Noradrenaline transporter gene transfer for radiation cell kill by 131 I meta-iodobenzylguanidine

Transcription

1 Gene Therapy (1999) 6, Stockton Press All rights reserved /99 $ Noradrenaline transporter gene transfer for radiation cell kill by 131 I meta-iodobenzylguanidine M Boyd 1, SH Cunningham 1, MM Brown 2, RJ Mairs 1,3 and TE Wheldon 1,4 1 Department of Radiation Oncology, Glasgow University, CRC Beatson Laboratories; 2 Department of Urology, Gartnavel General Hospital, Glasgow; 3 Department of Child Health, Royal Hospital for Sick Children, Yorkhill Hospital, Glasgow; and 4 Department of Clinical Physics, West of Glasgow Hospitals University Trust, Western Infirmary, Glasgow, UK Meta-iodobenzylguanidine conjugated to 131 I-iodine is an methods: (1) survival of clonogens derived from monolayer effective agent for the targeted radiotherapy of tumors of culture; (2) survival of clonogens derived from disaggreneural crest origin which express the noradrenaline trans- gated multicellular spheroids; and (3) spheroid growth porter (NAT). The therapeutic application of 131 I MIBG is delay. 131 I MIBG was twice as toxic to cells in spheroids presently limited to the treatment of phaeochromocytoma, compared with those in monolayers, consistent with a neuroblastoma, carcinoid and medullary thyroid carci- greater effect of radiation cross-fire (radiological bystander noma. To determine the feasibility of MIBG targeting for a effect) from 131 I -radiation in the three-dimensional tumor wider range of tumor types, we employed plasmid- spheroids. The highest concentration of 131 I MIBG tested mediated transfer of the NAT gene into a human glioblas- (1 MBq/ml) was nontoxic to UVW control cells or spheroids toma cell line (UVW) which does not express the NAT transfected with the NAT gene in reverse orientation. gene. This resulted in a 15-fold increase in uptake of MIBG These findings are encouraging for the development of by the host cells. A dose-dependent toxicity of 131 I MIBG NAT gene transfer-mediated 131 I MIBG therapy. to the transfectants was demonstrated using three Keywords: gene therapy; targeted radiotherapy; noradrenaline transporter; 131 I MIBG; radiation cross-fire; spheroids Introduction Meta-iodobenzylguanidine (MIBG) is an analogue of the adrenergic neurone blockers guanethidine and bretylium. 1 It is actively taken up by the noradrenaline transporter (NAT) 2 into cells of the adrenal medulla, adrenergic nerve cells and tumors such as phaechromocytoma and neuroblastoma which are derived from the neural crest. 3 Radioiodinated MIBG can be used for imaging 4,5 as well as targeted radiotherapy. 6,7 Early studies of patients with recurrent or refractory neuroblastoma treated with 131 I MIBG have been encouraging, 8 but it appears unlikely that 131 I MIBG used alone will be curative. MIBG is a small molecule which penetrates multicellular tumor spheroids and probably tumors more readily than antibodies. 9 An attractive feature of tumor targeting with 131 I MIBG is that cancer cells which fail to accumulate a lethal quantity of 131 I MIBG may still absorb -radiation from neighboring, targeted cells. 10 Although 131 I MIBG is currently one of the best available agents for targeted radiotherapy, its use is confined to a few NAT-expressing tumor types which selectively accumulate the radiopharmaceutical. In order to determine whether 131 I MIBG could be used to treat diverse tumor types for which no specific, targetable characteristic currently exists, we have used a recombinant prep9 Correspondence: M Boyd, Department of Radiation Oncology, Glasgow University, CRC Beatson Laboratories, Glasgow G61 1BD, UK Received 15 July 1998; accepted 15 January 1999 plasmid to transfect the bovine NAT cdna, 11 under the control of the RSV promoter/enhancer, into a human, glioblastoma cell line (UVW) which does not express the NAT gene. For this initial in vitro study, the prep9 plasmid was employed. This is an Epstein Barr virus (EBV)- based vector which contains an EBV origin of replication (orip) and nuclear antigen (EBNA-1) allowing high-copy episomal replication and expression in primate cell lines. Transfectants exhibited a 15-fold enhancement of 131 I MIBG uptake and NAT-transfected clonogens succumbed to 131 I MIBG administration in a dose-dependent manner. The superior toxicity of 131 I MIBG to cells grown as spheroids rather than as monolayers confirmed the efficacy of the -radiation cross-fire effect. These results demonstrate the potential of gene therapy-assisted, MIBG-targeted radiotherapy for the treatment of nonneuroectodermal tumors. Results MIBG uptake To assess whether the bovine NAT cdna was expressed and a functional protein produced in the transfected UVW cells, 131 I MIBG uptake was measured in UVW cells transfected with recombinant plasmids containing the bovine NAT cdna (bnat) in the sense and reverse orientation (pnat+ and pnat, respectively). The cells transfected with the cdna in the reverse orientation (UVW/pNAT ) were used throughout as a negative control to determine whether differences observed in the

