Engineered Biomimetic Platelet Membrane-Coated Nanoparticles Block Staphylococcus aureus Cytotoxicity and Protect Against Lethal Systemic Infection

Abstract

Staphylococcus aureus is a leading human pathogen capable of producing severe invasive infections such as bacteremia, sepsis and endocarditis with high morbidity and mortality, exacerbated by expanding antibiotic-resistance exemplified by methicillin-resistant strains (MRSA). S. aureus pathogenesis is fueled by secretion of secreted toxins including the membrane damaging pore-forming α-toxin that have diverse cellular targets including epithelium, endothelium, leukocytes and platelets. Here we examine human platelet membrane-coated nanoparticles (PNPs) as a biomimetic decoy strategy to neutralize S. aureus toxins and preserve host cell defense functions. PNPs blocked platelet damage induced by S. aureus secreted toxins, supporting platelet activation and bactericidal activity. Likewise, PNPs blocked macrophage damage induced by S. aureus secreted toxins, supporting macrophage oxidative burst and nitric oxide production and bactericidal activity, and diminishing MRSA-induced neutrophil extracellular trap release. In a mouse model of MRSA systemic infection, PNP administration reduced bacterial counts in the blood and protected against mortality. Taken together, the present work provides proof-of-principle of therapeutic benefit of PNPs in toxin neutralization, cytoprotection and increased host resistance to invasive S. aureus infection.

1. Introduction

Platelets are abundant, small anucleate cell fragments in the blood circulation. The classical function of platelets lies in the regulation of blood clotting and vascular integrity. However, emerging evidence has revealed a role of platelets as sentinel effector cells during infectious diseases [1–3]. Innate immune responses to invading pathogens are significantly influenced by crosstalk with platelets, as they can sense and react to danger signals and guide leukocytes to sites of injury, inflammation or pathogen invasion [4–7].

Platelets have multiple identified roles relevant to host defense [1,2,8,9]. They can directly kill microbes by the release of antimicrobial peptides including defensins [10], cathelicidins [11], thrombocidins [12] and kinocidins [13]. Platelets can also aggregate to entrap microbes and restrict pathogen spread [14]. Through a variety of different mechanisms, platelets regulate release of a variety of intracellular mediators that are stored in granules [3]. These molecules can induce inflammation and influence the recruitment and activity of effector cells of the immune system directly or indirectly [15]. The mechanisms governing platelet-leukocyte and platelet-microbe interactions are complex, reflecting the diversity of platelet receptors such as complement receptors [16], FcγRIIa [17], TLRs [18], GPIIb-IIIa [19], and GPIb [20]. In sum, platelets are more than clotting agents, but critical players in the fine equilibrium of host immune and inflammatory responses.

Staphylococcus aureus, including methicillin-resistant strains (MRSA), is a leading opportunistic Gram-positive bacterial pathogen producing a wide array of diseases including invasive bloodstream infections, sepsis and endocarditis [21,22]. Deep-seated S. aureus infections are usually accompanied by clear immune dysregulation provoked in part by an array of secreted toxin virulence factors, including alpha-toxin (α-toxin). S. aureus toxins can engage cognate surface receptors on host cells and compromise their integrity and function by formation of membrane pores, disruption of signal transduction pathways, or activating enzymes that degrade host molecules [23,24]. As secreted toxins play important roles in driving S. aureus disease pathogenesis, methods to remove or counteract toxins have become a potential therapeutic target to improve patient outcomes. In this regard, targeted antibody or nanomedicine approaches for toxin neutralization have gained attention [25–27]. One such approach for biodetoxification is natural cell membrane-coated nanoparticles that function through biomimicry [28,29]. Their unique structure can function as a decoys to absorb bacterial membrane toxins factors nonspecifically like a sponge, neutralizing their cytolytic activity regardless of their precise molecular architecture [30]. For instance, red blood cell (RBC) membrane-coated “nanosponges” developed by coating polymeric nanoparticles with natural erythrocyte membranes protected mice against lethal intoxication with purified α-toxin protein [31], and reduced lesion size in murine models of MRSA or group A Streptococcus skin infection [32,33]. In a mouse model of Escherichia coli sepsis, macrophage membrane-coated nanosponges bound bacterial lipopolysaccharide (LPS) and sequestered proinflammatory cytokines, reducing bacterial spread and conferring protection against mortality in treated mice [34].

