Down-regulation of the Sodium/Iodide Symporter Explains ¹³¹I-Induced Thyroid Stunning

Radioiodine therapy with ¹³¹I has two main indications: toxic diffuse goiter (i.e., Graves' disease) and differentiated thyroid cancer. Successful outcome of treatment relies on the ability of the thyroid tissue to concentrate and retain ¹³¹I⁻. In both normal and pathologic thyroid tissues, this is primarily achieved by the action of the sodium/iodide symporter (NIS) present at the thyroid cell surface (1). Loss of NIS expression leads to insufficient radioiodine uptake, which is the main reason why ¹³¹I therapy is not applicable to patients with poorly differentiated thyroid cancer.

Diagnostic scanning with low amounts of ¹³¹I is often used to estimate the mass of thyroid remnants after tumor surgery. Numerous studies have reported that diagnostic exposure to ¹³¹I may diminish subsequent uptake of the therapeutic dose, a phenomenon known as thyroid stunning, which potentially may compromise the therapeutic efficiency and outcome of radioiodine treatment (reviewed in refs. 2–5). The stunning mechanism has for long been obscure mainly because of lack of experimental data. However, recent studies in our laboratory, using an in vitro model of internal radiation of normal thyroid cells, show that ¹³¹I indeed reduces thyroidal iodide transport in a dose-dependent manner (6, 7) at absorbed doses equivalent to those estimated to induce stunning in vivo (3). Importantly, as quiescent (G₀) cells were irradiated and the cell number did not change, it could be excluded that the inhibited transport was related to compromised cell cycle progression or apoptosis otherwise readily induced when similar amounts of ¹³¹I were administered to proliferating thyrocytes (6, 7). Considering NIS as the key molecule responsible for the iodide-concentrating mechanism in the thyroid gland and in thyroid carcinoma cells with preserved iodide uptake, it can be hypothesized that the expression or function of NIS might be affected by ionizing radiation and that this could explain thyroid stunning. However, this possibility has not been experimentally investigated.

The NIS expression in thyroid cells is transcriptionally regulated by both positive and negative signals. Most importantly, thyroid-stimulating hormone (TSH) up-regulates the NIS mRNA level multifold via activation of a cyclic AMP (cAMP)-responsible element in the NIS promoter (8, 9). This favorable effect is likely responsible for the fundamental increase in tumor radioiodine uptake in thyroid cancer patients given recombinant human TSH before radioiodine therapy (10). Conversely, down-regulation of NIS mRNA expression can be experimentally induced by excess iodine (11, 12), which requires organification (i.e., iodide-protein binding) and thus explains the so-called escape of the Wolff-Chaikoff effect, a thyroid protective mechanism against toxicity of iodine overload. In this study, we investigated with real-time reverse transcription-PCR (RT-PCR) analysis whether ¹³¹I irradiation influences the NIS transcript levels in primary cultured normal thyrocytes stimulated with TSH and insulin-like growth factor-I (IGF-I).

Cell culture and hormone treatment. Pig thyroid glands obtained from the local abattoir were prepared for isolation and culture of thyroid epithelial cells as described (13). Briefly, purified thyroid follicle segments were plated on collagen type 1–coated filters in Transwell bicameral chamber inserts (3413 or 93413, Corning Costar) at 37°C in a CO₂ (5%) incubator. The cells were grown in Earle's MEM with 5% FCS, penicillin (200 units/mL), streptomycin (200 μg/mL), and fungizone (2.5 μg/mL; all reagents were purchased from PAA Laboratories GmbH) until forming a continuous monolayer resembling the native thyroid follicular epithelium after ∼1 week. In the culture chamber, the basal and apical compartments facing either side of the epithelium correspond to the extrafollicular space and the follicular lumen, respectively, which makes the working model highly suitable for studies of transepithelial iodide transport normally taking place in vivo. Epithelial barrier function indicating tight junction integrity was assessed before experiments by monitoring the transepithelial electrical resistance with a Millicell-ERS ohmmeter (Millipore Corp.).

