Radiation-Induced Glycogen Accumulation Detected by Single Cell Raman Spectroscopy Is Associated with Radioresistance that Can Be Reversed by Metformin

Abstract

Altered cellular metabolism is a hallmark of tumor cells and contributes to a host of properties associated with resistance to radiotherapy. Detection of radiation-induced biochemical changes can reveal unique metabolic pathways affecting radiosensitivity that may serve as attractive therapeutic targets. Using clinically relevant doses of radiation, we performed label-free single cell Raman spectroscopy on a series of human cancer cell lines and detected radiation-induced accumulation of intracellular glycogen. The increase in glycogen post-irradiation was highest in lung (H460) and breast (MCF7) tumor cells compared to prostate (LNCaP) tumor cells. In response to radiation, the appearance of this glycogen signature correlated with radiation resistance. Moreover, the buildup of glycogen was linked to the phosphorylation of GSK-3β, a canonical modulator of cell survival following radiation exposure and a key regulator of glycogen metabolism. When MCF7 cells were irradiated in the presence of the anti-diabetic drug metformin, there was a significant decrease in the amount of radiation-induced glycogen. The suppression of glycogen by metformin following radiation was associated with increased radiosensitivity. In contrast to MCF7 cells, metformin had minimal effects on both the level of glycogen in H460 cells following radiation and radiosensitivity. Our data demonstrate a novel approach of spectral monitoring by Raman spectroscopy to assess changes in the levels of intracellular glycogen as a potential marker and resistance mechanism to radiation therapy.

Fig 1. Single-cell Raman spectroscopy with PCA detects early radiation-induced glycogen synthesis in H460 and MCF7 cells, but not in radiosensitive LNCaP cells.

(A) Single-cell Raman spectra of an irradiated H460 cell (10 Gy) and an unirradiated H460 cell at 3 days post-irradiation demonstrate the Raman spectroscopic detection of increased intracellular glycogen in irradiated H460 and MCF7 cells. The difference spectrum (dashed trace) is shown for comparison with the first PCA component (solid gray trace) from the entire Raman spectroscopy data set presented in B. The Raman spectrum of glycogen (black trace) is shown for comparison. (B) H460, MCF7 and LNCaP cells were irradiated with single fractions of radiation, and single-cell Raman spectra were collected from 20 cells from each sample post-irradiation. Experiments were performed in triplicate, resulting in 60 single-cell spectra per sample condition, and 3240 spectra overall. The mean PCA scores (n = 60 spectra per point, error bars are ± SE) for the first PCA component indicate statistically significant (p<0.05 by unpaired two-tailed t-test) increases in intracellular glycogen, relative to same day unirradiated cells. (C) Clonogenic survival of irradiated H460, MCF7 and LNCaP cells. Data are the mean ± SE from three independent experiments, each cultured in triplicate. *** p<0.001 (extra sum-of-squares F test).

Fig 2. Western blotting for the effect of irradiation on signals associated with the regulation of glycogen synthesis and utilization.

Whole cell lysates were prepared 3 days post-irradiation and Western blot analysis was conducted using the indicated antibodies. Blots are representative of three experiments. Actin was used as a loading control. Each panel was generated as a composite from the same gel with non-adjacent bands separated by a vertical line.

Fig 3. Single-cell Raman spectroscopy detects early and sustained inhibition of radiation-induced glycogen accumulation in metformin treated MCF7 cells.

(A, B) H460 and MCF7 cells were treated with or without 5 mM metformin from 1 hour prior to irradiation with single fractions at the doses shown. Single-cell Raman spectra were collected for 20 cells from each sample at 1, 2 and 3 days post-irradiation. Experiments were performed in triplicate, resulting in 60 single-cell spectra per sample condition, and 2160 spectra overall per data set. The mean PCA scores (n = 60 spectra per point, error bars are ± SE) for the first PCA component are shown. (C) Single-cell Raman spectra of 10 Gy irradiated MCF7 cells at 3 days post-irradiation, with and without 5 mM metformin. The difference spectrum (dashed trace) is shown for comparison with the first PCA component (solid gray trace) from the entire Raman spectroscopy data set presented in B. The Raman spectrum of glycogen (black trace) is shown for comparison. (D) Intracellular glycogen levels measured enzymatically 3 days post-irradiation in the presence or absence of 5 mM metformin. Each bar represents the average of 3 independent experiments ± S.D where each sample is assayed in triplicate. Statistical significance was determined using unpaired two-tailed t-test, n.s. not significant, *p<0.05, ** p<0.01.

Fig 4. Metformin suppresses GSK-3β but not AMPKα.

Western blotting was performed on lysates collected 3 days post-irradiation using the indicated antibodies, n = 3 experiments. Actin was used as a loading control. Each panel was generated as a composite from the same gel with non-adjacent bands separated by a vertical line.

Fig 5. Preferential sensitization of MCF7 cells treated with radiation and metformin.

(A) Total numbers of live and (B) dead cells in metformin co-treated H460 and MCF7 cultures were counted in triplicate at 3-days post-irradiation. Viable cell counts in (A) are relative to the untreated control for each cell line. Statistical significance tests shown apply to each dose-matched sample pairing within each group. Cell cycle distributions from each culture were determined via propidium iodide staining, and S-phase fractions for metformin co-treated (C) H460 and (D) MCF7 cells are shown at 3-days post-irradiation. Data are the mean ± SE from 3 independent experiments. * p<0.05, ** p<0.01, and n.s. not significant (unpaired two-tailed t-test). Clonogenic survival of metformin co-treated (E) H460 and (F) MCF7 cells performed 1 day post-treatment as described in the Materials and Methods. Data are the mean ± SE from three independent experiments, each cultured in triplicate. n.s.—no significant difference between curves, *** p<0.001 (extra sum-of-squares F test).