The Na؉/I؊ Symporter (NIS): Imaging and Therapeutic Applications

Ekaterina Dadachova and Nancy Carrasco

.The Na؉/I؊ symporter (NIS) is the plasma mem- cancer, human glioma, and hepatoma cell lines brane glycoprotein that mediates the active uptake Most notably, significant radioiodine therapy of I؊ in the , ie, the crucial first step in results have been obtained in the NIS-transfected thyroid hormone biosynthesis. NIS also mediates I؊ human prostatic adenocarcinoma cell line LNCaP uptake in other tissues, such as salivary glands, and in NIS-transfected myeloma cells, both gastric mucosa, and lactating (but not nonlactating) of which exhibited prolonged retention of radio- .mammary gland. The ability of thyroid cancer cells even in the absence of I؊ organification -to actively transport I؊ via NIS provides a unique The therapeutic potential of alternative NIS and effective delivery system to detect and target transported radioisotopes with different decay these cells for destruction with therapeutic doses of properties and a shorter, physical half-life than ؊ ␤ 188 188 ؊ 131 radioiodide. Breast cancer is the only malignancy I , such as -emitter Rhenium ( ReO4 ) -other than thyroid cancer to have been shown to and ␣-emitter 211Astatine (211At؊), has been evalu functionally express NIS endogenously. The consid- ated. In conclusion, it is clear that the remark- erable potential diagnostic and therapeutic use of able progress made in the last few years in the radioiodide in breast cancer is currently being as- molecular characterization of NIS has created sessed. On the other hand, exogenous NIS gene new opportunities for the development of diagnos- transfer has successfully been carried out into a tic and therapeutic applications for NIS in nuclear variety of other cell lines and tumors, including medicine. A375 human melanoma tumors, and SiHa cervix © 2004 Elsevier Inc. All rights reserved.

ϩ Ϫ HE Na /I symporter (NIS) is an integral remnants after surgery and in metastases.4 There- T plasma membrane glycoprotein most com- fore, the study of NIS is of high relevance to monly studied in connection with the thyroid thyroid . Nevertheless, no molec- gland, where NIS mediates the active transport of ular information on NIS was available until very Ϫ I into the thyroid follicular cells as the crucial recently. In the absence of nucleotide and protein first step in thyroid hormone biosynthesis. Thyroid sequence information, the detailed molecular char- hormones T3 and T4 (tri-iodothyronine and thyrox- acterization of NIS started in 1996 when the ine [or tetra-iodothyronine], respectively) are the complimentary deoxyribonucleic acid (cDNA) en- only iodine-containing hormones in vertebrates. coding rat NIS was isolated by Dai and coworkers Ϫ Because I is an essential constituent of T3 and T4, by expression cloning in Xenopus laevis oocytes, both thyroid function and its systemic ramifica- Ϫ using cDNA libraries derived from a highly func- tions depend on an adequate supply of I to the tional rat thyroid-derived cell line (FRTL-5 cells).5 gland.1 This supply in turn depends on the suffi- Ϫ Another major development in the molecular cient dietary intake of I and proper NIS function. characterization of NIS was the generation of high NIS is also expressed endogenously functionally in ϳ affinity (Kd 1 nM), site-directed, polyclonal anti- other tissues, including salivary glands, gastric NIS Abs against the carboxyl terminus of the mucosa, and lactating mammary gland, in all of 6 Ϫ protein, by Levy and coworkers. Based on the which NIS mediates active I transport. While the cloned cDNA, rat NIS was determined to be a functional significance of NIS in the gastric mu- protein of 618 amino acids (relative molecular cosa and salivary glands is unknown, in the lactat- mass 65,196). The current secondary structure ing mammary gland, NIS mediates the transloca- model for NIS, based on extensive experimental tion of IϪ into the milk, making this anion testing, proposes 13 transmembrane segments, available for the nursing newborn to biosynthesize with the amino terminus facing extracellularly and his/her own .2,3 The ability of the thyroid to accumulate IϪ via NIS has long provided the basis for diagnostic From the Departments of Nuclear Medicine and Molecular scintigraphic imaging of the thyroid with radioio- Pharmacology, Albert Einstein College of Medicine, Bronx, NY. dide. It has served as an effective means for Address reprint requests to Dr. Nancy Carrasco, Department therapeutic doses of radioiodide to target and of Nuclear Medicine, Albert Einstein College of Medicine, 1300 Morris Park Ave., Bronx, NY 10461. destroy hyperfunctioning thyroid tissue, such as in Ϫ © 2004 Elsevier Inc. All rights reserved. Graves’ disease, and to destroy I -transporting 0001-2998/04/3401-0004$30.00/0 thyroid carcinoma cells, both within primary tumor doi:10.1053/j.semnuclmed.2003.09.004

