Decreasing undesirable absorbed radiation to the intestine after administration of radium-223 dichloride for treatment of bone metastases

[223Ra]RaCl2 is the first alpha-particle emitting radiopharmaceutical to be used for castration-resistant prostate cancer patients with bone metastases because of its excellent therapeutic effects. [223Ra]RaCl2 is excreted via the intestine into feces, and some is absorbed from the intestine into the blood, which may be undesirable in terms of the exposure to radiation. Recently, we showed that a complex of myo-inositol-hexakisphosphate (InsP6) with zinc is a useful decorporation agent against radiostrontium. In this study, we hypothesized that Zn-InsP6 could bind to not only strontium but also to radium, and could inhibit the absorption of radium from the intestine. In in vitro binding experiments, Zn-InsP6 showed a high binding affinity for radium. In in vivo biodistribution experiments by intravenous injection of [223Ra]RaCl2 after treatment of Zn-InsP6, mice treated with Zn-InsP6 showed significantly lower bone accumulation of radioactivity (34.82 ± 1.83%Dose/g) than the mice in the non-treatment control group (40.30 ± 2.78%Dose/g) at 48 h postinjection. These results indicate that Zn-InsP6 bound radium in the intestine and inhibited the absorption of radium into the blood. Therefore, the insoluble Zn-InsP6 complex has high potential to decrease the side effects of [223Ra]RaCl2.

Scientific RepoRtS | (2020) 10:11917 | https://doi.org/10.1038/s41598-020-68846-x www.nature.com/scientificreports/ We have previously conducted studies with the aim to decrease the dose of absorbed radiation in the case of intake incident of radionuclides [8][9][10] . Chlorella, a genus of single-cell green algae that grows in fresh water, is a known health food composed of approximately 1-4% chlorophyll, 55-67% protein, 9-18% dietary fiber, and large amounts of minerals and vitamins 11 . It has been previously reported that Chlorella enhances the excretion of heavy metals [12][13][14][15] . Thus, we hypothesized that Chlorella could also enhance the elimination of radiocesium (Cs + ) and radiostrontium (Sr 2+ ), which are major radionuclides released into the atmosphere and ocean by nuclear power plant accidents, from the body in the case of intake accidents. The study indicated that Chlorella could inhibit the absorption of 90 Sr into the blood and enhance the elimination of 90 Sr from the body through adsorption in intestine 8 . Moreover, Myo-inositol-hexakisphosphate (phytic acid: InsP6, Fig. 1), a natural product found in abundance in plants, especially in whole grains, cereals, legumes, seeds, and nuts 16 , possesses high chelation potential with many kinds of metal cations due to its structure [17][18][19] . Because Zn-InsP6 is insoluble in water and has enough room to potentially accommodate additional radiocesium or radiostrontium coordinated by chelation, we prepared and evaluated a complex of InsP6 with zinc ions (Zn-InsP6) for use as decorporation agents for radiocesium and radiostrontium. The results demonstrated that Zn-InsP6 adsorbed radiocesium or radiostrontium in vitro. Although Zn-InsP6 did not affect radiocesium in vivo, Zn-InsP6 adsorbed radiostrontium in the gastrointestinal tract, inhibited its absorption from the intestine into the blood, and enhanced its excretion into the feces 9,10 . Accordingly, these studies indicated that Chlorella and Zn-InsP6 could work as decorporation agents for radiostrontium.
Strontium and radium are classified as alkaline earth metals and have some similarities in their physical and chemical properties. Thus, we hypothesized that Chlorella and Zn-InsP6 could bind not only strontium but also radium, and could inhibit the absorption of radium from intestine into blood. If Chlorella and Zn-InsP6 can inhibit the absorption of [ 223 Ra]Ra 2+ from the intestine, it is expected that the side effects of [ 223 Ra]RaCl 2 in the intestine will be decreased by pretreatment with Chlorella or Zn-InsP6. Therefore, in this study, we evaluated the effects of Chlorella and Zn-InsP6 to [ 223 Ra]RaCl 2 both in vitro and in vivo.

