In vitro proof of concept studies of radiotoxicity from Auger electron-emitter thallium-201

Background Auger electron-emitting radionuclides have potential in targeted treatment of small tumors. Thallium-201 (201Tl), a gamma-emitting radionuclide used in myocardial perfusion scintigraphy, decays by electron capture, releasing around 37 Auger and Coster–Kronig electrons per decay. However, its therapeutic and toxic effects in cancer cells remain largely unexplored. Here, we assess 201Tl in vitro kinetics, radiotoxicity and potential for targeted molecular radionuclide therapy, and aim to test the hypothesis that 201Tl is radiotoxic only when internalized. Methods Breast cancer MDA-MB-231 and prostate cancer DU145 cells were incubated with 200–8000 kBq/mL [201Tl]TlCl. Potassium concentration varied between 0 and 25 mM to modulate cellular uptake of 201Tl. Cell uptake and efflux rates of 201Tl were measured by gamma counting. Clonogenic assays were used to assess cell survival after 90 min incubation with 201Tl. Nuclear DNA damage was measured with γH2AX fluorescence imaging. Controls included untreated cells and cells treated with decayed [201Tl]TlCl. Results 201Tl uptake in both cell lines reached equilibrium within 90 min and washed out exponentially (t1/2 15 min) after the radioactive medium was exchanged for fresh medium. Cellular uptake of 201Tl in DU145 cells ranged between 1.6 (25 mM potassium) and 25.9% (0 mM potassium). Colony formation by both cell lines decreased significantly as 201Tl activity in cells increased, whereas 201Tl excluded from cells by use of high potassium buffer caused no significant toxicity. Non-radioactive TlCl at comparable concentrations caused no toxicity. An estimated average 201Tl intracellular activity of 0.29 Bq/cell (DU145 cells) and 0.18 Bq/cell (MDA-MB-231 cells) during 90 min exposure time caused 90% reduction in clonogenicity. 201Tl at these levels caused on average 3.5–4.6 times more DNA damage per nucleus than control treatments. Conclusions 201Tl reduces clonogenic survival and increases nuclear DNA damage only when internalized. These findings justify further development and evaluation of 201Tl therapeutic radiopharmaceuticals. Supplementary Information The online version contains supplementary material available at 10.1186/s13550-021-00802-w.

However, because of low signal-to-noise ratio, significant tissue attenuation and high radiation absorbed dose [8], diagnostic use of 201 Tl has substantially diminished in recent years.
Apart from its gamma rays (135 keV-3%, 167 keV-10%) and X-rays (67-83 keV-94%) used in clinical imaging [9], 201 Tl decay by electron capture to stable mercury ( 201 Hg) also produces around 37 short-range Auger and Coster-Kronig electrons, one of the highest numbers among other Auger-electron emitters (Table 1) [10] with therapeutic potential. The average total energy of Auger and Coster-Kronig electrons per decay is calculated as 15.3 keV [10] and even though this is comparable to the energy of 125 I, its half-life is around 20 times shorter and hence more amenable to clinical use. Auger electronemitters such as 111 In, 125 I and 67 Ga (Table 1) [10], have already shown potential for targeted radionuclide therapy of cancer [11][12][13], especially when delivered close to sensitive cellular targets such as DNA and plasma membrane [10,14,15]. However, the therapeutic potential of 201 Tl has barely been studied, despite its potent shower of Auger electrons and good worldwide availability. Nearly 40 years ago, in the context of the safety of diagnostic use of 201 Tl, Rao et al. observed radiotoxic effects of 201 Tl in spermatogonial cells after intratesticular injection in a mouse model [16], and Kassis et al. compared the cellular kinetics and radiotoxicity of 201 Tl to those of β-emitters ( 42 K, 43 K and 86 Rb) in Chinese hamster lung fibroblasts [17]. Since then there has been little interest in 201 Tl radiobiology or radionuclide therapy.
In this preliminary evaluation of the potential of 201 Tl as a therapeutic radionuclide, we investigated the rate and extent of non-targeted 201 Tl uptake into cancer cells via potassium channels, such as the Na + /K + pump ( Fig. 1), and used this mechanism of accumulation to assess its radiotoxic effect in human cancer cells in vitro. The primary aim of this study was to measure the effect of internalized and non-internalized 201 Tl on clonogenic survival and DNA integrity in prostate and breast cancer cells. Non-targeted 201 TlCl was used in this study only to assess the radiotoxic potential of the radionuclide, and not to assess the potential of 201 TlCl itself. In contrast to other Auger electron-emitting electrons, such as 111 In or 67 Ga, no suitable chelators are available for 201 Tl and their development can be challenging due to its physical and chemical properties. A finding that 201 Tl has potential as a therapeutic radionuclide would stimulate development of effective thallium chelators for construction of targeted bioconjugates.