2 1148 UVW/pNAT+ cell line were due to the production of a functional NAT transporter protein rather than the prep9 plasmid or the transfection procedure. UVW/pNAT+, UVW/pNAT and nontransfected UVW cells (pnat0) were seeded at cells per well in sixwell plates and after 2 days the specific uptake of 131 I MIBG in each cell line was measured against desmethylimipramine (DMI) inhibited controls. DMI is a potent antagonist of noradrenaline transport. Optimal cell densities and methodologies for MIBG uptake studies were as previously determined. 10 Figure 1 shows the 131 I MIBG active uptake capacity of the two transfected cell lines and the nontransfected UVW cell line. Both untransfected UVW cells and UVW cells transfected with the pnat plasmid (reverse orientation) showed negligible active accumulation of 131 I MIBG. In contrast, UVW/pNAT+ cells, which contained the recombinant plasmid with the bnat cdna in the sense orientation, showed a 15-fold enhancement of 131 I MIBG uptake relative to the same cells treated with the specific moamine uptake inhibitor, DMI. These findings indicated that the enhanced 131 I MIBG uptake was a result of the expression of the bnat cdna in the correct orientation. Cell kill Having established a procedure for NAT gene expression which resulted in active 131 I MIBG uptake, we sought to evaluate the dose dependency of cell kill. These experiments were performed in cellular monolayers and also in a three-dimensional spheroid model. Colony formation after 131 I MIBG exposure of monolayer cultures: To establish the dose dependency of cell kill in monolayers resulting from 131 I MIBG uptake, UVW/pNAT+, UVW/pNAT and untransfected UVW cells were seeded into 25 cm 2 flasks and after 48 h a range of concentrations of 131 I MIBG was added. After 2 h, the cells were washed to remove free 131 I MIBG, seeded into 60 mm Petri dishes in triplicate and incubated for approximately 14 days or until discrete colonies had formed. The colonies were then stained and counted and the surviving fraction at each dose was calculated (Figure 2a). No clonogenic cell kill was observed in the nontransfected UVW cell line or in the UVW/pNAT cell line, indicating that neither the prep9 plasmid (lacking the bnat insert in the correct orientation), nor the transfection procedure was toxic. However, in the UVW/pNAT+ cell line, which had substantial 131 I MIBG uptake capacity (Figure 1), a dose-dependent reduction in surviving fraction was observed (Figure 2a). This demonstrated a clear Figure 1 Uptake of 131 I MIBG in transfected and nontransfected UVW cells. UVW represents untransfected UVW cells, UVW/pNAT+ represents UVW cells transfected with the prep9 plasmid containing the bovine NAT cdna in the sense orientation and UVW/pNAT represents UVW cells transfected with the prep9 plasmid containing the bovine NAT cdna in the reverse orientation. DMI is desmethylimipramine which inhibits the specific uptake of 131 I MIBG by the noradrenaline transporter. Figure 2 Clonogenic survival curves derived by colony formation for cells and spheroids exposed to graded doses of 131 MIBG in vitro. (a) Survival curves for UVW/PNAT0 (monolayer cells (untransfected), UVW/PNAT cells (UVW cells transfected by NAT gene in reverse orientation) and UVW/PNAT+ cells (UVW cells transfected by NAT gene in sense orientation)). The surviving fraction (SF) data are plotted as log mean values with error bars showing log standard deviations. The log scale (to base 10) is such that a value of 1 corresponds to a 10% surviving fraction. Only the PNAT+ cells showed a significant correlation with dose. (b) Survival curves for cells derived from disaggregated spheroids (UVW cells transfected with NAT gene in reverse or sense orientation; UVW/PNAT and UVW PNAT+, respectively). The surviving fraction (SF) data are plotted as log mean values with error bars showing log standard deviations. The log scale is such that a value of 1 corresponds to a 10% surviving fraction. Only the UVW/PNAT+ cells showed a significant correlation with dose.

