The Thermal Dose of Photothermal Therapy Generates Differential Immunogenicity in Human Neuroblastoma Cells

Abstract

Photothermal therapy (PTT) is an effective method for tumor eradication and has been successfully combined with immunotherapy. However, besides its cytotoxic effects, little is known about the effect of the PTT thermal dose on the immunogenicity of treated tumor cells. Therefore, we administered a range of thermal doses using Prussian blue nanoparticle-based photothermal therapy (PBNP-PTT) and assessed their effects on tumor cell death and concomitant immunogenicity correlates in two human neuroblastoma cell lines: SH-SY5Y (MYCN-non-amplified) and LAN-1 (MYCN-amplified). PBNP-PTT generated thermal dose-dependent tumor cell killing and immunogenic cell death (ICD) in both tumor lines in vitro. However, the effect of the thermal dose on ICD and the expression of costimulatory molecules, immune checkpoint molecules, major histocompatibility complexes, an NK cell-activating ligand, and a neuroblastoma-associated antigen were significantly more pronounced in SH-SY5Y cells compared with LAN-1 cells, consistent with the high-risk phenotype of LAN-1 cells. In functional co-culture studies in vitro, T cells exhibited significantly higher cytotoxicity toward SH-SY5Y cells relative to LAN-1 cells at equivalent thermal doses. This preliminary report suggests the importance of moving past the traditional focus of using PTT solely for tumor eradication to one that considers the immunogenic effects of PTT thermal dose to facilitate its success in cancer immunotherapy.

Keywords: MYCN amplification; Prussian blue nanoparticles; immunogenic cell death; immunogenicity; nanoimmunotherapy; neuroblastoma; photothermal therapy; thermal dose.

Figure 1

Schematic of the study. The thermal dose generated by PBNP-PTT was controlled by varying the PBNP concentration and near infrared laser power, independently. Subsequently, tumor cell death, immunogenic cell death (ICD), and cell surface markers associated with immunogenicity were measured on neuroblastoma cells as a function of PBNP-PTT-administered thermal dose.

Figure 2

PBNP-PTT-administered thermal dose is controlled by PBNP concentration and laser power. Three million (A,B) SH-SY5Y or (C,D) LAN-1 human neuroblastoma cells were exposed to 0.75, 1, and 1.5 W laser power in combination with 0 (Laser), 0.06, 0.1, or 0.15 mg/mL PBNPs. (A,C) Time-temperature graphs represents the cell temperatures recorded for each condition at 1 min intervals for a total of 10 min. (B,D) Thermal doses applied to cells were calculated from the time-temperature curves for each condition.

Figure 3

PBNP-PTT generates a thermal dose window of immunogenic cell death in SH-SY5Y and LAN-1 cells in vitro. Three million (A–C) SH-SY5Y or (D–F) LAN-1 cells were exposed to various thermal doses using PBNP-PTT. After 24 h, cells were analyzed for (A,D) intracellular ATP, (B,E) HMGB1 release, and (C,F) surface calreticulin expression, represented as median fluorescence intensity (MFI). Inset values in the histograms denote the thermal dose. The extent of ICD as measured by its correlates is more pronounced in SH-SY5Y cells compared with LAN-1 cells. Ordinary one-way ANOVA was used to calculate significance between vehicle and different thermal doses and laser alone for HMB1 analysis. n = 2/group; * p < 0.03, ** p < 0.002, *** p < 0.0002, **** p < 0.0001.

Figure 4

The effect of PBNP-PTT-administered thermal dose on co-stimulatory and HLA molecules is more pronounced in SH-SY5Y cells compared with LAN-1 cells. (A–D) SH-SY5Y and (E–H) LAN-1 cells were subjected to various thermal doses via PBNP-PTT. After treatment, cells were rested for 24 h and evaluated by flow cytometry analysis for (A,E) CD80, (B,F) CD86, (C,G) HLA-ABC, and (D,H) HLA-DR. MFI is represented as bar graphs representing averages, and corresponding representative histograms. Inset values in the histograms denote the thermal dose. Each parameter was compared to the vehicle control by one-way ANOVA, where * p < 0.03, ** p < 0.002, *** p < 0.0002, **** p < 0.0001; n = 2/group.

Figure 5

The effect of PBNP-PTT-administered thermal dose on immune checkpoint molecules, an NK cell-activating ligand, and a neuroblastoma-associated antigen is more pronounced in SH-SY5Y cells compared with LAN-1 cells. (A–D) SH-SY5Y and (E–H) LAN-1 cells were treated in vitro with varied thermal doses via PBNP-PTT. 24 h post-PBNP-PTT, cells were analyzed for expression of (A,E) PD-L1, (B,F) B7-H3, (C,G) PVR, and (D,H) GD2 using flow cytometry analysis. Inset values in the histograms denote the thermal dose. Each parameter was compared to the vehicle control by one-way ANOVA, where * p < 0.03, ** p < 0.002, *** p < 0.0002, **** p < 0.0001; n = 2/group.

Figure 6

PBNP-PTT triggers greater immunophenotypic changes in MYCN-non-amplified SH-SY5Y cells than MYCN-amplified LAN-1 neuroblastoma cell line in vitro. SH-SY5Y (blue) and LAN-1 (green) cells were treated with varied thermal doses via PBNP-PTT and analyzed for (A) % live cells (B) intracellular ATP, (C) secreted HMGB1, and cell surface expression levels of (D) calreticulin, (E) CD80, (F) CD86, (G) PD-L1, (H) B7-H3, (I) HLA-ABC, (J) HLA-DR, (K) PVR, and (L) GD2. Data represent mean ± SD (n = 2 independent samples).

Figure 7

Equivalent thermal doses of PBNP-PTT induces greater T cell cytotoxicity against SH-SY5Y tumor cells than LAN-1 cells. Counterclockwise (A) Schematic representing the workflow of the co-culture study. SH-SY5Y and LAN-1 cells were treated with either control treatments or PBNP-PTT treatments at varied thermal doses (low to high). Tumor cells were then co-incubated with their respective tumor-reactive T cells at a 1:1 effector:target cell (E:T) ratio for 4 h and were analyzed for T cell cytotoxicity toward the treated tumor cells using flow cytometry. (B) Scatter plots showing live target cells for a representative thermal dose treatment group as compared with the control. (C) Cytotoxicity of T cells toward tumor cells when exposed to various (low to high) thermal doses (SH-SY5Y- low: 5.1, medium: 7.2, medium-high: 9.9, high: 11.2; LAN-1- low: 4.2, medium: 6.9, medium-high: 10.1, high: 11.5). Error bars depict standard deviation between two donors. Studies with each donor were performed in duplicate. A two-way ANOVA was used to compare the statistical significance between the two tumor lines (* p < 0.03, *** p < 0.0002, **** p < 0.0001).