Non-invasive imaging of radiocesium dynamics in a living animal using a positron-emitting 127Cs tracer

Visualizing the dynamics of cesium (Cs) is desirable to understand the impact of radiocesium when accidentally ingested or inhaled by humans. However, visualization of radiocesium in vivo is currently limited to plants. Herein, we describe a method for the production and purification of 127Cs and its use in visualizing Cs dynamics in a living animal. The positron-emitting nuclide 127Cs was produced using the 127I (α, 4n) 127Cs reaction, which was induced by irradiation of sodium iodide with a 4He2+ beam from a cyclotron. We excluded sodium ions by using a material that specifically adsorbs Cs as a purification column and successfully eluted 127Cs by flowing a solution of ammonium sulfate into the column. We injected the purified 127Cs tracer solution into living rats and the dynamics of Cs were visualized using positron emission tomography; the distributional images showed the same tendency as the results of previous studies using disruptive methods. Thus, this method is useful for the non-invasive investigation of radiocesium in a living animal.

Visualizing the dynamics of cesium (cs) is desirable to understand the impact of radiocesium when accidentally ingested or inhaled by humans. However, visualization of radiocesium in vivo is currently limited to plants. Herein, we describe a method for the production and purification of 127 cs and its use in visualizing cs dynamics in a living animal. the positron-emitting nuclide 127 cs was produced using the 127 I (α, 4n) 127 cs reaction, which was induced by irradiation of sodium iodide with a 4 He 2+ beam from a cyclotron. We excluded sodium ions by using a material that specifically adsorbs Cs as a purification column and successfully eluted 127 Cs by flowing a solution of ammonium sulfate into the column. We injected the purified 127 cs tracer solution into living rats and the dynamics of cs were visualized using positron emission tomography; the distributional images showed the same tendency as the results of previous studies using disruptive methods. thus, this method is useful for the noninvasive investigation of radiocesium in a living animal.
The accident at Tokyo Electric Power Company's Fukushima Daiichi Nuclear Power Station in March 2011 caused various radioactive materials to fall over a wide area 1 . In particular, 137 Cs, which has a half-life of 30 years, had a serious impact on the affected area 2,3 . Subsequently, interest in the dynamics of radiocesium in animals, plants, and humans has increased. There are concerns about where in the body radiocesium ingested by humans accumulates, and how quickly it is excreted.
To more accurately calculate the internal exposure dose of radiocesium, it is necessary to analyze the dynamics of radiocesium in animals in more detail. The current calculation of internal exposure doses of radiocesium is based on data from animal studies, and human urine and feces. More than 90% of ingested soluble cesium (Cs) is absorbed in the gastrointestinal tract 4 , but the real-time behavior and kinetics immediately after ingestion remain unclear. For example, spherical Cs-bearing particles (Cs-balls) were emitted in the Fukushima nuclear accident 5 , and the dynamics of such Cs-balls in the human body are unknown.
Visualizing the dynamics of radiocesium in vivo is a powerful method for studying the dynamics of radiocesium. In the field of plant science, there have been reports of visualizing the dynamics of radiocesium in living plants. Kobayashi et al. have succeeded in visualizing the β rays of 137 Cs using a real-time radioisotope imaging system 6 . Kurita et al. have succeeded in visualizing the dynamics of 137 Cs by imaging Cherenkov light emitted from β rays with an ultra-sensitive CCD camera 7 . Although these methods are suitable for the visualization of radiocesium in plants with a small tissue thickness, they cannot visualize the dynamics of radiocesium in animals, which have a large tissue thickness. Kawachi  Positron imaging is widely used to non-destructively visualize the dynamics of positron-emitting radionuclides in vivo. The most common use of positron imaging is cancer screening by positron emission tomography (PET), which exploits the tendency of fluorodeoxyglucose labeled with the positron-emitting radionuclide 18 F to accumulate in cancer cells. Additionally, PET has been used in medical research to analyze the kinetics of drugs labeled with 11 C, 13 N, and 15 O. As for the positron-emitting nuclide of Cs, McKee et al. reported that 132 Cs (half-life of 6.48 d, β + intensity of 0.43%) was produced and administered to animals to verify the applicability of 132 Cs as a tumor marker 9 . However, because 132 Cs was produced by irradiating stable Cs, 132 Cs could not be purified, resulting in low specific activity. Therefore, a 132 Cs tracer is not suitable for non-invasive imaging of radiocesium in vivo where the tracer must exhibit high specific activity.