Platelets have been shown to contribute to host resistance against invasive S. aureus infection. Antibody-mediated platelet depletion in mice impaired S. aureus clearance as evidenced by higher bacterial burden in kidneys, more exaggerated cytokine responses, and decreased survival compared to control mice [35]. Another study indicated that platelets enhanced uptake and intracellular killing of S. aureus by peritoneal macrophages, perhaps through a mechanism dependent on platelet granule associated β1-defensin [8]. Platelets are an important target of S. aureus α-toxin as they express A Disintegrin And Metalloproteinase domain-containing protein 10 (ADAM10) on their surface membrane [36], the identified α-toxin receptor. In mice, α-toxin-mediated platelet damage and aggregation contributes to liver injury in S. aureus sepsis [37].

As platelets are both important in defense against S. aureus and the target of the pathogen’s membrane toxins, we hypothesized that biodegradable polymeric nanoparticle cores coated with biomimetic human platelet membranes, or platelet nanoparticles (PNPs), could be used to fortify platelet-mediated defense against the pathogen. The present work provides proof-of-principle of therapeutic benefit of PNPs in toxin neutralization, cytoprotection and increased host resistance to invasive S. aureus infection.

2. Materials and Methods

2.1. Platelet membrane derivation.

Blood bank-approved human O-negative (universal donor) platelet-rich plasma (PRP) stored in standard acid-citrate-dextrose (ACD) was obtained within 24–48 h of its expiration for clinical use from the San Diego Blood Bank; this PRP serves as a source of fully functional platelet membranes based on our prior investigations [38]. Phosphate-buffered saline (PBS) solution with 50 mM ethylenediaminetetraacetic acid (EDTA) and 300 μl protease inhibitor (PI) was added to the PRP preparation to restrict platelet activation. Platelets were centrifuged at 4000 × rpm for 15 min at room temperature, the supernatant discarded, and pelleted platelets resuspended in PBS + 1 mM EDTA and protease inhibitor tablets (Pierce).

2.2. Platelet nanosponge preparation.

Aliquots (1.2 ml) of platelet preparation (~ 3 × 10⁹ cells) were used for coating 1 mg of poly(lactic-co-glycolic acid) (PLGA) core. Platelet membranes suspensions were derived through three cycles of freeze-thawing: −80°C freezing, room temperature thaw, then centrifugation at 8000 relative centrifugal force (rcf) × 7 min for pelleting, with final resuspension in water and quantification via the Pierce BCA Protein Assay Kit (Thermo Scientific). PNPs were derived in two stages. First, ~80-nm polymeric cores were prepared by nanoprecipitation using 0.67 dL/g carboxyl-terminated 50:50 PLGA (LACTEL absorbable polymers), dissolved in acetone at 10 mg/ml, 1 ml added quickly to 3 ml of water, the open mixture stirred for 12 h to evaporate the acetone to final nanoparticle concentration of 2.5 mg/ml. Second, platelet membrane preparations were combined with nanoparticle cores at a 1:1 ratio (membrane protein-to-polymer weight). Platelet membrane vesicles were dispersed and fused PLGA particles via sonication using an at a frequency of 42 kHz and a power of 100 W for 5 min to achieve membrane coating.

2.3. Analysis of PNP size distribution and coating.

A 1:1 membrane protein-to-polymer weight ratio yields particles slightly larger than the PLGA core with a surface zeta potential approaching that of the platelet membrane-derived vesicles, confirming successful membrane coating. Indeed, the coating enhanced colloidal stability of the PLGA cores, which are prone to aggregate under physiological salt concentrations. The PNPs in this study were analyzed by dynamic light scattering (ZEN 3600 Zetasizer, Malvern) in triplicate for size and consistency. For transmission electron microscopy (TEM), PNPs were laid on a 400-mesh carbon-coated copper grid (Electron Microscopy Sciences), stained with 1% uranyl acetate (EM Sciences), and studies under a Zeiss Libra 120 PLUS EF-TEM.

