs

To evaluate MΦ-ND neutralization of cytokines, including IL-6, TNF-α, and IFN-γ, MΦ-NDs or RBC-NDs (5, 10, 25, 50, 100, and 200 μg/ml) were mixed with IL-6 (500 pg/ml), TNF-α (500 pg/ml), or IFN-γ (500 pg/ml) in 1× PBS containing 1% BSA. The mixtures were incubated at 37°C for 20 min. After the incubation, the samples were washed using ultrafiltration (Amicon, 100 kDa MWCO). The filtrate was collected, and unbound cytokines were quantified using enzyme-linked immunosorbent assay (ELISA, BioLegend).

Bacterial culture

MRSA (USA300, American Type Culture Collection) was cultured on tryptic soy broth (TSB) agar overnight at 37°C. Then, a single colony was inoculated into TSB medium and incubated at 37°C until the concentration reached ≈ 1 × 10⁹ CFU/ml (OD₆₀₀ = 1 corresponds to 1 × 10⁸ CFU/ml bacteria). The bacteria were then harvested by centrifugation at 9000g for 10 min and then washed with sterile PBS twice. The obtained bacteria were pelleted and resuspended in an appropriate amount of sterile PBS for future use.

Inhibitory and bactericidal activities of MΦ-NDs against MRSA

To evaluate the inhibitory and bactericidal activity of MΦ-NDs against MRSA, 5 × 10⁶ CFU of the bacteria was added to 1 ml of fresh TSB containing various concentrations (10, 25, 50, and 100 μg/ml) of MΦ-NDs or RBC-NDs. The mixture was cultured at 37°C with shaking for 24 hours. For determination of the minimal inhibitory concentration (MIC), the optical density at 600 nm (OD₆₀₀) of the bacterial suspension treated by MΦ-NDs or RBC-NDs was measured. For determination of the MBC, the above bacterial culture was subjected to serial 10-fold dilution and spotted onto TSB agar plates. The plates were cultured at 37°C for 24 hours before counting the colonies. In both experiments, MRSA treated with PBS served as the negative control.

Bacterial membrane integrity study

The integrity of the MRSA bacterial membrane after MΦ-ND treatment was evaluated using bacterial DNA leakage assays and bacterial PI uptake assay. For dose-dependent DNA leakage, MRSA (1.0 × 10⁶ CFU/ml) was incubated with MΦ-NDs or RBC-NDs (5, 10, 25, 50, and 100 μg/ml) in TSB for 4 hours at 37°C. The bacterial suspensions were pelleted at 9000g, and the supernatants were collected. The absorbance at 260 nm was measured to assess the leaked DNA concentrations (Tecan M200). To evaluate time-dependent DNA leakage, MRSA (1.0 × 10⁶ CFU/ml) was incubated with MΦ-NDs or RBC-NDs (50 μg/ml) in TSB for 0.5, 1, 2, 4, and 8 hours at 37°C, followed by centrifugation and measurement of supernatant absorbance at 260 nm. For bacterial PI uptake study, MRSA (1.0 × 10⁶ CFU/ml) was incubated with MΦ-NDs (5, 10, 25, 50, and 100 μg/ml) in TSB for 4 hours at 37°C. After the incubation, the suspensions were centrifuged at 9000g to remove the supernatant. Bacteria were washed by using sterile PBS and then resuspended. PI dye (Biolegend, CA) was added to the bacterial suspensions (final concentration, 10 μg/ml). The samples were incubated in the dark for 20 min at room temperature. After the incubation, the samples were measured for the PI fluorescence intensity (Ex/Em = 544/620 nm, Tecan M200).

Macrophage bactericidal assay

To study the dose-dependent macrophage bactericidal efficacy, MΦ-NDs or RBC-NDs (50 μl, 5, 10, 25, 50, and 100 μg/ml final concentration) were mixed with MRSA (50 μl, 2.0 × 10⁷ CFU/ml final concentration) in 1× PBS at 37°C for 10 min. Then, the mixture was added to the THP-1 macrophages (1.0 × 10⁶ cells/ml) and incubated at 37°C for 2 hours. After the incubation, the cells were pelleted and washed with 1× PBS. After washing, the cells were resuspended in an RPMI 1640 medium containing gentamicin (100 μg/ml; Millipore-Sigma). The samples were incubated for 3 hours to eliminate extracellular bacteria. The viable intracellular bacteria were quantified by first disrupting THP-1 cells using a hypotonic solution containing tris-HCl (10 mM, pH 7.4) and EDTA (1 mM). Then, a series of 10-fold dilutions from the disrupted cell suspension was inoculated onto TSB agar plates. The agar plates were cultured at 37°C for 24 hours, and the colony formation was counted. To study the MOI-dependent bactericidal efficacy of MΦ-NDs, MΦ-NDs or RBC-NDs (50 μl, 100 μg/ml final concentration) was mixed with MRSA (50 μl, 1.0 × 10⁶, 5.0 × 10⁶, 1.0 × 10⁷, 2.0 × 10⁷, and 5.0 × 10⁷ CFU/ml final concentration) in 1× PBS at 37°C for 10 min. Then, the mixture was added to the THP-1 macrophages (1.0 × 10⁶ cells/ml) and incubated at 37°C for 2 hours. After incubation, cells were pelleted, washed, and resuspended in RPMI 1640 medium containing gentamicin (100 μg/ml) for 3 hours to eliminate extracellular bacteria. The same disruption, dilution, and plating procedures were then performed, with MRSA incubated with THP-1 cells alone serving as the control.