In this study, the positron-emitting nuclide 127 Cs (half-life of 6.25 h, β + intensity of 3.04%) without stable Cs was produced by irradiating iodine with a helium ion beam accelerated by a cyclotron. Then, we developed a unique purification method to adsorb and elute only 127 Cs using a material that specifically collects Cs. Additionally, by administering the purified 127 Cs tracer to living animals, we verified that 127 Cs is useful for visualization of the short-term dynamics of radiocesium.

Results and discussion
production of 127 cs. 127 Cs was produced by the 127 I (α, 4n) 127 Cs reaction. A sodium iodide (NaI) target plate was irradiated with a 4 He 2+ beam from an AVF cyclotron at Takasaki Ion Accelerators for Advanced Radiation Application (TIARA). The incident energy on the target was set to 55 MeV with reference to the crosssection from the experimental data of Sathik et al. 10 and data from the ACSELAM library 11 ( Supplementary  Fig. 1). Under these irradiation conditions, 129 Cs was also produced by the 127 I (α, 2n) 129 Cs reaction. Supplemental Table 1 lists nuclides that were potentially generated. Figure 1A shows the gamma spectrum of the mother solution, in which the irradiated NaI target was dissolved. Under the irradiation conditions in this study, the production yields of 127 Cs and 129 Cs at the end of irradiation were 80.1 ± 6.5 and 1.3 ± 0.04 MBq/µAh, respectively (n = 6). Additionally, 24 Na, 124 I, and 126 I were also produced as by-products, which could be used effectively as radioactive tracers for sodium and iodide ions from the NaI target in the subsequent purification experiments. The decay data from the radionuclides produced in this study are summarized in Supplementaary Table 2. In the gamma-ray spectrum, the other peaks could not Figure 1. Gamma spectrum of the mother and purified solutions. (A) Gamma-ray spectrum of the mother solution containing a hundredth (10 µL) of the total volume (1 mL), which was measured for 20 min at 21.5 cm from the detector, 47 min after the end of irradiation. (B) Gamma-ray spectrum of the purified solution containing a thousandth (10 μL) of the total volume (10 mL), which was measured for 20 min at 3.5 cm from the detector, 124 min after the end of irradiation. Purification of 127 cs tracer. The procedure for the separation of 127 Cs from the irradiated NaI target is shown schematically in Fig. 2. Table 1 shows the radioactivity of the nuclides in each purification step when the irradiation condition was 0.5 µA × 2 h. First, we loaded the mother solution on to an anion-exchange column. The results showed that 124 I and 126 I were not detected in the effluent solution from the anion-exchange column, which indicated that iodide ions derived from the NaI target were removed. Next, to separate 127 Cs from sodium ions, we used fibrous Cs adsorbent, which was synthesized by radiationinduced graft polymerization to introduce ammonium 12-molybdophosphate (AMP) as an adsorption ligand onto a polyethylene trunk fabric. This material has been reported to adsorb Cs specifically 12 . We considered that this adsorbent is effective in purifying trace amounts of 127 Cs under the extreme condition in this study, where the Na + / 127 Cs ratio was 0.8 mmol/4.3 pmol (80 MBq), a 10 8 -fold molar number difference. For efficient adsorption of 127 Cs on the Cs adsorbent column, we reduced the concentration of the sodium ions in the effluent solution by increasing the volume of the solution from 5 to 100 mL. We then passed the solution through the Cs adsorbent column. The results showed that 98% ± 2% (n = 6) of 127 Cs was adsorbed on the Cs adsorbent packed into the column (Table 1). After washing the column with 100 mL of H 2 O, we finally passed 10 mL of 1 mM (NH 4 ) 2 SO 4 through the Cs adsorbent column. The results showed that 55% ± 4% (n = 6) of 127 Cs was eluted from the adsorbent but 24 Na was not detected in the eluted solution (Table 