Experimental validation of proton boron capture therapy for glioma cells

Proton boron capture therapy (PBCT) has emerged from particle acceleration research for enhancing the biological effectiveness of proton therapy. The mechanism responsible for the dose increase was supposed to be related to proton-boron fusion reactions (11B + p → 3α + 8.7 MeV). There has been some experimental evidence that the biological efficiency of protons is significantly higher for boron-11-containing prostate or breast cancer cells. The aim of this study was to evaluate the sensitizing potential of sodium borocaptate (BSH) under proton irradiation at the Bragg peak of cultured glioma cells. To address this problem, cells of two glioma lines were preincubated with 80 or 160 ppm boron-11, irradiated both at the middle of 200 MeV beam Spread-Out Bragg Peak (SOBP) and at the distal end of the 89.7 MeV beam SOBP and assessed for the viability, as well as their ability to form colonies. Our results clearly show that BSH provides for only a slight, if any, enhancement of the effect of proton radiation on the glioma cells in vitro. In addition, we repeated the experiments using the Du145 prostate cancer cell line, for which an increase in the biological efficiency of proton irradiation in the presence of sodium borocaptate was demonstrated previously. The data presented add new argument against the efficiency of proton boron capture therapy when based solely on direct dose-enhancement effect by the proton capture nuclear reaction, underlining the need to investigate the indirect effects of the secondary alpha irradiation depending on the state and treatment conditions of the irradiated tissue.

www.nature.com/scientificreports/ that can bring about a very significant dose enhancement in tissue enriched with 10 B isotope. According to the early clinical studies, BNCT has the potential to become effective as a targeted therapy 10 . Being initially limited by the need of usage of research nuclear reactors as the neutron source, nowadays BNCT approach is boosted by the availability of specialized compact neutron source specially designed for medical purposes 11 Several investigations are under way for the use of BNCT in high grade glioma treatment 12,13 . One of the critical components for successful deployment of BNCT is the choice of compounds for 10 B delivery to the tumor that would meet the requirements for high tumor uptake, low normal tissue uptake, low toxicity and rapid clearance after treatment [14][15][16] . Currently the two major compounds used for this purpose are sodium borocaptate ((B 12 H 11 SH) Na 2 ), BSH) and boronophenylalanine (C 9 H 12 BNO 4 , BPA) 10 . Approval of these boron-containing drugs in clinical trials of BNCT could give a clear advantage for their use in other types of radiotherapy. Proton boron capture therapy (PBCT) was proposed as a potential cancer treatment method a few decades ago. In PBCT the dose enhancement effect was assigned to the reaction of nuclear fusion between 11 B nuclei and the incident beam protons ( 11 B + p → 8 Be + α → 3α + 8.7 MeV). Emission of the high Linear Energy Transfer (LET) α-particles by proton-boron fusion reaction, that occurs primarily in the Bragg peak region and causes irreparable clustered DNA damages, was expected to increase the effectiveness of destruction of the tumor cell. There are two experimental reports showing a significant increase in the effectiveness of proton beam irradiation of 11 B-containing cancer cells 17,18 . However, other works containing simulation and analytical calculations seem to indicate that the total number of the generated alpha particles cannot explain the experimental results 19,20 . Thus, the reported evidence for the possibility to use of boron-containing compounds for the dose enhancement in PT as well as the possible mechanisms of their radio-sensitizing effect remain contradictory and, in the view of a very limited number of studies on PBCT, call for a further investigation.
In the present study we have examined the possibility to enhance biological effects of proton irradiation in boron-containing cultured glioma cells.