Methods
Cell culture consumables, unless specified, were purchased from Sigma-Aldrich, UK. [

Cell culture
Human breast adenocarcinoma cells MDA-MB-231 (ATCC ® HTB-26 ™ ) were cultured in Dulbecco's Modified Eagle Medium (DMEM, low glucose 1000 mg/L) and human prostate carcinoma cells DU145 (ATCC ® HTB-81 ™ ) in RPMI-1640 medium. Both media were supplemented with 10% fetal bovine serum, 5% L-glutamine, penicillin (100 units) and 100 μg/mL streptomycin. Cultured cells were trypsinized and seeded at 250,000 cells per well in 24-well plates 16 h before each experiment and grown at 37 °C in a humidified 5% CO 2 atmosphere. During experiments, cells were kept in full medium or phosphate buffered saline (PBS; D8537) supplemented with MgCl 2 and CaCl 2 in concentrations equivalent to those found in media, or PBS without potassium (Alfa Aesar, J60465) supplemented as described above for PBS. KCl solutions were prepared by dissolving KCl (BDH Laboratory) in 0.9% NaCl. Concentrations of K + and Na + in PBS, PBS without K + , RPMI and DMEM media are shown in Table 2. To measure the rate of 201 Tl washout, cells were incubated at 37 °C for 90 min with 50 kBq (1 MBq/mL) 201 Tl in 250 µL medium (total), which was then removed. Adherent cells were washed briefly with 250 µL PBS and 250 µL of fresh non-radioactive medium was added to each well. Cells were incubated for various times (from 15 to 180 min) at 37 °C, after which medium was collected and cells washed and lysed as described above. A second method, to assess continuous complete 201 Tl efflux, involved repeated medium changes on the same Fig. 1 Schematic representation of K + , Na + and Tl + transport across cellular membranes. Sodium-potassium pump (Na + /K + -ATPase) transports K + ions inside and Na + ions outside a cell against their concentration gradient using energy released by the hydrolysis of ATP. Tl + ions, which mimic K + physiological behaviour, can be transported inside the cell via potassium channels, such as the sodium-potassium pump. Tl + transport mechanism outside the cell is currently unknown Table 2 Concentration of potassium and sodium in incubation solutions used in experiments cells. In brief, after incubation with 50 kBq (1 MBq/mL) 201 Tl as above, replacing radioactive medium with fresh medium and allowing a further 15 min of efflux; medium was again removed, cells were washed once with PBS and fresh medium added. This process was repeated 4 times at intervals from 15 to 60 min. After 180 min cells were lysed and 201 Tl activity measured as described above, converting CPM to activity by means of a calibration curve (Additional file 1: Fig. S1). A nominal average intracellular volume of 1.6 pL for individual DU145 cells (as measured by diffusion NMR spectroscopy [18]) was used for intracellular-to-extracellular concentration ratio calculations. 203 Tl is a stable isotope used as a target material in the 201 Tl manufacturing process [19] and despite a rigorous purification process, some 203 Tl is still present in [ 201 Tl] TlCl solution (Additional file 1: Fig. S2 and Table S1). Inductively Coupled Plasma-Mass Spectrometry (ICP-MS) analysis was performed using a PerkinElmer Nex-ION350D ICP-Mass Spectrometer to estimate 203 Tl concentrations in [ 201 Tl]TlCl decayed samples, using a calibration curve prepared as described in Additional file 1.

