The present study was approved by the Committee of Medical Ethics of the Hirosaki University Graduate School of Health Sciences (approval no. 2021–050; Hirosaki, Japan). To protect the rights and privacy of the participants, written informed consent was obtained after providing a detailed verbal explanation, and all collected data were anonymized and handled in accordance with ethical guidelines. The present study included 3 patients (2 female patients and 1 male patient; median age, 58 years; age range, 45–81 years) with DTC who received ¹³¹I internal therapy with rhTSH (Thyrogen^®; Sanofi S.A.) between April 2022 and December 2024 at Aomori Rosai Hospital (Hachinohe, Japan) (Table I). The inclusion criteria were as follows: Age ≥18 years; histologically confirmed differentiated thyroid carcinoma; completion of total thyroidectomy; planned adjuvant ¹³¹I therapy with rhTSH stimulation; and provision of informed consent. The exclusion criteria were: Presence of medullary or anaplastic thyroid carcinoma; uncontrolled severe comorbidities (such as renal failure or severe cardiac disease); prior radioiodine therapy within 12 months; and inability to comply with the study procedures. During treatment, clinical information and blood samples were collected. Routine blood tests, including complete blood counts (red blood cells, white blood cells and platelets) and thyroid-related parameters [TSH, free triiodothyronine (FT3), free thyroxine (FT4), Tg and Tg/Ab], were performed at the clinical laboratory of Aomori Rosai Hospital at three timepoints: Before ¹³¹I administration (baseline), on the day of ¹³¹I administration (day 0) and 30 days after treatment (day 30), consistent with the measurements shown in Fig. 1.

Peripheral blood was collected from patients with DTC using serum separation tubes (BD Biosciences) at two time points: Before the administration of ¹³¹I (day 0) and 30 days after administration (day 30), both under Thyrogen stimulation. To minimize the influence of recent dietary intake, patients were instructed to fast for ≥6 h prior to blood collection. The collected sample tubes were stored in a deep freezer (−80°C) until analysis. Serum metabolomic analysis was performed using liquid chromatography-mass spectrometry (LC-MS) and flow injection analysis-mass spectrometry (FIA-MS), and the AbsoluteIDQ P180 kit (Biocrates, Inc.) was used to quantify metabolites. The analysis was performed using the method described in our previous study (18). The target metabolites were phenyl isothiocyanate-derivatized and analyzed based on internal standards for quantitation. LC-MS and FIA-MS were performed in positive ion mode using electrospray ionization using a high-performance liquid chromatography system (ExionLC^™ AD; SCIEX) combined with a QTRAP 6500+ triple quadruple ion trap hybrid mass spectrometer system (SCIEX) operated with Analyst^® 1.6.3 software (SCIEX). LC-MS was used to measure amino acids and biogenic amines, and the multiple reaction monitoring (MRM) conditions are shown in Table SI. The MS measurement conditions were as follows: Curtain gas, 45 psi; collision gas, 6 psi; ion spray voltage, 5,500 V; temperature, 500°C; ion source gas 1, 40 psi; and ion source gas 2, 50 psi. LC separation was performed on a Zorbax Eclipse XDB-C18 column (3×100 mm; 3.5 µm; Agilent Technologies, Inc.) at 40°C. The mobile phases consisted of solvent A (water with 0.2% formic acid) and solvent B (acetonitrile with 0.2% formic acid). The flow rate was set at 0.5 ml/min. The gradient program was as follows: 0–0.5 min, 0% B; 0.5–5.5 min, linear increase to 95% B; 5.5–6.5 min, held at 95% B; 6.5–7 min, returned to 0% B and equilibrated for the next injection for 1.5 min. FIA-MS measurements included carnitines and acylcarnitines, hydroxy- and dicarboxyacylcarnitines, sphingomyelins and hydroxysphingomyelins, diacyl phosphatidylcholine, acyl-alkyl phosphatidylcholine, lysophosphatidylcholine, and sugar as targets. The MRM conditions for measuring these metabolites are shown in Table SII. The MS settings were the same as those for LC-MS, except that the temperature was 175°C. The FIA-MS flow program was as follows: 0–1.6 min, 0.03 ml/min; 1.6–2.4 min, linear increase to 0.2 ml/min; 2.4–2.8 min, held at 0.2 ml/min; and 2.8–3.0 min, returned to 0.03 ml/min. The injection volume was 20 µl. As with LC-MS, FIA-MS quantification was performed using isotopically labeled internal standards provided in the AbsoluteIDQ P180 kit to ensure accuracy and reproducibility. All metabolites were identified and quantified using isotopically labeled internal standards and multiple reaction monitoring as optimized and provided by Biocrates, Inc. Table II shows the measurable metabolites and their abbreviations. Amino acids and biogenic amines were quantified using LC-MS, whereas acylcarnitines, phosphatidylcholines, lysophosphatidylcholines, sphingomyelins and hexose were analyzed by FIA-MS. Among these, Fig. 2 highlights representative metabolites from each category that showed statistically significant or consistent changes before and after ¹³¹I administration. The metabolomics datasets of LC-MS and FIA-MS generated in the present study have been deposited in MetaboBank (https://www.ddbj.nig.ac.jp/metabobank/index-e.html) under the accession numbers MTBKS257 and MTBKS258, respectively. For pathway enrichment analysis, metabolites that showed significant changes before and after ¹³¹I administration were analyzed using MetaboAnalyst 6.0 (https://www.metaboanalyst.ca/) with pathway annotation based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (https://www.genome.jp/kegg/).