2.4. Bacterial strains and α-toxin.

Methicillin-resistant Staphylococcus aureus (MRSA) strain USA300/TCH1516 and its isogenic HLA mutant (Sun BioRxiv) were used in this study. Strains were propagated in Todd Hewitt broth (THB) at 37°C to mid-logarithmic phase (optical density 600 nm (OD₆₀₀) = 0.4), pelleted by centrifugation at 4,000 RPM × 10 min, washed once and resuspended to the desired dilution in PBS, and bacterial inocula confirmed dilution plating for colony-forming units (CFU). Recombinant α-toxin was purchased from Sigma (#H9395).

2.5. Platelet isolation.

Human venous blood was collected by simple phlebotomy from healthy human donors under informed consent and anticoagulated with acid-citrate-dextrose buffer (ACD; Sigma, 1:6 v/v). PRP was prepared from blood centrifuged at 1,000 rpm × 10 min with no brake, using only the upper two thirds of the PRP fractions to avoid leukocyte contamination. Platelets were isolated from PRP by centrifugation for 10 min at 1,500 rpm, and resuspended in serum-free RPMI 1640 media (Fisher Scientific) at room temperature.

2.6. Platelet cytotoxicity assay.

Human platelets (1 × 10⁷ per well) pre-treated with PNPs (1.0 mg/ml) or vehicle control were placed at room temperature for 30 min, then exposed to 3 μl supernatant of MRSA culture supernatants for 1 h. Samples were centrifuged for 5 min at 500 × g for 5 min, and lactate dehydrogenase (LDH) released into the media determined using the Promega assay.

2.7. Platelet bactericidal assay.

We evaluated bacterial killing by isolated platelets pre-treated with 1.0 mg/ml of PNPs or vehicle control for 30 min at room temperature, then infected for 1 h with 10 μl of MRSA at MOI = 0.1 bacteria per platelet. For enumeration of CFU, the challenged platelets were sonicated (Fisher Sonic Dismembrator 550) for 3 sec, and dilution plated for CFU enumeration and calculation of percent MRSA killing in comparison to the original inoculum.

2.8. Platelet activation assay

MRSA supernatants were collected by centrifugation (4000 rpm × 15 min) of overnight bacterial culture grown in THB media at 37°C. Then, 1 × 10⁷ platelets were treated with 1.25 – 2.5 μl MRSA supernatant premixed with PNPs or vehicle control. After incubation t at 37°C, samples were stained with phycoerythrin (PE) anti-human CD62p (P-selectin) antibody (Biolegend) for 20 min at room temperature, diluted in 1 ml PBS, and expression of P-selectin measured by flow cytometry (FACSCalibur, BD Biosciences) analyzed FlowJo v10.2 software. Human platelets and PNPs were separated by size gating and the human platelet population analyzed for mean fluorescence (PE).

2.9. Macrophage preparation and cell viability assays.

Human THP-1 monocytes (ATCC) were cultured in RPMI + 10% fetal bovine serum (FBS). The monocytes were differentiated into macrophages with 25 nM PMA (Sigma) for 48 h follwed by a 24 h cool down period in RPMI + 10% FBS. For monocyte cytotoxicity, 5 × 10⁵ THP-1 cells were plated in each well and were pre-treated with 1.0 mg/ml of PNPs or vehicle control for 30 min at room temperature, then exposed to a range of MRSA supernatant doses (1.25 μl to 10 μl) for 1 h at 37°C. Viability of THP-1 monocytes and macrophages was measured using the MTT Cell Proliferation Assay Kit (ab211091) that quantifies ATP conversion of MTT to formazan as read by absorbance at 590 nm.

2.10. Macrophage bactericidal assays.

For THP-1 macrophage bactericidal assays, differentiated macrophages were treated with 1.0 mg/ml PNPs or vehicle control for 30 min, and infected with MRSA at MOI = 1.0 bacteria per cell. Cells were lysed using Triton X-100 (0.025%), and serially diluted for CFU enumeration and calculation of percent MRSA killing in comparison to the original inoculum.