DNA damage
γH2AX assays were used to quantify DNA doublestrand breaks (DSBs). Cells were seeded on coverslips coated with poly-L-lysine (50 µg/mL) placed in each well of a 24-well plate. Following a 90-min incubation with 50 µL (50-2000 kBq) of different 201 Tl activities (200-8000 kBq/mL), 203 Tl/ 201 Hg in equivalent concentrations, or 201 Tl (1000 kBq, 4000 kBq/mL) with added KCl (15 and 25 mM), medium was removed and coverslips were washed with PBS, fixed with 3.7% formalin in PBS, treated for 15 min with 0.5% Triton X-100 ® and 0.5% IGEPAL CA-630 ® solution, incubated with 1% goat serum/2% bovine serum albumin (BSA) in PBS for 1 h, washed with PBS, incubated overnight at 4 °C with mouse antiphospho-histone H2A.X monoclonal with PBS and attached to microscope slides using 2.5% DABCO/Mowiol ® (except for the experiment blocking 201 Tl uptake with KCl, where cells were stained and mounted with Prolong ™ Gold Antifade Reagent with DAPI (Invitrogen)). Slides were stored at 4 °C. A TCS SP5 confocal microscope with Leica software was used to obtain fluorescent images and CellProfiler was used to quantify foci numbers per nucleus. After 90 min, the radioactive incubation solution was removed, cells were washed three times with PBS, trypsinized, re-suspended in non-radioactive medium, seeded at 1000 cells/well in a 6-well plate and cultured for 8-10 days, changing medium every 2-3 days. Colonies were fixed and stained with 0.05% crystal violet in 50% methanol and counted manually with blinding, defining colonies as containing > 50 cells. All clonogenic assays were done alongside uptake assays to relate activity per cell with the surviving fraction.

Statistical analysis
Data were analysed in Excel Microsoft 2016 and Graph-Pad Prism 8, and expressed as mean ± SD. A nonparametric Mann-Whitney test for unpaired data was used to determine significance (P < 0.05). Each experiment was performed in triplicate or quadruplicate. Error bars represent standard deviations among independent experiments, except when n = 1 where the error is intra-experimental.
The intracellular activity was calculated comparing the gamma counter measurements (cpm) of the collected medium activity (with three PBS washes) to the lysate (NaOH) activity. The background count (BC) was subtracted from both values.
For the gamma counter measurements, fractions of the culture medium and lysate were taken (when needed) and then multiplied by the appropriate factor to obtain total values necessary for calculations.

Tl kinetics: uptake and efflux
In preliminary experiments in RPMI or DMEM medium (Table 2) Fig. S3a). The % uptake was independent of the activity used within the range examined (Fig. 2b, c) showing that the uptake mechanism was not saturable in this range.   S3b) with the same intracellular-to-extracellular ratio of 201 Tl concentration. When medium was repeatedly refreshed, we again observed an exponential drop in the activity bound per cell, from 0.017 to 0.019 Bq at the beginning of the efflux experiment to approach 0 after 180 min (Additional file 1: Fig. S3c).
To modulate 201 Tl uptake into cells, cardiac glycosides digoxin and ouabain were evaluated but rejected because of their toxic effects on cells as described in Supplementary Material (Additional file 1: Fig. S4a and b) [20]. Instead, the effect of varying potassium concentration was evaluated. DU145 cells were incubated with 80 kBq/ mL of 201 Tl for 90 min in PBS solution without potassium ( Table 2) and supplemented with different concentrations of KCl rising to 50 mM. 201 Tl uptake decreased from 23.6 ± 0.6% in 0 mM potassium to 1.1 ± 0.1% in 50 mM potassium (Additional file 1: Fig. S4c). The potassium concentrations used showed no measurable baseline effects as measured by clonogenic assay or yH2AX imaging (see below), therefore variation of potassium concentration was adopted as the method of choice to modulate 201 Tl uptake.

Tl-induced nuclear DNA damage (yH2AX assay)
Results from three experiments on DU145 and MDA-MB-231 cells are shown in (Fig. 3a, Table 3). Cells treated with 4000 kBq/mL of 201 Tl showed a significantly higher average number of foci per nucleus (3.5 and 4.6 times for DU145 and MDA-MB-231 cells, respectively) than in the 0.9% NaCl-treated control (P < 0.05, Mann-Whitney test) (Additional file 1: Fig. S5a and b). Comparing the nuclear DNA damage in DU145 cells caused by 4000-8000 kBq/mL of 201 Tl to that caused by an equivalent volume of decayed [ 201 Tl]TlCl sample, we observed that only the radioactive component of the samples significantly increased the average number of foci per nucleus (P < 0.05, Mann-Whitney test) (Fig. 3b). The decayed sample caused no visible DNA damage compared to untreated controls (Fig. 3c) as expected, since thallium concentration is very low and other heavy metals (such as Hg decay product) were not detectable by ICP-MS (see Additional file 1: Fig. S2a and b and Table S1).   (Fig. 4c, d) reduced clonogenic survival by at least 90% compared to untreated cells.

