Antibodies and cells
The mouse melanoma B16F10 cells, CT26 cells and 4T1 cells were tagged with luciferase for in vivo bioluminescence imaging. The B16F10, NIH/3T3, Sarcoma 180 (S180) and Raw 264.7 cells were purchased from ATCC. Luciferase-expressed B16F10 cells, CT26 cells and 4T1 cells were obtained from Imanis Life Sciences Inc. Cells were cultured in the CO2 incubator (Fisher) at 37 °C with 5% CO2 and 90% relative humidity. The cells were sub-cultured about every 2 days at 80% confluence. The antibodies used in this study were summarized as follows (company, clone, category number): GoInVivo Purified anti-mouse CD279 (PD-1) (BioLegend, RMP1-14, 114114), Fluorescein isothiocyanate (FITC)-anti-mouse CD45 (BioLegend, 30-F11, 103108), APC-anti-mouse F4/80 (BioLegend, BM8, 123116), PerCP/Cy5.5-anti-human/mouse CD11b (BioLegend, M1/70, 101227), APC-anti-mouse CD3 (BioLegend, 17A2, 100236), FITC-anti-mouse CD4 (BioLegend, GK1.5, 100406), PE-anti-mouse CD8a (BioLegend, 53-6.7, 100708), FITC-anti-mouse IFNγ (BioLegend, XMG1.2, 505806), PerCP/Cy5.5-anti-human/mouse Granzyme B (BioLegend, QA16A02, 372212), PE-anti-mouse CD45 (BioLegend, 30-F11, 103106), FITC-anti-human/mouse CD11b (BioLegend, M1/70, 101206), Alexa Fluro 594 anti-mouse CD8a (BioLegend, 53-6.7, 100758), Alexa Fluro 647 anti-mouse F4/80 (BioLegend, BM8, 123121), PE-anti-mouse CD62P (BioLegend, RMP-1, 124807), PE-anti-mouse CD41 (BioLegend, MWReg30, 133905), FITC-anti-mouse CD9 (BioLegend, MZ3, 148305), FITC-anti-mouse CD61 (BioLegend, 2C9.G2, 104305), Precision Count Beads (BioLegend, 424902). All antibody dilutions were performed following the manufacture’s guidance (diluted ~200 times for immediate use). FlowJo software was used to analyze flow cytometry data.
Preparation and characterization of PLX-NP
Dextran was modified with Pyridinium P-toluenesulfonate (PPTS) and 2-ethoxypropene for the preparation of nanoparticles. Briefly, 1 g dextran was dissolved in 10 ml anhydrous dimethyl sulfoxide, and 0.062 mmol PPTS (Sigma Aldrich) and 37 mmol 2-ethoxypropene (Matrix Scientific) were added to the dextran solution during stirring. After 30 min, the reaction was quenched by adding 1 ml triethylamine (Sigma Aldrich) during stirring at room temperature, resulting in 2-ethoxypropene-modified dextran (designated m-dextran). m-dextran was then precipitated in the basic water and collected by centrifugation.
To prepare PLX-NP, 10 mg m-dextran and 0.5 mg PLX were firstly dissolved in 2 ml dichloromethane (DCM). Afterwards, 4 ml 3% (w/v) poly (vinyl alcohol) (PVA) solution was then slowly added to the DCM solution followed by the sonication for emulsification. Then, the emulsion was added to 20 ml 0.3% (w/v) PVA solution and stirred for one hour for solvent evaporation. The nanoparticles were collected by centrifugation at 21,900 g for 45 min. The resulted PLX-NP was analyzed by DLS to determine the average size, and the morphology of nanoparticles was characterized by TEM. To study the in vitro release property of PLX, PLX-NP was suspended in 3 ml phosphate-buffered saline (PBS, pH 6.5) with 0.1% Tween 80 and loaded into a 3 ml 20,000 MWCO dialysis cassette (Thermo scientific). The cassette was placed into a container with 4 L PBS with 0.1% v/v Tween 80, and at predetermined time points, 10 µl supernatant was collected, dissolved by acetonitrile, and the concentration was analyzed by high-performance liquid chromatography (HPLC). Similarly, for the in vivo drug release of PLX at the tumor site, the surgical removal of tumor mass and the implantation of the hydrogel systems into the mice were performed. And then, at different time points, the hydrogel systems were taken out and de-crosslinked with EDTA, and the remaining drug concentration was measured by HPLC as described above. To further test the cytotoxicity of PLX-NP against macrophages and normal cells, the cell viability of Raw 264.7 and NIH/3T3 cells treated with different concentrations (0 to 125 µg/ml) of PLX-NP for 24 h was determined by MTT assay.