Seminars in Nuclear Medicine, Vol XXXIV, No 1 (January), 2004: pp 23-31 23 24 DADACHOVA AND CARRASCO the carboxyl terminus intracellularly (Fig 1).7 The cDNA encoding human NIS (hNIS) was subse- quently identified on the expectation that hNIS would be highly homologous to rat NIS (rNIS). Using primers to the cDNA rNIS sequence, Smanik and coworkers identified a cDNA clone encoding hNIS.8 The nucleotide sequence of hNIS revealed an open reading frame of 1,929 nucleo- tides, which encodes a protein of 643 amino acids. The hNIS shows an 83% identity and 93% simi- larity to rNIS. In all of the tissues and cells where it is functionally expressed, NIS couples the inward Fig 1. Secondary structure model of Na؉/I؊ symporter “downhill” translocation of Naϩ to the inward (NIS). Structure contains 13 putative transmembrane seg- Ϫ ments (TMS). Both the hydrophilic loop containing N225 and “uphill” translocation of I , establishing ϳ20-fold the NH terminus face extracellularly. All 3 N-linked glycosyl- Ϫ 2 to 40-fold I concentration gradients under steady- ation consensus sequences are indicated at positions 225, 2 state conditions. The driving force for the process 485, and 497. Reprinted with permission. (Color version of figure is available online.) is the inwardly directed Naϩ gradient generated by the Naϩ/Kϩ adenosinetriphosphatase (ATPase). ENDOGENOUS VERSUS EXOGENOUS NIS NIS activity in all these cells is blocked by the EXPRESSION IN CANCER: ENDOGENOUS IN well-known, “classic” competitive inhibitors, the THYROID AND BREAST CANCER, anions thiocyanate (SCNϪ) and Ϫ EXOGENOUS ELSEWHERE (ClO4 ). In thyroid follicular cells, NIS mediated IϪ accumulation is stimulated by thyroid-stimulat- Endogenous NIS, as indicated previously, is ing hormone (TSH). IϪ is then translocated from physiologically and functionally expressed in only the cytoplasm of these cells across the apical a few normal tissues, including salivary glands, plasma membrane towards the colloid in a process gastric mucosa, and lactating mammary gland. The called IϪ efflux, mediated by a different transporter only cancer known for decades to express endog- 9,10 enous NIS was thyroid cancer. The ability of that has yet to be identified unequivocally. In a Ϫ Ϫ cancerous thyroid cells to transport actively I via complex reaction called organification of I , cata- NIS provides a unique and effective delivery sys- lyzed by thyroid peroxidase (TPO) at the cell- Ϫ tem to detect and target these cells for destruction colloid interface, I is oxidized and incorporated with therapeutic doses of radioiodide, largely with- into some tyrosyl residues within the thyroglobulin out harming other tissues. Therefore, it seems (Tg) molecule, leading to the subsequent coupling feasible that radioiodide could be a diagnostic and of iodotyrosine residues. The term organification Ϫ therapeutic tool for the detection and destruction of refers to the covalent incorporation of I into other cancers in which endogenous NIS is func- organic molecules (in this case Tg), as opposed to Ϫ Ϫ tionally expressed. Pointing in this direction is a nonincorporated, inorganic, or free I .I organi- report by Tazebay and coworkers in which they fication in the thyroid results in the retention and Ϫ Ϫ showed that human breast carcinomas and experi- storage of I within the gland. The I organifica- mental mammary carcinomas in transgenic mice tion reaction can be blocked pharmacologically by express NIS.11 In vivo scintigraphic imaging of 6-n-propyl-2-thiouracil and 1-methyl-2-mercapto- experimental mammary adenocarcinomas in non- imidazole. In response to the demand for thyroid gestational and nonlactating female transgenic hormones, phagolysosomal hydrolysis of endocy- mice carrying either an activated ras oncogene tosed iodinated Tg ensues. T3 and T4 are secreted (c-Ha-ras) or overexpressing the Neu oncogene into the bloodstream. All of these steps, like (c-erbB-2) showed pronounced, active, specific, Ϫ Ϫ NIS-mediated I uptake, are stimulated by TSH. In and ClO4 -inhibitable NIS activity. This was also contrast, NIS in extrathyroidal tissues is not sub- observed later in transgenic mice overexpressing jected to regulation by TSH, and these tissues show the polyoma middle T antigen (PyV). Clearly, Ϫ minimal or no I organification.2,3 transgenic mice bearing experimental mammary THE Naϩ/IϪ SYMPORTER 25

99m Ϫ tumors provide an excellent model to study the have reported pertechnetate ( TcO4 ) accumu- potential role of endogenous NIS expression in lation in primary breast tumors in humans in 14 99m Ϫ mammary cancer and, particularly, the possible vivo. As discussed extensively later, TcO4 is effectiveness of radioiodide therapy in combating widely used for diagnostic imaging and is also this disease. transported by NIS, with the advantage of having a By immunohistochemistry, Tazebay and co- shorter half-life (6 hours) than 131I (8 days). These workers showed further that 87% of 23 human investigators studied 25 patients with cancer by invasive breast cancers and 83% of 6 ductal carci- scintigraphy and found active uptake by the tumors nomas in situ expressed NIS, as compared with in 4. This is a highly meaningful result, not only Ϫ only 23% of 13 extratumoral samples from the because it shows the existence of I transport vicinity of the tumors.11 Even more significantly, activity in a significant percentage of studied hu- none of the 8 normal samples from reductive man patients with breast cancer in vivo, but also mammaplasties they studied expressed NIS. Kogai because the observation was made in patients and colleagues reported an increase in NIS mes- whose were not downregulated (ie, thy- senger ribonucleic acid (mRNA), NIS protein, and roid NIS was still expressed, and, therefore, there Ϫ 99m Ϫ I uptake activity in a human mammary adenocar- was avid thyroidal TcO4 uptake; this de- cinoma cell line (MCF-7) in response to trans- creased the amount of radioisotope available for retinoic acid treatment.12 Doha«n and coworkers uptake