Ionizing radiation causes DNA damage and consequent apoptosis, mainly due to the production of hydroxyl radicals (HO•) that follows radiolytic splitting of water. However, superoxide (O2•−) and H2O2 also form and induce oxidative stress with resulting LMP (lysosomal membrane permeabilization) arising from iron-catalysed oxidative events. The latter will contribute significantly to radiation-induced cell death and its degree largely depends on the quantities of lysosomal redox-active iron present as a consequence of autophagy and endocytosis of iron-rich compounds. Therefore radiation sensitivity might be depressed by lysosome-targeted iron chelators. In the present study, we have shown that cells in culture are significantly protected from ionizing radiation damage if initially exposed to the lipophilic iron chelator SIH (salicylaldehyde isonicotinoyl hydrazone), and that this effect is based on SIH-dependent lysosomal stabilization against oxidative stress. According to its dose-response-modifying effect, SIH is a most powerful radioprotector and a promising candidate for clinical application, mainly to reduce the radiation sensitivity of normal tissue. We propose, as an example, that inhalation of SIH before each irradiation session by patients undergoing treatment for lung malignancies would protect normally aerated lung tissue against life-threatening pulmonary fibrosis, whereas the sensitivity of malignant lung tumours, which usually are non-aerated, will not be affected by inhaled SIH.
Non-surgical cancer therapy, e.g. chemo- and radio-therapy, is mainly based on the induction of apoptotic cell death following the production of ROS (reactive oxygen species). Proteins combating oxidative stress, such as members of the thioredoxin family of proteins, superoxide dismutases or catalases, are often up-regulated in tumour cells and associated with resistance to such therapies [1–5]. It is generally assumed that DNA damage, mediated by hydroxyl radicals (HO•) that are formed by radiolytic cleavage of water, is responsible for cell death caused by ionizing radiation .
It has been pointed out previously that, in addition to DNA damage and resultant p53-mediated cell death, LMP (lysosomal membrane permeabilization) induced by oxidative stress is a contributing factor in apoptotic cell death caused by ionizing radiation . Such LMP is dependent on intralysosomal redox-active iron that causes peroxidation and fragmentation of the lysosomal membrane secondary to the oxidative stress that radiation induces [7,8]. Fenton-type reactions between H2O2 and redox-active iron lead to the formation of HO• radicals inside the lysosomal compartment. It therefore follows that the lysosomal concentration of redox-active iron would be directly related to the extent of LMP. It has been found that irradiation-induced LMP can be abrogated by chelation of lysosomal redox-active iron using DFO (desferrioxamine) . DFO, however, stays within the lysosomal compartment following its endocytic uptake, causes iron starvation with ensuing cell death and is obviously not a well-suited chelator.
In the lung alveoli exist a large number of macrophages, many of which have engulfed erythrocytes and, consequently, contain iron-rich lysosomes that may burst as a consequence of ionizing radiation, induce macrophage death and contribute to the induction of radiation pneumonitis and pulmonary fibrosis. Furthermore, other pulmonary cell types may have iron-rich lysosomes and, interestingly, the reparative autophagy that is initiated by irradiation greatly enhances the amount of lysosomal redox-active iron . Reparative autophaphagy is a way for cells to degrade damaged constituents and involves the breakdown of cellular ferruginous materials, such as ferritin and mitochondria. As a result, autophagolysosomes transiently become rich in low-mass redox-active iron, although it is eventually transported out of the lysosomal compartment to be stored in ferritin, or exploited in a variety of anabolic processes within mitochondria and the cytosol [9,10].
Lung cancer is presently the leading cause for cancer-related death worldwide . Many cases of lung cancer require ionizing radiation as part of the management of this common group of diversified malignancies with a generally poor outcome. A major problem that limits the dose of radiation is the risk of inducing pulmonary fibrosis which may turn out to be life-threatening . Consequently, it is often necessary to apply a dose of ionizing radiation that is less than optimal for effective therapy. Improvement of therapeutic efficiency is therefore obviously needed. In the present paper, we suggest a somewhat unorthodox way of handling the situation. Most drugs or treatments aim to enhance the irradiation efficiency on tumours, whereas we instead suggest strategies for the protection of the surrounding normal tissue. So far, only one radioprotector, amifostine, which incidentally also happens to be a lysosomotropic iron chelator, has been explored in clinical trials [13–15].
