Efficient Heat Shock Response Affects Hyperthermia-Induced Radiosensitization in a Tumor Spheroid Control Probability Assay

Abstract

Hyperthermia (HT) combined with irradiation is a well-known concept to improve the curative potential of radiotherapy. Technological progress has opened new avenues for thermoradiotherapy, even for recurrent head and neck squamous cell carcinomas (HNSCC). Preclinical evaluation of the curative radiosensitizing potential of various HT regimens remains ethically, economically, and technically challenging. One key objective of our study was to refine an advanced 3-D assay setup for HT + RT research and treatment testing. For the first time, HT-induced radiosensitization was systematically examined in two differently radioresponsive HNSCC spheroid models using the unique in vitro "curative" analytical endpoint of spheroid control probability. We further investigated the cellular stress response mechanisms underlying the HT-related radiosensitization process with the aim to unravel the impact of HT-induced proteotoxic stress on the overall radioresponse. HT disrupted the proteome's thermal stability, causing severe proteotoxic stress. It strongly enhanced radiation efficacy and affected paramount survival and stress response signaling networks. Transcriptomics, q-PCR, and western blotting data revealed that HT + RT co-treatment critically triggers the heat shock response (HSR). Pre-treatment with chemical chaperones intensified the radiosensitizing effect, thereby suppressing HT-induced Hsp27 expression. Our data suggest that HT-induced radiosensitization is adversely affected by the proteotoxic stress response. Hence, we propose the inhibition of particular heat shock proteins as a targeting strategy to improve the outcome of combinatorial HT + RT.

Keywords: head and heck squamous cell carcinomas (HNSCC); heat shock proteins (Hsps); hyperthermia; proteotoxic stress; radiation therapy; spheroids.

Figure 1

HT induces thermal dose-dependent HNSCC spheroid volume growth delay: (a) Representative image series of FaDu and SAS spheroids upon exposure to HT demonstrate the impact of various thermal doses on spheroid growth. (b) Effect of different hyperthermia treatments and thermal doses, respectively, on FaDu and SAS spheroid volume growth up to 14 day of post-treatment. Data show means (±SD) of ∑n ≥ 56 spheroids per treatment arm from N = 2 independent experiments. The dotted line indicates the relative value for 5 × V₀. (c) Mean relative spheroid growth delay (±SD) induced by different thermal doses calculated from the data presented in (b) as the extended time to reach 5× pre-treatment spheroid volume (5 × V₀) is shown to enable the comparison of the different spheroid types; treatment arms were normalized to the respective untreated controls. The Mann-Whitney U test was applied to estimate statistical significance of the relative growth delay (treated versus control spheroids) based on the pooled data; * p < 0.05; ** p < 0.01; *** p < 0.001.

Figure 2

HT efficiently sensitizes both HNSCC spheroid types to single dose irradiation: (a) Representative images of 28 SAS spheroids exposed for 30 min or 60 min to 42.5 °C followed by 12.5 Gy single dose irradiation illustrate the radiosensitizing effect of different doses of HT; spheroids before treatment and without (0 min) HT pre-exposure are documented for comparison. (b) Proportions of controlled (non-regrown) FaDu and SAS spheroids documented as function of time post-treatment when exposed for 0, 30, or 60 min to 42.5 °C HT before single dose irradiation (FaDu—7.5 Gy; SAS—12.5 Gy); N = 2; ∑n ≥ 56 spheroids per condition; *** p < 0.001 as assessed with the Log-rank (Mantel-Cox) test for survival curves based on the pooled data. (c) Spheroid control dose response curves after HT pre-exposure and 0–25 Gy single dose irradiation; the proportion of spheroids that lost regrowth capacity (spheroid control probability, SCP) is recorded as a function of the irradiation dose. Every data point in each HT treatment arm represents the SCP of ∑n ≥ 56 individual spheroids from N = 2 independent experiments monitored up to 60 days post-treatment (≥560–620 spheroids per SCP curve). Horizontal bars present the 95% confidence interval of the SCD₅₀ (spheroid control dose 50%).