by breast tumors). It is possible that a larger 99m Ϫ have recently developed a method for early detec- proportion of TcO4 -accumulating breast can- tion (by flow cytometry) of NIS expression in cer tumors might have been detected by scintigra- human mammary adenocarcinoma cells collected phy if the availability of the radioisotope to tu- by fine needle aspiration.3 The results obtained moral tissue had been optimized by thyroid with this method correlate closely with NIS ex- suppression. pression detected by immunohistochemistry of the In conclusion, although more extensive studies corresponding biopsy specimens. in humans are still necessary, functional expression Wapnir and coworkers examined the immuno- of endogenous NIS in breast cancer has been histochemical profile of NIS in thyroid, breast, and documented in both experimental mice and hu- other carcinomas using high-density tissue mi- mans. This result strongly suggests that NIS is croarrays and conventional sections.13 Validating upregulated with high frequency during malignant the use of microarrays, the study showed that transformation in breast cancer. Therefore, only 2 results obtained with microarrays, which allow the cancers, those of the thyroid and the breast, have simultaneous analysis of multiple samples on the been shown to express endogenous NIS function- same slide, correlated closely with those obtained ally. The tremendous potential of this finding for with conventional sections. Insofar as breast in the use of radioiodide in the diagnosis and treat- particular was concerned, they analyzed as many ment of breast cancer remains to be fully assessed. as 371 breast specimens, and the results confirmed For other cancers, exogenous NIS gene transfer is the findings stated previously: NIS expression was actively being researched. observed in whole tissue sections in 76% of inva- sive breast carcinoma and 88% of ductal carci- EXOGENOUS NIS GENE TRANSFER FOR noma in situ samples. The majority of normal IMAGING breast cores were negative (87%), as were 70% of The cloning and characterization of NIS, in light normal/nonproliferative samples analyzed. Plasma of the ample experience accumulated over the last membrane immunoreactivity was observed in ges- 60 years of imaging and treating thyroid patients tational breast tissues, and in some in situ ductal with radioiodide, has led to the development of a carcinomas and invasive ductal carcinomas. Be- novel gene transfer strategy in other tissues: the cause these studies in human samples were per- targeted expression of functional exogenous NIS in formed by immunohistochemistry, the findings selected cells aimed at rendering them susceptible show only endogenous NIS expression but not to imaging and/or destruction with radioiodide. necessarily endogenous NIS functional expression Several in vitro experiments on NIS gene transfer in human breast cancer. for diagnostic and therapeutic purposes have been Addressing this issue, Moon and coworkers reported in which NIS-mediated radioiodide up- 26 DADACHOVA AND CARRASCO take was used to visualize and destroy malignant tumor tissue. Cho and colleagues showed func- tumor cells. The NIS gene has clear and compel- tional expression of hNIS in a xenografted human ling advantages as a reporter gene. Scintigraphic glioma as a result of intratumoral injection of imaging of NIS expression can easily be attained recombinant adenovirus rAd-CMV-hNIS.19 Naka- with commercially available, inexpensive radionu- moto and coworkers established a novel breast 131 Ϫ 123 Ϫ 99m Ϫ clide probes, such as I , I , and TcO4 , cancer cell line, MCF3B, by stably transfecting all of which have long been approved for human NIS into the widely used MCF-7 cells.20 In a use. Groot-Wassink and coworkers have studied biodistribution study using MCF3B-xenografted Ϫ the ectopic expression of the hNIS gene in vivo mice, high 125I uptake (16.7%) was observed in and monitored it in biodistribution studies on the tumors 1 hour after injection. In addition, high Ϫ intravenous injection of 125I .15 Adenovirus (Ad) tumor-to-normal tissue ratios were also observed delivery successfully induced NIS gene expression (ranging from 4 to 21), except in the stomach in the liver, adrenal glands, lungs, pancreas, and (0.47). Sieger and colleagues showed radioiodine spleen. In addition, NIS expression was also uptake in a hepatoma cell line in vitro and in vivo readily detected in tumor xenograft models when after transfer of the hNIS gene under the control of the virus was injected intratumorally. Finally, NIS a tumor-specific regulatory element, the promoter expression was monitored by positron emission of the glucose transporter 1 gene (GTI-1.3).21 The tomography (PET) after intravenous injection of same group (ie, Haberkorn and coworkers22) used 124IϪ, showing the potential of this approach for NIS-transduced prostate carcinoma cells to study noninvasive imaging. in vivo biodistribution. In rat, the hNIS-expressing Ϫ Although NIS can conceivably serve as a pure tumors accumulated up to 22 times more I than reporter gene for noninvasive imaging of the trans- contralateral transplanted wild-type tumors. In fer of other therapeutic transgenes, most efforts conclusion, an impressive body of evidence has have concentrated on using NIS for imaging of its already been produced showing the induction of own expression in selected transfected cells or NIS functional expression in nonthyroid tumors or