Chitosan nanoparticles as antigen vehicles to induce effective tumor specific T cell responses

Abstract

Cancer vaccinations sensitize the immune system to recognize tumor-specific antigens de novo or boosting preexisting immune responses. Dendritic cells (DCs) are regarded as the most potent antigen presenting cells (APCs) for induction of (cancer) antigen-specific CD8+ T cell responses. Chitosan nanoparticles (CNPs) used as delivery vehicle have been shown to improve anti-tumor responses. This study aimed at exploring the potential of CNPs as antigen delivery system by assessing activation and expansion of antigen-specific CD8+ T cells by DCs and subsequent T cell-mediated lysis of pancreatic ductal adenocarcinoma (PDAC) cells. As model antigen the ovalbumin-derived peptide SIINFEKL was chosen. Using imaging cytometry, intracellular uptake of FITC-labelled CNPs of three different sizes and qualities (90/10, 90/20 and 90/50) was demonstrated in DCs and in pro- and anti-inflammatory macrophages to different extents. While larger particles (90/50) impaired survival of all APCs, small CNPs (90/10) were not toxic for DCs. Internalization of SIINFEKL-loaded but not empty 90/10-CNPs promoted a pro-inflammatory phenotype of DCs indicated by elevated expression of pro-inflammatory cytokines. Treatment of murine DC2.4 cells with SIINFEKL-loaded 90/10-CNPs led to a marked MHC-related presentation of SIINFEKL and enabled DC2.4 cells to potently activate SIINFEKL-specific CD8+ OT-1 T cells finally leading to effective lysis of the PDAC cell line Panc-OVA. Overall, our study supports the suitability of CNPs as antigen vehicle to induce potent anti-tumor immune responses by activation and expansion of tumor antigen-specific CD8+ T cells.

Fig 1. Uptake of CNPs by human and murine APCs.

Different APC populations were incubated with 100 μg/ml CNPs of different size (90/10, 90/20 and 90/50) for 24 hours. A) Representative images of primary human dendritic cells (DCs) which were incubated with 90/10-CNPs and analyzed by Imaging cytometry; I shows DCs with internalized CNPs, II shows DCs with intra- and extracellular CNPs; BF = brightfield; red = CD11c-APC membrane staining of DCs; green = FITC-conjugated CNPs; overlay. B) Representative images of immunofluorescence (IF) and confocal laser scanning (CLS) microscopy of DCs incubated with 90/10-CNPs; blue = Hoechst nuclear counterstaining; red = CD11c-APC membrane staining; green = FITC-conjugated CNPs. C) Image cytometry analysis of CNP uptake by primary human DCs, murine DC line DC2.4 and human M1- or M2-macrophages (Mᶲ). D) Imaging cytometry analysis of CNP uptake by cocultured human DCs and human H441 lung epithelial cells after 24 hour incubation with CNPs. Data are expressed as % intracellularly stained cells and as mean + SEM of three independent experiments. *p < 0.05.

Fig 2. Uptake of large CNPs is toxic for APCs.

Primary human DCs, M1- and M2-macrophages (Mᶲ) were either left untreated (ut) or incubated with 100 μg/ml CNPs of different size (90/10, 90/20 and 90/50) for 24 hours. A) Graphs present relative caspase-3/7 activity. Data are presented as n-fold caspase-3/7 activity of untreated cells and as median with 75th and 25th percentile of four independent experiments (DCs) or as mean + SEM of three independent experiments (M1- and M2-Mᶲ). B) Representative phase contrast images of DCs either left untreated or incubated for 24 hours with the indicated CNPs. Images were acquired at 280-fold magnification. C-D) DCs were stained with propidium iodide (PI) after indicated treatment and (C) cell confluence as well as (D) PI+ area were determined. Data are expressed as n-fold of untreated cells and mean + SEM of three independent experiments. *p < 0.05.

Fig 3. Uptake of CNPs promotes a pro-inflammatory phenotype of DCs.

