Fabrication of semi-Burkholderia Pseudomallei, loading tumor lysates and CpG
Semi-B. Pseudomallei (SB) loading tumor lysates and CpG were constructed somewhat as previously reported but without hydrothermal process which was replaced by a deeply dialysis and then an electrophoretic process 8. In brief, B. pseudomallei (strain BPC006) 32was first attenuated by ionizing radiation at 10 Gy (RS–2000-PRO; Rad Source, GA, USA), and then treated with 95% ethyl alcohol for 1 h following another treatment by 1.5% (v/v) Triton X–100 for 5 h. After that, the bacteria were put in a 0.22 μm poly tetra fluoroethylene (PTFE) membrane filter (Jinteng, Shanghai, China) and then dialysed in ultrapure water at 4°C for 5 h. After that, the membrane filter was merged in the electrophoresis tank of Mini Transblot device (Bio-Rad, Hercules, CA) containing Tris-glycine-SDS buffer (T7777, Sigma-Aldrich), and then electrophoresis was performed by 100 V for 2 h at 4°C. After that, the bacteria were washed twice with ultrapure water. The tumor lysates used in this study were prepared as previously reported 33, and CpG1826 (5′-TCCATGACGTTCCTGACGTT–3′) was purchased from TAKARA (Dalian, China). To load the SBs with CpG (SB-C), tumor lysates (SB-L), a combination of tumor lysates and CpG (SB-LC), tumor lysates, or CpG, were subjected to vacuum negative pressure and then successively incubated with 50 μg SBs in 100 μl PBS at 4°C for 12 h. The free tumor lysates or CpG were washed with PBS by three centrifugations. The loading efficiency of tumor lysates or CpG was determined from the difference in mass between the initial incubation solution and the reserved supernatant. The unlabeled SBs, tumor lysates, and CpG were replaced by the FITC-labeled SBs, rhodamine-labeled tumor lysates, and DAPI-labeled CpG to confirm the success of loading. The single or merged fluorescent images were captured by confocal microscopy (FV1000; Olympus, Japan).
Generation of DCs and detection of DC maturation, activation, and antigen presentation
Murine bone-marrow-derived dendritic cells (BMDCs) were generated from female C57BL/6 or BALB/c mice 8–10 weeks old as previously reported 34. The erythrocytes in the isolated bone marrow cells (1 × 105/well) were depleted, and then cultured in RPMI–1640 media with recombinant murine IL–4 (20 ng/ml) and GM-CSF (100 U/ml). The non-adherent cells were gently removed on days 2 and 4. The adherent cell aggregates were dislodged and removed to another 6-well plate on day 6. The cells were cultured for a further 6 days, and the resultant non-adherent cells with typical morphological characteristics of BMDCs were harvested for analysis of DC maturation, activation, and cross-presentation. Briefly, BMDCs (1×106 cells/ml) were incubated with an OVA-B16 formulation of LC, SB-C, SB-L, or SB-LC for 24 h in 12-well plates. Thereafter, the BMDCs were collected and stained with antibodies for 30 min at 4°C to examine the presence of markers related to DC maturation (CD86, CD80, and CD40) and antigen presentation (SIINFEKL-MHC I and MHC II) by flow cytometry (CyFlow Cube 6; Sysmex, Japan). The antibodies used in this experiment were as follows: APC-Cy7-CD45, PE-CD86, FITC-CD80, PE-CD40, APC-SIINFEKL-MHC I, and PerCP-eFluor 450-MHC II (all BioLegend; CA, USA) in a 1:250–500 solution. To determine the state of DC activation, the titre of TNF-α, IFN-γ, or IL–6 secreted by BMDCs in the supernatant was detected by flow cytometry (CyFlow Cube 6) using a CBA Mouse Inflammation Kit (BD) following the manufacturer’s protocols. The data generated from the test samples were quantified and compared to a standard curve that was generated from standard samples using FCS EXPRESS V5 software (De Novo Software, Pasadena, CA).