The impact of 201 Tl internalisation on clonogenic survival and nuclear DNA damage in DU145 cells
Varying potassium concentrations in the incubation solution above and below the physiological level (3.5-5 mM [21]) was adopted as a method to modulate 201 Tl uptake in cells, for the purpose of comparing the radiotoxicity of intracellular and extracellular 201 Tl. Applying this method, we used the yH2AX assay to compare the average number of foci per nucleus in DU145 cells with high 201 Tl uptake ([K + ] 0 mM, 25.9 ± 1.7% uptake, 181:1 intra:extra ratio) and low 201 Tl uptake ([K + ] 15 and 25 mM, 1.6 ± 0.4 and 2.2 ± 0.2% uptake respectively, 8:1 and 12:1 intra:extra ratio) after incubation with the same radioactive concentration (4,000 kBq/mL) of 201 Tl (Fig. 5a, b) and (Additional file 1: Table S2). We observed a significant reduction in the average number of foci per nucleus induced by 201 Tl-from 31.9 ± 5.2 at 0 mM K + to 8.8  Table S2a and b) (very similar to literature estimates for K + ratios of 30:1-50:1 [16]). This ratio rises to 181:1 in potassium-free medium and collapses to 8:1 at elevated potassium levels (25 mM) (Additional file 1: Table S2c). These findings are consistent with the assumption that Tl + mimics potassium in a biological     of thallium is unchanged after entering the cell and the uptake mechanism is completely reversible, since the same ratios are preserved at every wash; and that only by the continuous efflux method can radioactivity be completely removed from the cells. Nevertheless, the high intracellular to extracellular concentration ratio achievable allowed evaluation of radiotoxicity of internalized 201 Tl without having a specific tumor-targeted uptake mechanism.
The radiotoxic effect of 201 Tl uptake in both cell lines was assessed by measuring clonogenic survival and the average number of DNA damage foci per nucleus. Treatment of both cell lines with 201 Tl in standard incubation media caused a substantial reduction in the surviving fraction compared to non-treated cells. Cells incubated with an equivalent volume of decayed [ 201 Tl]TlCl did not show this effect, indicating that the toxicity was solely due to the radioactivity. The average intracellular activity per cell required to achieve 90% reduction in clonogenicity over 90 min was estimated as 0.29 Bq (DU145) and 0.18 Bq (MDA-MB-231). These levels of exposure were accompanied by nuclear DNA damage measurable by the γH2AX assay, which showed a significant increase in the nuclear DNA DSBs caused by 201 Tl compared to negative controls. Larger standard deviations when measuring nuclear DNA damage could be linked to a heterogeneous uptake of 201 Tl among cells and/or its unequal subcellular distribution. Factors such as a different cell passage number or different cell cycle phases within the population, might explain large variation in results between experiments. Although nuclear DNA damage might be an important factor leading to a reduction in clonogenic survival, we cannot exclude 201 Tl impact on other susceptible cellular targets, such us cell membrane or mitochondria, both of which have been linked to the toxicity of other Auger electron-emitting radionuclides [22]. Currently, the precise sub-cellular localization of 201 Tl in cancer cells remains unknown.
In clonogenic survival studies on MDA-MB-231 cells we used 201 Tl activity of 8000 kBq/mL in one experiment only as the lower activities used were considered high enough to cause at least 90% reduction in  (Figs. 3b, 4b) showing that various components of decayed 201 Tl material are not affecting 201 Tl radiotoxicity, we aimed to use the maximum amount of the radioactive material (and its "cold" equivalent) to be able to observe the potential impact of non-radioactive thallium, if there was one. We therefore limited the number of experiments with higher 201 Tl activities (4000-8000 kBq/ mL) to DU145 (n = 3) only or n = 1 for MDA-MB-231. This is justified by repeatedly observing no impact of low thallium concentrations on cell toxicity and our efforts to reduce radiation expose caused by the higher doses of gamma radiation.