Macrophage oxidative burst assays

THP-1 cells (1.0 × 10⁶ cells/ml) were first incubated with 2,7-dichlorofluorescein diacetate (DCFH-DA, 25 μM, Thermo Fisher Scientific) in Ca²⁺ and Mg²⁺ free Hanks’ balanced salt solution (Mediatech) at room temperature for 30 min. After incubation, the cells were pelleted, resuspended, and plated in a 96-well plate (1.0 × 10⁶ cells/ml). For the time-dependent oxidative burst assay, the same initial incubation of THP-1 cells with DCFH-DA was performed. Then, MΦ-NDs (10 μl, 100 μg/ml final concentration) were mixed with MRSA (10 μl in 1× PBS, 1.0 × 10⁸ CFU/ml) at 37°C for 10 min. A control group containing MRSA without any pretreatment was also included. The mixtures were then added to the THP-1 cells and incubated at 37°C for 15, 30, 45, 60, 90, and 120 min, and fluorescence intensity was measured at each interval (Ex/Em = 485/520, Tecan M200). For the dose-dependent oxidative burst assay, MΦ-NDs or RBC-NDs (10 μl, resulting in final concentrations of 5, 10, 25, 50, and 100 μg/ml) were mixed with MRSA (10 μl in 1× PBS, 1.0 × 10⁸ CFU/ml) at 37°C for 10 min, then added to the THP-1 cells and incubated at 37°C for 30 min. The fluorescence intensity was measured at each time point (Ex/Em = 485/520, Tecan M200).

Macrophage NO production assays

THP-1 cells (180 μl per well, 1.0 × 10⁶ cells/ml) were seeded into a 96-well plate containing fresh medium. For the time-dependent NO production assay, MΦ-NDs (10 μl, 100 μg/ml final concentration) were similarly mixed with MRSA (10 μl in 1× PBS, 1.0 × 10⁸ CFU/ml) at 37°C for 10 min, with a control group containing MRSA alone. The mixtures were added to THP-1 cells and incubated for 1, 2, 4, 8, 12, or 24 hours, after which 100 μl of Griess reagent was added to each well, and absorbance at 550 nm was again measured for NO quantification (Tecan M200). For the dose-dependent NO production assay, MΦ-NDs or RBC-NDs (10 μl at final concentrations of 5, 10, 25, 50, and 100 μg/ml) were mixed with MRSA (10 μl in 1× PBS, 1.0 × 10⁸ CFU/ml) at 37°C for 10 min. Then, the mixture was added to the THP-1 cells and incubated at 37°C for 8 hours. After the incubation, 100 μl of Griess reagent (Promega) was added to each well, and absorbance at 550 nm (Tecan M200) was measured to quantify NO production.

In vivo efficacy against MRSA infection

For MRSA lethal dosage assessment, MRSA USA300 (0.8 × 10⁹, 1.2 × 10⁹, 1.4 × 10⁹, and 1.6 × 10⁹ CFU/kg, injection volume: 100 μl) was intravenously injected into mice via the tail veins. An MRSA dosage that induced 50% of death in mice (LD₅₀) was selected for the following studies. For the dose-dependent survival study, mice were injected with MRSA (1.2 × 10⁹ CFU/kg, LD₅₀) via tail veins. After 1 min, they were injected with MΦ-NDs (5, 10, and 15 mg/kg) through the tail vein. The nanodisc dosage that rescued 100% of mice was selected for the following studies. To study efficacy in a therapeutic regimen, 4-week-old ICR mice were intravenously injected with MRSA (1.2 × 10⁹ CFU/kg). Then, each group was injected with MΦ-NDs intravenously (15 mg/kg) at 5 and 15 min after the MRSA injection. To study efficacy in a preventative regimen, mice were first injected with MΦ-NDs (15 mg/kg) through the tail vein. Then, they were injected with MRSA (1.2 × 10⁹ CFU/kg) intravenously at 1, 5, and 15 min after the MΦ-ND injection. In both studies, mouse survival was monitored for 72 hours.

Evaluation of resistance development of MRSA

To investigate resistance development, MRSA (1 × 10⁶ CFU/ml) was incubated in 1 ml of TSB containing various concentrations of MΦ-NDs, linezolid, or vancomycin. The bacteria were cultured at 37°C with gentle shaking. After 24 hours, 100 μl of sample from each well was collected and added to 900 μl of fresh medium containing the same type of drug and drug concentration. Meanwhile, the bacteria from each treatment group were tested for sensitivity to the corresponding drug using the same protocol described in the bactericidal study. This process was repeated for 14 days.

In vivo acute toxicity study

MΦ-NDs were injected intravenously through the tail vein into the mice (20 mg/kg, injection volume: 100 μl). After 24 hours, approximately 350 μl of the blood was collected via submandibular puncture. The first 250 μl of blood was placed into Eppendorf tubes and allowed to coagulate. The remaining 100 μl of blood was collected into an EDTA-coated microtube (Hyclone). The first sample containing 250 μl of blood was centrifuged at 2500g for 3 min to collect the serum for comprehensive metabolic panel analysis. The second sample containing 100 μl of blood was used for complete blood count without processing. All blood samples were analyzed at the UCSD Animal Care Program Diagnostic Services Laboratory. Meanwhile, immediately after the blood collection, mice were euthanized. The major organs were collected, including the heart, liver, spleen, lung, and kidney. Organs were fixed with 10% formalin, sectioned, and stained with H&E for histological analysis. Histology slides were imaged with a Micromaster II microscope (Fisher Scientific). Mice that received PBS only served as controls.

Supplementary Materials

This PDF file includes:

Figs. S1 to S10

Tables S1 to S19