The TPC-1 human thyroid papillary carcinoma cell line was obtained from Merck KGaA. The cells were cultured in RPMI-1640 (Nacalai Tesque, Inc.) containing 10% heat-inactivated fetal bovine serum (Biowest), 2 mM L-glutamine (MilliporeSigma) and 1% penicillin/streptomycin in a humidified atmosphere at 37°C with 5% CO₂.

TPC-1 cells were irradiated using an X-ray generator (MBR-1520R-3; Hitachi Power Solutions Co., Ltd.) at 1 Gy/min (150 kVp; 20 mA) with 0.5-mm aluminum and 0.3-mm copper filters. The radiation dose was monitored using a thimble ionization chamber. Handling of ¹³¹I in vitro requires specialized radiation safety measures and licensing, which can impede exploratory studies. Therefore, X-rays were chosen as a practical and controllable radiation source to model general cellular responses to IR independently of sodium/iodide symporter (NIS)-mediated uptake. Based on clonogenic survival assays, an 8 Gy dose was selected for irradiation to ensure ~1% survival of TPC-1 cells. This dose was chosen to induce measurable biological effects relevant to therapeutic radiation exposure while maintaining sufficient cell viability for downstream analyses. Although clinical radioiodine therapy doses vary depending on remnant size and uptake, the 8 Gy dose provides a reproducible and biologically meaningful model to study cellular radiation responses. Clinical absorbed doses to thyroid remnants and metastases are typically much higher (commonly cited targets, ~300 Gy for remnants and ~80 Gy for metastases) (19,20). Cell viability was assessed using a trypan blue exclusion assay. Briefly, 20 µl cell suspension was mixed with 20 µl trypan blue solution (0.4%; 2-fold dilution; Nacalai Tesque, Inc.) at room temperature. The cells were gently mixed and incubated for <3 min to allow dye penetration into nonviable cells. The viable (unstained) and nonviable (blue-stained) cells were then counted using a Bürker-Türk hemocytometer (Erma, Inc.) under a light microscope (IX71; Olympus Corporation) without fixation.

Colony-forming cells were counted using a clonogenic potency assay. The cells were plated at a density of 500–8,000 cells/dish in a 35-mm diameter culture dish, irradiated with 0 to 8 Gy of IR and incubated for 6–9 days in a humidified atmosphere at 37°C with 5% CO₂. The irradiation was delivered at a dose rate of 1 Gy/min, and the exposure time was adjusted according to the prescribed dose. Subsequently, the dishes were fixed with −20°C cold methanol for 15 min and stained with Giemsa solution (Muto Pure Chemicals Co., Ltd.) at room temperature for 30 min. Colonies containing >50 cells were manually counted under an inverted microscope (IX71; Olympus Corporation). The cell survival rate was calculated relative to the nonirradiated controls and plating efficiency, as described previously (21).

The cell cycle distribution was analyzed using flow cytometry. The harvested cells were treated with precooled (−20°C) 70% ethanol for 10 min on ice and stored at −20°C until measured. RNase I was added to the sample tube at 37°C for 20 min (5 µg/ml; Merck KGaA) to remove internal RNAs. The cells were then stained with PI (40 µg/ml; FUJIFILM Wako Pure Chemical Corporation) in the dark at room temperature for 2 min. Cell cycle distribution analysis was performed using Cell Lab Quanta^™ Sc MPL (Beckman Coulter, Inc.) and analyzed with Kaluza software (version 2.1; Beckman Coulter, Inc.).