1), which indicated that sodium ions derived from the NaI target were removed. Figure 1B shows the gamma spectrum of the eluted solution, which also indicated that the peaks of 24 Na and 126 I had disappeared in the final solution. Supplementary Fig. 2 shows the gamma-ray spectrum of the mother solution and the final 127 Cs tracer solution measured for 3 h. Because the   127 Cs tracer solution at the end of irradiation are less than 5.1 and 3.8 kBq, respectively. Therefore, we calculated that even at the three half-lives of 127 Cs (18.75 h), the ratios of the positrons from 124 and 126 I to those from 127 Cs are maximally less than 0.68% and 0.03%, respectively. Because this study was performed by the no-carrier-added production method, the specific activity of 127 Cs was calculated as 1.5 × 10 17 Bq/g, which is much higher than the 1.0 × 10 6 Bq/g (28 µCi/g) calculated for the previous 132 Cs report 9 . We simultaneously visualized the movement of the positron-emitting nuclides in the purification step of the Cs adsorbent column using a planar positron imaging system (PPIS). Figure 3A shows the process of 127 Cs adsorption on the column. The effluent solution of the anion-exchange column diluted to 100 mL was flowed from the left bottle to the right bottle via the Cs adsorbent column in the direction from the bottom to the top over 30 min. We observed that the signal at the Cs adsorbent column position gradually increased. However, signals were also seen in the right bottle, even though the effluent solution from the Cs adsorbent column contained little radioactivity from 127 Cs ( Table 1). The half-life of the signal of the right bottle was measured with the planar positron imaging system and was estimated to be 109.9 min (Supplementary Fig. 3). It has been reported that 18 F is produced by the 23 Na (α, n + 2α) 18 F reaction when sodium is irradiated with a 4 He 2+ beam 13 . The calculated cross-section of 18 F at 55 MeV was 21.7 m barn (Supplementary Table 1) and the radioactivity of 18 F quantified from the data in Supplementary Fig. 3 was 2.2 MBq. Thus, we believe that the effluent solution from the Cs adsorbent column may contain 18 F, which has a half-life of 109.8 min. The estimated value (109.9 min) of the half-life of the signal of the right bottle was slightly longer than the half-life of 18 F and this was assumed to be because a small amount of 127 Cs was included in the effluent solution from the Cs adsorbent column (Table 1). Because the affinity for fluorine of the anion exchange column used in this study is significantly lower than the affinity for iodine, it is possible that 18 F could pass through the column. Figure 3B shows the process of 127 Cs elution from the Cs adsorbent column. First, to minimize re-adsorption of 127 Cs to the Cs adsorbent, the column was turned upside down. Then, 10 mL of 1 mM (NH 4 ) 2 SO 4 was flowed from the left vial to the right vial via the column in the direction from the bottom to the top over 30 min. In the last 10 min, the column was turned upside down and the solution in the column was pushed out with air.
To verify the contamination of 18 F into the final 127 Cs tracer solution, we calculated the decay time of the peak counts of annihilation gamma ray (511 keV) with the data obtained by measuring the final solution every 2 h with the Ge detector (Fig. 4). By one-component fitting, we calculated that the half-life of the peak count was 6.279 ± 0.015 h, which is almost consistent with the half-life of 127 Cs. Furthermore, using two-component fitting ( 127 Cs and 18 F) of this decay curve, the percentage of the 18 F component in this peak count at the start time was calculated as 0.663 ± 1.273%. These results indicate that the contamination of 18 F into the final 127 Cs tracer solution was extremely low. Additionally, the estimate indicates that the radionuclidic purity of 127 Cs in the 36 h after the end of irradiation was more than 99%.
