Materials. 4T1 cells, RAW264.7 cells and MC3T3-E1 cells were obtained from Shanghai Institute of Cell Biology, Chinese Academy of Science. S. aureus (ATCC43300) and E.coli (ATCC35218) were purchased from the American Type Culture Collection. BMDCs were isolated from the bone marrow of 6-week-old C57 mice according to a precious reported method49. Male male BALB/c mice (6–8 weeks) and C57BL/6 mice (6–8 weeks) were purchased from Laboratory Animal Management Department, Shanghai Family Planning Research Institute. All of the animal experiments performed in this research were approved by the Institutional Animal Care and Use Committee of the Shanghai Sixth People’s Hospital.
Dulbecco’s modified Eagle’s medium (DMEM), 1640 media, penicillin and streptomycin, fetal bovine serum (FBS), PBS, SYTOX/PI staining kits and ACK lysis buffer were obtained from Thermo Fisher Scientific. Antibodies, True-Nuclear™ Transcription Factor Buffer Set, Cyto-Fast™ Fix/Perm Buffer Set used for flow cytometry were purchased from Biolegend company, if not specially indicated. TNFα, IL6 Elisa kits were purchased from R&D Systems, IFNγ and IL10 Elisa Kits were purchased from Dakewei Biotech. IgG and IgM Elisa kits were purchased from Crystal Chem. The titanium implant (plateau, 2 mm×1.5 mm; stem diameter, 0.6 mm; stem length, 2 mm) used for osteomyelitis model was custom-made by Shanghai Sunshine-laser Company. Flow cytometric analysis was performed on FACSVerse flow cytometer (BD Biosciences). All experiments involving US treatment were performed with a US transducer (Chattanooga Co., USA).
Synthesis of HMMP. Hollow porous MnOx was synthesized with a previous reported method25. Briefly, MnOx@SiO2 was synthesized by simply mixing the as-made silica nanoparticles with KMnO4, utilizing the reduction reaction between organosilica and KMnO4. Then the silica template was removed with Na2CO3 solution, the obtained product was then washed with water thrice, yielding hollow porous MnOx. MnOx was mixed with PpIX in a ratio of 1:10, followed by magnetic rotary stirring for 12 h at room temperature. After washing with ethanol thrice, the product was then dispersed into water.
To extract membranes from Raw264.7 and 4T1 cells, the respective cells were suspended in hypotonic lysing buffer, followed by disrupting with a Dounce homogenizer. The homogenized solution was centrifuged at 20,000 g at 4°C for 30 min. The supernatant was harvested and centrifuged at 100,000 g at 4°C for 35 min. The pellets were harvested and washed with PBS containing protease inhibitor, and sonicated for 5 min. The membrane vesicles were then obtained after extrusion through 400 nm, 200 nm, and 100 nm polycarbonate porous membranes (Avanti Polar Lipids). The membrane vesicles from Raw264.7 and 4T1 cells were mixed in a weight ratio of 1:1, sonicated for 5 min, followed by extrusion through 100-nm polycarbonate porous membranes. For membrane coating, the hybrid membrane was mixed with MP core at a weight ratio of 2:1, then sonicated for 2 min, followed by serially extrusion through 400 nm and 200 nm polycarbonate porous membranes.
Characterization. TEM, HR-TEM and elemental mapping was conducted on a JEM-2100F electron microscope operated at 200 kV. UV − vis − NIR absorption spectra were performed on a UV-3101PC Shimadzu spectroscope. Zeta potential measurements were performed on a Zetasizer Nanoseries (Nano ZS90, Malvern Instrument Ltd.). The diameters of different nanoparticles were analyzed with Nano Measure software. The degradation behaviors of MnOx were studied in redox conditions with 4mM GSH under constant shaking. The samples were harvested at predetermined time (0, 3, 6, 12 and 24 h), and observed with TEM. To study the PpIX release, a solution of MP was dialyzed against PBS with or without 4mM GSH under constant shaking, the amounts of PpIX release were measured at preset time using UV–vis spectra.