Results
We examined the radiosensitizing effect of sodium borocaptate for proton beam irradiation applied to two glioma cell lines (A172 and Gl-Tr), as well prostate cancer cells Du145, for which such an effect of BSH was demonstrated earlier 17,18 . The boron-11 concentration in most experiments was 80 ppm, and the periods of cell incubation with BSH were 7 h and 18 h. After incubation with BSH the cells were irradiated with 89.7 MeV clinical proton beam, positioning the cells at the distal end of Spread-Out Bragg Peak (SOBP) (Supl. Figure 1). An additional experiment was carried out with irradiation of cells with a 200 MeV proton beam, positioning the cells in the middle of the SOBP (Supl. Figure 1). These incubation points and irradiation conditions were selected based on the earlier reports demonstrating increased cell death under proton irradiation in the presence of different boron-containing compounds 17,18,21 . Cytotoxicity of sodium borocaptate for glioma cells. To determine the threshold toxic concentration of 11 B, glioma cells, Du145 cells, as well as fibroblasts of human skin, DF-2, were incubated in the presence of BSH for 7 h and 18 h. The cytotoxic effects of BSH at 40-1000 ppm of 11 B concentration have been studied by the AlamarBlue cell viability assay. The cytostatic effect of sodium borocaptate on glioma cells, as well as Du145 prostate cancer cells, is manifested at concentrations from 500 ppm of 11 B and higher ( Fig. 1A-C). For normal cells, DF-2 fibroblasts, the cytotoxic effect was not observed over the entire range of analyzed boron concentrations (Fig. 1D). No significant decrease in the survival rate of all studied cell types was observed at 80-160 ppm boron concentration range used in the study.

Cell cycle parameters of glioma cells in the presence of sodium borocaptate. Measurements
of cell cycle parameters of A172 cells during incubation for 7 h (p = 0.12) or 18 h (p = 0.507) with BSH at a concentration of 80 ppm showed a typical pattern of cell distribution along the cell cycle, recorded in control cells ( Fig. 2A). For Gl-Tr cells, a very slight increase in the number of cells in the G0/G1 phase of the cell cycle (p = 0.05) was observed during 18 h of incubation with sodium borocaptate (Fig. 2B). Thus, in the analyzed  Effect of preincubation with 80 ppm boron-11 as part of BSH on the sensitivity of glioma cells to proton irradiation at the Bragg peak. Next, we investigated whether the presence of sodium borocaptate increased the sensitivity of glioma cells to proton irradiation. Cells were incubated in a medium containing 80 ppm of boron-11 for 7 h or 18 h and then irradiated at the distal end of SOBP of 89.7 MeV clinical proton beam in a dose range of 0-6 Gy. Irradiation was carried out in the presence of boron-11 in the cell culture medium. The efficiency of irradiation was determined using both the AlamarBlue Assay and the Clonogenic assay. Cell viability was fitted with a linear-quadratic function of the radiation dose. In addition, irradiation dose for half-maximal inhibition of cell viability (IC 50 ) and a dose-modifying factor for 20% cell survival (DMF 20 ) in the presence of borocaptate, as compared to irradiation alone, were determined. The Mann -Whitney test was used to assess the differences between groups (with/without BSH) and one sample t-test-to evaluate difference DMF 20 values from 1. Evaluation of the effectiveness of radiosensitization with sodium borocaptate of glioma cells, carried out using the AlamarBlue Cell Viability Reagent, did not demonstrate any significant effect from the presence of boron in the cell culture medium before and during irradiation (Fig. 3A,B). Fitting parameters, irradiation dose for half-maximal inhibition of viability, and dose-modifying factor for 20% survival of Gl-Tr cells with irradiation alone or after incubation with sodium borocaptate for 7 h are presented in Table 1. A similar lack of sensitizing effect was observed for Du145 cells (Fig. 3C) The AlamarBlue assay is commonly used to quantify the number of live cells in a sample, and to monitor cell viability / cytotoxicity. But only a fraction of live seeded cells retains the capacity to produce colonies after irradiation. Clonogenic assay is the method of choice to determine cell reproductive death after treatment with ionizing radiation. Therefore, at the next step of the study, we tested the ability of glioma cells incubated with 80 ppm of boron-11 to form colonies after irradiation with an 89.7 MeV clinical proton beam at the distal end of Bragg peak. For cells of both glioma lines we observed a very slight and not significant effect of reducing the number of viable colonies in the samples incubated and irradiated in the presence of sodium borocaptate   Fig. 4A,B). For example, for Gl-Tr cells, the irradiation doses corresponding to 50% survival of colonizing cells were 1.9 ± 0.8 and 1.4 ± 0.6 (p = 0.275) for irradiation alone and