Figure 3

Combination of HT with RT affects the main signaling networks and stress response pathways in HNSCC spheroids: (a) Venn diagrams illustrating the number of differentially expressed genes from the whole genome for RT, HT, and HT + RT-treated FaDu and SAS spheroids in triplicate (0.5 h after treatment). Genes were selected by DESeq2 with at least 2-fold changes in expression relative to the appropriate control spheroids (adjusted p ≤ 0.05). (b) The top 20 selected signaling pathways and processes that are significantly (2-fold change or more, adjusted p ≤ 0.05) over-represented in FaDu and SAS spheroids 0.5 h after exposure to HT + RT (42.5 °C/60 min; 7 Gy for FaDu and 10 Gy for SAS) when compared to controls using the GO enrichment analysis. The X-axis presents the corresponding adjusted p values according to Fischer exact’s test (in negative log10 scale). (c) Representative Western blot data sets showing the expression/activation of proteins of interest from the MAPK and PKB/Akt signaling pathways in the two HNSCC spheroid types upon treatment. GAPDH was used as loading control. Spheroid treatment conditions: Ctrl—control; RT—single dose irradiation (7 Gy for FaDu and 10 Gy for SAS); HT—hyperthermia (42.5 ˚C/60 min); HT + RT—hyperthermia and single dose irradiation according to mono-treatments.

Figure 4

HT triggers heat shock response and proteotoxic stress in HNSCC spheroids: (a) Heat map of differentially expressed HSR and UPR genes in FaDu and SAS spheroids within 0.5 h of treatments profiled by RNA-Seq analysis in triplicate. Genes were selected by DESeq2 with at least 2-fold changes in expression under HT + RT treatment as highlighted in Figure 3, relative to appropriate control spheroids (adjusted p ≤ 0.05). The color-coded data represent log2 values reflecting down-regulation in green and up-regulation in red. (b) q-PCR analysis of three selected genes identified as upregulated by HT and HT + RT treatments in FaDu and SAS spheroids in our RNA-Seq analysis. Data were normalized to ACTB gene expression and are shown as means (±SD) of N = 3 independent experiments; * p ≤ 0.05; ** p ≤ 0.01. (c) Western blot data sets showing the expression/activation of HSR and UPR proteins of interest in FaDu and SAS spheroids 0.5–24 h after exposure to HT and/or irradiation; GAPDH was used as loading control. (d) Representative PCR analysis documenting the splicing of XBP1 mRNA in FaDu and SAS spheroids 0.5 h after HT, RT and HT + RT treatments; the ACTB gene expression served as reference. Spheroid treatment conditions: Ctrl—control; RT—single dose irradiation (7 Gy for FaDu and 10 Gy for SAS); HT—hyperthermia (42.5 °C/60 min); HT + RT—hyperthermia and single dose irradiation according to mono-treatments.

Figure 5

Induction of proteotoxic stress plays a protective role in HNSCC spheroids against HT-induced radiosensitization: (a) 24 h treatment with the chemical chaperones TUDCA (0.4 mM) and 4-PBA (5 mM) alone does not affect SAS spheroid volume growth kinetics. Data points show mean spheroid volumes of ∑n ≥ 56 spheroids per treatment arm from N = 2 independent experiments (±SD). (b) Proportion of controlled SAS spheroids irradiated with 10 Gy directly after exposure to 42.5 °C for 60 min in the absence or presence of TUDCA (0.4 mM) or 4-PBA (5 mM); the time courses represent the monitoring of ∑n ≥ 56 spheroids per treatment arm from N = 2 independent experiments over a period of 60 days post-treatment. ** p < 0.01 and *** p < 0.001 as assessed with the Log-rank (Mantel-Cox) test for survival curves based on the pooled data. (c) Images of 28 representative SAS spheroids before and at day 60 after treatment with HT + RT with or without TUDCA or 4-PBA according to (b). (d) qPCR analysis of sXBP1 and HSPB1 gene expression 0.5 h after HT (42 °C, 60 min) with or without RT (7 Gy—FaDu, 10 Gy—SAS) in the absence and presence of TUDCA or 4-PBA according to (b,c); data were normalized to ACTB gene expression and are shown as means (+SD) of N = 3 independent experiments; * p < 0.05; ** p < 0.01.