Radiation Protection Dosimetry Advance Access published May 11, 2015 Radiation Protection Dosimetry (2015), pp. 1–5
doi:10.1093/rpd/ncv299
GLIOBLASTOMA STEM CELLS: RADIOBIOLOGICAL RESPONSE TO IONISING RADIATIONS OF DIFFERENT QUALITIES I. Pecchia1,*, V. Dini1,2, L. Ricci-Vitiani3, M. Biffoni3, M. Balduzzi2,4, E. Fratini2,5, M. Belli2, A. Campa1,2, G. Esposito1,2, G. Cirrone6, F. Romano6, C. Stancampiano6, F. Pelacchi3, R. Pallini7 and M. A. Tabocchini1,2 1 Department of Technology and Health, Istituto Superiore di Sanita`, viale Regina Elena 299, Rome 00161, Italy 2 Sez. Roma1-Gruppo collegato Sanita`, INFN, Rome, Italy 3 Department of Hematology, Oncology and Molecular Medicine, Istituto Superiore di Sanita`, Rome, Italy 4 ENEA-CASACCIA, Rome, Italy 5 Museo storico della fisica e Centro Studi e Ricerche ‘E. Fermi’, Rome, Italy 6 Laboratori Nazionali del Sud (LNS) – INFN, Catania, Italy 7 Department of Neurosurgery, Catholic University of Rome, Rome, Italy
Glioblastoma multiforme (GBM) is the most common and malignant primary brain tumour, with very poor prognosis. The high recurrence rate and failure of conventional treatments are expected to be related to the presence of radio-resistant cancer stem cells (CSCs) inside the tumour mass. CSCs can both self-renew and differentiate into the heterogeneous lineages of cancer cells. Recent evidence showed a higher effectiveness of C-ions and protons in inactivating CSCs, suggesting a potential advantage of Hadrontherapy compared with conventional radiotherapy for GBM treatment. To investigate the mechanisms involved in the molecular and cellular responses of CSCs to ionising radiations, two GBM stem cell (GSC) lines, named lines 1 and 83, which were derived from patients with different clinical outcomes and having different metabolic profiles (as shown by NMR spectroscopy), were irradiated with 137Cs photons and with protons or C-ions of 62 MeV u21 in the dose range of 5– 40 Gy. The biological effects investigated were: cell death, cell cycle progression, and DNA damage induction and repair. Preliminary results show a different response to ionising radiation between the two GSC lines for the different end points investigated. Further experiments are in progress to consolidate the data and to get more insights on the influence of radiation quality.
Glioblastoma multiforme (GBM, World Health Organization grade IV glioma) is one of the most common, malignant and lethal primary brain tumours and it is associated with very poor prognosis. Even after treatments like maximal surgical resection, systemic chemotherapy with temozolomide and local radiotherapy, the mean survival time is still in the range of just 12 –15 months. According to the cancer stem cell (CSC) hypothesis, the high recurrence rate and the failure of conventional treatments are expected to be related to the presence of radio-resistant CSCs inside the tumour mass. Therefore, cancer treatment should be aimed not only to eliminate the bulk of cycling and differentiated cancer cells but also especially to the eradication of the stem cell population. A GBM stem cell (GSCs) subpopulation expresses CD133 (Prominin-1), a trans-membrane glycoprotein marker of normal neural stem cells, associated with resistance to radiation and chemotherapy, thus indicating a more aggressive tumorigenic phenotype. Some literature data report that GSCs radio-resistance to gamma-rays is due to preferential activation of the DNA damage response followed by a better capability to repair DNA damage(1). On the other hand, recent experimental evidence showed a higher
effectiveness of C-ions with respect to photons in inactivating CSCs from colon carcinoma, likely related to the different quality of the induced DNA damage(2). These results suggest a potential advantage of Hadrontherapy compared with conventional radiotherapy. The aim of the present study was to increase knowledge on the radiobiological response of human CSCs from GBM to photons and charged particles, investigating the mechanisms involved at the molecular and cellular level. MATERIALS AND METHODS Cell lines GSCs isolated from surgical samples of patients with similar tumour location, gender and age but different clinical outcomes were used: line 1 and line 83, the former deriving from a patient with 6 months progression-free survival (PFS) and 12.5 months overall survival (OS) and the latter deriving from a patient with 3 months PFS and 6 months OS. According to the Stupp protocol, after surgery both patients received radiotherapy to limited fields (2 Gy per fraction, once a day, 5 d a week, 60 Gy total dose) and concomitant
# The Author 2015. Published by Oxford University Press. All rights reserved. For Permissions, please email:
[email protected] Downloaded from http://rpd.oxfordjournals.org/ at CSU Fresno on July 8, 2015
*Corresponding author:
[email protected] I. PECCHIA ET AL.