3 correlation between 131 I MIBG uptake and cell kill in NAT-transfected UVW glioma cells Colony formation after 131 I MIBG exposure of multicellular spheroids: Tumor spheroids are cellular aggregates which are useful in vitro models of micrometastases. Their three-dimensional structure provides greater opportunity for radiation cross-fire than two-dimensional cellular monolayers. Spheroids were grown as described and after 5 days when they were approximately 250 m in diameter, aliquots were removed and incubated in universal containers with various concentrations of I 131 MIBG. To allow the absorption of -decay energy by neighboring nontransfected cells, the following procedure was employed. Spheroids were left intact during the 2 h incubation with the radiopharmaceutical and during the washing procedure which removed free 131 I MIBG. The spheroids were then placed in fresh universal containers, equilibriated with 5% CO 2 and returned intact to a rotating mixer at 37 C to maintain the integrity of individual spheroids. After 24 h the spheroids were disaggregated by agitation after addition of trypsin and the cells from each universal container were seeded at in triplicate into 60 mm Petri dishes. The cultures were incubated at 37 C for 14 days or until the formation of large discrete colonies was observed. The surviving colonies were then stained, counted and analyzed as described for cellular monolayers. A dose-dependent reduction in surviving fraction was observed for cells from disaggregated spheroids comprised of UVW/pNAT+ cells compared with UVW/pNAT spheroids. The surviving fraction was almost half of that obtained for monolayers at the maximum dose of 131 I MIBG (Figure 2b). Once again, no cell kill was observed in the UVW/pNAT cell line. The difference between the UVW/pNAT+ monolayer and spheroid cell survival curves is consistent with a contribution of dosedependent radiation cross-fire to cell kill in this threedimensional culture. Growth delay after 131 I MIBG exposure of multicellular spheroids: Spheroid response to cytotoxic treatment can also be evaluated by the end-point of growth delay. This provides an assay of cytoxic effect which is independent of colony formation. In these experiments, spheroids were initially grown in stirrer flasks to m diameter. Groups of spheroids were exposed to a range of activities of 131 I MIBG for 2 h, then washed and transferred individually on to 24-well plates. Individual spheroid growth was measured using a microscope coupled to an image analysis system which allowed serial monitoring of spheroid volume and the construction of growth curves based on median volume for each group. Figure 3 shows log-linear plots of relative spheroid volume (normalized to median volume at time zero) as a function of time after treatment for (1) UVW/pNAT and (2) UVW/pNAT+ spheroids. ForUVW/pNAT (spheroids transfected with bnat cdna in the reverse orientation), the growth curves of all spheroid groups were similar, with no correlation between growth pattern and dose group (Figure 3a). By contrast, Figure 3b shows that for UVW/pNAT+ spheroids (sense orientation of NAT cdna) there was a distinct spread of growth curves: spheroids in the higher dose groups showed more delayed growth. This provides further evidence that Figure 3 Growth curves for untreated and 131 I MIBG-treated multicellular spheroids derived from: (a) UVW cells transfected with bnat in reverse orientation and (b) UVW cells transfected with bnat in sense orientation. Data points represent the logarithm of median relative volume V/V0 (normalised to volume V0 at time zero) and error bars correspond to median deviations. In (a) only the largest error bars are shown, for reasons of clarity, for the control groups. The growth curves are not well separated indicating little effect of dose. In (b) error bars are shown at all time-points for the control and highest dose groups. The separation of the growth curves of treated from control spheroids increases with dose with non-overlapping error bars (indicating significant separation) between the control and higher dose groups. This implies a retardation of spheroid growth as a function of dose only when the NAT gene is transfected in the sense orientation. bnat cdna transfection (in the sense but not the reverse orientation) conferred dose-dependent sensitivity to 131 I MIBG. Discussion We have shown that transfection of the bovine NAT cdna into human glioma cells induced the expression of a functional transporter which facilitated the active uptake of 131 I MIBG. This resulted in a dose-dependent toxicity to the host cells, demonstrating for the first time the potential application of 131 I MIBG therapy for tumors other than those derived from the neural crest. Untransfected cells and cells transfected with the NAT cdna in the reverse orientation exhibited only nonspecific, low level uptake of MIBG. Previous studies of human NAT gene transfection by