for irradiation of cells preincubated with BSH for 7 h. At the same time, the dose-modifying factor, DMF 20 for these cells was 1.32 ± 0.15 (p = 0.07). This weak effect was not reproduced stably in independent experiments and did not depend on the time of incubation with the BSH (Fig. 4A,B). Moreover, for Du145 cells, we did not observe a significant effect of sodium borocaptate introduced into the culture medium on the formation of viable colonies after clinical proton beam irradiation ( Experiments to determine the survival of the cells and their ability to form colonies after incubation with BSH were also carried out using a 200 MeV proton beam. The parameters of cell irradiation, as well as the conditions of cell incubation with boron-11, were chosen similar to our previous study, in which we observed the sensitizing effect of sodium tetraborate (Na 2 B 4 O 7 ) during proton irradiation of glioma cells 21 . Cells were incubated with 80 ppm of boron-11 for 18 h, transferred to test tubes in a complete culture medium without BSH, irradiated at the middle position of SOBP (Suppl. Fig. S1), and then analyzed for viability by AlamarBlue and colony forming assay. A comparative analysis of the metabolic activity of cells after irradiation with a 200 MeV proton beam did not reveal any radiosensitizing effect of boron-11 for both glioma cells and prostate cancer cells (Fig. 5, upper panels). DMF 20 for A172, Gl-Tr and DU145 cells preincubated with sodium borocaptate were less than unity 0.98 ± 0.29 (p = 0.948), 0.92 ± 0.08 (p = 0.02) and 0.98 ± 0.31 (p = 0.468), respectively. When analyzing the ability of cells to form colonies after proton irradiation, a slight decrease in the number of colonies of A172  Table 1. Parameters for the linear-quadratic fit of cell viability curve Y(X) = exp (-αX-βX 2 ), irradiation dose for half-maximal inhibition of viability (IC 50 ), and dose-modifying factor for 20% survival (DMF 20 ) of Gl-Tr cells with proton irradiation alone (Control) or after incubation with 80 ppm of boron-11 (BSH) for 7 h.   www.nature.com/scientificreports/ cells preincubated with BSH was observed compared with only irradiated cells (Fig. 5A, bottom panel). The determined IC 50 for these cells without and with sodium borocaptate was 0.75 ± 0.30 and 0.50 ± 0.30 (p = 0.05), and DMF 20 was 1.20 ± 0.01 (p = 0.05) (Suppl . Table S1). No significant radiosensitizing effect of boron-11 was observed for the second glioma cell line (Fig. 5B, bottom panel). Prostate cancer cells formed the same number of colonies after irradiation with a 200 MeV proton beam, both after cell incubation with and without sodium borocaptate (Fig. 5C, bottom panel). DMF 20 for Gl-Tr and DU145 cells preincubated with sodium borocaptate were 1.02 ± 0.12 (p = 0.415) and 1.19 ± 0.04 (p = 0.144), respectively (Suppl . Tables S2,S3). Examples of colony assays for proton boron capture treatment at the middle position of SOBP of 200 MeV proton beam are shown in Supplementary Fig. S4.
Possibilities for increasing the biological effect of the sodium borocaptate on the survival of cells when irradiated with clinical protons at the Bragg peak. Despite the high content of boron atoms per single BSH molecule ((B 12 H 11 SH)Na 2 ), sodium borocaptate has a disadvantage for clinical usage for BNCT of gliomas, namely, poor cellular uptake by brain tumors 15,22 . To test the hypothesis about the possible insufficiency of boron in the culture medium of glioma cells to obtain the radiosensitizing effect of BSH for proton boron capture therapy, we increased the concentration of boron-11 up to 160 ppm. We did not observe any sensitizing effect when the amount of BSH was doubled during incubation and 89.7 MeV clinical proton beam irradiation of glioma cells (Fig. 6A,B). Moreover, we also did not observe any effect of boron proton capture therapy for Du145 prostate cancer cells (Fig. 6C), for which the expected effect was described earlier 17,18 .  (Fig. 7A). In addition, in these experiments we used a triple concentration of borocaptate corresponding to 250 ppm of boron-11. Example of colony forming assays for 7 h of incubation of Du145 cells with BSH and subsequent proton beam irradiation along the proton SOBP is shown in Fig. 7B. No dose enhancement effect was observed when cells were irradiated in the presence of BSH both at the position I of the proton beam and at position II, at the distal end of the SOBP (Fig. 7C,D). The calculated DMF 20 for cell irradiation at position I with or without sodium borocaptate was close to 1 when using the AlamarBlue metabolic test and less than 1 for the clonogenic assay. The similar factors for the irradiation position II were indistinguishable from 1 (Suppl . Table S3). Summarizing the results, we found that irradiation-induced cellular lethality was either not enhanced by pre-treatment of cells with BSH, or the observed enhancement was rather small and not statistically significant.