The principle of Cs adsorption and elution from the Cs adsorbent can be described as follows: Cs is adsorbed by a substitution reaction with ammonia from the AMP in the column structure, and then the adsorbed Cs is replaced by ammonia from contact with the ammonium solution. Figure 5 shows the elution rates of Cs from the adsorbent with various concentrations of different ammonium compounds in a batch experiment using non-radioactive Cs. We selected an eluate solution using two criteria to make the final tracer solution suitable for physiological study for both animal and tracer experiments in general biology, including plant experiments: the ion concentration should be as low as possible, and the pH should be close to neutral. Also, it was desired that the elution rate be at least 50% or more for efficient purification. When the ion concentration was limited to 1 mM, (NH 4 ) 2 CO 3 and (NH 4 ) 2 SO 4 were identified as candidates. The elution rate of 1 mM (NH 4 ) 2 CO 3 was higher than that of (NH 4 ) 2 SO 4 , but the pH values were 9.2 and 5.7, www.nature.com/scientificreports/ respectively. Therefore, we used a solution of 1 mM (NH 4 ) 2 SO 4 , for eluting 127 Cs from the Cs absorbent column in the purification experiment.
To verify the desorption of AMP particles from the adsorbent, the molybdenum concentration of the solution eluted with 1 mM ammonium sulfate was measured and the desorption rate was calculated as 0.0086%. This indicates that the possibility of contamination of AMP particle into the final tracer solution is quite low when this adsorbent is used for purification. pet imaging. The purified 127 Cs tracer solution was injected into the tail vein of a living rat, and the dynamics of Cs were visualized using PET, in duplicate. The doses of 127 Cs for Experiments 1 and 2 were 2 and 11 MBq, and the imaging times were 3 and 4 h, respectively. To make the dose as low as reasonably achievable, we administered 2 MBq of 127 Cs in Experiment 1. To obtain more clear PET images, we administered 11 MBq of 127 Cs in Experiment 2, which we consider to be the more desirable dose. Figure 6A shows a 4-h integrated image of 127 Cs in the whole body of the rat in Experiment 2. This result indicated that 127 Cs passed or accumulated in the neck, heart, small intestine, and kidney. Figure 6B shows a cross-sectional view of the integrated image of 127 Cs in the neck. From comparison with the CT image, the signal for 127 Cs was found to be localized in the salivary glands in the neck. The localization pattern of the 127 Cs signal was the same in Experiment 1 (Supplementary Fig. 4). Figure 7 shows the time course of 127 Cs PET counts in the whole body and in each organ. All counts were decay corrected at the start time of PET imaging and normalized to the dose of 127 Cs and the volume of each region. Although there were differences in the normalized PET counts in the two experiments, the same trend was observed over each time course. The time course of the whole body indicated that the 127 Cs tracer injected into the tail vein moved into the field of view of the PET in 30 min. The accumulation of 127 Cs in the kidney reached a peak at 20-30 min and was rapidly excreted. In the salivary glands and heart, the accumulation of 127 Cs reached a peak at 30-60 min and was slowly excreted. In contrast, although the molecular number of   www.nature.com/scientificreports/ 127 Cs administered in Experiments 1 and 2 differed five-fold, the kinetics of 127 Cs exhibited the same trend and no saturation was observed even with the five-fold difference in molecular number, which reflects no region of specific affinity of 127 Cs in rats. Moreover, our method can produce no-carrier-added 127 Cs and no influence of injected radioactivity is expected. Kaikkonen et al. have measured the radioactivity using a gamma counter in different tissues of goats 30 min after administrating 134 Cs intravenously 14 . The results showed that the highest relative concentration of 134 Cs was found in the kidney cortex, followed by the salivary gland, cardiac muscle, and small intestine. Sasaki et al. have reported that cattle fed with 134 Cs and 137 Cs for several weeks had more radiocesium accumulation in the kidney than in other organs 15 . Although the tested animal species and methods of radiocesium administration are different, these studies are consistent with the results of our study, which suggests that 127 Cs can be used to correctly visualize Cs dynamics in the living animal body.