Preparation and characterization of anti-PD-1-conjugated platelets
The mouse platelets were purified from whole mouse blood collected by retro-orbital bleeding. Afterwards, the blood was centrifuged for 20 min at 100 g, followed by centrifugation for 20 min at 800 g. The platelets were collected and suspended in PBS with the addition of 1 µM PGE1 to prevent platelet activation. To conjugate aPD-1 antibodies on the surface of platelets, aPD-1 antibodies were first reacted with SMCC linkers for 2 h at 4 °C at a molar ratio of 1:1.2. And then, the excess SMCC linkers were discarded by centrifugation at 21,900 g for 10 min at 4 °C, using 3000 KDa MWCO ultrafiltration tubes. The synthesized SMCC-aPD-1 was added into Traut’s reagent-pretreated platelets and stirred at room temperature for 1 h to obtain P-aPD-1. The excess antibodies were removed by centrifugation at 1500 g for 20 min. To characterize the conjugation efficacy of aPD-1 with platelets, P-aPD-1 was subjected to 0.1% Triton-X100 buffer to release aPD-1 under unltrasonication, and the amount of aPD-1 was determined by ELISA kit (Rat IgG total ELISA Kit, Invitrogen).
To demonstrate the successful conjugation of aPD-1 on the surface of platelets, confocal microscopy (Nikon A1RS) imaging and flow cytometry (ThermoFisher Attune) analysis were performed. Briefly, aPD-1 was stained by FITC, and platelets were stained by Wheat Germ Agglutinin 594 (WGA 594). And then, the functionality of aPD-1-conjugated platelets was studied by two assays: collagen binding assay and surface antigen expression study. First, collagen from the human placenta (Sigma) was reconstituted to a concentration of 1.0 mg/ml and was added to a confocal dish for incubation overnight at 4 °C. The coated confocal dishes and uncoated confocal dishes were further blocked with 1 ml 2% (w/v) bovine serum albumin in PBS for two hours and washed with PBS. Rhodamine B-labeled naïve platelets and P-aPD-1 were then added to the dishes and incubated for 5 min. The unbinding platelets and P-aPD-1 were washed with PBS, and then the dishes were visualized under the confocal microscope and analyzed with NIS-Elements viewer. The surface protein expression of P-aPD-1 was investigated by flow cytometry by staining with various antibodies (CD61, CD41, CD9 diluted by ~200 times), compared with unmodified platelets. Furthermore, the platelet activation marker CD62P was characterized by flow cytometry after P-aPD-1 was treated with thrombin.
Preparation of alginate-based hydrogel
The alginate-based hydrogel was formed by adding 10 µl 100 mg/ml CaCl2 solution into 200 µl 10 mg/ml alginate solution in HEPES buffered saline. To study the in vivo degradation rate of alginate-based hydrogel, alginate was conjugated with Cy5.5. Briefly, 50 mg sodium alginate was dissolved in 5 ml of HEPES buffer (50 mM, pH = 5), 14 mg NHS, 116 mg EDC, and 18 mg NH2-PEG3-N3, and the mixture were stirred for 30 min at the room temperature. Then the pH of the solution was adjusted to 7.5–8.0 and reacted overnight at the room temperature. The synthesized alginate-N3 was purified by a 3-day dialysis against water. The 100 µl of 1% (w/v) solution of alginate-N3 was incubated with 30 µl of 1 mM Cy5.5-DBCO for four hours. The final product was purified using dialysis against water. The synthesized alginate-Cy5.5 was mixed with unreacted alginate at a volume ratio of 1:1 to form Cy5.5-labeled hydrogel. The Cy5.5-labeled hydrogel was implanted into the C57BL/6 mice subcutaneously, and the fluorescence signals were monitored by IVIS (Perkin Elmer).
Characterization of hydrogel loaded with PLX-NP and P-aPD-1
The predetermined amounts of PLX-NP and P-aPD-1 were suspended into 50 µl HEPES buffered saline and mixed with 10 µl 100 mg/ml CaCl2 solution, adding to 200 µl 10 mg/ml alginate solution in HEPES buffered saline to form PLX-NP-P-aPD-1@Gel. The Cryo SEM imaging was first performed to visualize the morphology of PLX-NP-P-aPD-1@Gel, and confocal microscopy was also used to characterize the successful loading of NP and P-aPD-1 in the hydrogel. WGA 594 labeled P-aPD-1 and FITC-loaded NP were prepared and loaded into the alginate hydrogel, and the hydrogel was then observed under the confocal microscope.
To examine the loading capacity of alginate-based hydrogel, different volumes of suspension containing a fixed amount of PLX-NP and P-aPD-1 were loaded into a 200 µl alginate solution in a 96-well plate. After 3 min, hydrogels were removed from wells, and the P-aPD-1 was counted using hemocytometers under the microscope, and the remaining amounts of PLX were determined by HPLC.