Irradiation conditions The same experimental protocols, both in terms of sample holders and timing, were maintained during charged particles and photons irradiations. Commercial 1-ml polystyrene cuvettes were used as sample holders. In these cuvettes, the cell suspensions were centrifuged and irradiated as a pellet, 4-mm thick, with doses of 5, 10, 20 and 40 Gy. Irradiation with charged-particle beams, namely protons and C-ions, was carried out at the superconducting cyclotron radiobiology facility of the INFNLNS in Catania. Before impinging the cellular sample, the beam passes through different types of materials. The energy spectra of the particles entering
and exiting from the 4-mm cell pellet were estimated using the Monte Carlo simulation program SRIM2013 (http://www.srim.org/). The energies at the entrance of the cell suspensions for protons and for C-ions were 61 and 52 MeV u21, respectively, corresponding to LET values (in water) of 1.1 and 43.2 keV mm21. The thickness of the 4-mm cell pellet does not affect the LET of protons that remains almost constant through the sample. On the contrary, for C-ions, there is a significant variation in the average LET from 43.2 to 59.0 keV mm21. 137 Cs gamma irradiation was performed at the ISS, Rome, at a dose rate of about 0.8 Gy min21. After irradiation, samples were tested for different biological end points. Gamma-H2AX assay At the end of various incubation times at 378C (i.e. 0.5 –2–4 and 24 h), unirradiated and irradiated samples were tripsinised, fixed in 70 % ethanol at room temperature (r.t.) and kept at þ48C until performing the assay. To this purpose, samples were seeded in 96-well U-bottom plates, washed twice with PBS and labelled with primary anti-phospho-histone H2AX, Ser 139 antibody (Upstate) diluted 1:1000, and subsequently with secondary Alexa Fluor 488 F(ab0 )2 fragment of goat anti-mouse IgG antibody (Invitrogen), diluted 1:100. Radiation induced phosphorylation levels of histone H2AX (g-H2AX), a commonly used in situ marker of DNA double-strand breaks (DSB), was determined by flow cytometry in terms of median of the fluorescence intensity distribution. Data analysis was carried out using FlowJo software (Tree Star). Cell cycle analysis To assess cell cycle progression after irradiation, cells were incubated at 378C in a humidified atmosphere of 5 % CO2 for 2, 24, 48, 72, 96 h and 7, 8, 9, 10 d. At the end of each time, cells were tripsinised, fixed in cold 70 % ethanol at r.t., and treated with propidium iodide and RNase solution for 40 min. Measurement of cellular DNA content and cell cycle analysis (i.e. fraction of G0/G1, S and G2/M cells) after gammarays and protons irradiation were performed by flow cytometry. Data analysis was carried out using FlowJo software (Tree Star). Apoptosis assay Unirradiated and irradiated cells were seeded in sixwell plates and incubated at 378C in a humidified atmosphere of 5 % CO2 for 48 and 72 h after irradiation. At the end of each incubation time, the samples were centrifuged, tripsinised, resuspended in buffer solution and stained with Annexin V and propidium
Page 2 of 5
Downloaded from http://rpd.oxfordjournals.org/ at CSU Fresno on July 8, 2015
temozolomide (75 mg m22 of body surface area per day) for 7 d a week from the first to the last day of radiotherapy followed by cycles of adjuvant temozolomide (at 200 mg m22 of body surface area on days 1–5) given at 4-week intervals. Differences in the outcomes of the two patients are likely to be related with the poorer response to radiotherapy and temozolomide by the donor patient of line 83, whose tumour showed unmethylated O 6-methylguanineDNA methyltransferase (MGMT) promoter. Specimens of tumour tissues were obtained by surgical resection at the Institute of Neurosurgery, Catholic University School of Medicine, Rome. GSC lines were isolated by mechanical dissociation of surgical specimens, and cultured in serum-free medium supplemented with growth factors (EGF and FGF) at the Department of Hematology, Oncology and Molecular Medicine, Istituto Superiore di Sanita`, Rome. The stemness of the GSC lines was defined by primary spheres formation in vitro, by self-renewal capacity, by co-expression, under serum stimulation, of astrocytic and neuronal phenotypic markers in vitro, and by the generation of tumours in immunodeficient mice(3). The two used GSC lines show differences in their growth characteristics: line 1 grows in suspension and the cells have a doubling time of nearly four days; line 83 presents a part of the population growing in suspension and a part adherent to the flasks’ surface, the two components being in balance, and the cells have a doubling time of about 2 d. Guidoni et al. (4) carried out the metabolic characterisation of several GSC lines identifying two major phenotypes: a mixed neural/astrocytic metabolic phenotype mainly based on oxidative phosphorylation, with a strong neuronal fingerprint, and an astrocytic/glioma-like metabolism relying on glycolysis. In the two GSC lines used in this study, both a neural-like and glioma-like metabolisms were identified with the prevalence of the former in line 1 and of the latter in line 83.