Detection of SB uptake and its influence on antigen cross-presentation by DCs
Uptake of SB and its influence on antigen cross-presentation by DCs were determined by flow cytometry (CyFlow Cube 6) in the presence or absence of blocking antibodies against PAR1 and MR (BioLegend) in a 1:250–500 solution, and were directly visualized by confocal microscopy (FV1000, Olympus). To evaluate PAR1- or MR-mediated uptake, BMDCs were incubated for 1 h with a sufficient concentration of anti-PAR1 and anti-MR antibodies alone or combination to completely block the receptors. Following two washes with PBS, the BMDCs were incubated with comparable FITC-labeled SBs or SEs, and the mean fluorescence intensity (MFI) of the BMDCs, indicating the level of SB or SE uptake, was detected by flow cytometry (CyFlow Cube 6); the Cyt D influence on the uptake of SB or SE was also detected in this manner. To understand whether different vaccine formulations (LC, SB-C, SB-L, and SB-LC) were internalized by BMDCs, the FITC-labeled SBs or SEs were replaced by different formulations of tumor lysates (L) labeled with FITC or CpG (C) with DAPI. To investigate the trafficking of SB in BMDCs, the BMDCs were incubated with different formulations of SB (SB-C, SB-L, and SB-LC) labeled the ingredient tumor lysates (L) with FITC for 5 h. Thereafter, the early endosomes were labeled with polyclonal rabbit anti-murine early endosome antigen–1 (EEA1, Abcam) in a 1:250–500 solution followed by Texas Red anti-rabbit IgG (KPL) in a 1:250–500 solution, and the lysosomes were stained with LysoTracker Red (Invitrogen) in a 1:250–500 solution. The resultant fluorescent images were captured by confocal microscopy (FV1000, Olympus). To investigate the pathways involved in cross-presentation, BMDCs were incubated with different vaccine formulations in the presence or absence of the proteasome inhibitor MG132 or the lysosome protease inhibitor leupeptin, and the expression of SIINFEKL-MHC I was detected by flow cytometry (CyFlow Cube 6).
Establishment of tumor models and vaccination strategies
In this study, OVA-B6 (B16), CT26, and 4T1-luciferase cell lines (ATCC, Manassas, VA) were used to establish tumor models. The use of mice in this study was approved by the Animal Use and Care Committee of Hainan Medical University. To establish therapeutic tumor models, 5 × 106 tumor cells were subcutaneously injected into the right flank of syngeneic female mice aged 6–8 weeks. Thereafter, the tumor-bearing mice were randomly divided into four groups (n = 10 mice per group) and subcutaneously injected with different vaccine formulations (LC, SB-C, SB-L, or SB-LC) containing comparable does of corresponding tumor lysates (10 μg) or CpG (3 μg) in 100 μl PBS on day 1 and day 8 after tumor cell inoculation as previously described 35. In contrast to establish prophylactic models, the mice in this study were first subcutaneously injected with different vaccine formulations on day −7 and 0, and the tumor cells were injected on day 1. The OVA-B16 and CT26 models were established in C57BL/6 and BALB/c syngeneic mice, respectively. The tumor images and volumes at different time points were obtained by a portable measuring device (TM900; Peira, Belgium). In addition, 4T1-luciferase cells were used to establish metastatic model in syngeneic BALB/c mice using a similar protocol as a previously established prophylactic model by tail vein injection of 5 × 106 cells. To judge the anti-metastatic effects of the vaccinations, the mice were intraperitoneally injected with D-fluorescein potassium (300 mg/kg), and the luminescence intensity images were captured by an IVIS lumina II imaging system (Caliper Life Sciences).
Observations of adverse side effects
To observation the potential side effects induced by the vaccination, normal mice aged 6–8 weeks were vaccinated only with various formulations as in the protocol of prophylactic models, but not establishment of tumor models, and the following parameters were observed as potential side effects for 180 days. The body weight and life span of the mice were recorded every month, and other gross measures, including ruffled fur, feeding, and behavior, were also observed. Fertility was also observed as a parameter of potential toxicity, as reported previously. In brief, 2 weeks after the second vaccination, female mice were placed in cages with males, and both the number of days until parturition and the number of pups were recorded. At the end of the observation period, tissues of major organs, including the heart, liver, spleen, lung, and kidney, were collected and fixed in 10% neutral-buffered formalin and embedded in paraffin. Thereafter, 3–5 μm sections were stained with haematoxylin and eosin (H&E), and the tissue structures and cellular morphology were observed under microscopy (). The sections of major organs were also stained with FITC-conjugated anti-mouse IgG, IgM, or IgA and observed under fluorescence microscopy to determine whether autoantibody deposition was present. Furthermore, blood samples were extracted by tail vein 1 day after terminating treatment. The white blood cells (RBCs), red blood cells (RBCs), and platelets were counted as a measure of bone marrow toxicity, and the levels of creatinine, aspartate transaminase (AST), alanine transaminase (ALT), and lactic dehydrogenase (LDH) were recorded as measures of heart, liver, kidney, or other pan-organ toxicity using a small animal biochemical analyser (AAA, BB, USA).