Discussion
A method to increase the biological effectiveness of proton beam is of high interest in radiation oncology. In thefirst publication on proton boron capture in radiation therapy authors used Monte Carlo simulations and showed an enhanced high dose in the areas where the boron uptake coincided with the Bragg peak 23 . In this work the increase of the maximum dose level was reported at about 50% for 80 MeV proton energy beam. Also, the increase of the maximum dose level of PBCT was reported at 192.4% for 75 MeV beam 24 . The mechanism responsible for a higher dose was suggested to be related to proton-boron fusion reactions, leading to the production ofhigh Linear Energy Transfer (LET) α-particles.
There are a few other studies which used Monte Carlo simulations, but which produced highly divergent results on the potentially achievable boron proton capture dose based on using different proton energies and 11 B concentrations 20,25,26 . These later simulations confirm that the number of induced reactions on boron-11 in proton therapy is negligible. For example, according to the results of Tayebeh A et al. the total dose of alpha particles generated from boron proton capture for 70-85 MeV proton beam is piddling, even in the situation that the entire tumor region was considered full of boron-11 20 . Also, according to simulations of Mazzone et al. at 80 ppm concentration of 11 B and 45-60 MeV proton energies, the effect of the p + 11 B → 3α reaction is negligible and the dose related to this reaction several orders of magnitude lower than the dose delivered by the primary proton beam 25 . In fact, both results indicate that the reaction probability is too low to bear any effect in the therapy.
On the basis of these considerations, it seems highly unlikely that the reported decrease in cellular survival in the presence of BSH 17,18 is related with the boron proton capture. Another explanation for the described radiosensitizing effect of boron-11 may be the biochemical properties of the boron-containing compound. For example, according to our previous results the radiosensitizing effect of boron-11 on the viability of the cancer cells subjected to proton irradiation was detected in vitro using two human malignant glioma cell lines, and sodium tetraborate (Na 2 B 4 O 7 ) as a boron containing substance 21 . But the fact that we observed a similar radiosensitizing effect of sodium tetraborate on A172 cells when they were subjected to gamma-radiation indicated that the decrease in the survival rate of cells irradiated after pre-incubation with the boron containing compound is likely due to its pharmacological and chemical properties rather than an increase in the absorbed dose due to 11 B + p → 3α nuclear reaction 21 . In this work we have also tested the possibility of increasing the biological effects of gamma radiation by the presence of sodium borocaptate and observed no effect (Suppl. Fig. S5, Suppl. Tables S1-S3).
In this study, we examined the radiosensitivity of two glioma cell lines by proton beam irradiation in the presence of sodium borocaptate. Our results clearly show that BSH, being non-toxic to the normal morphology DF-2 cells and to the glioma cells at the concentrations used, produces little, if any, enhancement of the www.nature.com/scientificreports/ effect of proton radiation on the glioma cell lines in vitro. Since the additional dose enhancement generated by boron proton capture depends on the proton energy and intracellular 11 B concentration, we have tested various variants of incubation of glioma cells with BSH, and also carried out a series of irradiation on two cyclotrons with different energies of the proton beam. Fitting parameters and dose-modifying factor for 20% survival for colony forming assays of glioma cells with irradiation alone or after incubation with sodium borocaptate are presented in Table 2. Despite the fact that the dose-modifying factor in some experiments was greater than one, in all experiments the radiosensitizing effect of sodium borocaptate was not statistically significant and did not depend on the time of preincubation with BSH (7 or 18 h), concentration of boron-11 (80 or 160 ppm), or on the initial energy of the proton beam (89.7 or 200 MeV). Given the published data 12,15,22 , it is possible that the apparent lack of radiosensitizing effect of BSH upon glioma cell irradiation with protons could be explained by the ineffective uptake of this compound by the studied glioblastoma cells. Testing this hypothesis, we repeated all experiments using the Du145 prostate cancer cell line, for which an increase in the biological efficiency of proton irradiation in the presence of sodium borocaptate was demonstrated 17,18 . In these two experimental studies the authors obtained a first pre-clinical demonstration of PBCT at the relatively low-energy clinical proton beamline (62 MeV) and at the high-energy proton beamline (250 MeV) reporting a significant reduction in the colony-forming ability of prostate cancer Du145 cells irradiated in the presence of the boron carrier BSH 17,18 . In both experiments, the authors preincubated cells for 6-8 h with 80 ppm of boron-11 and irradiated cells at various depths along the two clinical SOBPs, observing the radiosensitizing effect of BSH at the mid-and distal SOBP positions 17,18 .