tumors, with the purpose of quantifying NIS ex- its restoration in undifferentiated thyroid malig- pression and establishing its expression efficiency nancies on the application of gene transfer tech- for follow-up radioiodine therapy. In one of the niques of the NIS gene. first publications on NIS gene transfer for imaging purposes, Shimura and coworkers reported trans- Exogenous NIS for Therapy fection of the rNIS cDNA in malignantly trans- The successful induction of NIS functional ex- formed rat thyroid cells (FRTL-Tc) that ordinarily pression by NIS gene transfer in nonthyroid and Ϫ do not show I transport activity.16 The cell line undifferentiated thyroid tumors prompted many that resulted from the transfection (ie, Tc-rNIS) groups to study the effects of 131IϪ therapy in these Ϫ displayed a 60-fold increase in 125I accumulation tumors.16,18,20,22 However, no tumor shrinkage was over background in vitro. Tumors formed with detected with treatment, apparently as a result of Tc-rNIS cells accumulated up to 27.3% of the total the rapid radioisotope efflux observed. It appears Ϫ administered 125I , showing an 11-fold to 27-fold that as the retention time of 131IϪ in the tumors was Ϫ increase in 125I concentration, as compared with only a few hours, it was not enough to deliver the nontransfected cells. Mandell and colleagues scin- radiation dose necessary to have a discernible tigraphically imaged rNIS-transduced A375 hu- therapeutic effect. Moreover, under these condi- man melanoma tumors, which accumulated signif- tions, it is likely that the 8-day half-life and decay Ϫ icantly more 123I than non-transduced tumors.17 properties of 131IϪ, which result in the emission of ϭ ␤ For their part, Boland and coworkers used the low-energy (Eaverage 0.134 MeV) -particles, AdNIS vector to transfect human tumor cells, also may have played a role in the disappointing namely SiHa cervix cancer cells and MCF-7 breast results obtained. cancer cells, in nude mice.18 A quantitative analy- Hence, several different approaches have been sis revealed that uptake in AdNIS-injected tumors pursued to circumvent the problem of insufficient was 4 to 25 times higher than in nontreated tumors. radiation dose to NIS-expressing tumors. Boland On average, 11% of the total amount of injected and coworkers proposed to improve the efficiency 125IϪ was recovered per gram of AdNIS-treated of NIS gene transfer and, thus, the IϪ uptake THE Naϩ/IϪ SYMPORTER 27 capacity of the target tissue, by using modified in cells coinfected with AdNIS and AdTPO in the vectors and/or higher viral doses.23 However, it presence of exogenous hydrogen peroxide. However, remains to be seen whether this approach will the levels of IϪ organification obtained were too low prove safe and will result in increased radiation to increase significantly the IϪ retention time in the doses to the tumors. Nakamoto and coworkers20 target cells. and Daniels and Haber24 have suggested that the Smit and coworkers studied whether transfec- fast radioiodine efflux from breast cancer cells tion of hNIS into the hNIS-deficient follicular might be pharmacologically modulated by admin- thyroid carcinoma cell line FTC133 makes these istering lithium salts, which increase the half-life cells susceptible to radioiodine therapy.28 In addi- of radioiodide. Yet, in vitro experiments with tion, the effects of a low IϪ diet and thyroid NIS-transduced hepatoma cells performed by ablation on IϪ kinetics were investigated. Tumors Sieger and colleagues21 revealed that lithium had derived from NIS-transfected FTC133-NIS30 in no significant effect on IϪ efflux. mice kept on a normal diet showed a high IϪ Boland and coworkers have proposed to increase accumulation rate (17.4% of administered activity the retention time of radioiodine in tumor cells by versus 4.6% in nontransfected tumors). The IϪ coupling the transfer of the NIS gene with the retention time (ie, IϪ biological half-life) in delivery of a gene involved in the IϪ organification FTC133-NIS30 tumors was 3.8 hours. In mice kept Ϫ process, such as TPO.23 As explained previously, the on a low I diet, peak activity in FTC133-NIS30 thyroid is the only tissue known to organify IϪ to a tumors was diminished (8.1%), while IϪ accumu- significant extent. IϪ organification is the covalent lation in the thyroid gland was increased. In incorporation of IϪ into selected tyrosyl residues on thyroid-ablated mice kept on a low IϪ diet, the the Tg molecule.25 The organification process causes half-life of radioiodide was increased considerably administered radioiodine to be retained within the (26.3 hours), leading to a much higher area under gland for several days.26 This relatively long reten- the time-radioactivity curve than in FTC133- tion time, which matches the physical half-life of 131I NIS30 tumors in mice on a normal diet without (ie, 8 days), allows a significant radiation dose to be thyroid ablation. Experimental radiotherapy with 2 delivered to the tissue. However, the simultaneous mCi in thyroid-ablated nude mice kept on a low IϪ transfer of both the NIS and TPO genes is not easy to diet postponed tumor development for up to 4 accomplish because of the inherent complexity of weeks after therapy, although the tumors eventu- gene therapy procedures and the difficulty in trans- ally regrew. The investigators concluded that the fecting only the desired target tissue (ie, the tumor) in short half-life of IϪ in NIS-transfected tumors