Detection of cellular immunity
To test the efficiency of antigen-specific CTLs, splenocytes from the mice vaccinated with OVA-B16 formulations were double stained with SIINFEKL/H–2Kb peptide-MHC tetramers (Becton Dickinson) and PE-Cy7-anti-CD8α antibody (BioLegend) in a 1:250–500 solution, and then analysed by flow cytometry (CyFlow Cube 6). CTL-induced direct tumor lysis by splenocytes (effector) against B16 and CT 26 tumor cells (target) was detected using a CytoTox 96 cytotoxicity assay kit (Promega, Madison, WI) according to the manufacturer’s instructions and as previously reported 33. In brief, splenocytes and tumor cells at various ratios (5–40:1) were plated on 96-well plates in a final volume of 100 μl and then incubated in a humidified atmosphere (5% CO2 at 37ºC) for 6 h. The aliquots and reconstituted substrates (each 50 μl) were transferred to another flat-bottom 96-well plate, and incubated in the dark at room temperature for 50 min. The experiment was terminated by adding 50 μl stop solution, and the optical density (OD) values at 492 nm were obtained using an ELX808IU microplate reader (Bio-Tek, USA). The percentage of target cell death at each E:T ratio was calculated as reported previously 36.
In this study, we adapted an in vivo method from a previous study to detect the specific CTL response. Briefly, H22 cells were stained with CFSE (7 mM) for 20 min, and 5 × 106 cells were injected into the abdominal cavity of BALB/c mice vaccinated with various vaccine formulations. After 24 h, the mice were sacrificed, and the proliferated H22 and immune cells in the abdominal cavity were collected. These cells were further stained with Cy5-anti-CD45 antibody (Abcam), and the CD45−H22 cells were gated from the CD45+ immune cells to analyze cell proliferation. In addition, the total cell number in each group was used to indirectly calculate the percentage of H22 cell lysis using the following formula: Percentage of specific lysis = (1 − vaccinated group/unvaccinated group) × 100.
The splenocytes secreting IFN-γ, TNF-α, and IL–12 were isolated from the spleens of mice vaccinated with the various vaccine formulations 33. These splenocytes (1 × 106) were stained with the corresponding monoclonal antibodies conjugated with FITC-anti-IFN-γ, PE-anti-TNF-α, or Percp-Cy5.5-anti-IL12 antibody (BD, NJ, USA) for 30 min at 4°C. The stained cells were detected by flow cytometry (CyFlow Cube 6 and FACS Canto II), and the data were analysed with FlowJo Software version 10 (Tree Star, Ashland, OR).
Analysis of humoral immunity
IgG in the serum of mice vaccinated with various vaccine formulations in both B16 and CT26 models was detected by western blotting 36. The proteins of the B16 or CT26 cell lysates were separated in 12% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) gels and then transferred onto polyvinylidene difluoride membrane (Bio-Rad) using a Mini Transblot device (Bio-Rad, Hercules, CA). Thereafter, the membranes were stained with anti-mouse IgG antibodies (Abcam) at a 1:150 dilution after blocking by 10% non-fat milk at 4°C for 2 h. Finally, the membranes were incubated with chemiluminescence substrates A and B (1:1) for 30 min, and images were captured using an ECL system (Amersham Biosciences, UK).
The numbers of splenic monocytes secreting tumor-specific antibodies were detected by ELISPOT assay. In brief, B16 or CT26 tumor lysate (5 μg/well) was first fixed onto the polyvinylidene difluoride membrane fixed to the bottom of ELISPOT Plates (Millipore, Billerica, MA). The splenic monocytes from the mice vaccinated with various vaccine formulations were incubated in the plates for 6 h, in which the IgG antibody secreted by the splenocytes served as the primary antibody. Thereafter, the secondary anti-mouse IgG antibody was used to stain the ELISPOT membrane following a procedure similar to the aforementioned Western blotting assay, but the results appeared as spots and not bands. In addition, the titres of IgG and its subtypes were detected by ELISA kits (Wuhan Boster Biological Technology, China), according to the manufacturer’s protocol 36.