In this study, we used the previously reported experimental conditions 17,18 : we pre-incubated Du145 prostate cancer cells with BSH (boron-11 concentration 80 ppm) for 7 h; irradiated with a 200 MeV proton beam, positioning the cells at the middle SOBP position; irradiated with 89.7 MeV clinical proton beam, positioning the cells at the distal end of SOBP. The effectiveness of the combined exposure was analyzed using both the AlamarBlue cell viability assay and the Clonogenic assay. Fitting parameters and dose-modifying factor for 20% survival for colony forming assays of Du145 cells with irradiation alone or after incubation with sodium borocaptate under various experimental conditions are summarized in Table 2. In all those experiments, proton irradiation in the presence of BSH did not appear to reduce Du145 cell viability or clonogenic survival to the degree comparable to that reported in the works 17,18 . Moreover, neither an increase in the concentration of boron-11 in the culture medium (160 or 250 ppm), nor an increase in the incubation time with sodium borocaptate (18 h) or the positioning of the cells at the most distal end of the clinical SOBP did not increase the sensitizing effect of BSH.
Based on the published experimental data and the results of calculations and simulations 17,18,20,21,25,26 , our results appear to indicate that the mechanisms of the radiosensitizing effects of boron during proton irradiation involves a complex process that is not limited to the dose enhancement by the alpha-particles produced in 11 B + p → 3α nuclear reaction. As already indicated, the additional dose produced by this reaction is not sufficient to bring to any substantial cell damage at any reasonable BSH concentrations 25 . Moreover, considering a relative sparseness of the nuclear reaction events that follows from the simulation and the short range of the alpha particles released, these particles appear to directly affect only a small fraction of the irradiated cells. Our Table 2. Parameters for the linear-quadratic fit of clonogenic survival data Y(X) = exp (-αX-βX 2 ), and dosemodifying factor for 20% survival (DMF 20 ) of Du145, A172, and Gl-Tr cells with proton irradiation alone (Control) or after incubation with boron-11 (BSH). www.nature.com/scientificreports/ results therefore seem to support the hypothesis that radiosensitizing effect of BSH could be to the large extent attributed to the indirect effects of the alpha irradiation 18 , such as radiation induced bystander effects [27][28][29] . These effects could largely depend on the state and treatment conditions of the cell culture or irradiated tissue, calling for further studies to establish the applicability of the proton boron capture therapy technique.