In the present study, we have assessed the effect of lysosomal iron chelation by the lipophilic chelator SIH (salicylaldehyde isonicotinoyl hydrazone) on cell survival following irradiation in a variety of cultured cells. Since SIH enters and leaves cells rapidly, being in equilibrium with its concentration in the surrounding medium, it can easily be rinsed away and, in contrast with DFO, has no long-lasting effects .
AO (Acridine Orange) base was from Gurr. SIH (a gift from Professor Des Richardson, University of Sydney, Sydney, New South Wales, Australia) was dissolved in DMSO and then diluted in ethanol in such a way that the final stock solution contained SIH at a concentration of 10 mM in a 10% DMSO/90% ethanol vehicle. Aliquots of this stock solution were added to cell culture medium to obtain final concentrations of 10–100 μM SIH. Since DMSO is a well-known scavenger of HO• radicals, and protects against ionizing radiation , initial experiments were carried out to ensure that the low final concentration of the DMSO/ethanol vehicle had no influence on the cellular sensitivity to radiation or H2O2 (results not shown). All other chemicals were from Sigma–Aldrich.
Cell lines were originally from the A.T.C.C. (Manassas, VA, U.S.A.) or Uppsala University. HeLa and J774 cells were grown in DMEM (Dulbecco's modified Eagle's medium) (Gibco), U1690 cells were grown in MEM (minimal essential medium), and the cell lines U2020, U1810 and U1906e were grown in RPMI 1640 (Gibco). All media were supplemented with 10% (v/v) heat-inactivated FBS (fetal bovine serum), 2 mM glutamine and 100 units·ml−1 penicillin/streptomycin (PAA). Cells were grown in plastic flasks and 35-mm-diameter Petri dishes (Corning) at 37 °C in a 90% humidified atmosphere containing 5% CO2. They were subcultivated once or twice a week.
γ-I radiation was performed with a 137Cs source (Scanditronix) at the Karolinska Institute, Stockholm, at a photon dose rate of 0.5 Gy·min−1. Dosimetry was performed using an ionization chamber as well as with ferrous sulfate. According to the sensitivity of the cell lines used, doses were in the range 0–8 Gy. Cells were transported in insulated boxes and irradiated at room temperature (22 °C). The irradiation was carried out in fresh medium, with or without SIH. When applied, SIH was added 30 min before irradiation. The irradiated medium was replaced by fresh growth medium (without SIH) when cells were returned to standard culture conditions.
Estimation of clonogenic cell survival
Appropriate cell numbers were plated for survival using the clonogenic assay technique described previously . Single-cell suspensions were plated in 35-mm-diameter plastic Petri dishes or six-well plates in triplicate or quadruplicate in a final medium volume of 3 ml/dish or well and then left in the incubator for 3–4 h to attach before irradiation, which was performed as described above.
Following irradiation, the cultures were incubated for 10–14 days, with a change of medium after 5–7 days. Thereafter, colonies were fixed, stained and counted. Radiation-survival curves were constructed from one to four independent experiments.
Dose–response models for clonogenic cell survival
The doses for 10% survival levels were calculated to estimate the dose-modifying fraction, in this case the PF (protection factor).
