The Impact of Non-Lethal Single-Dose Radiation on Tumor Invasion and Cytoskeletal Properties

Abstract

Irradiation is the standard therapy for glioblastoma multiforme. Glioblastoma are highly resistant to radiotherapy and the underlying mechanisms remain unclear. To better understand the biological effects of irradiation on glioblastoma cells, we tested whether nonlethal irradiation influences the invasiveness, cell stiffness, and actin cytoskeleton properties. Two different glioblastoma cell lines were irradiated with 2 Gy and changes in mechanical and migratory properties and alterations in the actin structure were measured. The invasiveness of cell lines was determined using a co-culture model with organotypic hippocampal slice cultures. Irradiation led to changes in motility and a less invasive phenotype in both investigated cell lines that were associated with an increase in a "generalized stiffness" and changes in the actin structure. In this study we demonstrate that irradiation can induce changes in the actin cytoskeleton and motility, which probably results in reduced invasiveness of glioblastoma cell lines. Furthermore, "generalized stiffness" was shown to be a profound marker of the invasiveness of a tumor cell population in our model.

Keywords: actin; cell mechanics; cytoskeleton; glioblastoma; phalloidin; radiation; slice cultures.

Figure 1

Influence of irradiation on cell division and death. (A,B) representative images of the Ki67 staining of LN229 cells with and without irradiation with 2 Gy. Most of the cells were found to be proliferating; (C,D) depicts a representative sample of the Ki67 staining of U87 cells with and without irradiation with 2 Gy. One can observe a slightly decreased number of Ki67 positive cells compared to the control group. LN229 (A,B) cells are more proliferative than U87 (C,D) cells; (E) illustrates the respective Ki67 index (Ki67 positive cells/cell number) of LN229 (sample size: n_CTL = 20; n_2Gy = 16) and U87 (n_CTL = 21; n_2Gy = 27) cells before and after irradiation. Irradiation decreased the proliferation rate of U87 only; (F) shows the PCNA expression of LN229 (n_CTL = 5; n_2Gy = 5) and U87 (n_CTL = 4; n_2Gy = 4) cells, normalized to the control groups and a representative blot with GAPDH at 37 kDa and PCNA at 36 kDa. No significant effects were observed; (G) visualizes the ratio of propidium iodide positive (dead) cells to the total number of cells for LN229 (n_CTL = 10; n_2Gy = 10) and U87 (n_CTL = 10; n_2Gy = 10). No significant effect was observed. The graphs show the mean value together with the standard error of the mean (sem). Statistics was performed using t-test and significance was chosen for p < 0.05. The asterisk denotes significant results regarding the control measurement of the same cell line. Scale bar corresponds to 10 µm.

Figure 2

Results of the atomic force microscopy (AFM) and live cell imaging measurements for LN229 and U87 (A) Shows the Young’s modulus. Statistical significance was found after irradiation of LN229 cells (n_CTL = 60; n_2Gy = 35) only. U87 (n_CTL = 60; n_2Gy = 35) cells did not show a change in elasticity; (B) Calculated mean adhesion energies using the Derjarguin-Muller-Topolov model. Irradiation led to a decrease in adhesion for both cell lines. The number of analyzed cells used for determination of adhesion energy were identical to those for the Young’s modulus; (C) Derived speeds from the time laps images. Inverse effects could be observed for the two cell lines. LN229 reacted with a decrease (n_CTL = 118; n_2Gy = 247), while irradiation of U87 led to an increase in cell speed (n_CTL = 158; n_2Gy = 118); (D) Regarding the contact area of the cells after irradiation both cell lines reacted with a decrease in area. The numbers of experimental values for the cell area were identical to the ones for the cell speed. The graphs show the mean value together with the standard error of the mean. Statistical analysis was performed using the Mann–Whitney test and significance was chosen for p < 0.05. The asterisk denotes significant results regarding the control measurement of the same cell line.

Figure 3

Measurements of the invasion for LN229 and U87. (A) represents a typical invasion pattern generated by LN229 and U87 with or without irradiation after four days of invasion. In red the propidium iodide dyed cytoarchitecture of the OHSC is visualized (labeled PI), while green depicts the tumor labeled using carboxyfluorescin diacetate (labeled CFDA); (B) For both, LN229 (n_3dCTL = 37; n_3d2Gy = 22; n_4dCTL = 41; n_4d2Gy = 20) and U87 (n_3dCTL = 53; n_3d2Gy = 30; n_4dCTL = 51; n_4d2Gy = 16) cells, and 3 or 4d invasion time the irradiation led to a significant decrease in the invasiveness. Statistics was performed using the Mann–Whitney test and significance was chosen for p < 0.05. The asterisk denotes significant results regarding the control measurement of the same cell line. The scale bar corresponds to 400 µm.

Figure 4

Barplot of the composite parameter “stiffness” and actin structure measurements. (A) The composite parameter “stiffness” is in both cases strongly increased after irradiation; (B) depicts the structure image as a heat map on the left and the respective phalloidin staining on the right for U87 control cells. A correspondence of highly structured regions in the actin staining with the respective structure image is visible. Furthermore, unstructured, homogeneous regions do not contribute to the structure image, as for example in the center of the image.; (C) displays sample images of actin staining for U87 and LN229 cells with and without irradiation. A dense actin network is visible for both cell lines and the respective treatments; (D) shows the quantification of the quality of actin structures. A decrease after irradiation is found in the case of LN229 (n_CTL = 52; n_2Gy = 66) cells, while an increase is observed for U87 (n_CTL = 124; n_2Gy = 94) cells; (E) illustrates the changes of actin structure density after irradiation. Only in U87 cells an increase in structure density could be observed. The sample size is identical to the one of the quality measurements. Statistics was performed using the Mann–Whitney test and significance was chosen for p < 0.05. The asterisk denotes significant results regarding the control measurement of the same cell line. The scaling corresponds to 30 µm.