2.11. Macrophage oxidative burst and nitric oxide production assays.

For oxidative burst assays, differentiated THP-1 macrophages were loaded with 25 μM 2,7-dichlorofluorescein diacetate (DCFH-DA; Fisher) in Hank’s balanced salt solution (HBSS, Cellgro) lacking Ca²⁺ and Mg²⁺ and rotated at room temperature for 30 min. Macrophages were then infected with MRSA (MOI 1.0 bacteria/cell) with or without PNPs, and incubated at 37°C. Every 15 – 30 min the fluorescence intensity at 485 nm excitation/520 nm emission was compared (SpectraMax M3). For quantifying nitrite production in THP-1 differentiated macrophages in the same exposure conditions, the Greiss reagent (Promega) was used per the manufacturer’s protocol.

2.12. Human neutrophil extracellular trap (NET) staining and quantification.

Neutrophils were isolated from blood collected from healthy adult human donors under informed consent using the PolymorphPrep^™ (Progen) per manufacturer’s instructions. Next 5 × 10⁵ neutrophils were placed in wells of a 24-well plate, stimulated with vehicle control, 25 nM phorbol 12-myristate 13-acetate (PMA), MRSA alone, or MRSA premixed with PNPs for 20 min (MOI = 10 bacteria per neutrophil), incubated for 3 h at 37°C. For visualization, cells fixed in 4% paraformaldehyde were stained with an antibody against myeloperoxidase (MPO) (1:300, Calbiochem) in PBS + 2% BSA for 1h, then Alexa Fluor 488 goat anti-rabbit (Life Technologies 1:500) for 45 min, and finally counterstained with 1 μM Hoechst-3342-trihydrochloride diluted in 2% PBS-BSA for 10 min before imaging on a fluorescence microscope. In parallel, NETs were quantified using the Quant-iTTM Picogreen^® -dsDNA kit (Invitrogen) on wells in which micrococcal nuclease solution was added to release DNA of NETs into the supernatant, and 500 mM EDTA added to the solution to stop the micrococcal nuclease reaction. Picogreen solution was prepared per manufacturer’s instructions and incubated for 5 min, and fluorescence signals were measured with filter settings of 480 nm (excitation) and 520 nm (emission).

2.13. Cytokine quantification assays.

Cytokines IL-8 and IL-1β were quantified from infected THP-1 differentiated macrophages cells (4 h, 8 h, and 24 h post-infection) supernatants using ELISA kits per manufacturer’s protocol (R&D systems). Experiments were conducted in triplicate or quadruplicate.

2.14. In vivo murine S. aureus infection experiments

For survival studies, MRSA cultures were grown in THB to mid-log phase, washed once in PBS, and 3 × 10⁸ CFU intraperitoneally (i.p.) injected in outbred, 10 to 12 week-old, CD1 mice (Charles River, Wilmington, MA, USA). For the PNP treatment group, 100 μL of a 5.0 mg/ml PNPs were injected i.v. twice, immediately after MRSA infection and 3 h after the first injection. Survival was monitored daily for 6 days. In a separate challenge experiment using the same MRSA challenge dose and PNP treatment protocol, mice were sacrificed at 6 h for determination of bacterial CFU units and quantification of tumor necrosis factor (TNF) and IL-6 in serum and spleen.

2.15. Statistical analyses

All studies were conduced in duplicates or triplicates and repeated independently at least twice. All data are graphed as means with standard error of the mean (S.E.M.) or standard deviation (SD). Statistical evaluation was done by one-way ANOVA or unpaired two-tailed Student’s t-test (Graph Pad Prism 5.03). *P < 0.05, **P < 0.01 and ***P < 0.001.