Estimation of growth curves
Survival of HeLa cells was estimated as described above. Cultures were prepared in numbers that allowed daily counting for 3 days following irradiation. An alternative method to measure cell survival following irradiation was applied to the cell lines U1906e and U1810 as these cell typesdo not readily form colonies. In those cases, cells were seeded and grown in 25 cm2 culture flasks for 24 h before irradiation that was performed under conditions described above. Cells were then routinely subcultured in a 1:4 ratio and counted three times during a period of 14–16 days following irradiation. Estimation of cell numbers (cells/ml) was obtained by assaying attenuance (D) at 600 nm on trypsinized single-cell suspensions. The D600 values were compared with a standard curve that was constructed previously by counting a series of diluted cell suspensions in a Bürker chamber. Finally, growth curves were obtained by comparing cell numbers at a number of time points in relation to the cell number at the previous subcultivation.
Cell survival following exposure to H2O2
Cells were seeded in 96-well plates at 104 cells/well. After 16 h, the cells were incubated for 1 h with different concentrations of H2O2 (0–100 mM) in HBSS (Hanks balanced salt solution) with or without 100 μM SIH present. Some cells were incubated with 30 μM FeCl3 for 5 h before the H2O2 treatment [when added to culture medium, Fe(III) forms insoluble iron phosphates/hydroxides that are taken up by endocytosis and transported to the lysosomal compartment]. Following 1 h of H2O2 exposure (during this period of time, most of the H2O2 was degraded by the cells) cells were washed and returned to standard culture conditions. The number of viable cells was determined 24 h later using the Cell Proliferation Kit II (Roche Applied Science). This assay is based on formation of a coloured formazan following mitochondrial oxidation of the tetrazolium salt XTT [2,3-bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide] by metabolically active cells. The dye was quantified using a microplate reader (SpectraMax 340PC, Molecular Devices) at 490 and 650 nm.
Lysosomal membrane stability assay
AO is a metachromatic fluorophore and a lysosomotropic base (pKa=10.3), which becomes charged (AOH+) and retained by proton trapping within acidic compartments, mainly secondary lysosomes (pH 4.5–5.5). Using blue light excitation, normal cells show bright red lysosomes (indicating high AO concentration) and weak green cytoplasmic and nuclear fluorescence (indicating low AO concentration). The AO relocation technique [16,21] was used to show early lysosomal damage. The lysosomes of cells are pre-loaded with AO before exposure to any treatment that is supposed to cause LMP, which is registered by flow cytofluorimetry as an increase in green AO fluorescence that results from AO relocation to the cytoplasm.
Approx. 106 U1690 cells in 2 ml of complete medium were exposed to 10 μg/ml AO for 15 min under otherwise standard conditions. Cells were then washed with complete medium and equilibrated under standard conditions for another 15 min, before they were exposed to 100 μM H2O2 in HBSS, with or without 100 μM SIH, for 30 min at 37 °C. At the end of the oxidative stress period, cells were kept under standard culture conditions for another 30 min before they were trypsinized, and green AO fluorescence was analysed by flow cytofluorimetry (FACScan, Becton-Dickinson) using the FL1 channel.
Iron chelation protects cells against radiation-induced cell death
In order to find out whether iron chelation protects against cell death, several cell lines were irradiated with or without the iron chelator SIH present. We tested the mouse macrophage cell line J774, the cervical cancer cell line HeLa, and a number of lung cancer cell lines, U1690, U1906e and U1810. U1690 is a small-cell lung cancer cell line , U1810 is a radioresistant non-small-cell lung cancer cell line , and U1906e is a radiosensitive small-cell lung cancer subcell line . First, growth curves of HeLa, U1906e and U1810 were recorded (Figure 1). Cells were exposed to a single fraction of ionizing radiation at 2 Gy (Figure 1B), 3 Gy (Figure 1A) or 5 Gy (Figures 1C and 1D) with or without 10 μM SIH. In HeLa cells, we investigated the direct effect of radiation on cell survival, and in the lung cancer cell lines, we investigated the ability to repopulate after irradiation. Both immediate protection and repopulation were