vivo. could only be partially improved by conventional Huang and coworkers observed that, although the conditioning with thyroid ablation and low IϪ diet. expression of NIS resulted in significant radioiodide Cho and coworkers have extended their studies uptake in transfected non-small cell lung cancer on the imaging of NIS-transduced glioma tumors (NSCLC) cell lines, rapid radioiodide efflux limited to the analysis of the effect of radioiodine therapy tumor cell death.27 The transfection of NSCLC cells on these tumors.29 Gliomas are known to be with human NIS and TPO genes led to increases in aggressive and radioresistant tumors. Three doses radioiodide uptake and retention, as well as to en- of 4 mCi 131IϪ were administered to rats bearing hanced tumor cell apoptosis. Therefore, the investi- hNIS-transduced F98 glioma tumors. Some in- gators concluded that although single gene therapy crease in survival was observed, with the average with the NIS gene alone may have limited efficacy survival time for the animals with vector alone- because of rapid radioiodide efflux, the codelivery of transduced F98/LXSN tumors with 131IϪ treat- the TPO and NIS genes may be an effective way to ment, F98/hNIS tumors without 131IϪ treatment, circumvent the problem. Boland and coworkers con- and F98/hNIS tumors with 131IϪ treatment being structed a recombinant adenovirus encoding the hu- 30.4 Ϯ 3.2, 39.0 Ϯ 4.1, and 45 Ϯ 8.6 days, man TPO gene under the control of the cytomegalo- respectively. Tumor volume seemed to be reduced Ϫ virus early promoter (AdTPO).23 Infection of SiHa for a certain period by 131I treatment, but the human cervix tumor cells with this virus led to tumors eventually re-grew. production of an enzymatically active protein. A Significantly, NIS-transfected cell lines and tu- significant increase in IϪ organification was observed mors generated in animal models with these cells 28 DADACHOVA AND CARRASCO

Table 1. Decay Properties of Radioisotopes Used in NIS Research

Radioisotope Half-Life Therapeutic Emissions Photon Energy (keV) and Abundance (%)

131 ␤Ϫ ϭ -Iodine 8d ,Eav 0.134 MeV 364 (81) 99m-Technetium 6 h none 140 (89) 188 ␤Ϫ ϭ -Rhenium 16.9 h ,Eav 0.764 MeV 155 (15) 211-Astatine 7.2 h ␣,Eϭ 5.87 MeV 79 (21) have an ability to retain IϪ in significant amounts theless very impressive. Noting that treatment was for sufficient periods without organification. For successful even though hNIS was not transduced example, no IϪ organification occurs in either the into every myeloma cell, the investigators ex- MCF-7 or T-47D mammary tumor cell lines, but plained that nontransduced cells could be killed by the IϪ retention time in both of these cell lines is electrons emitted by 131IϪ from distant cells. Such longer than in cells from other origins.30,18 This emitted electrons can travel several cell diameters result suggests that the IϪ efflux system is less in vivo. This is a classical example of “cross-fire” efficient or more rapidly saturated in some cell radiation damage by ␤-emitters, such as 131IϪ, types than in others. In turn, this could explain the which was referred to in the article as the “by- encouraging therapeutic results reported by several stander effect.”33 groups in various NIS-transduced tumors that do not organify IϪ. The first reports of this kind were Alternative NIS-Transported Radioisotopes published by Spitzweg and coworkers, who ob- In light of the relatively short biological retention tained excellent therapeutic results by stably trans- time of 131IϪ in NIS-expressing tumors, alternative fecting the human prostatic adenocarcinoma cell NIS-transported radioisotopes with superior decay line LNCaP with the NIS cDNA under the control properties and a shorter physical half-life than 131IϪ of the prostate-specific antigen promoter.31,32 NIS- (Table 1) must be considered as potentially better transfected NP-1 tumor xenografts in mice accu- therapeutic options. Two isotopes have been pro- mulated 25% to 30% of the total administered IϪ, posed for this purpose, ␤-emitter 188Rhenium showing a long retention time (45 hours) of the (188ReϪ) and ␣-emitter 211Astatine (211AtϪ) (Table radioisotope. After administration of a single 3 1). It has long been recognized by nuclear medicine mCi dose of 131I, significant tumor reduction was practitioners that due to their common ionic charac- Ϫ 99m 99m Ϫ achieved in NP-1 tumors in comparison with non- teristics, I and Tc-pertechnetate ( TcO4 )be- treated controls. It would have been informative to have similarly following intravenous administra- 34 99m Ϫ include biodistribution and dose-escalation studies tion. In humans, TcO4 localizes in the thyroid, in this report to clarify the investigators’ reason for salivary glands, gastric mucosa, and choroid plexus Ϫ 99m Ϫ using such a high dose. of the brain, just like I . TcO4 is concentrated The same group recently has recently reported but not organified in the thyroid gland and is used in 131IϪ therapy of NIS-transfected myeloma in SCID nuclear medicine as an alternative to Na131I for mice.33 They used self-inactivating lentivector assessing thyroid conditions. Rhenium (Re) is a with enhanced green fluorescent protein expression chemical analogue of technetium (Tc) and shows under the control of a minimal immunoglobulin practically identical chemical and biodistribution promoter to transduce hNIS into myeloma cells. properties to those of Tc.35 Ϫ Tumor xenografts in severe combined immunode- The