Adoptive transfer of splenocytes and antibodies
Adoptive transfer of splenocytes from mice vaccinated with different vaccine formulations to newly B16-bearing and CT26-bearing mice was performed as reported previously 36. In brief, the splenocytes from the mice vaccinated with the different formulations were isolated on day 14 after vaccination. The B16 or CT26 tumor models were established in new recipient mice (n = 5) by injection of corresponding tumor cells in the same manner as the previously mentioned therapeutic protocol. Subsequently, 5 × 106 splenocytes per mouse were transferred into the recipient mice by tail vein on day 0 after the tumor models were established. The transfers were repeated two more times on days 5 and 10. On day 18, the tumor images and volumes were captured using a portable TM900 system (Peira, Belgium). To adoptively transfer antibodies, the pooled sera from mice vaccinated with different formulations were passed through a chromatography column (CM Affi-gel Blue Gel Kit; Bio-Rad) to acquire the purified immunoglobulins. These antibodies (50 mg/kg in each mouse) were injected into tumor-bearing recipient mice three times at the same time points as the mice that received the splenocytes. The tumor images and volumes were then captured by a TM900 portable measuring device.
Calculation of synergistic indexes
In this study, we investigated whether the cellular and antibody immunities synergized to combat tumor growth. Splenocytes and antibodies isolated from H22 SB-LC-vaccinated mice using the same protocols as adoptive transfer were used in this study. Twenty BALB/c mice were injected with CFSE-labeled H22 cells and then randomly divided into four groups. The mice were then intraperitoneally injected with PBS, splenocytes (5 × 106), and antibodies (50 mg/kg) either alone or in combination. The cells in the abdominal cavity were collected, stained, and gated by flow cytometry (CyFlow Cube 6) as outlined previously. The synergistic indexes (SI) were calculated from the number of CD45− H22 cells 23. In brief, the observed relative ratio (ORR) of the number of cells in the splenocyte, antibody, or combination group was calculated by dividing the number of cells in each group by that in the PBS group. The expected relative ratio (ERR) of the combination group was calculated by multiplying the ORR of the splenocyte group with that of the antibody group; then, the SI of the combination was calculated as SI = ERR/ORR, where SI > 1 indicates a synergistic effect.
Depletion of CD4+ or CD8+ lymphocytes
Depletion of CD4+ or CD8+ lymphocytes was performed as reported previously 36. In brief, rat monoclonal anti-CD4 IgG (clone GK 1.5) or an anti-CD8 IgG (clone 2.43) antibody isolated from the corresponding hybridoma (ATCC, Manassas, VA, USA) was intraperitoneally injected into B16- or CT26-bearing mice (1 mg/mouse) four times in 4-day intervals, with the first injection on day 0. The splenocytes were analysed by flow cytometry (CyFlow Cube 6) to ensure that the efficacy of cell depletion was greater than 98%. Thereafter, the depleted mice were used to establish a B16 or CT26 model with subcutaneous injection of 5 × 105 tumor cells. The tumor images and volumes were captured by a TM900 portable measuring device on day 18 after tumor cell injection as mentioned previously. Meanwhile, the titres of IgG antibodies specific to B16 or CT26 cells in the sera were collected and detected by ELISA 18 days after tumor cell injection as outlined above.