3. Results

3.1. Preparation and analysis of PNPs.

In this study, our goal was to construct PNPs composed of a PLGA polymer core wrapped in natural platelet bilayer membranes. The platelet membrane shell provides a faithful mimicry of the parent platelet surface to absorb toxin virulence factors of MRSA of diverse molecular structures, with the aim of reducing platelet toxicity and preserving platelet defense function against MRSA infection (Fig. 1a). Cell membranes from recently (< 24–48 h) expired, blood bank-approved platelets concentrates from hypotonic lysis, mechanical disruption, and differential centrifugation, retain membrane composition and attendant functions [38]. EDTA was used as the anticoagulant to chelate divalent cations including calcium and avoid activating coagulation processes while stabilizing the platelets [38]. Furthermore, a protease inhibitor was added to block platelet aggregation [38]. A sonication procedure yielded membrane vesicles for fusion onto PLGA cores to create the final PNPs. The inner polymeric core stabilizes the outer membrane from collapsing and fusing with each other, optimizing in vitro and in vivo stability. After membrane coating, PNP diamster increased from 88.4 ± 5.6 nm to 120.0 ± 4.8 nm as assessed by dynamic light scattering, reflecting encapsulation of the the polymeric cores with a cell membrane bilayer (Fig. 1b, c). TEM clearly showed PLGA polymer cores uniformly coated by a unilamellar membrane, indicating successful PNP formation (Fig. 1d).

Fig. 1. Formulation and analysis of platelet membrane-coated nanoparticles (PNPs).

(a) Model showeing our rationale of applying PNPs to modulate immune cells, pathogen binding, and control in vivo therapeutic efficacy. (b, c) Dynamic light scattering to evaluate hydrodynamic size (diameter, nanometers) of the PLGA polymeric cores before and after platelet membrane coating. (d) Transmission electron microscopy images of derived PNPs using uranyl acetate counterstain.

3.2. PNPs prevent human platelet damage and dysfunction induced by MRSA supernatants.

Bacterial toxins including S. aureus α-toxin can damage and inhibit the functions of host cells [35,39]. We first investigated if PNPs could protect intact human platelets against the cytotoxic effect of MRSA supernatants. When human platelets were incubated with MRSA supernatant for 1 h, a significant LDH release reflecting platelet damage was seen; however, this was significantly diminished in the presence of PNPs: 0.52 ± 0.13 in MRSA supernatant alone, 0.27±0.08 with PNP treatment, p = 0.002 (Fig. 2a). Activation of platelets associated with release of antimicrobial peptide rich alpha granules [2] can be quantified by surface expression of P-selectin. Upon exposure of human platelets to MRSA supernatants, we found that co-incubation with PNPs significantly reduced the time to P-selectin expression, and significantly increased the intensity of its expression (Fig. 2b, c), indicating PNPs could help preserve this key platelet response function. To examine if platelet cytoprotection and functional activation improved innate immune activity, MRSA killing by platelets were assessed in the presence or absence of PNPs. A significant enhancement of platelet MRSA killing was seen upon PNP treatment (Fig. 2d), suggesting the protective effect of PNPs on MRSA-induced platelet injury might improve host defense function.

Fig. 2.

PNPs prevented human platelet damage and dysfunction induced by MRSA supernatants. (a) PNP reduced MRSA supernatant-induced platelet cytotoxicity as measured by LDH release; 0.52 ± 0.13 in MRSA supernatant alone, 0.27±0.08 with PNP treatment, p = 0.002 (b, c) P-selectin expression on platelets exposed to MRSA supernatant in the presence or absence of PNPs. PNP treatment led to a striking increase of P-selectin expression beginning at the early time point (10 min). (d, e) Increased platelet viability upon PNP treatment is accompanied by improved MRSA killing. **P < 0.01; ***P < 0.001.

3.3. PNPs prevent human macrophage damage and dysfunction induced by MRSA supernatants.

The effect of MRSA supernatants on viability of human monocyte cell line THP-1 was examined with or without preincubation of the MRSA supernatant with PNPs for 30 min beforehand. As shown in Fig. 3a, a nearly 75% reduction in LDH release was observed in the presence of PNPs. Cell viability further assessed by MTT assay, showing that viability of THP-1 monocytes after 1 h treatment was markedly increased when MRSA supernatant was premixed with PNPs compared to control (Fig. 3b). LDH release was also measured in macrophages differentiated from THP-1 cells, revealing a significant protective effect of PNPs (Fig. 3c). Finally, the THP-1 macrophages were then infected with live MRSA in the presence or absence of PNPs, showing that nanosponge treatment significantly increased macrophage killing of the pathogen (Fig. 3d).

Fig. 3.