Materials and methods
Reagents. The  Irradiation. Cell irradiation was performed in two ways. For the first way, two days before irradiation A172, Gl-Tr, and Du145 cells were seeded in T25 flasks at 5 × 10 5 cells/flask and were cultured for 24 h. Then the cell growth medium was aspirated and replaced with 5 ml of fresh medium (control) or with BSH-containing medium. The working concentration of 80 ppm of 11 Supplementary  Fig. S1A. Cell samples were positioned in the phantom near the middle of SOBP and irradiated with increasing doses of 0.5-6 Gy. After irradiation the cells were seeded at 1 × 10 4 cells/well in a 24 well tissue culture plate and incubated at normal conditions during 48 h. The irradiation experiments were carried out in triplicate. For the second way, two days before irradiation A172, Gl-Tr, and Du145 cells were plated at 4 × 10 4 cells/well in a six well tissue culture plate, and were cultured for 24 h. Then the cell growth medium was aspirated and replaced with 3 ml of fresh medium (control) or with BSH-containing medium. In most experiments, we used a working concentration of BSH of 80 ppm, but in a number of experiments, concentrations of BSH of 160 ppm and 250 ppm were also used. The working concentrations of 80 ppm, 160 ppm and 250 ppm of 11 B corresponded to 0.17 mg/ml, 0.34 mg/ml, and 0.53 mg/ml of BSH, respectively. After 18 or 7 h of incubation in the presence of 11 B, the cells were irradiated in 6-well plates in a medium containing boron-11 with a clinical proton beam at Bragg peak in a dose range of 2-6 Gray. During irradiation, the plates with cells were positioned horizontally in front of the proton beam. Proton irradiation was performed on a Varian ProBeam version 3.5 at the MIBS proton therapy center ( www.nature.com/scientificreports/ beam in layers, with the beam proton energy from 89.735 MeV in the first layer to 110.735 MeV in the eighth layer. Spot weight ranged from 0 to 119.98 MU. Irradiation plans were created using Eclipce software, version 13.6. The dose profile is shown in Supplementary Fig. S1B. Water-equivalent plates 30 mm thick were placed above the samples to shift the Bragg peak to the cell layer. In most experiments, the cells were positioned at a depth of 30 mm, and in several-at a depth of 32 mm (Fig. 7A). One hour after irradiation, the culture medium in the plates was replaced with complete culture medium and the cells were incubated under standard conditions for 48 h. All irradiations with a clinical proton beam were carried out in triplicate.
To test whether the physical nuclear reaction 11 B(p,3α) affects the cancer cell death by proton beam irradiation, cell lines were also exposure to 80 ppm boron-11 and gamma radiation at different absorbed dose. After 18 h of incubation in the presence of 11 B, the cells were trypsinized, washed with medium, resuspended in complete culture medium without BSH, and counted by LUNA-II™ Automated Cell Counter (Logos Biosystems, South Korea). Then each cell suspension was split into five portions (0.1 ml each), transferred into sterile Eppendorf tubes at 10 5 cells/tube, irradiated with graded doses 0-8 Gy using 60 Co γ-ray source "Issledovatel" (NRC «Kurchatov Institute»-PNPI, Gatchina, Russia).
Cell viability and colony formation assays after irradiation. For cell viability assay irradiated and control cells were passaged at a concentration of 1 × 10 3 cells/well in 96 well tissue culture plate in five replicates of each experimental point. To determine the survival rate, the AlamarBlue® Assay was applied 6 days after irradiation. AlamarBlue Cell Viability Reagent was added to the plate according to the manufacturer's protocol and incubated for two hours. The fluorescence was detected using EnSpire Multimode Plate Reader (PerkinElmer, USA).
Clonogenic assay is the method of choice to determine cell reproductive death after treatment with ionizing radiation. After irradiation, the cells were seeded at 1 × 10 3 cells/well in a six well tissue culture plate, to form colonies in 1-2 weeks. Cells were incubated in standard conditions until cells in control dishes have formed sufficiently large clones. Then colonies were fixed with 96% ethanol, stained with crystal violet (0.5% w/v), washed with water, dried in air, photographed and counted by free ImageJ software.
Statistical analysis. Raphical visualization and statistical analysis were performed using SPSS 22.0 and GraphPad Prism 9.4.0. Dose-response curves were constructed fitting the cell viability values to the linearquadratic equation Y(X) = exp (-αX-βX 2 ) by GraphPad Prism 9.4.0 (nonlinear regression with the parameter β constrained to non-negative values), and α/β parameters were used to calculate Dose-Modifying Factor at 20% survival level (DMF 20 ). Experimental data are expressed as the mean ± the standard deviation (SD). To assess differences between groups (with or without BSH), the Mann-Whitney test was used. One sample t-test was used to evaluate difference of DMF 20 values from 1. Level of significance was set at p < 0.05.

Data availability
All data generated or analyzed during this study are included in this published article and its supplementary information files.