Human and murine DCs (DC2.4) were either left untreated or incubated with 100 μg/ml empty (90/10) or SIINFEKL-loaded 90/10-CNPs for 5, 24 and 48 hours. A) Flow cytometric analysis of CD80, CD86, HLA-DR and PD-L1 cell surface levels in human DCs. Median fluorescence intensity (MFI) ratio was calculated by normalizing MFI of each surface marker to the MFI of its respective isotype control. Then, MFI ratio of CNP treated cells was normalized to untreated cells and expressed as n-fold MFI of untreated cells. B+C) Relative mRNA levels of TGF-β1, IL-10, IL-1β, IL-6 and TNF-α in B) primary human DCs and C) murine DC2.4 cells was determined by RT-qPCR. Expression levels were normalized to expression of the housekeeping gene TBP/GAPDH and normalized to values determined for untreated cells. Data are presented as mean + SEM or median with 75^th and 25^th percentile of three independent experiments. *p < 0.05.

Fig 4. Uptake of SIINFEKL loaded CNPs by DCs efficiently stimulates CD8+ OT-1 T cell activation and expansion.

A) DC2.4 cells were either left untreated or treated with 100 μg/ml SIINFEKL-loaded 90/10-CNPs for 1, 5, 9 and 24 hours or B) DC2.4 cells were treated with 100 μg/ml empty (90/10) or SIINFEKL-loaded 90/10-CNPs or 1 μg/ml SIINFEKL for 5 hours. SIINFEKL bound to H-2Kb molecules was determined by flow cytometric analysis. Median fluorescence intensity (MFI) ratio was calculated by normalizing MFI of staining with anti-SIINFEKL-H-2Kb antibody to MFI detected in staining with its respective isotype control (n-fold MFI). In A) MFI ratio of treated cells was also normalized to MFI of 1h treated cells and expressed as n-fold MFI of untreated cells. C) CD8+ OT-1 T lymphocytes were isolated from spleens of OT-1 mice. Purity of CD8+ OT-1 T cells was determined after negative MACS selection by CD8 staining and flow cytometric analysis. One representative dot blot is shown. D-F) DC2.4 cells were either left untreated or were treated with 100 μg/ml empty (90/10) or 100 μg/ml SIINFEKL-loaded 90/10-CNPs or 1 μg/ml SIINFEKL for 5 hours. Afterwards, differentially treated DC2.4 cells were cocultured with CD8+ OT-1 T lymphocytes for 72 hours. D) Representative phase contrast images of CD8+ OT-1 T cells cocultured with DC2.4 cells that had been either left untreated or subjected to the indicated treatment. Scale bar = 1000 μm. E) Numbers of vital CD8+ OT-1 T lymphocytes were determined after the indicated coculture setting. Data are presented as mean + SEM of three independent experiments. F) Flow cytometric analysis of T cell activation by staining of cell surface CD25, CD44 and CD69 of CD8+ OT-1 T cells directly after isolation and after coculture with DC.24 cells that had been pretreated with SIINFEKL or 90/10-SIINFEKL CNPs. Representative histograms from one out of 3 independent experiments are shown and bar charts present the % of CD8+ OT-1 T cells positive for cell surface expression of CD25, CD44 and CD69, respectively. Data are presented as mean +SEM of three independent experiments. *p < 0.05.

Fig 5. CD8+ OT-1 T cells activated by SIINFEKL presenting DCs after uptake of SIINFEKL-loaded CNPs efficiently kill SIINFEKL expressing tumor cells.

Panc02 and Panc-OVA cells were cocultured with CD8+ OT-1 T lymphocytes derived from previous coculture with DC2.4 cells that had been either left untreated or treated with 100 μg/ml empty (90/10) or SIINFEKL-loaded 90/10-CNPs or 1 μg/ml SIINFEKL peptide. After 24 hours, CD8+ OT-1 T cells were removed from the wells and A) phase contrast images were taken (Scale bar = 500 μm) and B) cellular confluence of tumor cells was determined. Relative confluence of Panc02 and Panc-OVA cells after either indicated treatment is presented as n-fold of cellular confluence determined in cultures of respective Panc02 and Panc-OVA cells that were cultured with CD8+ OT-1 T lymphocytes from previous coculture with untreated DC2.4 cells. Data are presented as mean +SEM of three independent experiments. *p < 0.05.