RAW264.7 membrane vesicles, 4T1 membrane vesicles and HMMP containing equivalent total proteins were prepared using BCA assay (Beyotime). The proteins were resolved on SDS–polyacrylamide gel electrophoresis (SDS–PAGE) gels, followed by Coomassie Brilliant Blue staining or standard WB. The primary antibodies used were anti-TLR6 (Proteintech, 22240-1-AP) and anti-E-cadherin (Proteintech, 20874-1-AP). Blots were incubated with anti-Rabbit IgG, HRP-linked Antibody (Proteintech, SA00001-2). The signals were detected by the HRP-based chemiluminescence analysis on ChemiDoc (BioRad). To evaluate ROS generation, 20 µL DPBF (1 mg mL− 1) was added in 1 mL HMMP (20 µg mL− 1), the mixture was then exposed to US stimulation (40 kHz, 1.5 W cm− 2). The absorbance changes of DPBF at 420nm were detected and recorded. PBS and PpIX solution (20 µg mL− 1) were used as controls.
Study design. For in vitro studies, cells or bacteria treated with PBS, PpIX, MnOx@PpIX and hybridized membrane@MnOx@PpIX were termed as Control, P, MP and HMMP, respectively. The naming method in vivo is in analogous to that in vitro. To load PpIX, MnOx was mixed with PpIX in a ratio of 1:10 under ultrasonication, follow by magnetic rotary stirring, thus yielding MnOx@PpIX.
Biosafety test. MC3T3-E1 cells (100 µL, 5,000 cells per well) were cultured in 96-well microplates overnight. Then, the culture media were replaced with 100 µL of fresh media containing PBS or HMMP with gradient concentrations. After further culture for 24 h, the cells were washed with PBS thrice, then CCK-8 solution was added. 90 min after the co-incubation, the absorbance at a wavelength of 450 nm was determined with a microplate reader (Bio-TekELx800, USA).
In vivo biodistribution of HMMP. Mice osteomyelitis model was first constructed. The method to establish osteomyelitis model was introduced in detail in the following “In situ” osteomyelitis model establishment. HMMP nanoparticles labelled with Cy5.5 (100 µL, 2 mg·mL− 1) were intravenously injected 3 days after the induction of infection. MP nanoparticles labelled with Cy5.5 were used as control. Signals of HMMP nanoparticles were monitored with an IVIS fluorescence imaging system (VISQUE InVivo Elite) over time. 12 h after the injection, the major organs were harvested for IVIS observation.
In vivo detection of oxygen saturation by PA. Subcutaneous implant-related infection model and sterile model were constructed as follows. Six-week-old BALB/c mice were anesthetized with pentobarbital via intraperitoneal injection. After anesthesia, the back was shaved and disinfected, an incision was then made to expose the subcutaneous cavity. Next, a sterile polyethylene gasket (diameter: 6 mm, thickness: 1 mm) was placed into the cavity followed by wound suturing. Finally, 100 µL of 106 CFU mL− 1 S. aureus or PBS was injected onto the polyethylene gasket to establish infectious or sterile implant model, respectively. Three days after the surgery, mice were injected intravenously with HMMP (100 µL, 2 mg mL− 1). The oxygen levels in the subcutaneous region were monitored with PA imaging (Fuji).
In vitro antibacterial tests. 1 mL of S. aureus and E. coli at the concentration of 106 CFU mL− 1 were co-incubated with PBS, P, MP or HMMP (20 µg mL− 1) for 8 h, followed by US irradiation (40 kHz, 1.5 W cm− 2, 1 min per cycle, 5 cycles). The bacterial suspension and adhered biofilm in the well were harvested, respectively. Bacterial viability from each sample was tested with CFU assay. The bacterial membrane integrity was tested with Live/Dead assay. Briefly, the bacterial suspension was spun down and wash with PBS for three times before the SYTO9/PI staining was added. After co-incubation for 15 min in the dark, the solution was further washed with PBS for three times. An aliquot was aspirated and tested with flow cytometry. The procedure for the detection of membrane potential resembled the SYTO9/PI staining protocol with the following exceptions. Bacterial samples were stained with 3 mM 3,3′-diethyloxacarbocyanine (DiOC2) dye. The DiOC2 dye forms tetramers that fluoresce at 630 nm when the membrane potential is intact, while it forms dimers that fluoresce at 530 nm when the membrane potential is lost. The stained solution was also analyzed with flow cytometry. To detect the intracellular ROS level, the bacteria was stained with DCFH-DA (20 µM), and analyzed using flow cytometry.