significantly improved by SIH. SIH-treated non-irradiated cells grew better than control cells (Figure 1). Since DMSO is known as a potent scavenger of HO• radicals , we ensured that DMSO in the 0.01–0.1% range had no protective effect (results not shown). Next, we determined the surviving fractions based on the clonogenic cell survival assay using HeLa, J774 and U1690 cells (Figure 2). The ability to undergo five or more cell divisions following irradiation is used as an indication of cell survival. A survivor that has retained its reproductive integrity and is able to proliferate continuously to produce a large clone or colony is said to be clonogenic. SIH increased the surviving fractions in all cell lines studied. In line with this result, exposure to an Fe(III) phosphate/hydroxide precipitate (obtained by adding 10 μM FeCl3 to the medium) that was endocytosed by the cells for 4 h before irradiation decreased the surviving fractions (Figure 2B). Compared with the control cells (irradiated without prior iron exposure), only approx. 30% of the iron-loaded cells survived the radiation doses of 6 Gy (Figure 2B) and 8 Gy (results not shown). Protection of cells against radiation was partly dependent on the SIH concentration. The PFs were calculated as the ratio of the doses that gave 10% survival with and without SIH protection respectively. As shown in Table 1, the PF for HeLa cells increased from 1.20 to 1.78 following doubling of the SIH concentration from 10 to 20 μM. This means that, as a consequence of SIH protection, the radiation doses can be increased by 20 and 80% respectively without a change in the survival rate. The PF for J774 cells was 1.30 at 10 μM SIH, whereas it was 1.20 for U1690 at 20 μM SIH. When cells were exposed to 2 and 4 Gy, which are reasonable daily doses in the treatment of lung cancers, the PF for U1690 was found to be between 1.40 and 1.80 in the presence of 20 μM SIH.
The iron chelator SIH preserves cell growth following ionizing radiation
The iron chelator SIH protects cells against ionizing-radiation-induced cell death
|Cell type||SIH concentration (μM)||PF||Additional radiation dose without changes in cell survival (%)|
|Cell type||SIH concentration (μM)||PF||Additional radiation dose without changes in cell survival (%)|
Iron chelation protects cells against H2O2-induced cell death
Since it is believed that the effect of ionizing radiation partly depends on intracellular formation of H2O2  with ensuing LMP , we investigated protection by SIH against H2O2-induced cell death. The small-cell lung cancer lines U2020  and U1690 were exposed to various concentrations of H2O2 with or without 100 μM SIH present (Figures 3A and 3B) and cell survival was calculated 24 h later. For U2020 cells, the EC50 H2O2 value increased from 0.22 mM to 7.85 mM (Figure 3A) and for the U1690 cells from 93 μM to 580 μM (Figure 3B). As was found previously for ionizing radiation (Figure 2B), survival decreased if cells were exposed to an iron phosphate complex before the induction of oxidative stress (Figure 3B). The EC50 value for U1690 cells fell to 53 μM following incubation with 30 μM FeCl3 for 5 h before ensuing H2O2 treatment. It should be pointed out that the addition of FeCl3 to culture medium results in the formation of an iron phosphate/hydroxide precipitate that is endocytosed by the cells. The lysosomal compartment is thereby enriched with iron.
The iron chelator SIH protects U2020 and U1690 cells against H2O2-induced cell death
Using again the clonogenic cell-survival assay, we calculated how many U1690 cells survived exposure to different concentrations of H2O2 compared with untreated cells. We found that, without SIH protection, no U1690 cells survived the 1 h period of H2O2 exposure at 50–150 μM initially, whereas 50–100% of the cells that were protected by 20 μM SIH did so (Figure 3C).
Iron chelation influences lysosomal stability under conditions of oxidative stress
U1690 cells were subjected to the AO-relocation test. Following AO loading, cells were exposed for 30 min to 100 μM H2O2 with or without 100 μM SIH in HBSS, and green fluorescence was assayed by flow cytofluorimetry after another 30 min (Figure 4A). Compared with the control cells, the mean green fluorescence increased up to 156% following exposure to H2O2 only, whereas cells exposed to 100 μM H2O2 under the protection of 100 μM SIH showed only a small increase in the mean green fluorescence; up to 110% of the control cells (Figure 4B).