PNPs prevented macrophage damage and dysfunction induced by MRSA supernatants (A) LDH cytotoxicity test and (B) MTT assay showed that PNP treatment showed about 75% reduction of THP-1 monocyte cell cytotoxicity. (C) PNP also improved the viability of THP-1 differentiated macrophages after MRSA supernatant treatment. (D, E) Pre-treatment of macrophages with PNP showed higher bactericidal efficiency regardless of bacterial load and incubation time. *P < 0.05, **P < 0.01 and ***P < 0.001.

3.4. PNP protection against cytotoxicity preserves macrophage and neutrophil effector responses.

We further examined the effect of PNPs on certain key macrophage functions involved in the antibacterial response. Generation of reactive oxygen species (ROS), known as the oxidative burst, is a key element of macrophage antibacterial defense, as evidenced by the high rate of S. aureus infections in NADPH oxidase deficiency (chronic granulomatous disease) patients with impaired oxidative burst function [40]. PNP treatment increased THP-1 macrophage ROS production in response to MRSA exposure (Fig. 4a). Nitric oxide generation by inducible nitric oxide synthase (iNOS) also contributes to antibacterial defense, as corroborated by more severe S. aureus infection in iNOS-deficient mice [41]. PNP treatment likewise boosted THP-1 macrophage nitric oxide production in response to MRSA exposure (Fig. 4b). S. aureus infection induces macrophage pyroptosis, a cell death program dependent on inflammasome activation and associated with IL-1β release [42]. The generation of IL-1β was modestly reduced at the early time point (Fig. 4c), coincident with reduced LDH release in cytotoxicity assays (Fig. 3a), indicating that pyroptosis is one element of the macrophage response to S. aureus challenge, and that the degree of associated cytotoxicity and cytokine release can be partially mitigated by PNP toxin absorption. Platelets express the IL-1β receptor (IL-1R) [43], so sequestration of the IL-1β cytokine by PNP may also contribute to the observed difference. Finally, S. aureus toxins [44,45] and activated platelets [46] also elicit neutrophil extracellular traps (NETs), with potential proinflammatory and procoagulant effects that may aggravate sepsis [47] or endocarditis [48]. PNP treatment significantly inhibited MRSA-induced NET formation by human neutrophils in vitro as visualized by immunostaining and measured by PICO-green quantification of released DNA (fig. 4d, e).

Fig. 4. The effect of PNPs on activation of macrophage and neutrophil bactericidal mechanisms.

(a) PNP treatment increases macrophage oxidative burst as measured by DCFH-DA assay for superoxide production (b) Increased nitrite production reflecting nitric oxide production was observed in PNP treated macrophages. (c) Reduced cytokine IL-1β production at the earliest (4h) time point in PNP-treated macrophages. (d) Immunostaining of human neutrophil extracellular traps (NETs) elicited by MRSA in the presence or absence of PNP treatment; PMA serves as a positive control. (e) Quantification of NETs by PICO-green assay. *P < 0.05, **P < 0.01 and ***P < 0.001.

3.5. PNPs improve survival in a murine MRSA systemic infection model.

Our in vitro studies above show that PNPs blocked MRSA-induced platelet and macrophage cytotoxicity, enhancing the antibacterial effectiveness of both cell types. These findings suggested a potential therapeutic benefit of PNPs against S. aureus infection in vivo. In prior rodent i.v. injection studies for pharmacokinetic and biodistribution assessment, more than 90% of PNPs were distributed to tissues within 30 min, most prominently in liver and spleen [38]. Importantly, abnormal blood coagulation has not been seen in prior uses of PNPs in murine models of noninfectious indications including immune thrombocytopenia [49] and atherosclerosis [50]. We challenged groups of 8 mice with 3 × 10⁸ CFU MRSA in 100 μl volume intraperitoneally to induce systemic infection. For the PNP treatment group, we injected 5 mg/ml PNPs directly into the bloodstream IV immediately after MRSA infection and again 3 h after the first injection (Fig. 5a). The goal of the two doses was to provide a modest extended therapeutic coverage window seeking to mitigate pathogen mediated toxic damage, propagation of cytokine storm, and prevent early death from sepsis. Survival of mice treated with PNP was significantly improved, with 37.5% of mice surviving through 5 days despite the absence of antibiotic therapy, whereas all mice in the control group died within the first 48 h (Fig. 5b). In a separate experiment using the same challenge and treatment conditions, we collected blood at 6 h post-MRSA challenge for bacterial CFU determination and serum cytokine levels. The two dose PNP treatment was associated with a significant (P = 0.036) reduction on MRSA CFU in blood (Fig. 5c), and a slight trend to reduced counts in spleen that did not reach statistical significance. Serum IL-6 levels in response to MRSA challenge were also reduced in the PNP-treated group compared to the control group (Fig. 5d).