perrhenate anion (ReO4 ) is concentrated in ficiency mice expressing hNIS were imaged with the thyroid and stomach by endogenous NIS because Ϫ Ϫ 123I and shown to retain 4% to 5% of I for up to of its chemical similarity to pertechnetate.36 Dada- 48 hours. Based on biodistribution results, the chova and coworkers have recently proposed the use therapeutic dose of 131IϪ was calculated to be 1 of 188Re for the treatment of NIS-expressing tu- mCi. A single administration of this dose com- mors.37 The 188Re is a powerful ␤-emitting radionu- pletely eradicated tumor xenografts without evi- clide with a 16.7-hour half-life, and it is conveniently dence of recurrence for up to 5 months after obtained from a 188W/188Re generator.38 The emis- therapy. Although myeloma is inherently a very sion characteristics and physical properties of 188Re radiosensitive malignancy, the success of this ther- are superior to those of 131I. The 188Re higher energy Ϫ ␤ ϭ apy in the absence of I organification is never- particles (Eav 0.764 MeV versus 0.134 MeV for THE Naϩ/IϪ SYMPORTER 29

131I) are effective over a higher range, sufficient to couraging. Carlin and coworkers showed that eradicate medium or large tumors by a “cross-fire” [211AtϪ]astatide uptake in the UVW human glioma effect,39 while its lower energy and low abundance cell line transfected to express NIS was NIS- gamma photons (155 keV, 15% abundance) are dependent, with characteristics similar to 131IϪ suitable for imaging yet easier to shield than the 364 uptake.44 keV photons of 131I. As a consequence of its emission Petrich and colleagues constructed a mammalian characteristics, the optimal tissue range for 188Re is NIS expression vector and generated 6 stable 188 23 to 32 mm, which means that Re can be used for NIS-expressing cancer cell lines, 3 derived from the treatment of relatively large solid tumors. The thyroid carcinoma, 2 from colon carcinoma, and 1 number of atoms of the radioisotope per gram of from glioblastoma.45 Compared with the respective tumor needed to produce a cure probability of 90% at control cell lines, steady-state radionuclide uptake ϫ 12 131 ϫ the optimal range is 6.38 10 for I and 5.34 in NIS-expressing cell lines increased up to 350- 11 188 10 for Re.39 These figures clearly show that fold for 123IϪ, 340-fold for 99mTc-pertechnetate, 188 Re would have the potential to deliver larger doses and 60-fold for 211AtϪ. Cellular 211AtϪ accumula- 131 Ϫ of radiation to tumors expressing NIS than I , tion was dependent on extracellular Naϩ and even if the uptake of 188Re-perrhenate were signifi- ϩ 131 Ϫ displayed a similar sensitivity towards Na per- cantly lower than that of I . Our dosimetry calcu- chlorate inhibition to that of both radioiodide and lations indicate that in the absence of organification, 99m Ϫ 188 TcO4 uptake. Heterologous competition with Re-perrhenate delivers a 4.5-fold higher dose to a Ϫ unlabeled NaI decreased NIS-mediated 211At up- 2-g tumor in a human than 131I. Administration of 188 take to the levels observed in control cells devoid Re-perrhenate is safe for the thyroid and stomach Ϫ of NIS. Following uptake, radioiodide and 211At in mice,40 and preliminary encouraging therapy re- were rapidly released (apparent half-life 3 to 15 sults have been reported in hNIS-transfected hepato- minutes) by the cells, as determined by wash out cellular carcinoma cells41 and in hNIS-expressing experiments. Data on scintigraphic tumor imaging F98 glioma tumors in animals.42 in a xenograft nude mouse model of transplanted The 211At is an ␣-emitting radiohalide, and its very high linear energy transfer of 98.8 keV/␮m NIS-modified thyroid cells indicate that radionu- clide uptake in NIS-expressing tumors was up to makes its relative biological efficiency close to 123 Ϫ 99m Ϫ 70 times I , 25 times TcO4 , and 10 times maximal. The drawbacks of this isotope are its 211 Ϫ scarce availability (ie, a cyclotron with a deutron At higher than in control tumors or normal beam is needed for 211At production and its short tissues, except stomach (3 to 5 times) and thyroid physical half-life of 7 hours precludes long-dis- gland (5 to 10 times). Compared with cell culture tance transportation) and nontrivial safety issues in experiments, the effective half-life in vivo was 123 Ϫ production and handling (211At is a gas). The first highly prolonged (ie, 6.5 hours for I , 5.2 hours 211 Ϫ report on the use of 211AtϪ as a substrate for NIS for At ). showed that although 211AtϪ was transported by an The remarkable progress made in the molecular IϪ accumulating mechanism, probably involving characterization of NIS in the last few years has NIS, neither ouabain nor perchlorate, both of clearly created new opportunities for the develop- which nearly abolish the transport of 125IϪ, were ment of diagnostic and therapeutic applications for able to fully suppress the basal-to-apical transfer of NIS.2,3,46-52 As discussed throughout this article, Ϫ 211At .43 It remains to be established whether this these applications are especially relevant in the realm is an effect of active transport or is merely due to of nuclear medicine, creating the prospect of using nonspecific adsorption of free 211AtϪ to a “sticky” administered radioisotopes in a highly targeted and apical cell surface. Later reports on the use of effective fashion to destroy a wide variety of cancer 211AtϪ for NIS-mediated therapy were more en- cells by internal, rather than external, radiation.