Confluent cells were stimulated alone or in combination with TSH (1 milliunit/mL; Sigma) and insulin-like growth factor 1 (IGF-I; 10 ng/mL; Roche Diagnostics GmbH). TSH and IGF-I were added to the medium at the start of ¹³¹I exposure, and hormonal stimulation (renewed every 2–3 days along with medium exchange) continued during the entire experiment (i.e., also after irradiation was stopped) as detailed below. As previously shown (6, 7), the majority of cells in the confluent state were growth arrested due to contact inhibition; hence, the cells were not responsive to cell cycle effects of irradiation. However, as IGF-I (but not TSH) is mitogenic also to confluent thyroid cells (14), we measured the total DNA content of individual filter cultures by a fluorimetric assay (15) to correlate data to cell number changes likely appearing in response to IGF-I stimulation with and without irradiation.

Irradiation with ¹³¹I. Both the apical (200 μL) and basal (400 μL) culture compartments were replaced with medium containing the same activity concentrations of ¹³¹I (Nycomed, Amersham). Calculations to obtain a mean absorbed dose of 7.5 Gy after irradiation for 48 h were carried out using Monte Carlo simulation as described (7), taking into account the fact that ¹³¹I⁻ gradually accumulated in the apical medium due to ongoing iodide transport during ¹³¹I exposure. The absorbed dose of 7.5 Gy was chosen to mimic a high diagnostic dose assumed to be received in thyroid remnants after whole-body scanning with ¹³¹I. Methimazole (1 mmol/L; Sigma) was present in the medium to prevent radioiodide-protein binding that might extend ¹³¹I exposure after removal of the free activity and thereby increase the absorbed dose in cells subjected to further culture. Culture plates were shielded with lead blocks to prevent external irradiation of control groups. Irradiation was stopped after 48 h by removing all ¹³¹I by repeated washings with activity-free medium.

Measurement of ¹²⁵I⁻ transport. Transepithelial ¹²⁵I⁻ transport was assessed 2, 5, or 7 days after the start of ¹³¹I irradiation. Triplicate cultures in each treated group were simultaneously replaced with basal medium (400 μL) containing 150 kBq/mL ¹²⁵I⁻ (Nycomed, Amersham; appreciated final iodide concentration, 10⁻⁹ mol/L) and apical medium (200 μL) that was free of radiotracer. After 30 min, 50 or 100 μL were taken from the apical and basal medium and the activities were measured in a gamma counter (Wallac 1480 WIZARD 3″, Wallac Oy). The relative amount of ¹²⁵I⁻ being transported to the apical compartment was calculated after correction for detector background.

Real-time RT-PCR analysis of NIS mRNA expression. Total RNA was extracted from cultures at the indicated time points with the Qiagen RNeasy Micro kit (Qiagen GmbH), according to the manufacturer's instructions, and quantified by UV spectrophotometry at 260 nm. For each sample, cDNA was synthesized from 0.5 μg RNA using random hexamers and Taqman reverse transcription reagents (Applied Biosystems) according to the manufacturer's instructions. Oligonucleotide primers for the porcine NIS (pNIS) gene and for the reference gene (18S) were designed using the Primer Express software (Applied Biosystems) and purchased from TAG Copenhagen A/S. Primer sequences were as follows: pNIS, 5′-ctctcctggcagggcatatct-3′ (forward) and 5′-gctgagggtgccgctgta-3′ (reverse); 18S, 5′-gtaacccgttgaaccccatt-3′ (forward) and 5′-ccatccaatcggtagtagcg-3′ (reverse). The relative quantitative PCRs were done by using the QuantiTect SYBR Green PCR kit (Promega) and the ABI PRISM 7900HT Sequence Detection System (Applied Biosystems). Thermal cycling conditions consisted of an initial cycle of 2 min at 50°C, a cycle of 15 min at 95°C, and 40 cycles of 15 s of denaturation (94°C) followed by 1 min of annealing and extension (60°C). All amplification reactions were done in triplicates. The threshold cycle (C_t) values were used for calculation of the relative expression ratios between control and treated cells using the formula described by Pfaffl (16).