To study the platelet and aPD-1 release from hydrogel, hydrogel containing 1 × 108 P-aPD-1 was placed into a 40 µm cell strainer in a 6-well plate and submerged by 5 ml medium. Afterwards, 1 U/ml thrombin was added to trigger the activation of platelets. At predetermined time points, 50 µl samples were collected, and the same amount of medium was added back to the wells. The platelets in collected samples were counted using hemocytometers under the microscope. And then, the collected samples were centrifuged for 20 min at 800 g, and the concentration of aPD-1 in the supernatant was detected by rat total IgG ELISA kit.
In vivo macrophage depletion ability of PLX-NP@Gel
The C57BL/6 mice (Male, aged 6–8 weeks) were purchased from Jackson laboratory. The animal study protocol was approved by the Institutional Animal Care and Use Committee (IACUC) at the University of Wisconsin-Madison. To build the in vivo mouse B16F10 melanoma recurrence model, B16F10 cells were injected into the right flank of C57BL/6 mice subcutaneously. Once the tumor size reached about 150 mm3, tumors were surgically removed as much as possible under a microscope, and different treatments were then applied to the tumor cavities, including saline, NP@Gel, free PLX, PLX-NP, and PLX-NP@Gel. The free PLX and PLX-NP were slowly dripped onto the tumor cavity after surgery. Tumors were collected on day 12 for following studies. Collected tumor tissues were digested by collagenase and then were dissociated by tissue dissociator (gentalMACS) to obtain single-cell suspension. The cell suspension was stained with FITC-anti-mouse CD45, APC-anti-mouse F4/80, APC-anti-mouse CD3, PerCP/Cy5.5-anti-human/mouse CD11b, FITC-anti-mouse CD4, PE-anti-mouse CD8a, FITC-anti-mouse IFNγ antibodies and analyzed using flow cytometry with Attune NxT Flow Cytometer software (All these antibodies were diluted by ~200 times). Furthermore, the collected tumors were embedded in optimal cutting temperature (OCT) compound and frozen in a −80 °C freezer for sectioning. To directly visualize the macrophages and T cells, the section slides were stained by the Alexa Fluro 594 anti-mouse CD8a, Alexa Fluro 647 anti-mouse F4/80 antibodies, and Hoechst 33342 trihydrochloride (Invitrogen, H3570), and then were imaged by the confocal microscope.
In vivo anti-tumor efficacy of PLX-NP-P-aPD-1@Gel
The in vivo mouse B16F10 melanoma recurrence model was established as described before. The in vivo mouse CT26 colon cancer recurrence model was established by subcutaneously injecting CT26 cells in the right flank of the BALB/c mouse. The in vivo mouse 4T1 breast cancer recurrence model was established by injecting 4T1 cells into the BALB/c mouse mammary fat pad. Both B16F10, CT26 and 4T1 cells were tagged with luciferase for in vivo bioluminescence imaging. For the above tumor models, different treatments were applied to the tumor cavities, including saline, NP-P@Gel (blank nanoparticle and unmodified platelets co-loaded hydrogel), PLX-aPD-1@Gel (free PLX and aPD-1 co-loaded hydrogel), PLX-NP@Gel (PLX-NP loaded hydrogel), P-aPD-1@Gel (P-aPD-1 loaded hydrogel), PLX-NP+P-aPD-1 (free PLX-NP and P-aPD-1), PLX-NP-P-aPD-1@Gel (PLX-NP and P-aPD-1 co-loaded hydrogel). The free PLX and PLX-NP were slowly dripped onto the tumor cavity after surgery. The dose of aPD-1 was 0.1 mg/kg, and the dose of PLX was 5 mg/kg. From day 0, at predetermined days, the bioluminescence signals of the resected tumor tissues were monitored by IVIS, after intraperitoneally injecting 150 mg/kg D-luciferin per mice in 100 μl PBS. Mice were imaged after 5 min with 0.5-second exposure. Bioluminescence images were analyzed using Living Image Software v.4.3.1 (Perkin Elmer). The weight and survival of mice were monitored during the time-course of treatment. Once the tumor volume was larger than 1.5 cm3 (calculated based on the equation: length × width2 × 0.5), mice were euthanized following the animal protocols. We further established a 4T1 metastasis model to investigate whether PLX-NP-P-aPD-1@Gel could prevent lung metastasis after local implantation. The 4T1 local recurrence model was built and treated with the hydrogel-based delivery systems. On day 21, the lungs of the mice in different treatment groups were harvested and fixed with Bouin’s solution. The lung metastasis in saline treatment group was validated as evidenced by visible metastatic nodules and further H&E staining. After the lung tissue was photographed, the number of tumor metastases on the surface was counted. H&E assay was performed to observe the tumor metastases inside the lungs. Also, the S180 sarcoma tumor recurrence model was established and treated with the similar method as mentioned above. The recurrent tumor volume and survival of the mice were monitored accordingly. The in vivo mouse B16F10 melanoma recurrence model in T deficient mice (Rag−/− mice) was established and the tumor growth and survival were monitored accordingly. To evaluate the treatment efficacy of PLX-NP-P-aPD-1@Gel to the distant tumor, we first established a primary B16F10 tumor model in the right flank of the mice and 6 days later, the distant tumor model was established by injecting B16F10 tumor cells into the left flank of the mice. Then one day later, the primary tumor was resected, and the hydrogel-based treatment systems were implanted. The growth and tumor volume of the distant tumor were monitored accordingly.
