Understanding radiation response and cell cycle variation in brain tumour cells using Raman spectroscopy

Abstract

Radiation therapy is currently utilised in the treatment of approximately 50% of cancer patients. A move towards patient tailored radiation therapy would help to improve the treatment outcome for patients as the inter-patient and intra-patient heterogeneity of cancer leads to large differences in treatment responses. In radiation therapy, a typical treatment outcome is cell cycle arrest which leads to cell cycle synchronisation. As treatment is typically given over multiple fractions it is important to understand how variation in the cell cycle can affect treatment response. Raman spectroscopy has previously been assessed as a method for monitoring radiation response in cancer cells and has shown promise in detecting the subtle biochemical changes following radiation exposure. This study evaluated Raman spectroscopy as a potential tool for monitoring cellular response to radiation in synchronised versus unsynchronised UVW human glioma cells in vitro. Specifically, it was hypothesised that the UVW cells would demonstrate a greater radiation resistance if the cell cycle phase of the cells was synchronised to the G₁/S boundary prior to radiation exposure. Here we evaluated whether Raman spectroscopy, combined with cell cycle analysis and DNA damage and repair analysis (γ-H2AX assay), could discriminate the subtle cellular changes associated with radiation response. Raman spectroscopy combined with principal component analysis (PCA) was able to show the changes in radiation response over 24 hours following radiation exposure. Spectral changes were assigned to variations in protein, specifically changes in protein signals from amides as well as changes in lipid expression. A different response was observed between cells synchronised in the cell cycle and unsynchronised cells. After 24 hours following irradiation, the unsynchronised cells showed greater spectral changes compared to the synchronised cells demonstrating that the cell cycle plays an important role in the radiation resistance or sensitivity of the UVW cells, and that radiation resistance could be induced by controlling the cell cycle. One of the main aims of cancer treatment is to stop the proliferation of cells by controlling or halting progression through the cell cycle, thereby highlighting the importance of controlling the cell cycle when studying the effects of cancer treatments such as radiation therapy. Raman spectroscopy has been shown to be a useful tool for evaluating the changes in radiation response when the cell cycle phase is controlled and therefore highlighting its potential for assessing radiation response and resistance.

Fig. 1. Radiation resistance and sensitivity changes through the cell cycle.

Fig. 2. Cell cycle analysis of (A). unsynchronised UVW cells, and (B). synchronised UVW cells. FACS results comparing cell cycle distribution of control and 6 Gy irradiated cells at 1 hour, 4 hours and 24 hours following irradiation. The coloured bars represent percentage of cells in sG₁ phase (pink), G₀/G₁ phase (blue), S phase (grey) and G₂/M phase (yellow). Data presented as an average of three independent replicates (mean ± standard deviation). Two-way ANOVA compared the mean cell cycle phase population at each time point following irradiation to the untreated control cells. Statistical analysis was performed using two-way ANOVA with Bonferroni post-tests at 95% confidence interval (p > 0.05 = ns (not significant) and p < 0.0001 = ****).

Fig. 3. Average Raman spectra comparing control UVW cells (blue) and 6 Gy irradiated UVW cells (pink). Unsynchronised cells 1-hour, 4-hour and 24 hours following radiation exposure (left panel). Synchronised cells 1 hour, 4 hours and 24 hours following radiation exposure (right panel). Results are presented as an average spectrum of three independent replicates including 43 single cell maps in total per sample and each replicate contained at least 10 cell maps.

Fig. 4. Principal component analysis (PCA) of average Raman spectra for 24-hour time point. (A). Box plot of PC3 median loading scores at the 24-hour time point. Box plot compares control cells and 6 Gy irradiated cells for unsynchronised UVW cells (blue) and synchronised UVW cells (pink). Centre point of box represents median value, notches represent the 25th and 75th percentile, whiskers represent the 5th and 95th percentile and stars represent outliers. (B). PC3 loading from PCA comparing control and 6 Gy samples for all time points following 6 Gy XBR exposure for unsynchronised and synchronised UVW cells. Statistical analysis was performed using a two-way ANOVA with Wilcoxon rank sum test at 99% confidence interval (p > 0.05 = ns (not significant), p < 0.001 = *** and p < 0.0001 = ****).

Fig. 5. Principal component analysis (PCA) of average Raman spectra for just the 4-hour time point showing the box plots and PC loadings for PC5 and PC1. Box plot compares control cells and 6 Gy irradiated cells for unsynchronised UVW cells (blue) and synchronised UVW cells (pink). Centre point of box represents median value, notches represent the 25th and 75th percentile, whiskers represent the 5th and 95th percentile and stars represent outliers. PC5 and PC1 loading from individual time point PCA comparing control and 6 Gy samples for the 24-hour time point following 6 Gy XBR exposure for unsynchronised and synchronised UVW cells. Statistical analysis was performed using a two-way ANOVA with Wilcoxon rank sum test at 99% confidence interval (p > 0.05 = ns (not significant), p < 0.01 = ** and p < 0.001 = ***).

Fig. 6. Principal component analysis (PCA) of average Raman spectra for just the 24-hour time point showing the box plots and PC loadings for PC5 and PC1. Box plot compares control cells and 6 Gy irradiated cells for unsynchronised UVW cells (blue) and synchronised UVW cells (pink). Centre point of box represents median value, notches represent the 25th and 75th percentile, whiskers represent the 5th and 95th percentile and stars represent outliers. PC5 and PC1 loading from individual time point PCA comparing control and 6 Gy samples for the 24-hour time point following 6 Gy XBR exposure for unsynchronised and synchronised UVW cells. Statistical analysis was performed using a two-way ANOVA with Wilcoxon rank sum test at 99% confidence interval (p > 0.05 = ns (not significant) band p < 0.0001 = ****).

Fig. 7. The fold change in γ-H2AX levels in irradiated UVW cells compared to an untreated control at each time point. (A). Unsynchronised UVW cells. (B). Synchronised UVW cells. Results are presented as an average of three independent experiments (mean ± standard deviation). One-way ANOVA compared the mean g-H2AX level fold change at each time point following irradiation to the untreated control cells. Statistical analysis was performed using one-way ANOVA with Bonferroni post-tests at 95% confidence interval (p > 0.05 = ns (not significant) and p < 0.01 = **).