Results and discussion
Experiments of in vitro 223 Ra adsorption at pH 1.2 and pH 6.8 by Chlorella and Zn-InsP6. The binding ratios of 223 Ra to Chlorella and Zn-InsP6 are shown in Fig. 2. 223 Ra adsorbed to Chlorella in a quantitydependent manner and the absorption ratios of 223 Ra (e.g. 97.7% ± 0.0% for 30 mg Chlorella/mL) were much higher than those of 85 Sr (e.g. 35.9% ± 0.9% for 30 mg Chlorella/mL) under a neutral pH condition 8 . 223 Ra highly adsorbed to Zn-InsP6 even at low concentrations under a neutral pH condition. Moreover, 223 Ra only adsorbed to Chlorella and Zn-InsP6 at high concentrations under an acidic pH condition. In this study, the binding affinity for 223 Ra under an acidic condition that mimics gastric acid is not so important because 223 Ra does not pass through the stomach after intravenous injection of [ 223 Ra]RaCl 2 as a therapeutic radiopharmaceutical. However, it was necessary to evaluate whether compounds bind radionuclides under acidic conditions in the application for decorporation agents of radionuclides in the case of intake accidents. pH dependence of in vitro adsorption of 223 Ra to Chlorella or Zn-InsP6. The open circles in Fig. 3a, b show the pH dependence of 223 Ra adsorption to Chlorella and Zn-InsP6, respectively. The binding ratios of 223 Ra to Chlorella were low at low pH, and the binding ratios increased with increasing pH, and were high (> 90%) at pH between 6.6 and 12.2. Furthermore, the binding ratios of 223 Ra to Zn-InsP6 were higher at a wider range of pH compared to those to Chlorella.
The closed circles in Fig. 3a, b show the binding ratios of 223 Ra to Chlorella and Zn-InsP6 at adjusted neutral pH after the Chlorella and Zn-InsP6 samples were exposed to an acidic condition. The results indicated that even after exposure to an acidic solution, Chlorella and Zn-InsP6 still highly bound 223 Ra at neutral pH. Namely, Chlorella and Zn-InsP6 after passing through the acidic conditions in the stomach could bind 223 Ra in the intestine.
Effects of cations (Na + , K + , or Ca 2+ ) on in vitro adsorption of [ 223 Ra]Ra 2+ by Chlorella and Zn-InsP6. Figure 4a, b show how the binding of Ra 2+ to Chlorella and Zn-InsP6 is affected by other cations, Na + , K + , or Ca 2+ . The presence of Na + and K + slightly decreased the amount of Ra 2+ bound to Chlorella, while there was no decrease in the binding rate of Ra 2+ in the case of Zn-InsP6. In addition, the presence of Ca 2+ decreased the binding ratios of Ra 2+ to both Chlorella and Zn-IP6, and this effect was dependent on Ca 2+ concentration. As Ca 2+ and Ra 2+ are divalent cations and classified as alkaline earth metals, it is reasonable that the high concentration of Ca 2+ inhibits the binding of Ra 2+ .
Based on the abovementioned results of the in vitro experiments, in which Zn-InsP6 showed higher affinity for Ra 2+ and was expected to be more effective than Chlorella, we performed in vivo experiments using only Zn-InsP6 (and not Chlorella).   Table 2. After intravenous injection of [ 223 Ra]RaCl 2 , a large portion of the radioactivity immediately accumulated in the bone, while the remainder was delivered to the kidney and intestine. At 1 h postinjection, the radioactivity in all tissues was almost the same between the Zn-InsP6 treated group and the non-treatment control group. Thus, although the decrease in the bone accumulation following pretreatment with Zn-InsP6 was not so substantial, it was significant at 48 h postinjection. Moreover, the excreted radioactivity into the feces was significantly increased. These results are in line with the abovementioned biodistribution after oral administration of [ 223 Ra]RaCl 2 and indicate that Zn-InsP6 bound [ 223 Ra]Ra 2+ in the intestine and inhibited the absorption of [ 223 Ra]Ra 2+ from the intestine.