REFERENCES 1. Carrasco N: Iodide transport in the thyroid gland. Biochim 3. Doha«n O, De La Vieja A, Paroder V, et al: The sodium/ Biophys Acta 1154:65-82, 1993 iodide symporter (NIS): Characterization, regulation, and med- 2. De La Vieja A, Dohan O, Levy O, et al: Molecular ical significance. End Rev Endocr Rev 24:48-77, 2003 analysis of the sodium/iodide symporter: Impact on thyroid and 4. Mazaferri EL: Carcinoma of the follicular epithelium, in extrathyroid pathophysiology. Physiol Rev 80:1083-1105, 2000 Braverman LE, Utiger R (eds): The Thyroid: A Fundamental 30 DADACHOVA AND CARRASCO and Clinical Text (ed 8). Philadelphia, PA, Lippincott, 2000, pp 21. Sieger S, Jiang S, Schonsiegel F, et al: Tumour-specific 904-930 activation of the sodium/iodide symporter gene under control of 5. Dai G, Levy O, Carrasco N: Cloning and characterization the glucose transporter gene 1 promoter (GTI-1.3). Eur J Nucl of the thyroid iodide transporter. Nature 379:458-460, 1996 Med Mol Imaging 30:748-756, 2003 6. Levy O, Dai G, Riedel C, et al: Characterization of the 22. Haberkorn U, Kinscherf R, Kissel M, et al: Enhanced ϩ Ϫ thyroid Na /I symporter with an anti-COOH terminus anti- iodide transport after transfer of the human sodium iodide body. Proc Natl Acad SciUSA94:5568-5573, 1997 symporter gene is associated with lack of retention and low 7. Levy O, De la Vieja A, Ginter CS, et al: N-linked absorbed dose. Gene Ther 10:774-780, 2003 ϩ Ϫ glycosylation of the thyroid Na /I symporter (NIS). Implica- 23. Boland A, Magnon C, Filetti S, et al: Transposition of tions for its secondary structure model. J Biol Chem 273:22657- the thyroid iodide uptake and organification system in nonthy- 22663, 1998 roid tumor cells by adenoviral vector-mediated gene transfers. 8. Smanik PA, Liu Q, Furminger TL, et al: Cloning of the Thyroid 12:19-26, 2002 human sodium lodide symporter. Biochem Biophys Res Com- 24. Daniels GH, Haber DA: Will radioiodine be useful in mun 226:339-345, 1996 treatment of breast cancer? Nature Med 6:859-860, 2000 9. Everett LA, Belyantseva IA, Noben-Trauth K, et al: 25. Taurog AM: Hormone synthesis: Thyroid iodine metab- Targeted disruption of mouse Pds provides insight about the olism, in Braverman LE, Utiger RD (eds): The Thyroid: A inner-ear defects encountered in Pendred syndrome. Hum Mol Fundamental and Clinical Text (ed 8). Philadelphia, PA, Lip- Genet 10:153-161, 2001 pincott, Williams and Wilkins, 2000, pp 61-85 10. Rodriguez AM, Perron B, Lacroix L, et al: Identification 26. Maxon HR, Thomas SR, Hertzberg VS, et al: Relation and characterization of a putative human iodide transporter between effective radiation dose and outcome of radioiodine located at the apical membrane of thyrocytes. J Clin Endocrinol therapy for thyroid cancer. N Engl J Med 309:937-941, 1983 Metab 87:3500-3503, 2002 27. Huang M, Batra RK, Kogai T, et al: Ectopic expression 11. Tazebay UH, Wapnir IL, Levy O, et al: The mammary of the thyroperoxidase gene augments radioiodide uptake and gland iodide transporter is expressed during lactation and in retention mediated by the sodium iodide symporter in non-small breast cancer. Nat Med 6:871-878, 2000 cell lung cancer. Cancer Gene Ther 8:612-618, 2001 12. Kogai T, Schultz JJ, Johnson LS, et al: Retinoic acid 28. Smit JWA, Schro¬der-van der Elst JP, Karperien M, et al: induces sodium/iodide symporter gene expression and radioio- Iodide kinetics and experimental 131-I therapy in a xenotrans- dide uptake in the MCF-7 breast cancer cell line. Proc Natl planted human sodium-iodide symporterÐTransfected human Acad SciUSA97:8519-8524, 2000 follicular thyroid carcinoma cell line. J Clin Endocrin Metabol 13. Wapnir IL, van de Rijn M, Nowels K, et al: Immuno- 87:1247-1253, 2002 histochemical profile of the sodium/iodide symporter in thyroid, 29. Cho JY, Shen DH, Yang W, et al: In vivo imaging and breast, and other carcinomas using high density tissue microar- radioiodine therapy following sodium iodide symporter gene rays and conventional sections. J Clin Endocrinol Metab. transfer in animal model of intracerebral gliomas. Gene Ther 88:1880-1888, 2003 9:1139-1145, 2002 14. Moon DH, Lee SJ, Park KY, et al: Correlation between 30. Kogai T, Schultz J, Johnson LS, et al: Retinoic acid 99mTc-pertechnetate uptakes and expressions of human sodium induces sodium/iodide symporter gene expression and radioio- iodide symporter gene in breast tumor tissues. Nucl Med Biol 28:829-834, 2001 dide uptake in the MCF-7 breast cancer cell line. Proc Natl 15. Groot-Wassink T, Aboagye EO, Glaser M, et al: Ade- Acad SciUSA97:8519-8524, 2000 novirus biodistribution and noninvasive imaging of gene ex- 31. Spitzweg C, O’Connor MK, Bergert ER, et al: Treatment pression in vivo by positron emission tomography using