To the best of our knowledge, this is the first study to visualize the dynamics of radiocesium in living animals. In this study, we administered 127 Cs to animals by intravenous injection. In the future, it should be possible to non-invasively calculate the transfer rate of radiocesium to each organ by administering 127 Cs through ingestion or inhalation. The calculated results will contribute to the construction of more accurate models of Cs dynamics and improvements in the accuracy of effective dose coefficients. Furthermore, because of the short half-life of 127 Cs (6.25 h), repeated experiments using the same animal are possible. For example, the effects of various types of internal decontamination agents could be evaluated accurately.
Methods production of 127 cs. 127 Cs was produced by the 127 I (α, 4n) 127 Cs reaction. A target plate with a diameter of 10 mm and a thickness of 0.5 mm was made by pressing 125 mg of NaI reagent (Fujifilm Wako Pure Chemicals, Tokyo, Japan). The NaI target plate was irradiated with a 4 He 2+ beam for 2-3.25 h, at a beam current of 0.5 µA and an incident energy of 55 MeV, delivered from an AVF cyclotron at TIARA.
Column purification of 127 cs tracer. After the irradiation, the NaI target was dissolved in 1 mL of H 2 O.
To remove iodide ions, the solution was loaded onto an anion-exchange column (Poly-Prep #731-6211; Bio-Rad, Hercules, CA, USA) and an additional 4 mL of H 2 O was loaded.
To separate 127 Cs from the sodium ions, we used a fibrous Cs adsorbent (KURANGRAFT-Cs, Kurashiki Textile Manufacturing Co. Ltd., Kurashiki, Japan), which comprised fibrous polyethylene trunk material crosslinked with a Cs adsorption ligand, ammonium 12-molybdophosphate (AMP). Supplementary Fig. 5 shows the structure of the Cs adsorbent column. The fibrous Cs adsorbent was cut out into 0.7-cm-diameter disks, and 40 disks were packed into a glass chromatography column (Econo-Column Chromatography Columns #737-0707, Bio-Rad) with quartz wool. The volume of Cs adsorbent in the column was 1.0 cm 3 (0.7-cm diameter × 2.5-cm height). This glass chromatography column was used as the Cs adsorbent column. The effluent solution of the anion-exchange column was diluted to 100 mL with H 2 O and passed through the Cs adsorbent column at a flow rate of approximately 4 mL/min using a peristaltic pump (Model: 13-876-2; Thermo Fisher Scientific, Waltham, MA, USA). The Cs adsorbent column was washed by passing 100 mL of H 2 O at the same flow rate, and then 127 Cs was eluted by passing 10 mL of 1 mM (NH 4 ) 2 SO 4 (Fujifilm Wako Pure Chemicals) through the column at a flow rate of approximately 0.4 mL/min using a peristaltic pump (Model: 13-876-1; Thermo Fisher Scientific). The purification process using the Cs adsorbent column was imaged using a planar positron imaging system (PPIS-4800; Hamamatsu Photonics, Hamamatsu, Japan).
Decay fitting analysis. The half-life of the radioactivity in the effluent solution ( Supplementary Fig. 3) and the eluted solution (Fig. 4) from the Cs adsorbent column was estimated by fitting the data with the following one-component formula: where t and N represent the measurement time and the count, respectively. The start time of the measurement and the count at the start time are represented by t 0 and A , respectively. The half-life is represented by τ . Two parameters, A and τ , were estimated by fitting the data ( t and N ) with the formula with fixing t 0 at the experimental value of the start time.