GLIOBLASTOMA STEM CELLS
iodide (Bio Vision Annexin V-EGFP Apoptosis detection Kit). The fraction of viable, necrotic and/or apoptotic cells was evaluated by flow cytometry using a biparametric analysis carried out using FlowJo software (Tree Star). Cell killing assay
RESULTS AND DISCUSSION Quantification of DNA damage induction and repair, in terms of DSB as evaluated by g-H2AX assay using flow cytometry (Normalized Relative Fluorescence Intensity), shows no significant differences between line 1 and line 83 after exposure to gamma-rays (Figure 1). In particular, after 24 h repair, the two cell lines show a similar fraction of residual DNA damage. However, lines 1 and 83 show different backgrounds (6.3+3.2 vs 3.1+0.3) as well as different radiation-induced g-H2AX levels (57.3+1.1 vs 27.5+8.0 30 min after 40 Gy of gamma-rays). These differences could be related, for example, to a different chromatin organization and/or to different levels of endogenous radio-protective compounds (e.g. GSH). Proton data, coming from a single experiment performed in duplicate, show a similar response when compared with photons, in particular after 24 h repair. The study of the cell cycle has highlighted a marked G2/M arrest in line 83 after both photons and protons irradiation with doses higher than 5 Gy. The accumulation of cells in G2/M phase is dose dependent; it begins 24 h after irradiation and lasts at least up to 7 d. Line 1 grows more slowly with respect to line 83 and therefore the presence of the cell cycle
CONCLUSIONS According to the CSC hypothesis, CSCs have the ability to self-renew and maintain the tumour growth; therefore, the complete eradication of tumours depends on the elimination of the minority CSCs population. The resistance to chemical and radiation treatments makes CSCs the main target of more effective therapies, among them Hadrontherapy. In this study, for some of the end points analysed, preliminary results showed a different response in the two GSC lines to different types of ionising radiations. In particular, when compared with photons, charged particles seem to be more effective in CSCs killing. This finding can be related to the different quality of the induced DNA damage, and it is in line
Page 3 of 5
Downloaded from http://rpd.oxfordjournals.org/ at CSU Fresno on July 8, 2015
Radiation-induced cell killing was evaluated after microtiter plating and scoring for wells positive for survivors(5). In these experiments, positive wells were defined as those containing both neurospheres and single viable cells. After verification that no differences were observed between delayed and immediate plating, for convenience (samples transport from Catania to Rome), the assay was carried out 2 d after irradiation. To this purpose, irradiated and control samples were tripsinised, diluted at the appropriate concentration and seeded in 96-well flat-bottom plates, two plates per sample. Dilutions of cell suspensions were made in such a way as to have one cell per well for unirradiated and 5 Gy irradiated samples, and to have three cells per well for 10-Gy, 20-Gy and 40-Gy irradiated samples. Fresh medium was added to the wells once a week and scoring of survivors was performed by microscopy about a month after seeding. The surviving fraction was calculated as the ratio between the plating efficiency at dose D and the plating efficiency of unirradiated cells.
arrest was investigated for a longer time, up to 10 d (data not shown). Nevertheless, line 1 does not show any evidence of cell cycle arrest. Further experiments are in progress to better investigate the underlying mechanisms and the fate of line 83 cells accumulating in the G2/M phase, in particular, whether these cells undergo mitotic catastrophe or whether they are eventually able to overcome the damage and re-enter the cell cycle Cell death by apoptosis, evaluated by bi-parametric flow cytometry analysis in both cell lines at 48 and 72 h after gamma-rays, protons and C-ions irradiation, did not show any significant induction of apoptosis or necrosis under the experimental conditions so far investigated (data not shown). Experiments aimed at investigating radiation induced cell death showed the presence of both neurospheres and single cells in the positive wells, being single cells especially present in line 1. In general, for both lines, the neurospheres formation decreases with increasing dose. Overall, the results (Figure 2) show that after gamma-rays irradiation, the survival of line 83 decreases more sharply than that of line 1. C-ions (and arguably protons), at doses of 10 and 20 Gy, seem to be more effective than gamma-rays in decreasing survival in line 83. Also for line 1 charged particles seem to be more effective than gamma-rays, although no significant differences were observed between protons and C-ions. This last finding is quite surprising; however, it must be taken into account that the protons and C-ions data come from a single experiment so they must be considered very preliminary. In both lines 1 and 83 at doses higher than 10 Gy, a plateau can be observed in the survival levels after charged particles and photons irradiation, suggesting the presence of radio-resistant subpopulations of CSCs. NMR analysis of clones deriving from irradiated cells is planned to investigate possible differences in the metabolic profiles(6).