SIH-protection against the effect of H2O2 is a function of lysosomal stabilization
Apart from radiolytic cleavage of water leading to formation of HO• radicals, the simultaneous production of H2O2 is a well-known effect of exposure of tissues to ionizing radiation [7,25]. However, the possible influence of H2O2 on radiation-induced cellular damage does not usually seem to be fully taken into account. This is somewhat surprising, since in a paper from 1962, Otto Warburg pointed out that the cellular effects of exposure to ionizing radiation or to H2O2 show substantial similarities :
Following studies of the damaging effects of randomly formed HO• radicals, it has been postulated that these short-lived (10−9 s) and extremely aggressive radicals react with nuclear DNA on the very spot where they are formed, causing adducts, mutations and single- and double-strand breaks with resulting cellular damage. Even if it is not definitively proven that HO• radical-induced DNA damage is the main cause of cellular injury following irradiation, there is an overwhelming amount of indirect evidence that this is indeed the case, and there seems to be little reason to question this dogma. However, apart from radiolytic cleavage of water, HO• radicals can also be produced by Fenton-type (transition-metal-mediated) reactions, which gives an incentive to examine the occurrence of such reactions during ionizing radiation:
Obviously, the presence of redox-active iron in direct contact with DNA would give rise to massive site-specific Fenton-type chemistry, given the radiation-induced presence of H2O2 and superoxide (O2•−). Under normal conditions, there are no indications of any significant amount of low-mass redox-active iron that is in juxtaposition to DNA [32–34]. However, as has been demonstrated, under conditions of oxidative stress, lysosomal rupture will occur, iron will be relocated and DNA damage initiated [7,32–34].
Because the lysosomal compartment is the centre for normal autophagic turnover of all organelles and most long-lived proteins, many of which are ferruginous compounds, lysosomes of all cells contain low-mass redox-active iron, explaining their vulnerability to oxidative stress [9,10]. An additional way of loading lysososomes with iron is of importance when scavenger cells, e.g. alveolar macrophages, endocytose erythrocytes and thereby enrich their lysosomal compartment with redox-active iron. The lysosomal compartment is acidic and rich in reducing equivalents, such as cysteine and glutathione, ensuring that any low-mass iron present would largely be in Fe2+ form [8,35]. That in turn would promote the generation of HO• radicals from H2O2 diffusing into this compartment.
Lysosomes show widely different sensitivity to oxidative stress . Using vital staining with lysosomotropic fluorochromes, e.g. AO or other available lysotrackers, it was found that, after heavy oxidative stress, some lysosomes always remain intact, while even low oxidative stress results in the rupture of a small, but obviously very sensitive, population of lysosomes . The explanation for this phenomenon is probably that lysosomes that are actively engaged in degradation of iron-containg macromolecules are rich in iron, whereas resting lysosomes may contain little or nothing of this transition metal .
Since the H2O2 that forms throughout the cell during irradiation is highly diffusible, it will enter the lysosomal compartment, meet redox-active iron and induce violent Fenton-type reactions with resultant LMP and release of lysosomal contents to the surrounding cytosol (Figure 5). LMP will thus allow not only the escape of low-mass iron from lysosomes, but also the relocation of potent lysosomal cathepsins. Dependent on the magnitude of lysosomal rupture, cell proliferation is stimulated or arrested by a minor or a somewhat more pronounced lysosomal destabilization respectively, whereas apoptosis or necrosis has been found to follow moderate or major destabilization respectively [37,38]. Consequently, the amelioration of LMP by chelating lysosomal redox-active iron in a non-redox-active form ought to reduce radiation sensitivity.