Statistical analysis. In total, two to six separate experiments with triplicates in each were carried out for each kind of assay and time point. The Student's t test or analysis of confidence intervals was used to determine the statistical significance (P ≤ 0.05) of differences between data obtained in measurements of total DNA and ¹²⁵I⁻ transport. The statistical significance of real-time RT-PCR data (P ≤ 0.001) obtained in a single analysis was calculated by using the Relative Expression Software Tool (REST; ref. 17). The Student's t test (P ≤ 0.05) was used for comparing log₂ values pooled together from several identical PCR experiments. Results presented in graphs show mean ± SE values.

¹³¹I irradiation inhibits TSH-stimulated NIS mRNA expression and iodide transport. Before ¹³¹I irradiation experiments were carried out, it was necessary to characterize the magnitude and kinetics over time of NIS mRNA expression in the applied thyroid cell culture model. Primers against defined sequences of the pNIS gene were found to give reliable quantitative data on the transcript level as analyzed with real-time RT-PCR. As shown in Fig. 1A, 1 milliunit/mL TSH increased the NIS mRNA expression ∼130-fold (log₂ 7.04) after 48 h. The stimulatory effect sustained after prolonged exposure to TSH for up to 7 days, although there was a significant reduction after the peak value at 2 days. This pattern correlated well with changes in the transepithelial transport of iodide from the basal to the apical medium (i.e., TSH enhanced the ¹²⁵I⁻ transport at all time points, although most prominently 2 days after stimulation was started; data not shown). As previously reported (13, 18), no transport of ¹²⁵I⁻ was measured in the opposite apical-to-basal direction (data not shown), illustrating the fact that NIS is solely expressed at the basolateral surface of fully polarized thyroid epithelial cells.

Next, cultures were stimulated with TSH and simultaneously exposed to ¹³¹I for 48 h to obtain the absorbed dose of 7.5 Gy, after which NIS mRNA was measured either directly or after culture for additional 3 days in activity-free medium. This showed that irradiation significantly reduced the NIS transcript levels measured at both time points (Fig. 1B). In average, ¹³¹I inhibited the NIS expression by 60%, but in two of five included experiments, the reduction was ∼80%. Importantly, down-regulation of NIS was accompanied by a corresponding decrease of the iodide transport capacity in the irradiated cultures (Fig. 1C).

IGF-I–stimulated NIS transcription is also radiosensitive to ¹³¹I. NIS is up-regulated by TSH mainly via a cAMP-dependent mechanism (19). It was therefore of interest to investigate whether NIS gene transcription as stimulated by the activation of another intracellular signaling pathway also might be radiosensitive. Earlier studies from our laboratory have shown that the transepithelial transport of iodide in pig thyrocytes is stimulated by IGF-I, although this occurs with a delayed onset in comparison with the effect of TSH (14). We confirmed this in the present study (data not shown) and found in addition that 10 ng/mL IGF-I increased the NIS mRNA levels at the corresponding investigated time points (i.e., 5–7 days after IGF-I was added; Fig. 2A). The magnitude of enhanced NIS expression was lower than that monitored after TSH stimulation, and a significant stimulatory effect of IGF-I on NIS mRNA was not always measured. Nevertheless, in experiments showing a clear response (two of six), the IGF-I–induced transcription of NIS was nearly completely inhibited by the simultaneous exposure to ¹³¹I (Fig. 2B). Likewise, irradiation inhibited the IGF-I–stimulated iodide transport recorded in parallel cultures (Fig. 2C).

According to previous studies (14), IGF-I is mitogenic to confluent thyroid cells, which thus escape from contact inhibition, enter the cell cycle, and start to proliferate, resulting in a gradual conversion of the cell monolayer into a multilayered epithelium. However, as the NIS mRNA quantification was normalized to 18S in all samples, it is conceivable to assume that the observed effects of IGF-I and ¹³¹I were not biased by altered cell numbers. To further corroborate this, we measured the total DNA content, which was found to be increased by 25% after stimulation with 10 ng/mL IGF-I, a submaximal mitogenic dose, after 5 days (Fig. 2D). Although simultaneous irradiation with¹³¹I (7.5 Gy) expectedly blocked the IGF-I–stimulated DNA synthesis (Fig. 2D), the magnitude of altered NIS expression and iodide transport was much higher than can be explained by the relatively small changes in cell number.

IGF-I and TSH synergistically counteract the loss of NIS expression and iodide transport in ¹³¹I-irradiated cells. As TSH and IGF-I stimulate the expression of NIS by different modes of action, it was of interest to study the combined effect and determine whether NIS in costimulated cells were radiosensitive to ¹³¹I to the same extent as in single-stimulated cells. Incubation with TSH and IGF-I for 5 days increased the NIS transcript levels in average 3-fold (log₂ 1.48) in comparison with TSH alone (Fig. 3A). This correlated with a corresponding acceleration of iodide transport (Fig. 3B), indicating that the two stimuli acted synergistically. ¹³¹I irradiation at an absorbed dose of 7.5 Gy reduced both NIS transcription and iodide transport in the costimulated cells (Fig. 3A and B). However, the inhibitory effect was unexpectedly small and in fact both the expression and function of NIS remained at higher levels than those measured in unirradiated TSH-stimulated cells. Thus, TSH and IGF-I cooperatively counteracted the down-regulation of NIS appearing at a given absorbed dose of ¹³¹I.

Thyroid stunning reflecting a reduced radioiodine uptake in thyroid remnants after whole-body diagnostic scanning with ¹³¹I has been an elusive phenomenon ever because it was first recognized more than 50 years ago (20). Although most experts in the field agree that it is a real side effect of the diagnostic use of ¹³¹I that needs to be considered in radioiodine therapy (2–5), the significance and even the existence of stunning have been questioned (21). Recently, we showed in cultured thyroid cells that ¹³¹I irradiation indeed inhibits iodide transport at absorbed doses equivalent to those estimated to be received in thyroid or tumor tissues after diagnostic administration of the same isotope (6, 7). In this article, we provide the first evidence of a molecular mechanism that may explain radiation-induced thyroid stunning (i.e., suppressed NIS gene transcription).

The experimental model was adopted to mimic in vitro a situation of internal radiation of the thyroid with ¹³¹I. For this purpose, primary cultured thyroid epithelial cells were continuously exposed to ¹³¹I and thereby allowed to take up and concentrate the isotope in the apical compartment (corresponding to the follicular lumen in vivo) for a period of 48 h. The received absorbed dose was calculated to be 7.5 Gy, which is comparable with a large diagnostic scanning dose clinically. Importantly, the irradiated cells were confluent and thus growth arrested already at start of the experiment, and no significant changes in cell number were noted in cultures in which TSH-stimulated NIS expression was markedly inhibited. In addition, NIS transcriptional activity was compared with 18S rRNA, which remained stable after irradiation, and the total RNA pool did not differ between irradiated and nonirradiated samples. Together, this indicates that stunning is not caused by a therapeutic effect due to compromised cell cycle machinery or loss of radiation-damaged cells by apoptosis, which would have been necessary to consider if the cells were proliferating. The distinction is not trivial, as stunning has been suggested to result from thyroid or tumor tissue destruction appearing early after administration of the ablation dose of ¹³¹I (22). In further support of stunning as a cell cycle–independent effect of ¹³¹I irradiation, we recently found that the dose response of growth arrest and inhibited iodide transport in proliferating and nonproliferating cells, respectively, are not entirely overlapping, the latter being more radiosensitive (7). Nevertheless, stunning might still be a consequence of nuclear events related to nonlethal DNA damage from ionizing radiation. Preliminary observations in our laboratory show that also nonproliferating G₀ thyrocytes display definitive signs of DNA double-strand breaks [i.e., activation and recruitment of the histone variant H2AX to nuclear foci of DNA repair (23) after ¹³¹I exposure using the same experimental design as in this study].³