For TEM observation, bacterial samples in the control group and HMMP group were fixed with 2.5% glutaraldehyde at 4°C for 2 h, then washed with PBS three times. Then, samples were serially dehydrated with ethanol at increased concentrations, permeated in the embedding medium and embedded for 48 h at 60°C. Ultrathin sections of the embedded samples were prepared on the grids and stained with uranyl acetate. Finally, the grids were observed with TEM.
Phenotyping in vitro cultures of macrophages and BMDCs. Macrophages were seeded in 24-well plate at a cell density of 10,000 cells per well and cultured overnight. Then, the cells were further co-cultured with PBS, P, MP or HMMP (20 µg mL− 1) for 8 hours, followed by US irradiation (40 kHz, 1.5 W cm− 2, 1 min per cycle, 5 cycles), then they were washed and harvested for further analysis. For macrophage polarization analysis, pellets were resuspended in 200 µL of cold PBS with 1 µL CD16/32 antibodies for Fc blocking, and were incubated at 4°C for 10 min. Afterwards, the cells were incubated with PE/Cy7-labeled anti-CD80 (16–10A1) for surface staining at 4°C in a dark enclosed space. After 30 min. cells were washed and resuspended with PBS three times. To stain the intracellular marker, cells were fixed and permeabilized with Cyto-Fast™ Fix/Perm Buffer Set following of the manufacturer’s instructions, then stained with PE-labeled anti-CD206 (C068C2f) or another 30 min, followed by flow cytometric analysis.
To assess the in vitro phagocytic clearance of bacteria by macrophages, the harvested macrophages were cultured with bacteria labelled with CFDA-SE in a ratio of 10:1 at 37°C in 5% CO2 atmosphere. After coculture for 2h, the macrophages were spun down, and washed with PBS three times to remove bacteria in the supernatant. The phagocytosis of bacteria was then evaluated by flow cytometric analysis. For transcriptional analysis, total macrophage RNA was extracted with an Tizol (Invitrogen) followed by sequencing with DNBseq Platform (BGI-Wuhan, China). BMDCs were exposed to different treatments using the same protocol as macrophages except for the following tests. The BMDCs were incubated with CD16/32 for 10 min and further incubated with antibodies against CD11c (N418), CD80 (16–10A1) and CD86 (GL-1) under the guidance of the manufacturer’s protocols. Lastly, BMDCs were washed with PBS three times and analyzed with flow cytometry. The gene expression levels of inflammatory related genes were analyzed with EZBioscience Qpcr kit (USA). RNA extraction and reverse transcription were performed according to the manufacturer’s recommendations. The final quantitative PCR were performed on Roche LightCycler 480.
“ In situ ” osteomyelitis model establishment. To evaluate the direct therapeutic effect of our nanoplatform, we constructed “in situ” mouse osteomyelitis model using previous reported protocol6 with some modifications. The male C57 mice (6 weeks old) were allocated to four groups randomly, naming as control, P, MP and HMMP. The mice were deeply anesthetized by intraperitoneal injection of pentobarbital. After anesthesia induction, the left leg of the mice was manually depilated and disinfected. The knee joint was exposed after layer-by-layer incisions, then a bone channel was created in the tibial plateau with a medical electric drill. 50 µL of 106 CFU mL− 1 S. aureus were injected into the channel followed by the placement of a custom-made implant. The wound was then carefully sutured layer-by-layer. Then the mouse was housed carefully for 1 day for the development of osteomyelitis. Treatment was initiated 24 h after the induction of infection, the mice were treated with 100 µL of either PBS, P, MP or HMMP (2 mg mL− 1) via intravenous injection, and then irradiated with US (40 kHz, 1.5 W cm− 2, 1 min per cycle, 5 cycles). To assess the therapeutic effect of different treatments, general observations, clinical imaging assessment, and bacterial viability test were performed at predetermined time points. Knee perimeter was recorded over time with a caliper and quantified using the formula of knee perimeter = π × a×b, where a and b are the large and small diameter, respectively. Infection severity was monitored using 3.0 T MR scanner MRI (CG NOVILA 7.0T). 21 days after the surgery, the implant were extracted and peri-implant tissues were dissected for bacterial CFU assay. The explanted tibias were fixed and decalcified in EDTA for 2 weeks, pathological sections were then prepared for Gimesa staining. Immunofluorescence staining was also performed for CD45, iNOS and CD206.