Fig. 5. PNPs improve survival in a murine MRSA systemic infection model.

(a) Schematic diagram of setup of in vivo experiments. (b) Survival rates of mice over 144 h following an intraperitoneal injection of MRSA (3×10⁸ bacteria per mouse). 100 μg PNP at 5.0 mg/ml was injected twice, at 0 h and 3 h after bacterial inoculation (n = 8 in each group). Treatment with PNPs provided a significant survival benefit. (c) Enhanced bacterial clearance in blood during upon PNP treatment.

4. Discussion

In this study, we provided proof-of-principle of the therapeutic potential of PNPs in systemic MRSA infection. This benefit is likely multifactorial, and may include (1) reduction of toxin-associated platelet damage, which allows for more rapid and effective deployment of platelet antimicrobial activity, (2) protection of phagocytic cells such as macrophage from MRSA cytolytic injury, allowing them to more efficiently deploy ROS and NO and achieve effective bacterial killing, and (3) sequestration of certain bacterial toxins (e.g. pore-forming toxins) in the platelet membrane. Also, since platelets express Toll-like receptors (e.g. TLR2) and cytokine receptors (e.g. IL-1R), they could scavenge pro-inflammatory bacterial cell wall components and cytokines to modulate the hyperinflammation of sepsis. These latter mechanisms would parallel in vivo therapeutic effects of macrophage membrane-coated nanoparticles that conferred protection against lethality in murine models of LPS endotoxemia and Escherichia coli sepsis [34]. While prior work showed that RBC nanosponges protected mice from lethal challenge with whole secreted proteins preparations from S. aureus supernatants [31], the present work with PNPs is the first report of protection by biomimetic membrane-coated nanoparticles against systemic infection with live S. aureus.

A recent study suggests that sepsis contributes to as much as 19.7% of human mortality worldwide [51], and there remains no drug specifically approved for treatment for sepsis [52]. Expanding numbers of antibiotic-resistant bacterial strains such as MRSA further complicate effective sepsis therapeutics [53]. In this manner, the use of PNPs or other cell membrane-coated nanosponges in sepsis provides a more “universal” approach to absorption and neutralization of bacterial toxins, in contrast to monoclonal antibodies or other platforms targeting the particular molecular structure of an individual toxin or host inflammatory factor. Other factors that might support an effective therapeutic profile of PNPs include evidence that they have reduced macrophage cellular uptake, do not induced human complement activation in autologous plasma, selectively bind to damaged vascular endothelial cells from human and rodents, and likewise bind to platelet-adherent pathogens [31], potentially favoring their accumulation at common sights of endovascular S. aureus infection.

In sum, our data suggested that PNPs may protect platelets during systemic MRSA infection by absorbing circulating toxins so that platelets sustain less damage, survive longer, activate more quickly, and have stronger antibacterial effects. PNPs might also protect and support the innate immune function of macrophages or other phagocytic cell types. This biomimetic detoxification strategy merits further exploration as an adjunctive therapy to improve clinicals in patients with MRSA bloodstream, thereby expanding the current approach to clinical management.

5. Limitations of this study

Murine studies of systemic MRSA infection cannot fully recapitulate the tempo and kinetics of human infection, as mice are relatively resistant to the pathogen and very high challenge doses are required to cause organ dissemination and mortality risk. Also in patients with presumed sepsis, rapid clinical intervention is required often before microbiological confirmation of the pathogen is established, and may have one or more comorbidities that increase risk of adverse outcome. Nanosponge therapeutics are thus envisioned as an addition to standard of care.