Amino acid uptake into cells was analyzed using a boronophenylalanine (BPA) uptake assay kit (cat. no. 342-09893; Dojindo Laboratories, Inc.). BPA is an amino acid analog. The cells exposed to IR were plated in a 96-well microplate and incubated for 12 h at 37°C in a 5% CO₂ incubator. Reagents were added according to the manufacturer's instructions. The fluorescence of the samples (Ex₃₅₀/Em₄₅₅) was first measured using a microplate reader (TriStar LB 941; Berthold Technologies GmbH & Co. KG). After this, the incubated cells were harvested and analyzed by flow cytometry (Aria SORP; BD Biosciences) using FACS Diva^™ software (ver.6.0; BD Biosciences) to assess BPA uptake at the single-cell level. This assay utilizes a cell-permeable fluorescent probe that binds to intracellular amino acid analogs. Upon uptake of BPA via amino acid transporters such as solute carrier family 7 member 5 [L-type amino acid transporter type 1 (LAT1)], the probe emits fluorescence (λ_ex, 360 nm; λ_em, 460 nm), allowing quantification of transporter activity in viable individual cells.

The expression levels of the CD98 cell surface antigen were analyzed using double staining with PI and CD98-FITC. The harvested cells were suspended in Hanks' balanced salt solution at 4°C. CD98-FITC (5 µl per 1×10⁶ cells in 100 µl staining volume; cat. no. 315603; BioLegend, Inc.) was added to the sample tubes and these were incubated for 20 min on ice (~0°C) in the dark. After washing with Hanks' balanced salt solution, the samples were stained with PI (40 µg/ml; FUJIFILM Wako Pure Chemical Corporation) for 3 min at room temperature. Cellular fluorescence was measured using Cell Lab Quanta^™ Sc MPL (Beckman Coulter, Inc.) and analyzed with Kaluza software (version 2.1; Beckman Coulter, Inc.).

Total RNA was extracted from the cells using the RNeasy^® Plus Mini Kit (Qiagen, Inc.). The concentration and quality were confirmed using the NanoDrop system (Thermo Fisher Scientific, Inc.). cDNAs were synthesized to analyze mRNA expression using ReverTra Ace qPCR RT Master Mix (Toyobo Co., Ltd.) according to the manufacturer's instructions. The primer sequences for LAT1, solute carrier family 7 member 8 [L-type amino acid transporter type 2 (LAT2)] and solute carrier family 3 member 2 (CD98hc) were designed based on the gene sequences obtained from the National Center for Biotechnology Information database (https://www.ncbi.nlm.nih.gov/gene/). The mRNA expression levels were evaluated using RT-qPCR with the Power SYBR^™ Green PCR Master Mix (Thermo Fisher Scientific, Inc.) and SmartCycler^® II (Takara Bio, Inc.). The thermocycling conditions were: 95°C for 10 min, followed by 40 cycles of 95°C for 15 sec and 60°C for 1 min. The relative expression of each target mRNA was determined using the 2^−ΔΔCq method (22). The accession numbers were as follows: LAT1, NM_003486.7; LAT2, NM_001267036.1; CD98hc, NM_001012662.3; and β-2-microglobulin (B2M), NM_004048.4. The oligonucleotide primer sets used for RT-qPCR were designed using Primer3 software (version 4.1.0) (23) and supplied by Eurofins Genomics (Table III). B2M was selected as the housekeeping gene for normalization based on the Minimum Information for Publication of Quantitative Real-Time PCR Experiments guidelines (24).

Statistical analysis was performed using R software (version 4.4.0; http://www.r-project.org/). Cell cycle analysis was conducted using Kaluza software (version 2.1; Beckman Coulter, Inc.) to identify sub-G₁, G₀/G₁, S and G₂/M phases. The clonogenic survival fraction (F) was fitted to the equation F(D)=exp (−αD-βD²), where D is the dose in Gy and α and β are fitting coefficients, using the Levenberg-Marquardt algorithm. For small sample sizes (n=3), the Wilcoxon signed-rank test was used for paired comparisons involving two timepoints (such as metabolomic data before vs. after ¹³¹I administration). For TSH levels measured at three timepoints (before rhTSH, day 0 and day 30), the Friedman test was applied to account for repeated measures across multiple timepoints. This test was chosen as a non-parametric alternative to repeated-measures ANOVA suitable for small sample sizes and non-normal distributions. For comparisons between two independent groups with normal distribution and equal variance, an unpaired Student's t-test was used. For data with unequal variance, Welch's t-test was applied. For comparisons among multiple groups with normal distribution and equal variance, one-way ANOVA was used, followed by the Tukey-Kramer post hoc test for pairwise comparisons. For comparisons of mRNA expression levels among multiple radiation dose groups, one-way ANOVA followed by the Tukey-Kramer post hoc test was used. Data are presented as the mean ± SD. The number of independent experiments ranged between 3 to 10. P<0.05 or false discovery rate (FDR) <0.05 was considered to indicate a statistically significant difference.