The both ends of the tibia were cut, and bone marrow suspensions were collected using an injector. The obtained bone marrow suspensions were digested in dissociation buffer (100 µg mL− 1 deoxyribonuclease and 1 mg mL− 1 collagenase IV) at 37°C for 30 min. The cell suspensions were then mashed over 70 µm strainers to obtain single cell suspension. Erythrocytes were depleted with ACK lysis buffer (Gibco) and washed with ice-cold PBS three times. For RNA-seq analysis, the total RNA was extracted with Trizol and followed by sequencing with DNBseq Platform. For the analysis of immune cells composition, the cell suspensions were incubated with CD16/32 for 10 min and then stained with antibodies against CD45 (30-F11), CD3 (145–2C11), CD4 (GK1.5), CD8 (53–6.7), F4/80 (BM8), CD80 (16–10A1), CD86 (GL-1), I-A/I-E (M5/114.15.2), CD11c (N418), CD19 (6D5), CD138 (281-2), NK1.1 (PK136), CD11b (M1/70), and Gr1 (RB6–8C5) on the ice for 30 min. The CD206 staining protocol is the same as that used in vitro. After surface staining, the cell suspensions were fixed and permeabilized with True-Nuclear™ Transcription Factor Buffer Set, then stained with antibodies against Foxp3 for 30 min on the ice in a dark enclosed place. After staining, the cell suspensions were centrifuged and washed with ice-cold PBS thrice, then resuspended in 500 µL ice-cold PBS for flow analysis. The cytokines (IL-10, TNFα, IL-12 and IFN-γ) and antibodies (IgM and IgG) in the bone marrow were tested with ELISA kits according to the manufacturer’s recommendations.
In vivo monitoring of systemic immune responses. Contralateral mouse osteomyelitis model was established to evaluate the systematic antibacterial immune responses as follows. First, we established an “in situ” osteomyelitis model in the left knee and treated with different methods as above mentioned. 14 days later, osteomyelitis model was established on the right knee using the same method as above. Again, the evaluation approaches for infection severity such as general observation, MRI imaging, bacterial viability test and Gimesa staining described above was used. Procedures for isolating and staining spleen-associated immune cells were similar to the protocols used for bone marrow. Briefly, the spleens were dissected and cut into small pieces, enzymatically degraded, and mechanically disrupted 70 µm cell strainers. After red blood cell removal with ACK lysis buffer, the cell suspensions were incubated with CD16/32 for 10 min and then stained with antibodies against CD45 (30-F11), CD3 (145–2C11), CD4 (GK1.5), CD8 (53–6.7), CD19 (6D5), CD138 (281-2), NK1.1 (PK136), CD11b (M1/70), and Gr1 (RB6–8C5) on the ice according to the manufacturer’s instructions. The intracellular staining of Foxp3 (MF-14) was the same as the procedures for bone marrow. Serum cytokines (IL-10 and TNFα) and antibody levels (IgM and IgG) were determined with Elisa kits following the manufacturer’s instruction.
Prophylactic vaccine study. To determine whether HMMP vaccine could elicit long-term immune memory responses or not, in 21 days after the establishment of in situ osteomyelitis model, we undertook thorough debridement, implant removal and local vancomycin (1mg, Sangon Biotech) administration. After the surgery, 15 mg kg− 1 of vancomycin was administrated intravenously per day. The procedures fully mimic the clinical management of osteomyelitis. 7 days later, after the infection was controlled, the mice received a re-challenge surgery in a manner identical to that in the “in situ” osteomyelitis model. The data of general photos, knee perimeter, MRI imaging were collected at preset time after the surgery. In 21 days after the surgery, the mice were euthanized. Bacterial viability test in vivo and Gimesa staining of the infected bone were conducted with above mentioned methods. The bone and spleen were harvested, and processed into single-cell suspension as above. Then bone marrow single-cell suspensions were stained with antibodies against B220 (RA3-6B2), IgG (Poly4053) and IgD (11-26c.2a), while the spleen single-cell suspensions were stained with antibodies against CD3 (145–2C11), CD4 (GK1.5), CD8 (53–6.7), CD44 (IM7), and CD62L (MEL-14), according to the manufacturer’s instructions.