human of prostate cancer by radioiodine therapy after tissue-specific sodium/iodide symporter as reporter gene. Hum Gene Ther expression of the sodium iodide symporter. Cancer Res 60: 13:1723-1735, 2002 6526-6530, 2000 16. Shimura H, Haraguchi K, Miyazaki A, et al: Iodide 32. Spitzweg C, Dietz AB, O’Connor MK, et al: In vivo uptake and experimental 131-I therapy in transplanted undiffer- sodium iodide symporter gene therapy of prostate cancer. Gene entiated thyroid cancer cells expressing the Naϩ/IϪ symporter Ther 8:1524-1531, 2001 gene. 138:4493-4496, 1997 33. Dingli D, Diaz RM, Bergert EK, et al: Genetically 17. Mandell RB, Mandell LZ, Link CJ Jr: Radioisotope targeted radiotherapy for multiple myeloma. Blood 102:489- concentrator gene therapy using the sodium/iodide symporter 496, 2003 gene. Cancer Res 59:661-668, 1999 34. Saha GP: Fundamentals of nuclear pharmacy. New 18. Boland A, Ricard M, Opolon P, et al: Adenovirus- York, NY, Springer, 1997, p 249 mediated transfer of the thyroid sodium/iodide symporter gene 35. Deutsch E, Libson K, Vanderheyden JL, et al: The into tumors for a targeted radiotherapy. Cancer Res 60:3484- chemistry of rhenium and technetium as related to the use of 3492, 2000 isotopes of these elements in therapeutic and diagnostic nuclear 19. Cho JY, Xing S, Liu X, et al: Expression and activity of medicine. Int J Rad Appl Instrum B 13:465-477, 1986 human Naϩ/IϪ symporter in human glioma cells by adenovirus- 36. Zuckier LS, Dadachova E, Li Y, et al: Comparative mediated gene delivery. Gene Ther 7:740-749, 2000 biodistribution of perrhenate, pertechnetate and iodide in NIS- 20. Nakamoto Y, Saga T, Misaki T, et al: Establishment and expressing and non-expressing tissues of mice. J Nucl Med characterization of a breast cancer cell line expressing Naϩ/IϪ 42:325, 2001 (abstr suppl) symporters for radioiodide concentrator gene therapy. J Nucl 37. Dadachova E, Bouzahzah B, Zuckier LS, et al: Rhenium- Med 41:1898-1904, 2000 188 as an alternative to iodine-131 for treatment of breast THE Naϩ/IϪ SYMPORTER 31 tumors expressing the sodium/iodide symporter (NIS). Nucl NIS-transfected human cancer cells. Nucl Med Biol 29:729- Med Biol 29:13-18, 2002 739, 2002 38. Knapp FF Jr: Rhenium-188ÐA generator-derived radio- 45. Petrich T, Helmeke HJ, Meyer GJ, et al: Establishment isotope for cancer therapy. Cancer Biother Radiopharm 13:337- of radioactive astatine and iodine uptake in cancer cell lines 349, 1998 expressing the human sodium/iodide symporter. Eur J Nucl 39. O’Donoghue JA, Bardie`s M, Wheldon TE: Relationship Med Mol Imaging 29:842-854, 2002 between tumor size and curability for uniformly targeted therapy 46. Schmutzler C, Ko¬hrle J: Implications of the molecular with beta-emitting radionuclides. J Nucl Med 36:1902-1909, 1995 characterizations on the sodium-iodide symporter (NIS). Exp 40. Dadachova E, Russell R, Zuckier LS: Assessment of Clin Endocrinol Diabetes 106:S1-S10, 1998 maximal tolerated dose of 188Re-perrhenate, a potential agent 47. Spitzweg C, Heufelder AE: The sodium iodide sym- porter: Its emerging relevance to clinical thyroidology. Eur J for treatment of NIS-expressing tumors, in a mouse model. Endocrinol 138:374-375, 1998 J Nucl Med 43:120, 2002 (abstr suppl) 48. Doha«n O, De La Vieja A, Carrasco N: Molecular study 41. Kang JH, Chung J-K, Lee YJ, et al: Application of of the sodium-iodide symporter (NIS): A new field in thyroid- radionuclide gene therapy to human hepatocellular carcinoma ology. Trends Endocrinol Metab 11:99-105, 2000 cell line using sodium/iodide symporter. J Nucl Med 44:336, 49. Riedel C, Dohan O, De la Vieja A, et al: Journey of the 2002 (abstr suppl) iodide transporter NIS: From its molecular identification to its 188 42. Shen DH, Marsee DK, Yang W, et al: Re-perrhenate clinical role in cancer. Trends Biochem Sci 26:490-496, 2001 treatment to enhance the survival of rats bearing intracerebral 50. Spitzweg C, Morris JC: Approaches to gene therapy with sodium-iodide symporter transduced gliomas. J Nucl Med sodium/iodide symporter. Exp Clin Endocrinol Diabetes 109: 44:35, 2003 (abstr suppl) 56-59, 2001 43. Lindencrona A, Nilsson M, Forssell-Aronsson E: Simi- 51. Spitzweg C, Harrington KJ, Pinke LA, et al: Clinical larities and differences between free 211-At and 125-I transport review 132: The sodium iodide symporter and its potential role in porcine thyroid epithelial cells cultured in bicameral cham- in cancer therapy. J Clin Endocrinol Metab 86:3327-3335, 2001 bers. Nucl Med Biol 28:41-50, 2001 52. Spitzweg C, Morris JC: The sodium iodide symporter: 44. Carlin S, Mairs RJ, Welsh P, et al: Sodium-iodide Its pathophysiological and therapeutic implications. Clin Endo- symporter (NIS)-mediated accumulation of [(211)At]astatide in crinol 57:559-574, 2002