4 1150 plasmid-mediated transfer, employed COS-1 cells 12 and HeLa cells 13 as hosts. These achieved three- and nine-fold enhancement of MIBG uptake, respectively. In the present study, by introduction of the bovine NAT cdna, 11 we increased the host cells capacity for active accumulation of MIBG by at least a factor of 15 compared with transporter-inhibited controls. Although the experimental results reported here are preliminary, and clinical applications some way ahead, it is of interest to consider what might be the clinical potential of this novel approach. The 15-fold enhancement of MIBG uptake, reported here for NAT gene transfected tumor cells, if reproduced in a clinical setting, corresponds to a tumor uptake of radiolabeled MIBG of 0.021%/g injected activity (assuming 70 kg standard man with only passive uptake of MIBG in normal tissues). This is within the range reported for MIBG uptake of neuroblastomas in the clinic, and is superior to the uptakes typically achieved with radiolabeled monoclonal antibodies, 14 although less than the uptake of radio-iodide (commonly about 0.1%/g) observed for thyroid carcinomas. 15 This suggests that MIBG uptake levels capable of being clinically useful have been achieved in these first experiments. It is possible that even greater concentration of MIBG could be achieved in vitro by manipulation of factors which enhance expression of the transgene, enhance stability of its mrna or facilitate integration of the protein into the cell membrane. Our procedures were designed to evaluate the enhancement of MIBG therapy by gene manipulation using a simple in vitro model. This incorporated a forced expression of the NAT gene by neomycin selection. While such a method of driving transgene expression is nonphysiological, it may none the less be possible to employ our recombinant plasmid vector for initial assessment of this therapeutic strategy using in vivo model systems. This is due to the fact that after the removal from cell cultures of the selection agent, the transfectants capacity for active uptake of MIBG remained constant over a period of 6 weeks (results not shown). The most popular gene manipulation strategy applied to experimental cancer treatment attempts to confer drug sensitivity on malignant host cells. 16 Because the gene transfer process is inefficient, a bystander effect (eg diffusion of cytotoxic drug) is an essential component of this approach. Here, we report the first evidence of the feasibility of combining gene transfection with radionuclide therapy using MIBG as the molecular vehicle. Other investigators have employed gene transfer methods to induce tumor cells to express elevated levels of surface antigens 17,18 and receptors 19,20 to enhance targeting by radiolabeled antibodies and peptides, respectively. An alternative approach could involve transfection of the sodium iodide symporter combined with administration of Na 13 I. 21 This symporter facilitates the accumulation of iodide by the thyroid gland 20- to 40-fold with respect to the iodide concentration in plasma. 22 A particular advantage of strategies which seek to achieve the concentration in tumor cells of radionuclides with long range emissions, is the presence of a radiological bystander effect. That is the bombardment of nontransfectants by radioactive decay particles emanating from neighboring, successfully transfected cells which have been induced to accumulate radiopharmaceutical actively. This type of collateral cell kill has previously been demonstrated using the tumor spheroid culture system. 10,23 The three-dimensional spheroid model was employed in the current study to determine experimentally the effectiveness of cross-fire radiation. 