In a previous Phase I study for dosimetry of [ 223 Ra]RaCl 2 , the absorbed doses of the small intestinal wall, the upper large intestinal wall, and the lower large intestinal wall were set to 0, derived from alpha-particles 6 . It Ra though the intestine was not considered in the Phase I study. Actually, in a proposed compartmental model for 223 Ra by Taprogge et al., 223 Ra was excreted into faces without the reabsorption from the small intestine 21 . However, some 223 Ra in the intestine must be absorbed into the blood, and the alpha-particles emitted from 223 Ra on absorption should, in theory, harm the intestinal cells. Indeed, diarrhea has been reported in some patients of clinical studies of [ 223 Ra]RaCl 2 6,22-24 . Classically, the contamination by people painting watch dials with radium from the 1910′s to the 1920′s suggests the absorption of radium because the distribution of radium to the bone of the dial painters had been found 25 . Epidemiological studies of radium dial painters also found an association between exposure to high-dose radium and osteosarcoma 26 . ICRP Task Group on Alkaline Earth Metabolism in Adult Man indicated that the fraction of radium absorbed from food or drinking-water is 0.15-0.21 27 . The absorbed fraction of radium was set to 0.2 in some reports 28 . Moreover, in the above-mentioned Phase I study, Chittenden et al. reported that cumulative urine excretion of 223 Ra was 2% ± 2% of the injected activity 6 . In another Phase I study in Japan, Yoshida et al. reported that the cumulative urine excretion reached 2% up to 48 h postinjection 29 . These results showed that the urine excretion of 223 Ra in humans negligible. Meanwhile, some radioactivity in the kidney at 1 h postinjection and in the urine at 48 h postinjection were observed in this study using mice. The difference should be caused by species differences. Thus, the strategy of this study with Zn-InsP6 complex could be more useful to humans because the fecal excretion rate of 223 Ra in humans must be higher than that in mice. Therefore, the inhibition of the absorption by binding of the insoluble Zn-InsP6 complex with [ 223 Ra]Ra 2+ is considered to be useful in decreasing side effects.
We supposed that the toxicity of Zn-InsP6 is low since the Zn-InsP6 complex is not absorbed due to its insolubility in water, similarly to barium sulfate, which is used as a contrast agent for X-ray diagnosis. If a small part of Zn-InsP6 is decomposed and absorbed into the body, InsP6 and Zn are known to be low toxicity compounds. InsP6 also exists in mammals, and works as a coenzyme 30 . InsP6 is orally taken as an over-the-counter supplement, and brings several health benefits without toxicity, and its anticancer efficacy has also been reported 31 . Zinc is an essential mineral and is also used as an over-the-counter supplement that is recommended for individuals at a high risk of zinc deficiency, which is associated with a number of diseases 32 . Namely, because Zn-InsP6 is a combination of a well-known element and a compound that are both commonly taken as supplements, it is likely that Zn-InsP6 will be suitable for use in clinical research.    In vitro adsorption of 223 Ra to Chlorella and Zn-InsP6. Zn-InsP6 was prepared by the method previously reported with slight modification 9 . Zn-InsP6 (Zn : InsP6 = 2 : 1) was used for both in vitro and in vivo experiments. This in vitro adsorption experiment was performed according to our previous study 8 . Namely, Chlorella (10, 30, or 100 mg) or Zn-InsP6 (1, 3, 10, or 30 mg) was suspended and 223 Ra (925 Bq) was added in 1 mL of the first test solution (artificial gastric juice, pH 1.2) or the second test solution (artificial intestinal juice, pH 6.8) defined in the Japanese Pharmacopoeia. After shaking the suspension at 1,000 rpm at 37ºC for 1 h using a shaking incubator (SI-300C; AS ONE Corp., Osaka, Japan), the samples were centrifuged at 10,000 g at room temperature for 10 min. The radioactivity of the supernatant was measured using an auto-well gamma counter (ARC-7010B; Hitachi Ltd., Tokyo, Japan) and the counts were corrected for background radiation. A window from 50 to 300 keV was used for the counting. The measurement time of each sample was set to 1 min. Control experiments were performed using the same procedure but without Chlorella or Zn-InsP6. The binding ratios were determined as follows: Binding ratio to Chlorella or Zn-InsP6 (%) = [1 − (radioactivity of the supernatant of each sample) / (radioactivity of the supernatant of the respective control)] × 100. Table 2. Biodistribution of radioactivity in mice at 1, 24, and 48 h after intravenous administration of [ 223 Ra] RaCl 2 with pretreatment of 5% glucose (control) or Zn-InsP6. Data are expressed as % injected dose per gram tissue. Each value represents the mean (SD) for four or five animals. † Data are expressed as % injected dose. Significance was determined using an unpaired Student's t-test (*p < 0.05 vs. control). was also performed according to our previous study with slight modification 8 . Namely, Chlorella (30 mg) or Zn-InsP6 (10 mg) was suspended and 223 Ra (925 Bq) was added in 1 mL of 0.01 M HEPES buffer solution (pH 2-13) and shaking, centrifugation, and radioactivity measurements were performed as described above. The pH of each suspension was measured after shaking. Reversibility of the adsorption potential between Chlorella or Zn-InsP6 and 223 Ra with pH variation was evaluated. Chlorella (30 mg) or Zn-InsP6 (10 mg) was suspended in 1 mL of the first test solution (pH 1.2) defined in the Japanese Pharmacopoeia and then shaken at 1,000 rpm at 37ºC for 1 h. After centrifugation at 10,000 g at room temperature for 10 min, 800 μL of the supernatant was removed. Following this, 23, 25, or 27 µL of 1 M NaOH solution and 777, 775, or 773 µL of 0.01 M HEPES buffer solution (pH 8) were added to the Chlorella suspension. Meanwhile, 9, 11, or 13 µL of 1 M NaOH solution and 791, 789, or 787 µL of 0.01 M HEPES buffer solution (pH 8) were added to the InsP6 suspension. Thereafter, 223 Ra (925 Bq) was added to the Chlorella or InsP6 suspension, and the suspension was shaken at 1,000 rpm and 37ºC for 1 h. After centrifugation at 10,000 g at room temperature for 10 min, the radioactivity and pH of the supernatant were measured as described above.