I. PECCHIA ET AL.
Figure 2. Cell inactivation by different radiation qualities. The surviving fraction was calculated as the ratio between the plating efficiency at dose D and the plating efficiency of unirradiated cells. The results of photons irradiation are the mean of three independent experiments for line 1 and two independent experiments for line 83, with its standard error. The protons and C-ions data come from a single experiment.
with a potential advantage of Hadrontherapy in GBM patients. Further experiments are in progress to consolidate data and to get more insights on DNA damage response, cell cycle checkpoints and repair pathways in GSC lines. Understanding the molecular
mechanisms underlying both radio and chemoresistant phenotypes will provide important insight for developing novel and cell line-specific therapeutic approaches. DNA repair mechanisms and drug resistance must be thoroughly investigated and clarified for
Page 4 of 5
Downloaded from http://rpd.oxfordjournals.org/ at CSU Fresno on July 8, 2015
Figure 1. Gamma-H2AX fluorescence intensity bar plots. NRFI, normalized relative fluorescence intensity. The histograms for each dose were obtained subtracting the fluorescence values of the unirradiated samples from the irradiated ones in GSC lines 1 and 83, at the end of each incubation time, and normalizing the data to the maximum fluorescence value, for each experiment. Gamma irradiation: upper plots; protons irradiation: bottom plots. The error bars are evaluated as fluctuations among different experiments. Having a single experiment, we chose not to plot any error bar for proton data.
GLIOBLASTOMA STEM CELLS
each line in order to identify the characteristics that allow an effective treatment for each one. Many studies suggest that MGMT methylation, although not completely predictive of tumour response, might be an important prognostic factor associated with better OS and PFS in GBM patients treated with radiotherapy and temozolomide. This makes chemotherapy less effective in cell lines that do not express MGMT DNA repair enzyme. Similarly a better understanding of DNA repair mechanisms is needed as a preferential activation of DNA damage checkpoint and radiation induced repair(1) was shown in GMB CSCs. A comprehensive analysis of DNA repair proteins, e.g. by transcriptomics, could shed light on the repair mechanisms involved.
The authors are indebted with the technical staff of the INFN-Laboratori Nazionali del Sud and with Pasqualino Anello for his skilful technical assistance during the experiments at the ISS. FUNDINGS This work is partially supported by the National Institute of Nuclear Physics (INFN), ‘RADIOSTEM’ experiment.
1. Bao, S., Wu, Q., McLendon, R. E., Hao, Y., Shi, Q., Hjelmeland, A. B., Dewhirst, M. W., Bigner, D. D. and Rich, J. N. Glioma stem cells promote radioresistance by preferential activation of the DNA damage response. Nature. 444, 756–760 (2006). 2. Cui, X., Oonishi, K., Tsujii, H., Yasuda, T., Matsumoto, Y., Furusawa, Y., Akashi, M., Kamada, T. and Okayasu, R. Effects of carbon ion beam on putative colon cancer stem cells and its comparison with X-rays. Cancer Res. 71, 3676–3687 (2011). 3. Pallini, R. et al. Cancer stem cell analysis and clinical outcome in patients with Glioblastoma multiforme. Clin. Cancer Res. 14, 8205–8212 (2008). 4. Guidoni, L., Ricci-Vitiani, L., Rosi, A., Palma, A., Grande, S., Luciani, A. M., Pelacchi, F., di Martino, S., Colosimo, C. and Biffoni, M. 1 H NMR detects different metabolic profiles in Glioblastoma stem-like cells. NMR Biomed. (2013) doi:10.1002/nbm.3044. 5. Thilly, W. G. et al. Gene-locus mutation assays in diploid human lymphoblast lines. In: Chemical Mutagens: Principles and Methods for their Detection, vol. 6. Hollander, A. and de Serres, F., Eds. Plenum Press (1980) ISBN 0306371014. 6. Rosi, A., Palma, A., Grande, S., Luciani, A. M., RicciVitiani, L., Runci, D., Biffoni, M., Pallini, R., Guidoni, L. and Viti, V. Glioblastoma stem cell metabolism modulated by different quality radiation. In: 40th Annual Meeting of the European Radiation Research Society (ERRS). 1 –5 September 2013, Ireland, Book of abstract T-6.
Page 5 of 5
Downloaded from http://rpd.oxfordjournals.org/ at CSU Fresno on July 8, 2015
ACKNOWLEDGEMENTS
REFERENCES