131 I MIBG was almost twice as toxic to spheroids compared with monolayers, suggesting that a substantial contribution to efficacy in the former system was due to the radiological bystander effect. Spheroids composed of a range of proportions of transfectants to nontransfectants will constitute a useful model system for the assessment of different types of bystander effect and to determine the minimum proportion of cells required to be transfected in spheroids, to result in total spheroid cell kill. These studies, which have direct relevance to the eradication of micrometastatic disease, are currently underway. A further advantage of the strategy of NAT gene transfer with administration of radiolabeled MIBG, is the scope for flexibility of design of the targeting agent. For example, an analogue of MIBG ( 211 At MABG) has been synthesized by replacing the -emitting radionuclide 131 I- iodine with the -emitting halogen 211 At-astatine. This compound has similar uptake specificity to MIBG but much greater radiobiological effectiveness of cell killing. 24 A promising application of these tactics to glioma therapy, for example, could include virally mediated transfer of the NAT gene, under the control of a tumourspecific promoter (such as the E2F-1 promoter 25 ), followed by intra-cranial infusion of the -emitting agent 211 At MABG. A suitable vector for such transfection may be a replication-compromised, variant herpes simplex virus whose safety has been demonstrated in a recent, Glasgow, phase 1, glioma therapy trial. 26 Future studies will examine the possibility of applying NAT gene transfer with administration of 131 I MIBG or 211 At MABG to the experimental therapy of various tumor types. This may indicate whether MIBG could eventually be used to treat the 15% of neuroblastoma patients whose tumors fail to express the NAT. It may also be feasible to enhance MIBG uptake in neuroblastomas which actively accumulate MIBG, thereby improving the liklihood of cure. This application will require the control of transgene expression by a neuroblastomaspecific sequence such as the tyrosine hydroxylase promoter Materials and methods Plasmids The bovine NAT cdna inserted into the EcoRI site of the eukaryotic expression vector psg-5 (Stratagene, Cambridge, UK), was kindly provided by Dr Michael Bruss and Professor Heinz Bonisch (University of Bonn, Germany). Owing to the lack of selection markers in this plasmid and the presence of a SV40 promoter, which expresses genes optimally in cells expressing the large T antigen, the 3.2 kb bovine NAT cdna (bnat) was initially subcloned into the EcoR1 site of the pind vector (Invitrogen, De Schelp, The Netherlands). As only one restriction enzyme was used for subcloning into the pind plasmid, the bnat cdna was present in different clones, in both sense and antisense orientations. The latter plasmid has suitable restriction enzyme sites for subcloning the bnat cdna into the KpnI/Xho sites of the

5 prep9 episomal expression vector (Invitrogen), which was chosen for this gene transfer study. Restriction mapping of the different clones identified prep9/nat recombinant plasmids with the insert in both orientations (results not shown). Subcloning into the pind and prep9 plasmids was carried out by standard methods. Plasmid purifications were carried out using Plasmid Maxi Kit (Qiagen, Germany). Target cells The transgene host was a radiation-resistant human glioma cell line (UVW). The UVW cell line was chosen for this study in preference to other available glioma cell lines because of its capacity for spheroid formation in stirrer culture. The UVW cell line was maintained in Eagle s minimum essential medium (MEM) (Life Technologies, Paisley, UK) with 10% (v/v) fetal bovine serum, penicillin/streptomycin (100 U/ml), fungizone (2 g/ml) and glutamine (200 mm). The cells were incubated at 37 C in an atmosphere of 5% CO 2. UVW transfection Fifty per cent confluent UVW monolayers seeded at cells per well in six-well plates 48 h before use, were transfected with 3 g plasmid DNA (prep9/bnat in both orientations) using DOTAP Liposomal Transfection Reagent (Boehringer Mannheim, Mannheim, Germany) in accordance with the manufacturer s instructions. After 24 h, geneticin G-418 sulphate (Life Technologies) was added at a concentration of 0.5 mg/ml to select for transfected cells. The transfectants were maintained as described for UVW cells with the addition of geneticin at each passage. Transfectants were assessed monthly for mycoplasma infection, which was