To investigate whether P-aPD-1 could be activated at the tumor site, WGA 594-labeled P-aPD-1 was loaded into the hydrogel. One week after subcutaneous B16F10 tumor implantation, the tumor was surgically resected with ~5% tumor mass left, and hydrogel was put into the tumor cavity beside the residue tumor. After three days, the residue tumor was collected for the frozen section. The section slides were imaged under the confocal microscope.
To study in vivo macrophage depletion ability and enhanced T cell infiltration of NP-PLX-P-aPD-1@Gel, the resection tumor model was established as above-mentioned and treated with saline, NP-P@Gel, PLX-aPD-1@Gel, PLX-NP@Gel, P-aPD-1@Gel, PLX-NP+P-aPD-1, PLX-NP-P-aPD-1@Gel. The free PLX and PLX-NP were slowly dripped onto the tumor cavity after surgery. On day 12, tumors were collected and digested by collagenase and then were dissociated by a tissue dissociator to obtain single-cell suspension. The cell suspension was stained with PE-anti-mouse CD45, FITC-anti-human/mouse CD11b, APC-anti-mouse F4/80, APC-anti-mouse CD3, FITC-anti-mouse CD4, PE-anti-mouse CD8a, and PerCP/Cy5.5-anti-human/mouse Granzyme B antibodies, and analyzed using flow cytometry (All these antibodies were diluted by ~200 times). Furthermore, the collected tumors were also embedded in OCT for frozen section. The section slides were stained by the Alexa Fluro 594 anti-mouse CD8a, Alexa Fluro 647 anti-mouse F4/80 antibodies, and Hoechst 33342 trihydrochloride for observation under the confocal microscope (All these antibodies were diluted by ~200 times). Moreover, to study the cytokine generation after treatments, tumor tissues were collected after 12-day treatments. The tumor tissues were resuspended in NP40 Cell Lysis Buffer (Alfa Aesar) at 4 °C and then were mechanically grounded. The homogenate was centrifuged for 10 min at 4,000 g, at 4 °C to collect the supernatant. Afterwards, 10 µl supernatant was used to be detected using corresponding cytokine ELISA kits following the manufacture’s guidance.
In vivo anti-tumor efficacy of PLX-NP@Gel and systemic injection of P-aPD-1
To build the in vivo mouse B16F10 melanoma recurrence model, B16F10 cells were injected into the right flank of C57BL/6 mice subcutaneously. Once the tumor size reached about 150 mm3, surgery was performed to remove tumor mass as much as possible under a microscope, and different treatments were applied to the tumor cavities, including saline, PLX-NP@Gel, PLX-NP@Gel with intravenous injection of free aPD-1 antibodies every other day for three times starting from day 0, and PLX-NP@Gel with intravenous injection of P-aPD-1 every other day for three times starting from day 0. The dose for aPD-1 was 0.5 mg/kg per mice, and the dose for PLX was 5 mg/kg per mice. From day 0, the volumes of the tumor tissues were measured using a digital caliper and calculated based on the equation: length × width2 × 0.5. The survival of mice was recorded for 60 days. The maximal tumor size permitted by our animal protocol approved by the IACUC at the University of Wisconsin-Madison is 2 cm3. Once the tumor volume is larger than 1.5 cm3, mice will be euthanized following the animal protocols to prevent further suffering of the mice in this study. Moreover, the tumor tissues were collected for weight measure at three weeks after the hydrogel implantation, and the representative tumor tissues from each group were imaged. To study in vivo macrophage depletion ability and enhanced T cell infiltration of different groups, the flow cytometry, and the ELISA assays for IFNγ and TNFα were performed as mentioned above.
Statistics
All the results are shown as mean ± s.d. The GraphPad Prism software was used to perform statistical analysis and ANOVA was used to compare multiple groups (>two groups) statistically. Log-rank test was performed for the statistical analysis of the survival study. A P value lower than 0.05 (*P < 0.05) was considered as the threshold for statistical significance among control groups and experimental groups.
Reporting summary
Further information on research design is available in the Nature Research Reporting Summary linked to this article.