Schematic representation of suggested mechanisms behind SIH-mediated protection against radiation-induced cell death
This hypothesis was supported previously by findings following treatment with DFO at high doses for several hours before irradiation . Unfortunately, this hydrophilic and high-molecular-mass drug has the disadvantage of being taken up only by endocytosis [39,40] and is retained in lysosomes where it causes iron-starvation and, ultimately, cell death [9,10]. Therefore DFO is not an ideal iron chelator for cellular protection against oxidative stress. In the present study, we tested the radioprotective effect of the lipophilic iron chelator SIH that is rapidly distributed throughout the cell, but can also easily be washed away . That the protective effect reported in the present paper is due to the iron-chelating effect of SIH is supported by the experiments showing that addition of iron had a sensitizing effect (Figures 2B and 3B). Although SIH has been shown to give excellent protection from H2O2-induced oxidative stress [10,16], the findings of the present study suggest that SIH also can be used to protect normal tissues from radiation damage and may allow exposure to a higher than normal dose of ionizing radiation without causing damage in the normal tissue that is adjacent to a malignancy.
As an example, in the specific case of lung cancers, the tendency of normal lung tissue to develop radiation-induced pulmonary fibrosis severely limits the use of radiotherapy. Lung cancers usually compress a branch of the bronchus system, leaving most of the lung aerated, whereas the tumour itself and the lung tissue distal to it are not (Figure 5). An aerosol containing a powerful iron chelator might therefore protect normal lung tissue against radiation, while the tumour itself should not be affected (Figure 5). Our findings indicate that even low concentrations of SIH (10 or 20 μM) would allow the radiation dose to be increased by 80% without the induction of additional damage to normal tissue. This dose-modifying effect makes SIH one of the most powerful radioprotectors tested so far. Interestingly, cells exposed to SIH only actually grew better than the control cells, suggesting that SIH protects against damage caused by having cells outside the incubator. Inasmuch as SIH can be removed readily, allowing high concentrations to be used, one might expect striking effects. Indeed, SIH at 100 μM protected between 6- and 35-fold against H2O2-induced cell death (Figure 3). Moreover, doubling the SIH concentration increased its radiation-dose-modifying effect 4-fold (Table 1). All other radioprotective substances, e.g. thiol (sulfhydryl) compounds, phytochemicals and aminothiols, which are the most effective of the presently known radioprotectors, must be applied in much higher concentrations (0.5–10 mM) in order to reach similar PFs [41–47]. To confirm the high effectiveness of SIH, and to compare it with other radioprotectors, animal experiments are needed.
The additive effects of LMP, a consequence of intralysosomal Fenton-type reactions secondary to enhanced cellular amounts of H2O2, on top of the effects induced by direct formation of HO• radicals following radiolytic cleavage of water, are dependent on the presence of oxygen that allows formation of O2•− and H2O2 (see the formulae at the beginning of the Discussion). The importance of this additive effect is illustrated by the well-known fact that hypoxic malignancies, e.g. those that infiltrate bone tissue, respond less well to ionizing radiation. In hypoxic tissues, there will be limited formation of O2•− and H2O2 and, consequently, little LMP will take place.
In the present paper, we propose a new strategy for protection of cells against ionizing radiation and explain its underlying molecular mechanisms. Our results indicate that application of SIH as an aerosol before each irradiation session would allow exposure to a higher than normal irradiation dose and may increase the survival chance for lung cancer patients, which now show the highest mortality of all cancer patients , by protecting normally aerated, and therefore accessible to an aerosol, lung tissue, but not the solid malignancy without airways.
We thank Professor John Eaton, University of Louisville, Louisville, KY, USA, for valuable suggestions.
Carsten Berndt, Tino Kurz and Ulf Brunk wrote the manuscript; Carsten Berndt, Tino Kurz, Aristi Fernandes, Margareta Edgren and Ulf Brunk designed the experiments; Carsten Berndt, Tino Kurz, Markus Selenius and Margareta Edgren performed the experiments.
We thank the Deutsche Forschungsgemeinschaft [grant number BE3259–2 (to C.B.)], the Karolinska Institute (to C.B. and A.F.), Cancer och Allergifonden (to A.F.), Radiumhemmets Forskningsfonder (to A.F.), and Hjärt-Lungfonden (to M.S.) for financial support.
These authors contributed equally to this work.