consistently absent, and for MIBG uptake (persistent expression of the bnat gene). MIBG uptake studies UVW/pNAT+ and UVW/pNAT cells were seeded in six-well plates at an initial density of cells per well and cultured for 48 h. MIBG incorporation was measured by incubating the cells for 2 h with 7 kbq of 131 I MIBG of specific activity MBq/mg, (Dupont Radiopharmaceuticals, Hertfordshire, UK). Nonspecific uptake was measured in the presence of 1.5 mm desmethylimipramine (DMI) (Sigma-Aldrich, Dorset, UK). After incubation, medium was removed, the cells were washed with PBS and radioactivity was extracted using two aliquots of 10% (w/v) trichloroacetic acid. The activities of the extracts were then measured in a gamma-well counter. Uptake was expressed as c.p.m. per 10 5 cells. Clonogenic assays UVW/pNAT+ and UVW/pNAT cells were seeded in 25 cm 2 flasks at cells per flask. After 2 days, medium was removed and replaced with fresh medium alone or medium containing 131 I MIBG at 100, 500, 750 or 1000 kbq/ml. After incubation for 2 h, medium was removed and the cells were washed three times with PBS. Cells were then trypsinized and counted. For each radioactivity concentration, three 35 mm Petri dishes (Nunc Plastics, Oslo, Denmark) were seeded with cells. The cultures were then incubated at 37 C. After days, medium was removed and the colonies were fixed and stained with Carbol Fuchsin (R A Lamb, Middlesex, UK). Colonies of more than 50 cells were counted using an automated colony counter (Synoptics, Cambridge, UK). Spheroid experiments To investigate the contribution of -particle cross-fire to cell kill, a three-dimensional multicellular spheroid model was used. Spheroids were grown by continuously stirring UVW/pNAT+ or UVW/pNAT cells in Techne stirrer flasks (Techne, Cambridge, UK). Spheroids of approximately m diameter were obtained after 4 5 days growth. Aliquots were transferred to 20 ml universal containers and suspended in 1 ml of MEM containing the appropriate activity of 131 I MIBG. After incubation for 2 h at 37 C, the medium was removed and the spheroids washed three times with PBS. Two types of spheroid experiment were undertaken with different endpoints spheroid clonogenic assay and spheroid growth delay assay. Spheroid clonogenic assay: Fresh culture medium was added to the washed spheroids which were then incubated at 37 C for 24 h. To prevent clumping, the spheroids were shaken continuously. They were then incubated for 10 min at 37 C with PBS containing 0.25% (v/v) trypsin and 1 mm EDTA and mechanically disaggregated using a syringe. Microscopic examination confirmed that the cell preparations were free from clumps. Cells were then counted and seeded into 60 mm Petri dishes for clonogenic assay, as described above. Spheroid growth delay: Spheroids were transferred to Petri dishes and individually transferred into agar-coated wells containing 1 ml of MEM medium. One 24-well plate was used per treatment. The spheroids were incubated at 37 C in a 5% CO 2 atmosphere. Growth of the spheroids was monitored over 2 3 weeks by measurement of their cross-sectional area, using a semiautomated image analysis system coupled via a television camera to an inverted optical microscope. From these measurements, the median volume of the spheroids was calculated to allow the construction of spheroid regrowth curves. Acknowledgements We would like to thank Professor Heinz Bonisch and Dr Michael Bruss, University of Bonn, for the kind gift of the bovine noradrenaline transporter cdna. This project was funded in part by the Department of Health, the Neuroblastoma Society, the British Urological Foundation, The European Community and the Cancer Research Campaign. Thanks also to Professor D Kirk, Department of Urology, Gartnavel Hospital and Professor A Barrett, Beatson Oncology Centre, Western Infirmary, for their support. References 1 Wieland DM et al. Radiolabelled adrenergic neuron-blocking agents: adrenomedullary imaging with [ 131 I]iodobenzylguanidine. J Nucl Med 1980; 21: Jacques S Jr et al. Comparison of the sodium dependence of uptake of metaiodo-benzylguanidine and norephrine into cultured bovine adrenomedullary cells. Mol Pharmacol 1984; 26:

6 Troncone L, Rufini V. I-131-MIBG therapy of neural crest tumours (review). Anticancer Res 1997; 17: Feine U, Muller-Schauenburg W, Treuner J, Klingebiel T. Metaiodobenzylguanidine (MIBG) labelled with 123 I/ 123 I in neuroblastoma diagnosis and follow up treatment, with a review of the diagnostic results of the international workshop of paediatric oncology held in Rome, September Med Pediatr Oncol 1987; 15: Shapiro B et al. Iodine- 131 I metaiodobenzylguanidine for the locating of suspected phaeochromocytoma: experience in 400 cases. J Nucl Med 1985; 26: Konings JE et al. Diagnosis and treatment of malignant phaeochromocytoma with [ 131 I]MIBG. Radiother Oncol 1990; 17: Voute PA et al. Results of treatment with [ 131 I]metaiodobenzylguanidine in patients with neuroblastoma. Future prospects of zetotherapy. In: Evans AE, D Angio GJ, Knudson AG, Seeger RC (eds). Advances in Neuroblastoma Research, 3rd edn. Wiley- Liss: New York, 1984, pp Lashford LS et al. Phase I/II study of iodine-131-metaiodobenzylguanidine in chemoresistant neuroblastoma: a United Kingdom Childrens Cancer Study Group investigation. J Clin Oncol 1992; 10: Mairs RJ et al. The distribution of alternative agents for targeted radiotherapy within human neuroblastoma spheroids. Br J Cancer 1991; 63: Cunningham SH et al. Radiotoxicity to neuroblastoma cells and spheroids of beta-, alpha- and Auger electron-emitting conjugates of benzylguanidine. Br J Cancer 1998; 77: Lingen B, Bruss M, Bonisch H. Cloning and expression of the bovine sodium-and chlorine-dependent noradrenaline transporter. FEBS Lett 1994; 342: Pacholczyk T, Blakely RD, Amara SG. Expression cloning of a cocaine- and anti-depressant-sensitive human noradrenaline transporter. Nature 1991; 350: Glowniak JV et al. Evaluation of metaiodobenzylguanidine uptake by the noradrenaline, dopamine and serotonin transporters. J Nucl Med 1993; 34: O Donoghue JA et al. The implications of uptake of 131 -metaiodobonzylguanidine (MIBG) for the targeted radiotherapy of neuroblastoma. Br J Radiol 1991; 64: Murray T, Hilditch TE. Therapeutic radionuclides. In: C Sampson (ed). Textbook of Radiopharmacy: Theory and Practice. Chapman and Hall: London, 1999 (in press). 16 Roth JA, Cristiano RJ. Gene therapy for cancer: what have we done and where are we going? J Natl Cancer Inst 1997; 89: Buchsbaum DJ et al. Approaches to enhance cancer radiotherapy employing gene-transfer methods. Gene Therapy 1996; 3: Raben D et al. Enhancement of radiolabeled antibody-binding and tumor-localization through adenoviral transduction of the human carcinoembryonic antigen gene. Gene Therapy 1996; 3: Rogers BE et al. Tumor localization of a radiolabeled bombesin analogue in mice bearing human ovarian tumors induced to express the gastrin-releasing peptide receptor by an adenoviral vector. Cancer 1997; 80: Rogers BE et al. Localization of iodine-125-mip-des-met(14) bombesin (7-13)NH2 in ovarian carcinoma induced to express the gastrin releasing peptide receptor by adenoviral vectormediated gene transfer. J Nucl Med 1997; 38: Shimura H et al. Iodide uptake and experimental 131I thaerapy in transplanted undifferentiated thyroid cancer cells expressing the Na+/I symporter gene. Endocrinology 1997; 138: Taurog A. Hormone synthesis: thyroid iodide metabolism. In: LE Braverman, Utiger RD (eds). The Thyroid: A Fundimental and Clinical Text. Lippincott-Raven: New York, 1996, pp Neshasteh-Riz A et al. The implications of proliferative heterogeneity for the design of DNA-targeted radiotherapy: differential cytotoxicity of alternative radioiodoanalogues of IUdR to human glioma cells in monolayer or spheroid culture. Br J Cancer 1998; 77: Strickland DK, Vaidyanathan G, Zalutsky MR. Cytotoxicity of alpha-particle-emitting m-[at-211] astatobenzylguanidine on human neuroblastoma-cells. Cancer Res 1994; 54: Parr MJ et al. Tumor-selective transgene expression in vivo mediated by an E2F-responsive adenoviral vector. Nature Med 1997; 3: Rampling R, Cruickshank G, MacLean A, Brown M. Therapeutic replication-competent herpes virus. Nature Med 1998; 4: Jin BK et al. Prolonged in vivo gene expression driven by a tyrosine hydroxylase promoter in a defective herpes simplex virus amplicon vector. Hum Gene Ther 1996; 7: Robert JJ, Geoffroy MC, Finiels F, Mallet J. An adenoviral vectorbased system to study neuronal gene expression: analysis of the rat tyrosine hydroxylase promoter in cultured neurons. J Neurochem 1997; 68: Song S et al. An HSV-1 vector containing the rat tyrosine hydroxylase promoter enhances both long-term and cell typespecific expression in the midbrain. J Neurochem 1997; 68: