Enhanced Iodide Transport After Transfer of the Human Sodium Iodide Symporter Gene Is Associated with Lack of Retention and Low Absorbed Dose

Home , Sodium perchlorate

Gene Therapy (2003) 10, 774–780 & 2003 Nature Publishing Group All rights reserved 0969-7128/03 $25.00 www.nature.com/gt RESEARCH ARTICLE Enhanced iodide transport after transfer of the human sodium iodide symporter gene is associated with lack of retention and low absorbed dose

U Haberkorn1,2, R Kinscherf3, M Kissel4,WKu¨ bler5, M Mahmut2, S Sieger2, M Eisenhut6, P Peschke4 and A Altmann2 1Department of Nuclear Medicine, University of Heidelberg, Heidelberg, Germany; 2Clinical Cooperation Unit Nuclear Medicine, DKFZ, Heidelberg, Germany; 3Department of Anatomy and Cell Biology III, University of Heidelberg, Heidelberg, Germany; 4Clinical Cooperation Unit Radiation Therapy, DKFZ, Heidelberg, Germany; 5Department of Radiation Protection, DKFZ, Heidelberg, Germany; and 6Department of Radiochemistry and Radiopharmacology, DKFZ, Heidelberg, Germany

Transfer of the sodium iodide symporter (hNIS) has been hNIS-expressing tumors lost 89, 89 and 91% of the initial proposed as a new principle of cancer gene therapy. Using activity, respectively. Dosimetric calculations showed that clinically relevant doses of 131I for the treatment of NIS- 1200 MBq/m2 resulted in 370.5 Gy (wild-type tumor expressing prostate carcinoma cells, we investigated the 0.1570.1 Gy) and 2400 MBq/m2 resulted in 3.170.9 Gy kinetics and the absorbed doses obtained in these tumors. (wild-type tumor 0.2670.02 Gy). Although transduction of hNIS-expressing cell lines accumulated up to 200 times more the hNIS gene induces iodide transport in rat prostate iodide when compared to wild-type cells. However, a rapid adenocarcinoma a rapid efflux occurs, which leads to a low efflux of the radioactivity (80%) occurred during the first absorbed dose in genetically modified tumors. With regard to 20 min after replacement of the medium. In rats, the hNIS- a therapeutic application additional conditions need to be expressing tumors accumulated up to 20 times more iodide defined leading to iodide trapping. when compared to contralateral transplanted wild-type Gene Therapy (2003) 10, 774–780. doi:10.1038/sj.gt.3301943 tumors. After 24 h and doses of 550, 1200 or 2400 MBq/m2

Keywords: human sodium iodide transporter; gene therapy; prostate carcinoma

Introduction human sodium-iodide symporter as a paradigm for the genetically modified uptake of 131I in a rat prostate Thyroid follicular cells accumulate iodide, which is used carcinoma model, and evaluated the iodide uptake for radioiodine therapy in benign thyroid diseases as kinetics and dosimetry after doses that are commonly well as for therapy of differentiated thyroid carcinoma. administered to patients. The first step in the complex process of iodide trapping in the thyroid is the active transport of iodide together with sodium ions into the cell, which is mediated by the Results sodium-iodide symporter. This process against an electrochemical gradient requires energy, is coupled to Immunocytochemistry the action of Na+/K+-ATPase and is also stimulated by Immunocytochemistry using two different clones of anti- TSH.1–5 Since the cloning of the human and rat cDNA NIS antibodies (BrA 6-1 and BrA 10-11) revealed a strong sequences several experimental studies have been red fluorescence (Cy3) on the cell surface of hNIS performed, which investigated the recombinant expres- expressing AT1 cells, whereas nontransfected cells sion of the hNIS gene in malignant tumors by viral (wild-type AT1) showed no NIS immunoreactivity transfer of the hNIS gene under the control of different neither using anti-NIS antibodies of the clone BrA 6-1 promoter elements.6–20 Although all of them reported nor of the clone BrA 10-11 (not shown). Omission of the high initial uptake in the genetically modified tumors, primary antibody abolished the immunoreactivity com- differing results have been obtained concerning the pletely. efficiency of radioiodine treatment based on NIS gene transfer with generally very high doses given to tumor- Na125I uptake and efflux in recombinant AT1 cell lines bearing mice. In this study, we used the gene for the With respect to the wild-type counterparts, up to 200- fold more iodide was transported into the recombinant prostate carcinoma cells (Figure 1). Given a median cell Correspondence: Dr U Haberkorn, Department of Nuclear Medicine, University of Heidelberg, Im Neuenheimer Feld 400, Germany Heidelberg volume of 1.3 fl as determined by Coulter counter an up 5 69120 to 10 -fold higher concentration of radioactive iodide in Received 28 August 2002; accepted 19 October 2002 the cells than in the medium was observed after 1 h Transfer of the hNIS gene in rat prostate carcinoma U Haberkorn et al 775 1e+5

wild type 40000 hNIS 8e+4 cells) 5 cells) 5 30000 6e+4

4e+4 20000

Iodide uptake (cpm/10 wild type Iodide uptake (cpm/10 2e+4 hNIS-expressing 10000

0 0 0 50 100 150 200 250 con FCCP SP10 SP50 DIDS DIDS ouab minutes 100 300

Figure 1 Time dependence of 125IÀ uptake in wild-type rat prostate Figure 2 Effect of DIDS, FCCP, sodium perchlorate and ouabain on the 125 À adenocarcinoma cells and in the hNIS-expressing cell line AThNIShyg9. I uptake in wild-type rat prostate adenocarcinoma cells (a) and in the Mean values and standard deviation (n¼3). hNIS-expressing cell line AThNIShyg9 (b) after 1 h incubation. Mean values and standard deviation (n¼3). incubation. The initial uptake of iodide was dependent tumor tissue significantly accumulated 131IÀ leading to on the incubation time and cells accumulated 125IÀ scintigraphic visualization, whereas only low-iodide maximally after 30–60 min incubation. The radioactivity uptake was observed in the wild-type prostate carcinoma measured after 2 and 4 h 125IÀ incubation was at a (Figure 4). In the genetically modified tumor, the tracer plateau level implicating the steady-state uptake. accumulation increased to a maximum level during the A functional characterization of the cells was per- first hour after administration, followed by a decrease of formed by competition and efflux experiments. Figure 2 the intratumoral radioactivity (Figure 5 and Table 1). presents the effect of DIDS, FCCP, sodium perchlorate After 24 h and a dose of 550 MBq/m2 hNIS-expressing and ouabain on the Na125I uptake in the wild-type (a) tumors lost 89% of the initial activity (wild type 88%); and the hNIS (b) expressing cells. In the presence of 10 or after a dose of 1200 MBq/m2 89% (wild type 73%) and 50 mM sodium perchlorate, we observed a dose-depen- after a dose of 2400 MBq/m2 91% (wild type 85%) of the dent inhibition of iodide accumulation in the ATh- maximum value were lost. Scintigraphically, the intratu- NIShygg cells, which was 89.9 and 94.5%, respectively. moral half-life of 131I was 7.470.9 h (hNIS) and 7.570.7 h Ouabain caused a 94.2% inhibition in uptake. The (wild type) for 550 MBq/m2, 7.370.9 h and 8.471.7 h addition of the proton conductor FCCP led to a loss of (1200 MBq/m2) and 9.072.1 h and 8.872.9 h 82.1% of the accumulated IÀ. The anion channel blocker (2400 MBq/m2). Table 1 presents the ex vivo quantitation DIDS had an opposite effect, as it increased the IÀ uptake of the 131IÀ uptake per g tissue of the tumors, which was into the cells to 136% at a concentration of 300 mM and evaluated at different time periods after tracer adminis- 115% with 100 mM (Figure 3). tration. In the hNIS-expressing tumor tissues an up to 20- In order to evaluate iodide efflux 125I uptake was fold (1200 MBq/m2; 15-fold after 2400 MBq/m2) higher allowed to proceed for 1 h, at which time a steady-state iodide accumulation was detected when compared to the level of iodide accumulation was achieved. After the wild-type tumor. At 24 h after 131I administration only medium had been replaced by nonradioactive medium, 0.470.2 (1200 MBq/m2) and 0.2470.02% (2400 MBq/m2) the amount of 125IÀ present in the AThNIShyg9 cell of the injected dose per g tissue were found in the hNIS- lysates was determined as a function of time. Up to 59% expressing tumors (0.0270.01 and 0.0470.01 for wild- of the cellular radioactivity was released into the type tumors). According to dosimetric calculations medium after 10 min and 81% efflux was observed after 1200 MBq/m2 resulted in 370.5 Gy (wild-type tumor 20 min, indicating that the radiotracer was not trapped 0.1570.1 Gy); 2400 MBq/m2 resulted in 3.170.9 Gy in the recombinant prostate carcinoma cells (Figure 3). (wild-type tumor 0.2670.02 Gy). The addition of 300 mM DIDS to the culture led to an increase of the initial 125IÀ uptake; however, a prolonged intracellular retention of the tracer was not achieved. Discussion 131Iodide uptake in tumor-bearing rats Currently used viral vectors for gene therapy of cancer To investigate the iodide uptake in hNIS expressing or have a low infection efficiency leading to moderate or nonexpressing Dunning prostate adenocarcinoma in vivo low therapy effects. This problem could be solved using AThNIShyg9 cells or wild-type AT1 cells were trans- an approach that leads to accumulation of radioactive planted into the right and left high of male young adult isotopes with beta emission. In this case, isotope trapping Copenhagen rats. After the animals had been injected centers in the tumor could create a crossfiring of beta with 131IÀ, a dynamic scintigraphic image was taken to particles thereby efficiently killing transduced and evaluate radioactivity in the blood circulation, the tumor nontransduced tumor cells. Transfer of the sodium- tissue and in different organs. Consistent with the data iodide symporter gene has been suggested as a new obtained from the in vitro studies the hNIS-expressing therapeutic principle to accumulate 131I in malignant

Gene Therapy Transfer of the hNIS gene in rat prostate carcinoma U Haberkorn et al 776 80000

70000 control µ

cells) 60000 300 m DIDS 5

50000

40000

30000

20000 intracellular iodide (cpm/10 intracellular 10000

0 0 5 10 15 20 minutes

Figure 3 Iodide efﬂux out the AThNIShyg9 cells after 1 h incubation with Na125I with and without 300 mM DIDS. Mean values and standard deviation (n¼3).

values reported in other models.8,10,12,13 However, as observed in a previous study with a rat hepatoma a rapid efflux occurred with 81% of the radioactivity released into the medium after 20 min. Similar data have been obtained by other groups in different tumor models.12,15,17,18,21 Since the effectiveness of radioiodine therapy depends not only on the type and amount, but also on the biological half-life of the isotope in the tumor, a therapeutically useful absorbed dose seems unlikely for that type of experiments. A significant efflux was also seen in vivo: only 0.470.2 (1200 MBq/m2) and 0.2470.02% (2400 MBq/m2) of the injected dose per g in the hNIS-expressing tumors were observed at 24 h after tracer administration. Similarly, Nakamoto et al found less than 1% of the injected radioactivity at 24 h after 131I administration in modified MCF7 mammary carcinomas, although initially a high uptake was seen.15 This corresponds to a very short half- life of 131I (7.5, 7.3 or 9.0 h) in rat prostate carcinomas, which has also been described by Nakamoto et al15 for human mammary carcinomas with a calculated biological half-life of 3.6 h. In contrast, differentiated thyroid carcinoma showed a biological half-life of less than 10 d and normal thyroid of approximately 60 d.22 However, in vitro clonogenic assays revealed selective killing of NIS-expressing cells in some studies.8–10,14,19 The activities applied were 29.6 MBq for 7 h, 3.7 MBq for 7 h or 4 MBq/ml.19,14,10 Also, bystander effects have been suggested in three-dimensional spheroid cultures.10 In vivo experiments in stably transfected human prostate carcinoma cells revealed a long biological half-life of Figure 4 Scintigraphic image of a tumor-bearing male young adult 45 h.19 This resulted in a significant tumor reduction Copenhagen rat subcutaneously transplanted with hNIS-expressing (right (84712%) after a single intraperitoneal application of a thigh) or wild-type prostate adenocarcinoma cells (left thigh) at 1 h after 131 19,20 injection of 131IÀ. very high I dose of 111 MBq. In order to improve therapy outcome, Smit et al investigated the effects of low-iodide diets and thyroid ablation on iodide ki- tumors. NIS gene transfer resulted in a high uptake in a netics.18 The half-life in NIS-expressing human follicu- variety of tumor models.8–20 In the rat prostate carcinoma lary thyroid carcinomas without thyroid ablation and we observed an up to 200-fold increased uptake as under a regular diet was very short with 3.8 h. In compared to the wild-type cell line which is similar to thyroid-ablated mice kept on a low-iodide diet, half-life

Gene Therapy Transfer of the hNIS gene in rat prostate carcinoma U Haberkorn et al 777 of radioiodide was increased to 26.3 h, which may be on a low-iodide diet, postponed tumor development. caused by diminished renal clearance of radioiodine and However, 9 weeks after therapy, tumors had developed lack of iodide trapping by the thyroid. Subcutaneous in four of the seven animals. The estimated tumor dose in injection of 74 MBq in thyroid-ablated nude mice, kept these animals was 32.2 Gy.18 Our study investigated the absorbed dose in genetically modified tumors after treatment with amounts of 131I corresponding to those given to patients. In order to determine the administered doses to the rats, we used factors relating the body weight to the body surface as described by Freireich et al.23 For a patient with 60 kg, an administered dose of 4 GBq 131I corresponds to 2466 MBq/m2. Therefore, we applied 1200 MBq 131I/m2 and 2400 MBq 131I/m2, which resulted in only 3 Gy absorbed dose in the genetically modified tumors. Since approximately 80 Gy has been described as necessary to achieve elimination of metastases in patients with thyroid cancer, this is not likely to induce a significant therapeutic effect in the tumors.24 Furthermore, the experiments were performed under ideal conditions with 100% NIS-expressing cells in the tumors. Given the low infection efficiency of currently viral vectors in vivo, the absorbed dose in a clinical study would be considerably lower. There are several differences between our study and the experiments of Spitzweg and Smit: administered dose, mode of 131I application and the tumor and animal systems used.17–20 In a mouse, doses of 74 and 111 MBq correspond to administered doses of 11 100 and 16 650 MBq/m2, respectively. This is far more than the doses used in patients. There is also a difference in tracer administration: intraperitoneal or subcutaneous versus intravenously. Intraperitoneal or subcutaneous administration may result in prolonged retention in the animal because of prolonged resorbtion, which may partly explain the longer half-life observed in these studies. Furthermore, Smit et al injected 131I at 3 weeks after transplantation when tumors were not visible. Tumors of small size may have a better chance to respond because there are no hypoperfused regions with limited tracer availability. However, this is not a realistic simulation of the tumor patient. There are also differences in the animal and tumor models, which may explain the varying results: Spitzweg et al and Smit et al used human tumors in nude mice, we used a syngeneic rat model. Therefore, differences in the biodistribution of iodide and the biochemical properties of the tumor cells may lead to differences in iodide retention. However, Smit et al observed similar values in tumor-bearing mice Figure 5 131IÀ uptake (imp/pixel) in hNIS-expressing and wild-type prostate adenocarcinomas after administration of 555, 1200 or 2400 MBq/ with intact thyroids. Finally, it is possible that thyroid m2 at different time periods after radiotracer application in rats. The sum of ablation as described by Smit et al would also result two exponential decay functions was fitted to the data points. Mean values in a longer intratumoral half-life of iodide in our tumor and standard deviations (n¼6). model.

Table 1 Iodide activity concentration (kBq/g tumor and % injected dose per g tumor) in hNIS-expressing and wild-type (wt) prostate adenocarinoma at different periods after injection of different doses of 131I

Hours p.i. wt (kBq/g) % dose/g hNIS (kBq/g) % dose/g Dose (MBq/m2)

1 197.4742.4 0.2470.05 2873.37898.6 3.4670.96 2400 8 88.4719.9 0.170.02 1393.17447.6 1.7270.56 2400 24 32.479.9 0.0470.01 181.8726.9 0.2470.02 2400 1 117.6711.9 0.2770.03 2411.27531.5 5.5771.2 1200 24 10.274.5 0.0270.01 172.9771.0 0.470.17 1200

Mean values and standard deviation (n=4). The weight represents the weight of the tissue sample measured including blood.

Gene Therapy Transfer of the hNIS gene in rat prostate carcinoma U Haberkorn et al 778 99m À 211 In order to prolong the iodine retention time in tumors 340-fold for TcO 4 and 60-fold for At. Although all some authors tried to simultaneously transfer the NIS radioisotopes showed a rapid efflux, higher absorbed and the thyroperoxidase gene.9,21 Boland et al observed doses in the tumor were found for 211At as compared iodide organification in cells coinfected with both the to 131I.34 NIS and the TPO gene in the presence of exogenous In conclusion, transfer of the sodium-iodide transpor- hydrogen peroxide. However, the levels of iodide ter gene into rat prostate adenocarcinoma cells together organification obtained were too low to significantly with the application of 131I in clinically relevant amounts increase the iodide retention time. In a variety of did not result in therapeutically useful absorbed doses in different cell lines including human anaplastic thyroid the NIS-expressing tumors in vivo. Further studies have carcinoma and rat hepatoma cells, we were not able to to examine pharmacological modulation of iodide measure TPO enzyme activity or enhanced accumulation efflux or the use of the hNIS gene as an in vivo reporter of iodide irrespective of very high amounts of hTPO gene.35 protein after retroviral transfer of the human TPO gene.25 In contrast, Huang et al observed an increased radioiodide uptake (by a factor of 2.5) and retention (by a Materials and methods factor of 3) and enhanced tumor cell apoptosis after transfection of nonsmall cell lung cancer cells with both Retroviral transfer of the hNIS gene human NIS and TPO genes.21 However, a 72% efflux For transfer of the hNIS gene a bicistronic retroviral occurred in vitro during the first 30 min, indicating a very vector was used consisting of the hNIS gene and the low hTPO activity in the genetically modified cells. hygromycin resistance gene cloned downstream of the Therefore, other modulations of iodide retention in elongation factor 1a (EF1a) promoter. To ensure simulta- tumor cells should be evaluated in future studies. The neous expression of the genes coding for the hNIS and determinants of TPO activity are incorporation of a heme for the hygromycin resistance and stabilization of the prosthetic group, glycosylation, localization, conforma- mRNA, a synthetic intron and an internal ribosomal tion, multimerization of the protein and the presence of entry site from encephalomyocarditis virus (EMCV) 12 thyroglobulin as acceptor. Therefore, future approaches were inserted between the genes. in our laboratory are directed towards stabilization of TPO by cotransfer of calreticulin and also the induction Cell culture, retroviral infection and generation of of thyroid specific gene expression by transfer of thyroid- recombinant cell lines specific transcription factors. Alternatively, we are using The rat prostate adenocarcinoma cell line (Dunning other peroxidases such as myeloperoxidase and lacto- R3327 subline AT1) and the transient packaging cell line peroxidase. BOSC23 employed for the production of ecotropic Lithium has been reported to reduce the release of retroviral particles were cultured in RPMI1640 medium iodine from the thyroid and, therefore, was used to (Gibco BRL, Eggenstein, Germany) supplemented with enhance the efficacy of radioiodine treatment of differ- 292 mg/l glutamine, 100 000 IE/l penicillin, 100 mg/l entiated thyroid cancer.26 When the biological half-life streptomycin and 10% fetal calf serum.12,36 The cell lines was less than 3 days lithium prolonged the effective half- were cultured at 371C, in an atmosphere of 95% air and 26 life by more than 50%. In FRTL-5 rat thyroid cells and 5% CO2. in primary cultures of porcine thyroid follicles 2 mM For transient packaging of the retroviral particles a lithium suppressed TSH-induced iodide uptake, iodide lipofection of BOSC23 cells was carried out. After 2 days uptake stimulated by 8-bromo-cAMP, iodine organifica- the medium was centrifuged to remove detached tion and de novo thyroid hormone formation.27,28 Lithium BOSC23 cells and used for the infection of the AT1 cells is concentrated by the thyroid and inhibits thyroidal in the presence of 8 mg polybrene/ml overnight. The cells iodine uptake, iodotyrosine coupling, alters thyroglobu- were treated with 425 mg/ml hygromycin for 4 weeks lin structure and inhibits thyroid hormone secretion.27–31 until 16 resistent cell lines were established. These were Therefore, if enhanced iodide trapping in the thyroid by used in iodide uptake experiments. The cell line which lithium relies on interaction with iodine coupling to presented the maximal 125IÀ uptake (AThNIShyg9) was tyrosine residues or inhibition of thyroid hormone employed for the following experiments. The tumor secretion, an organification process is still needed to doubling time in vitro was 11.271.3 h for AT1 and obtain a sufficient iodine accumulation in the tumor. First 12.471.4 h for AT1NIShyg9. experiments in our laboratory showed only a marginal effect of lithium in hNIS-expressing hepatoma cells.32 Immunocytochemistry A further option to increase therapy outcome is the use Immunocytochemistry was performed according to the of biologically more effective isotopes. Dadachova et al33 procedures previously described with minor modifica- compared 188Re-perrhenate with 131I for the treatment of tions.37 In detail, cells were fixed with 2% formaldehyde NIS-expressing mammary tumors.33 In a xenografted in 10 mM PBS (20 min, on ice) and 0.2% Triton X-100 in breast cancer model in nude mice, 188Re-perrhenate PBS (10 min, room temperature [RT]). For immunostain- exhibited NIS-dependent uptake into the mammary ing monoclonal mouse anti-human sodium-iodide sym- tumor. Dosimetry showed that 188Re-perrhenate deliv- porter antibodies BrA 6-1 and BrA 10-11 with 80 mg/ml ered a 4.5 times higher dose than 131I and, therefore, may in PBS for 1 h at RT (BRAHMS Diagnostics, Berlin, provide enhanced therapeutic efficacy. Furthermore, the Germany) were used. After exposure to these antibodies, high LET -emitter astatine-211 has been suggested as an the cells were incubated with Cy3-conjugated goat isotope with high radiobiologic effectiveness.15,34 First anti-mouse IgG (1:100, 1 h, RT; Dianova, Hamburg, experiments showed that the tracer uptake in NIS- Germany). Finally, nuclei were counterstained with expressing cell lines increased up to 350-fold for 123I, Hoechst 33342 dye (1 mM; Molecular Probes, Leiden,

Gene Therapy Transfer of the hNIS gene in rat prostate carcinoma U Haberkorn et al 779 The Netherlands). After every incubation period rinsing lateral tail vein of the rats a dynamic scintigraphic image steps (three times, 5 min; RT) with 10 mM PBS were was taken for the following 30 min to evaluate the early included. As a negative control, the primary antibody tumor uptake and the resulting body distribution of 131I was omitted. Cells were analyzed and photographed using a 10 in scintillation-camera (Searle-Siemens, Ger- using an Axiophot fluorescence microscope (Zeiss many). The time-dependent relative accumulation of GmbH, Jena, Germany). Digitalized figures were pro- radioactivity in different regions of interest (ROI), for cessed, arranged and lettered by computer software example, the heart, the liver, the tumor, the bladder and (Adobe Photoshop 5.0/Illustrator 8.0). the whole animal was monitored in six animals at 1, 2, 4, 8 and 24 h postinjection. Measurement and modulation of the 125I uptake and The dose given to the animals was determined using the formula efflux The iodide uptake was determined in triplicates as injected dose per rat ¼ðdose per m2=FÞ described previously.2,12 In the presence of 74 kBq Ã body weight of the rat in kg Na125Iodide (Amersham Buchler Germany; specific where F is a weight-dependent factor that was taken activity 625.3 MBq/mg; radioactive concentration according to Freireich.23 The F values for a 150, 200 or 3.7 GBq/ml; radiochemical purity 99.3%) the recombi- 250 g weighing rat are 5.9, 6.5 and 7, respectively (F¼3 nant cell line AT1hNIShyg9 and wild-type AT1 cells were for mice and F¼37 for patients). cultured for 1, 2, 5, 10, 30 min, or 1, 2 and 4 h in six-well The absolute amount of radioactivity (% injected dose/ plates. In order to simulate the clinical situation of g wet tissue) was determined in 20 animals after 1200 radioiodine treatment, we added no cold NaI. After and 2400 MBq/m2 131I, which were killed 1, 8 and 24 h washing twice with ice-cold phosphate-buffered saline (2400 MBq/m2) or 1 and 24 h (1200 MBq/m2) postinjec- (PBS), the cells were lysed with sodium hydroxide on ice. tion to analyze the tumors using an automated NaI(Tl) Using an automated NaI(Tl) well counter (Cobra II, well counter (Cobra II, Canberra Packard, Meriden, CT, Canberra Packard, Meriden, CT, USA) the radioactivity USA). Using MIRDOSE 3*38 the cumulated activity and was measured in cell lysates and in the medium. The the absorbed dose in wild-type and hNIS-expressing viable cell number was determined in a Coulter counter tumors after administration of 1200 and 2400 MBq/m2 (Coulter Electronics, Dunstable, UK) and by trypan blue 131I was calculated for a unit density sphere with a mass staining (more than 94% viable cells). Using the Coulter of 1 g considering only the value of the absorbed dose counter a median cell volume of 1.3 fl (between 0.6 and in the sphere. Therefore, the sum of two exponential 3.1 fl) with no significant difference for the genetically decay functions was fitted to the data points. The modified and wild-type AT1 cells was measured. MIRDOSE3 calculation S-value for self-irradiation was For the modulation of the iodide uptake hNIS- 29.5 mGy/MBq s. expressing and wild-type cells were incubated for 1 h in Na125I medium (74 kBq) or Na125I-medium supplemented with 10 or 50 mM sodium perchlorate (Sigma- Aldrich, Taufkirchen, Germany), 0.5 mM ouabain, 100 Acknowledgements or 300 mM of the anion channel blocker 4,40-diisothio- 0 We thank Mrs Irmgard Preugschat-Gumprecht and Silke cyano-2,2 -disulfonic acid stilbene (DIDS; Sigma-Aldrich) Vorwald for expert assistance with immunocytochemis- or 10 mM of the proton conductor carbonyl cyanide try and R Ku¨ hnlein for his help in performing the animal p-trifluoromethoxyphenylhydrazone (FCCP; Sigma-Al- 2 experiments. We also thank N Morgenthaler and J Struck drich) according to Weiss et al. Thereafter, the cells were (BRAHMS Diagnostics, Berlin, Germany) for supplying washed, lysed and counted as described. 125 hNIS antibodies. This study was supported by a grant To determine the Iodide efflux recombinant and from the ’Tumorzentrum Heidelberg/Mannheim’. wild-type cells were incubated for 1 h in the presence or This article is dedicated to Harald zur Hausen on the absence of 300 mM DIDS with a medium containing 125 occasion of his retirement as head of the German Cancer 74 kBq Na Iodide in six-well plates. After the cells had Research Center with gratitude and appreciation for 20 been washed twice three wells were lysed immediately. years of leadership. To the other wells fresh nonradioactive medium was added and the cells were again incubated for 2, 4, 6, 8, 10, 12, 16 or 20 min, and washed and lysed as described. References

Measurement of the 131I uptake in tumor tissue of rats 1 Marcocci C, Cohen JL, Grollman EF. Effect of actinomycin D on 115 The experiments were performed in compliance with iodide transport in FRTL-5 thyroid cells. Endocrinology 1984; : 2123–2132. German laws relating to the conduct of animal experi- 2 Weiss SJ, Philp NJ, Grollman EF. Iodide transport in a 6 m mentation. 2.5 Â 10 tumor cells in 100 l were trans- continuous line of cultured cells from rat thyroid. Endocrinology planted subcutaneously into the right (AT1hNIShyg9) or 1984; 114: 1090–1098. left (wild-type AT1) thigh of male young adult Copenha- 7 3 Paire A, Bernier-Valentin F, Selmi-Ruby S, Rousset B. Character- gen rats weighing 268.1 29.8 g. For imaging studies that ization of the rat thyroid iodide transporter using anti-peptide were performed under general gaseous anesthesia (40% antibodies. J Biol Chem 1997; 272: 18245–18249. 2 2 O /60% N O/1% halothane) only animals bearing 4 Nakamura Y, Ohtaki S, Yamazaki I. Molecular mechanism of tumors with a minimum size of 15 mm diameters were iodide transport by thyroid plasmalemmal vesicles: cooperative accepted. Immediately after injection of 200 mlof131Iin sodium activation and asymmetrical afﬁnities for the ions on the 0.9% NaCl (doses 550, 1200 and 2400 MBq/m2) into the outside and inside of the vesicles. J Biochem 1988; 104: 544–549.

Gene Therapy Transfer of the hNIS gene in rat prostate carcinoma U Haberkorn et al 780 5 Nakamura Y, Kotani T, Ohtaki S. Transcellular ioide transport 21 Huang M et al. Ectopic expression of the thyroperoxidase gene and iodination on the apical plasma membrane by mono- augments radioiodide uptake and retention mediated by the layer porcine thyroid cells cultured on collagen-coated fibers. sodium iodide symporter in non-small cell lung cancer. Cancer J Endocrinol 1990; 126: 275–281. Gene Ther 2001; 8: 612–618. 6 Smanik PA et al. Cloning of the human sodium iodide 22 Berman M, Hoff E, Barandes M. Iodine kinetics in man: a model. symporter. Biochem Biophys Res Commun 1996; 226: 339–345. J Clin Endocrinol Metab 1968; 28: 1–14. 7 Dai G, Levy O, Carrasco N. Cloning and characterization of the 23 Freireich EJ et al. Quantitative comparison of toxicity of antic- thyroid iodide transporter. Nature 1996; 379: 458–460. ancer agents in mouse, rat, hamster, dog, monkey, and man. 8 Boland A et al. Adenovirus-mediated transfer of the thyroid Cancer Chemotherapy Rep 1966; 50: 219–244. sodium/iodide symporter gene into tumors for a targeted 24 Maxon HR et al. Relation between effective radiation dose and radiotherapy. Cancer Res 2000; 60: 3484–3492. outcome of radioiodine therapy for thyroid cancer. N Engl J Med 9 Boland A et al. Transposition of the thyroid iodide uptake and 1983; 309: 937–941. organification system in nonthyroid tumor cells by adenoviral 25 Haberkorn U et al. Iodide uptake in human anaplastic thyroid vector-mediated gene transfers. Thyroid 2002; 12: 19–26. carcinoma cells after transfer of the human thyroid peroxidase 10 Carlin S et al. Experimental targeted radioiodide therapy gene. Eur J Nucl Med 2001; 28: 633–638. following transfection of the sodium iodide symporter gene: 26 Koong S et al. Lithium as a potential adjuvant to 131I therapy effect on clonogenicity in both two- and three-dimensional of metastatic, well differentiated thyroid carcinoma. J Clin models. Cancer Gene Ther 2000; 7: 1529–1536. Endocrinol Metab 1999; 84: 912–916. 11 Cho JY et al. Expression and activity of human Na+/IÀ symporter 27 Urabe M et al. Effect of lithium on function and growth of in human glioma cells by adenovirus-mediated gene delivery. thyroid cells in vitro. Endocrinology 1991; 129: 807–814. Gene Ther 2000; 7: 740–749. 28 Lazarus JH. The effects of lithium therapy on thyroid and 12 Haberkorn U et al. Transfer of the human sodium iodide thyrotropin-releasing hormone. Thyroid 1998; 8: 909–913. symporter gene enhances iodide uptake in hepatoma cells. 29 Sedvall G, Jonsson B, Petterson U, Levin K. Effects of lithium J Nucl Med 2001; 42: 317–325. salts on plasma protein bound iodine and uptake of 131Iin 13 La Perle KMD et al. In vivo expression and function of the thyroid gland of man and rat. Life Sci 1968; 7: 1257–1264. sodium iodide symporter following gene transfer in the 30 Temple R, Berman M, Robbins J, Wolff J. The use of lithium in MATLyLu rat model of metastatic prostate cancer. Prostate the treatment of thyrotoxicosis. J Clin Invest 1972; 51: 2746–2756. 2002; 50: 170–178. 31 Gershengorn MC, Izumi M, Robbins J. Use of lithium as an 14 Mandell RB, Mandell LZ, Link CJ. Radioisotope concentrator adjunct to radioiodine therapy of thyroid carcinoma. J Clin gene therapy using the sodium/iodide symporter gene. Cancer Endocrinol Metab 1976; 42: 105–111. Res 1999; 59: 661–668. 32 Sieger S et al. Tumour specific activation of the sodium/iodide 15 Nakamoto Y et al. Establishment and characterization of a symporter gene under control of the glucose transporter gene 1 breast cancer cell line expressing Na+/IÀ symporters for promoter (GTI-1.3). Eur J Nucl Med: in press. radioiodide concentrator gene therapy. J Nucl Med 2000; 41: 33 Dadachova E et al. Rhenium-188 as an alternative to Iodine-131 1898–1904. for treatment of breast tumors expressing the sodium/iodide 16 Shimura H et al. Iodide uptake and experimental 131J therapy in symporter (NIS). Nucl Med Biol 2002; 29: 13–18. transplanted undifferentiated thyroid cancer cells expressing the 34 Petrich T et al. Establishment of radioactive astatine and iodine Na+/IÀ symporter gene. Endocrinology 1997; 138: 4493–4496. uptake in cancer cell lines expressing the human sodium iodide 17 Smit JW et al. Reestablishment of in vitro and in vivo iodide symporter. Eur J Nucl Med 2002; 29: 842–854. uptake by transfection of the human sodium iodide symporter 35 Haberkorn U, Altmann A, Eisenhut M. Functional genomics and (hNIS) in a hNIS defective human thyroid carcinoma cell line. proteomics – the role of nuclear medicine. Eur J Nucl Med 2002; Thyroid 2000; 10: 939–943. 29: 115–132. 18 Smit JWA et al. Iodide kinetics and experimental 131I therapy in a 36 Pear WS, Nolan GP, Scott ML, Baltimorre D. Production of high xenotransplanted human sodium-iodide symporter-transfected titer helper-free retroviruses by transient transfection. Proc Natl human follicular thyroid carcinoma cell line. J Clin Endocrinol Acad Sci USA 1993; 90: 8392–8396. Metab 2002; 87: 1247–1253. 37 Kinscherf R et al. Induction of manganese superoxide dismutase 19 Spitzweg C et al. Treatment of prostate cancer by radioiodine by oxidized-LDL – its relevance in atherosclerosis of humans and therapy after tissue-specific expression of the sodium iodide heritable hyperlipidemic rabbits. FASEB J 1997b; 11: 1317–1328. symporter. Cancer Res 2000; 60: 6526–6530. 38 Stabin M. MIRDOSE: personal computer software for internal 20 Spitzweg C et al. In vivo sodium iodide symporter gene therapy dose assessment in nuclear medicine. J Nucl Med 1996; 37: of prostate cancer. Gene Ther 2001; 8: 1524–1531. 538–546.

Gene Therapy