The effect of the 18 F contamination in the eluted solution from the Cs adsorbent was also evaluated, by fitting the data with the following two-component formula: where A Cs and A F and τ Cs and τ F represent the counts at the start time and the half-lives for Cs and F, respectively. Two parameters, A Cs and A F , were estimated by fitting the data with the formula by fixing τ Cs and τ F at 6.25 h and 109.8 min, respectively, and then fixing the start time t 0 at 0. www.nature.com/scientificreports/ cs elution rate analysis. The Cs-adsorbed adsorbents were prepared as follows. The fibrous Cs adsorbent was cut into 0.5-cm-diameter disk (8.3 mg). The adsorbent disk was placed in 10 mL of stable Cs solution with a concentration of 7.5 nM (1 ppb). After shaking for 4 h at 25 °C, the Cs adsorbent disk was removed from the Cs solution and the concentration of the remaining Cs was measured using inductively coupled plasma-mass spectrometry (7700 × ICP-MS, Agilent Technologies, Santa Clara, CA, USA). The calculated adsorption rate of Cs to the adsorbent was 84% ± 7% (n = 40).
The elution was carried out with aqueous solutions of (NH 4 ) 3 C 6 H 5 O 7 , NH 4 HCO 3 , (NH 4 ) 2 CO 3 , (NH 4 ) 2 SO 4 , NH 4 HCO 2 , NH 4 NO 3 , CH 3 COONH 4 , and NH 4 Cl for the batch test. The concentrations of the ammonium compound solutions were adjusted to 0.5, 1.0, 2.0, 5.0, and 10 mM. The adsorbent disk was soaked in 5 mL of aqueous solutions of each ammonium compound, and was continuously shaken for 20 h at 25 °C. After elution, the concentration of Cs in the ammonium solutions was measured using ICP-MS. The elution rate of Cs was calculated by the following equation: where C A and C E are the amount of adsorbed Cs in the adsorbent and eluted Cs in the ammonium solutions, respectively. pet imaging. PET imaging experiments using two living rats were performed (Experiments 1 and 2).
The eluted solution containing 127 Cs was evaporated to dryness in a glass vial and transported from TIARA to Tohoku University. Approximately 300 µL of physiological saline solution (154 mM NaCl) was added to the glass vial, and the 127 Cs was recovered. The final concentration of (NH 4 ) 2 SO 4 was 33 mM. A tracer solution containing 2 or 11 MBq of 127 Cs was injected to a 6-week-old male Wistar rats (132 or 136 g) via tail vein under inhalation anesthesia with 2% isoflurane in air. The distribution of 127 Cs in the living rat was imaged using a small-animal PET instrument (Clairvivo PET 16 ; Shimadzu, Japan) for 3 or 4 h. The tracer was injected 15.3 or 14.2 h after the end of irradiation. A list-mode DRAMA algorithm 17 was used for image reconstruction with an iteration of 1 and sub-iterations of 100. The PET image had a matrix size of 128 × 128 × 213 and each voxel size was 0.7 mm 3 . No attenuation correction was performed, but we did conduct scatter correction using emission scatter convolution subtraction 18 . 18 F solution was used for scanner calibration. The energy window for the scanner was set to 250-750 keV. A CT scan of the rat was also performed using a small-animal CT instrument (Clairvivo CT; Shimadzu) before the injection of the 127 Cs tracer solution. Quantitative analysis of the PET images was performed with Amide software 19 . ethics approval. All experimental procedures conformed to "Regulations for Animal Experiments and Related Activities at Tohoku University", and were reviewed by the Institutional Laboratory Animal Care and Use Committee of Tohoku University, and finally approved by the President of University. Ethical approval in this study has been obtained from the Institutional Laboratory Animal Care and Use Committee of Tohoku University.