Of the 3 patients with DTC included in the present study, 2 were female and 1 was male. The patients were diagnosed with PTC (Table I). Additionally, the patients had no previous medical history of other thyroid diseases. Thyrogen was administered twice, 2 days and 1 day before ¹³¹I administration, and its accumulation in the residual tumor was observed using scintillation scanning during treatment. Antitumor effects were observed in all patients 6 months after the initial prescription of Thyrogen, as evaluated based on the medical records of the patients (data not shown). No significant hemocytopenia was observed in terms of the numbers of red blood cells, white blood cells and platelets within 30 days after ¹³¹I therapy (Fig. 1A-C). Serum TSH levels increased in all patients, reaching a maximum of 105–455 µIU/ml, from before Thyrogen administration to just prior to ¹³¹I administration. By day 30 after administration, TSH levels had returned to the same range as before ¹³¹I administration (Fig. 1D). FT3 and FT4 before and after treatment remained at 2.6–3.6 pg/ml and 1.4–2.4×10⁻² ng/ml, respectively, which were confirmed to be normal values (Fig. 1E and F). Conversely, the levels of Tg/Ab and Tg in the serum showed individual variations between patients (patient 1: Tg/Ab, 10–10.6 IU/ml; Tg, 2.24 ng/ml; patient 2: Tg/Ab, 12.6–12.9 IU/ml; Tg, 23.5–35.1 ng/ml; patient 3: Tg/Ab, 10 IU/ml; Tg, 0.07–0.39 ng/ml; Fig. 1G and H).

Metabolomic analysis was performed using patient serum to explore serum metabolites specifically altered by ¹³¹I radiotherapy following Thyrogen stimulation for DTC. Of the 188 types of target metabolomic analysis data obtained from mass spectrometry, 168 were highly reliable. Among them, the levels of nine metabolites that changed before and after ¹³¹I administration were increased, and the levels of one metabolite were decreased (Fig. 2). Furthermore, enrichment analysis based on the KEGG database using MeatboAnalyst6.0 identified valine, leucine and isoleucine biosynthesis as a significant metabolic pathway (FDR, 0.0392; data not shown). Valine, leucine and isoleucine are essential amino acids referred to as branched-chain amino acids (BCAAs) and are involved in the nitrogen supply to cancer (25). In all cases, BCAA levels tended to increase after ¹³¹I administration compared with before administration (Fig. 2). However, statistical significance was observed only for valine, whereas the increases in leucine and isoleucine were not significant. Given the small sample size (n=3), statistical comparisons were performed using the Wilcoxon signed-rank test. The consistent upward trend across patients supports the robustness of this exploratory observation.

The TPC-1 cell line was used to examine whether the serum metabolic changes observed in patients with DTC could be recapitulated at the cellular level. Clonogenic potency and the cell cycle were analyzed to confirm the radiosensitivity of TPC-1 cells. The plating efficiency for TPC-1 cell colony formation was 13±1.2% (Fig. 3A), the parameters in the linear-quadratic model were α=0.32±0.05 Gy⁻¹, β=0.03±0.02 Gy⁻² and α/β=12.12±0.66, and the survival rate was <1% at 8 Gy (Fig. 3B). The cell cycle phase distribution analysis of cells exposed to 8 Gy IR showed a time-dependent increase in cells in the G₂/M phase for 12 h. A significant increase was observed at 6–24 h compared with the non-irradiated control (Fig. 3C and D). The number of viable cells after 12 h of exposure decreased to ~50% compared with that of the non-irradiated group (Fig. 3E).

Similar to BCAAs, BPA is taken up by cells via LAT1 (26,27). Changes in BPA uptake were evaluated to determine the BCAA uptake ability of irradiated TPC-1 cells. After 12 h of exposure to 8 Gy IR, intracellular BPA levels significantly decreased to ~80% of those in the non-irradiated control group, corresponding to an ~20% reduction (Fig. 4A). Conversely, single-cell analysis by flow cytometry revealed that the levels of intracellular BPA following exposure to IR were significantly upregulated ~1.58 times compared with the non-irradiated control (Fig. 4B and C).

LAT1 forms a complex with CD98hc on the cell membrane and triggers amino acid uptake (28). The flow cytometry analysis of TPC-1 cells exposed to 8 Gy IR revealed significant upregulation of the CD98 cell surface antigen compared with the non-irradiated control (Fig. 5). However, RT-qPCR analysis revealed that the expression levels of the related mRNAs LAT1, LAT2 and CD98hc were similar to those in the non-irradiated control group (Fig. 6).