Based on the subcellular range of Auger electrons (< 1 µm) [11] and their typically low energy, we hypothesized that 201 Tl has to be internalized to exhibit its radiotoxic effect in cells and that the gamma and X-ray emissions do not contribute significantly to toxicity. To test this hypothesis, we needed a method to modulate 201 Tl cellular uptake from the medium. Na + /K + -pump inhibitors ouabain or digoxin, known to be effective in blocking potassium uptake [20], in our experiments caused substantial baseline cytotoxicity in tested cells. However, by varying the K + concentration in the incubation solution between 0 and 25 mM, we were able to achieve a range of 201 Tl uptake values from 25.9 (0.8 Bq/ cell) to 1.6% (0.05 Bq/cell), corresponding to intracellular to extracellular concentration ratios in the range 8:1 to 181:1-a wider range than could be achieved with the cardiac glycosides-without affecting baseline clonogenic survival. When 201 Tl uptake in cells was substantially suppressed by high K + concentration in the medium, both the loss in clonogenic survival and the increase in DNA DSB caused by 201 Tl were completely suppressed. This suggests that 201 Tl needs to be transported inside cells to cause significant radiotoxicity. In the future, if 201 Tl can be successfully incorporated into a targeted radiopharmaceutical, an optimal dose for the radionuclide administration can be established to achieve an effective therapeutic effect. Furthermore, an impact of low energy gamma and X-ray emission on patients and carers should be investigated to assess feasibility of this treatment. The same principle would apply to any other Auger electron-emitting radiopharmaceutical.
The average of 0.18 Bq/cell for MDA-MB-231 and 0.29 Bq/cell for DU145 of intracellular 201 Tl required to achieve 90% reduction in clonogenicity over 90 min is equivalent to around 1000-1600 decays per cell. This is less than half the number of decays per cell calculated for 111 In and 67 Ga-oxine complexes (3240 and 3600 decays per cell, respectively, over 60-min incubation) required to obtain the same 90% reduction in clonogenicity in DU145 cells [23]. Although these experiments were not conducted under fully identical conditions, this result is consistent with the higher total energy and number of Auger and Coster-Kronig electrons per decay released by 201 Tl (15.3 keV, 37 electrons) compared to 111 In and 67 Ga (6.7 keV, 14.7 electrons and 6.3 keV, 4.7 electrons, respectively) [24]. Because 201 Tl effluxes quickly from the cell after the incubation period, whereas 111 In and 67 Ga do not, the nominal incubation time in the latter cases severely underestimates the actual exposure time, suggesting that 201 Tl has a much higher radiotoxic potential per Bq.
Our results are consistent with early work by Rao et al. and Kassis et al. [16,17] addressing potential radiobiological risk associated with diagnostic use of 201 Tl, showing that decay of intracellular 201 Tl is toxic, reducing clonogenicity and inducing DNA DSBs. At the same time, 201 Tl decay presents little hazard to bystander cells that have not taken up the radionuclide. Although we have not compared this potency directly with other Auger electron-emitting radionuclides under consideration as radiotherapeutics, it appears that 201 Tl is more potent than 111 In and 67 Ga, consistent with the respective known number and energy of the emissions. Consequently, 201 Tl is an attractive candidate radionuclide for further research in targeted Auger electron-emitter therapy. In addressing this goal, it will be necessary to overcome several limitations of the present work. First, while we have shown that 201 Tl must be internalized to elicit significant radiotoxic effects, it is likely that different subcellular distributions will have greatly differing effects, since the range of the emitted electrons is much smaller than typical cell dimensions. Nothing is known about the subcellular distribution of 201 Tl in the cell lines investigated here, nor of 67 Ga and 111 In in similar studies [23]; indeed, in general the role of subcellular location of radionuclides in determining toxic effects is poorly understood. A radiopharmaceutical containing 201 Tl that is targeted both to cancer cells and to specific subcellular locations within them might be significantly more potent than [ 201 Tl]TlCl. Moreover, 'trapping' 201 Tl inside the cell will slow down its efflux and might further increase its radiotoxic potential. Despite the high potency, actualising therapeutic potential with 201 Tl is currently more challenging than with other potential Auger electron emitters ( 67 Ga, 111 In, 125 I etc.) because the chemistry required to incorporate 201 Tl into targeting molecules is underdeveloped-there are no satisfactory bifunctional chelators for thallium(I) or thallium(III). Here we have used physiological mimicry of potassium to achieve uptake in cells, but this has no specificity for cancer cells and probably has no specific targeting to subcellular structures.