IGF-I was found to partly counteract the ¹³¹I-induced decrease of the NIS mRNA and iodide transport levels in TSH-stimulated thyroid cells, although the weak stimulatory effect of IGF-I given alone was fully blocked by the same radiation dose. This strongly suggests that radiation affects NIS by interfering with its transcriptional regulation and argues against the unlikely scenario that the NIS gene would be, other than infrequently, damaged by a direct hit of randomly distributed ionizing events in the nucleus. Due to the complexity of NIS gene regulation, involving both protein kinase A–dependent and protein kinase A–independent signaling pathways (24, 25), there are several possible candidate mechanisms. Of particular interest may be the recently reported cooperative regulation of the NIS promoter activity by the thyroid transcription factor Pax8 and a redox factor, Ref-1 (26). Ref-1, synonymously called Ape1, is an endonuclease with multifunctional properties that can be triggered by genotoxic stress and participates in DNA repair (27). Speculatively, activated Ref-1/Ape1 might be recruited to sites of DNA damage whereby the transcriptional enhancer of NIS expression might suffer from relative lack of a necessary cofactor. Another possible mechanism leading to stunning might be that signals involved in physiologic or pathophysiologic down-regulation of NIS (e.g., as induced by growth factors) are activated by irradiation (28). In fact, both ligand-dependent and ligand-independent activation of growth factor receptor signaling pathways supposed to participate in the radioprotection response ultimately controlling cell survival are commonly observed in irradiated cells (29).

IGF-I alone stimulated NIS expression and iodide transport, although much less efficiently than TSH. Also contrasting the action of TSH, there was a delay of several days until any increased NIS transcript levels could be measured, suggesting that IGF-I probably regulates NIS indirectly. Notably, in combination, IGF-I and TSH stimulated NIS synergistically. In fact, in several experiments, IGF-I had a stimulating effect only in the presence of TSH. Therefore, it is likely that IGF-I enforced TSH-stimulated NIS expression by a permissive mode of action similar to that previously recognized in various experimental thyroid models (30). This might be conducted, for instance, by up-regulating the expression of thyroid transcription factors of importance to maintain the differentiated state of the thyroid cell (31, 32). However, it should be emphasized that IGF-I can also modulate the radiation response that in turn likely influences cell behavior. Most strikingly, IGF-I receptor activation has been found to confer radioresistance (33), and conversely, antagonism of IGF-I receptor signaling increases the radiosensitivity (34–36) of tumor cells. IGF-I seems to be one of several players linking the early DNA damage response mediated by the ataxia-telangiectasia mutated kinase, which in fact stimulates the IGF-I receptor expression (37, 38), to efficient DNA repair (39). The hypothesis that IGF-I might support, conjointly with TSH, the expression of NIS in irradiated thyroid cells by accelerating DNA repair and thereby restore the transcriptional machinery is an intriguing possibility that warrants further experimental studies.

In conclusion, our findings indicate that ¹³¹I irradiation down-regulates the expression of NIS at the transcription level, leading to inhibited iodide transport in primary cultured thyroid cells. This effect most likely explains the phenomenon thyroid stunning observed after pretherapeutic diagnostic examination of ¹³¹I uptake in patients with thyroid cancer. The ability of IGF-I and TSH to conjointly accelerate the NIS expression and thereby antagonize the ¹³¹I-induced loss of NIS suggests that thyroid stunning might be treatable. However, in view of the possibility that IGF-I regulates NIS in a species-specific manner [i.e., IGF-I suppresses NIS transcription in a rat thyroid cell line (40)], it will be important to extend these studies to human thyroid cells and also carcinoma cells before a possible clinical protocol can be discussed. Likewise, referring to the currently debated application of ¹³¹I in future treatment modalities of mammary adenocarcinomas endogenously overexpressing NIS (41) and of other cancers subjected to NIS gene transfer (42), it is pertinent to elucidate whether stunning involving down-regulation of NIS is thyroid specific or comprise a general response to radiation independently of cell type.

Grant support: Swedish Research Council grant 537 and Swedish Cancer Foundation grants 3427 and 4567.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank the Department of Oncology and the Division of Medical Physics and Medical Engineering at the Sahlgrenska University Hospital (Göteborg, Sweden) for help in supplying ¹³¹I.