Evaluation of the immunosuppressive microenvironment
Tumor-infiltrating lymphocytes (TILs), including CD8+ T cells, MDSC, and Treg, were first observed in tumor sections. Tumour tissues from the mice vaccinated with the different formulations were removed 18 days after tumor cell injection, and the frozen sections (3–5 μm) were stained with the following fluorescence-labeled antibodies: APC-Cy7-anti-CD8, PE-anti-CD11b, FITC-anti-Gr–1, and PerCP-anti-Foxp3 (BioLegend) at 1:250–500 dilutions. The fluorescent images were captured and merged by confocal microscopy (FV1000, Olympus). In addition, TILs were generated from different tumor masses and analysed by flow cytometry (CyFlow Cube 6). In brief, tumor masses from mice vaccinated with various formulations were removed by forceps and cut into pieces approximately 1 mm3 in size, and subsequently treated with DMEM media containing collagenase (1 mg/ml) and DNase I (Sigma) for 3 h at 37°C. The resultant cell suspension was treated with Lysing Buffer (BD Pharm) and passed through a 70-mm cell strainer. Thereafter, the cells were re-suspended in 5 ml Percoll (33%) and centrifuged at 3,000 rpm for 15 min. In addition to the abovementioned fluorescent antibodies, other antibodies, including FITC-anti-CD45, APC-anti-CD4, APC-Cy7-anti-CD25, PE-anti-CD206, and FITC-anti-F4/80, were used. The CD45+ single or combined with CD4+ or CD11b+ double-positive TILs were gated out from the mixture of CD45+ TILs and CD45− tumor cells; these cell phenotypes were further analysed by flow cytometry to identify IFN-γ-secreting CD8+ T cells, MDSC, Treg, and M2-type ATMs.
Detection of tumor cell ferroptosis
Lipid reactive oxygen species (ROS) were assessed by BODIPY-C11 staining (Thermo Fisher) in B16- or CT26-bearing mice or in B16 or CT26 cells in vivo as previously reported 30. To quantify the relative ROS in tumor tissues from mice vaccinated with SB-C or SB-LC treated with or without 10 mg/kg ferroptosis-inhibitor liproxstatin–1 (Cayman), the images and tumor volumes were monitored using a TM900 portable measuring device, and the mice were sacrificed to collect the tumor tissues on day 15 after tumor cell injection. The tumor tissues were resected and cut into small pieces, which were then mechanically minced in a cell strainer (100 μM) and washed twice with PBS. The mixture of the tumor and TILs was enriched by Ficoll (Sigma-Aldrich) centrifugation. The cell pellet was stained with anti-CD45 antibody (BioLegend) followed by BODIPY 581/591 C11 (Thermo Fisher) at 1:250–500 dilutions and analysed by flow cytometry (CyFlow Cube 6) after straining through a 40-μM cell strainer. The CD45−tumor cells were distinguished from CD45+ TILs or splenocytes by the flow cytometry gating strategy. For BODIPY 581/591 C11 staining, the fluorescence intensity from both oxidized C11 (FITC signal) and non-oxidized C11 (PE signal) was captured. The MFI ratio of the FITC signal to the PE signal was calculated for each sample. The data were normalized to the SB-C samples as shown by the relative lipid ROS.
To detect the relative ROS in vivo, B16 or CT26 cells (10,000 cells) were seeded with the same amount of splenocytes from the mice vaccinated with SB-C or SB-LC in the presence or absence of ferrostatin–1 (10 μM) or the ferroptosis inducer RSL3 (1 μM, Cayman) in 24-well plates for 20 h. Subsequently, the cells were trypsinized and collected for staining and flow cytometric analysis using the abovementioned method to analysed the relationship between the CD45- tumor cells and CD45+ TILs. These cells were also used to analyse ferroptosis-related cell death. In brief, after 20 h co-culture, the cells were collected and re-suspended in PBS containing 1 μg/ml 7-aminoactinomycin D (7-AAD) for 10 min. The CD45−tumor cells were gated from the CD45+ splenocytes, and the dead cells were calculated as the proportion of 7-ADD+ cells by flow cytometric analysis.
To investigate the relationship between ferroptosis and immunity in vivo, SB-C- or SB-LC-vaccinated B16- or CT26-bearing mice were treated with liproxstatin–1, and tumor tissues were collected. The TIL phenotypes related to IFN-γ-secreting CD8+ T cells, MDSC, Treg, and M2-type ATMs were analysed by flow cytometry as outlined above.
Statistical Analysis
The data in this study were analyzed by GraphPad-Prism Software version 9.0.0 for Windows (San Diego, California USA), and presented as the mean ± standard deviation (SD). As all the data concern more than two groups, one- or two-way analysis of variance (ANOVA) followed by a Tukey’s honest significant difference test was performed to compare the differences between groups. Animal survival was graphed as Kaplan–Meier survival curves and analysed with the log-rank (Mantel–Cox) test. * < 0.05, ** < 0.01, *** < 0.001, **** < 0.0001.