Effects of cations (Na + , K + , or Ca 2+ ) on the in vitro adsorption of 223 Ra to Chlorella or Zn-InsP6. This experiment was performed according to our previous study with slight modification 10 . 223 Ra solutions (925 Bq/mL) in 20 mM HEPES buffer (pH 7.4) containing 10 mM Na + , 20 mM Na + , 50 mM Na + , 110 mM Na + (in the case of Zn-InsP6), 10 mM Na + + 10 mM K + , 10 mM Na + + 40 mM K + , 10 mM Na + + 100 mM K + (in the case of Zn-InsP6), 10 mM Na + + 10 mM Ca 2+ , 10 mM Na + + 40 mM Ca 2+ , and 10 mM Na + + 100 mM Ca 2+ (in the case of Zn-InsP6) were prepared by dissolution of NaCl, KCl, or CaCl 2 . Chlorella (10 mg) or Zn-InsP6 (10 mg) was suspended in 1 mL of each 223 Ra solution. Additionally, after shaking the suspensions at 1,000 rpm at 37ºC for 1 h, the binding ratio of each sample to Chlorella or Zn-InsP6 was determined using the methods described above.
Biodistribution experiments after oral administration of 223 Ra with pretreatment of Zn-IP6. Animal experiments were conducted in strict accordance with the Guidelines for the Care and Use of Laboratory Animals of Kanazawa University. The animal experimental protocols were approved by the Committee on Animal Experimentation of Kanazawa University. The animals were housed with free access to food and water at 23ºC with a 12-h alternating light/dark schedule unless otherwise specified.
6-week-old male ddY mice (Japan SLC Inc., Hamamatsu, Japan) were used for all animal experiments. In the Zn-InsP6 administration group, Zn-InsP6 suspension (30 mg/0.5 mL) in 5% glucose aqueous solution was orally administrated into mice. In the InsP6 administration group, InsP6 solution (20 mg/0.5 mL) in 5% glucose was orally administrated into mice. In the control group, 0.5 mL of 5% glucose was orally administrated into mice. Then, just after the administration of Zn-InsP6, InsP6, or 5% glucose, a saline solution of [ 223 Ra]RaCl 2 (9.25 kBq/100 µL) was orally administrated. Mice were sacrificed at 48 h post-administration of [ 223 Ra]RaCl 2 . The tissues of interest were removed and weighed, and radioactivity counts were determined. Fasting was carried out from 12 h pre-administration to 24 h post-administration to exclude the effects of diet.
Biodistribution experiments after intravenous administration of 223 Ra after pretreatment with Zn-IP6. In the Zn-InsP6 administration group, Zn-InsP6 suspension (30 mg/0.5 mL) in 5% glucose aqueous solution was orally administrated into 6-week-old male ddY mice. In the control group, 0.5 mL of 5% glucose was orally administrated. Then, 1 h after the administration of Zn-InsP6 suspension or 5% glucose, a saline solution of [ 223 Ra]RaCl 2 (9.25 kBq/100 µL) was injected via the tail vein. To determine the amount and routes of the radioactivity excreted from the body, mice were housed in metabolic cages (Metabolica, Sugiyama gene, Tokyo, Japan) for 48 h after administration. Mice were sacrificed at 1, 24, and 48 h post-administration of [ 223 Ra]RaCl 2 . The tissues of interest were removed and weighed, and radioactivity counts were determined with an auto well gamma counter as described above. Fasting was carried out from 12 h pre-administration to 24 h post-administration to exclude the effects of diet.
Statistical analysis. Significance in biodistribution experiments was determined by a one-way analysis of variance followed by Tukey's post hoc test or unpaired Student's t-test using Prism 8 (GraphPad Software Inc., San Diego, CA, USA). Results were considered statistically significant at p < 0.05.