Vorinostat Increases Expression of Functional Norepinephrine Transporter in Neuroblastoma In Vitro and In Vivo Model Systems

Neuroblastoma (NB) is the most common extracranial solid tumor of childhood and is often diagnosed after metastasis has commenced (1). Originating from primitive sympathetic ganglion cells, NB can arise at any site in the sympathetic nervous system, with the adrenal gland being the most common site of origin. About 50% of all NB patients are classified as at “high risk,” with survival rates frequently being less than 40% with existing therapies (2). Patients with high-risk disease are at significant risk for resistance to initial therapy and have a high incidence of relapse subsequent to successful chemo- or radiotherapy (3). The unfavorable outcome for patients with relapsed or refractory disease creates an enormous need for novel therapeutic strategies with particularly high therapeutic indices for these patients.

As a sympathetic-derived tumor, NB cells typically possess capabilities of monoamine uptake, decarboxylation pathways for monoamine synthesis, and monoamine secretion (4). Of particular diagnostic and therapeutic value is the expression of the norepinephrine transporter (NET) (5). NET serves as an important uptake protein for the norepinephrine analog, meta-iodobenzylguanidine (MIBG), an agent used for the diagnosis and treatment of NB. The use of low-dose radiolabeled MIBG (typically ¹²³I-MIBG) enables imaging to identify the sites and extent of neuroblastoma spread (6, 7). The use of high-dose radiolabeled MIBG (¹³¹I-MIBG) enables targeted radiotherapy for patients with neuroblastoma. Single agent studies have shown that ¹³¹I-MIBG is one of the most active agents for the treatment of patients with relapsed or refractory neuroblastoma (8).

The use of ¹³¹I-MIBG in NB is predicated on the concept that the radionuclide will selectively target NET expressing tumors and metastases and will not harm normal tissues, particularly those not expressing NET. A strategy built upon this concept would fundamentally carry the potential to deliver the high therapeutic index needed for successful NB treatment. The aims of the current study consisted of (i) mapping of NET in various tissues, (ii) chemical induction of NET expression, and (iii) study of ¹³¹I-MIBG uptake into tissues under the influence of a chemical NET expression modulator. The working hypothesis was that a chemical entity that results in enhanced expression of NET in NB cells would increase the concentration of ¹³¹I-MIBG in those sites and lead to preferential radiotherapy. Tissues that do not express NET could be expected to resist damage by ¹³¹I-MIBG.

One of the possible avenues of influencing the expression levels of NET was the inhibition of histone deacetylation by an agent like vorinostat (suberoylanilide hydroxamic acid, SAHA). Histone deacetylase (HDAC) inhibition in turn could be expected to cause generalized perturbation of cellular protein expression, possibly affecting NET expression. In this paper, we describe the results of proof-of-concept in vitro and in vivo experiments in human NB cell lines and NB xenografts. The results support our hypothesis, and have enabled us to obtain approval for the initiation of clinical trials that examine a combination therapy of vorinostat and ¹³¹I-MIBG for the treatment of neuroblastoma (NANT N2007-03).

Radiolabeled ¹²³I-MIBG was purchased from GE Healthcare. Radiolabeled ³H-norepinephrine was purchased from Perkin Elmer Life Sciences. All other chemicals were purchased from Sigma. Vorinostat was synthesized in our laboratory using a combination of previously published methods (9, 10). All the experiments described herein were also conducted utilizing a clinical grade sample of vorinostat obtained from Merck for comparison purposes. Antibodies for Western blotting were obtained from the following sources: rat polyclonal anti-NET antibody from Alpha Diagnostic International (NET11-A), horseradish peroxidase (HRP)-conjugated anti-rabbit (sc-2004; Santa Cruz Biotechnology) immunoglobulin was used as the secondary antibody and mouse monoclonal anti-β-actin antibody from Santa Cruz Biotechnology. Lipofectamine 2000 and hygromycin B were from Invitrogen. The cell culture media DMEM H-21, RPMI 1640, and FBS were obtained from the Cell Culture Facility of the University of California at San Francisco (San Francisco, CA).

Cell lines stably transfected with human NET, HEK-hNET, and empty vector, HEK-EV, were established by transfection of pcDNA5/FRT vector (Invitrogen) containing the full-length hNET cDNA, and pcDNA5/FRT vector alone, respectively, into human embryonic kidney 293 (HEK293) Flp-In cells using LipofectAMINE 2000 (Invitrogen) as per manufacturer's instructions. The stable clones were selected with 75 μg/mL hygromycin B. All the NB cell lines (LAN5, SHEP, SH-SY-5Y, Kelly, SK-N-SH, SK-N-MC, NB1691, and NB1691-luc) used in the present study were obtained from the UCSF Cell Culture Facility.

Stably transfected HEK293 cells were maintained in DMEM H-21 medium supplemented with 10% FBS, 100 units/mL penicillin and 100 units/mL streptomycin, and 75 μg/mL hygromycin B. The culture medium for all the NB cell lines was RPMI 1640 containing 15% FBS, 1% modified Eagle media (MEM) nonessential amino acids, 100 units/mL penicillin, and 100 μg/mL streptomycin. All cell lines were grown at 37°C in a humidified atmosphere with 5% CO₂/95% air.

HEK-hNET cells were seeded in 24 well poly-d-lysine coated cell culture plates (BD Biosciences) in standard cell culture media and the uptake studies were conducted 24 hours after cell seeding. Cells were incubated in PBS buffer containing 15 nmol/L ³H-NE in the presence or absence of a NET specific inhibitor, desipramine (10 μmol/L), at 37°C for 3 minutes. After 3 minutes, uptake was terminated by rapid removal of the buffer containing the compounds and 3 quick washings with ice-cold PBS buffer. The cells were lysed with 1 mL of 0.1% SDS/0.1N NaOH solution for about 1 hour and the intracellular levels of radioactive NE were determined using liquid scintillation counting. Intracellular concentrations of norepinephrine were normalized to the total protein in each well using bicinchoninic acid (BCA) protein assay (Pierce).

The same procedure was followed for determination of ³H-NE or ¹²³I-MIBG uptake rate in NB cell lines, SH-SY-5Y and Kelly. In this case, cells were exposed to increasing concentrations of vorinostat ranging from 0 to 5 μmol/L for 24 hours, followed by exposure to ³H-NE or ¹²³I-MIBG in PBS buffer, at 37°C for 3 minutes. The intracellular levels of radioactivity corresponding to ³H-NE were determined as described earlier and were normalized to the total protein content in each well using BCA protein assay. The total concentration of ¹²³I-MIBG was measured in an automated Wallac 1480 Wizard gamma counter using the program RiaCalc Wiz (Wallac Oy).

Total RNA was isolated from each of the NB cell line on 6-well plates using the Invitrogen Micro-to-Midi Total RNA Purification System according to the manufacturer's protocol. Extracted RNA was stored at −80°C until use. Similarly, total RNA was isolated from 13 NB tumor samples. Samples of NB tumors were obtained from the UCSF Pediatric Solid Tumor Tissue Bank following procedures approved by the UCSF Committee on Human Research. Four samples were collected at initial diagnosis, 8 samples were collected after the initiation of therapy, and the timing of sample collection was unknown for 1 sample. Five of the tumor samples were from initial diagnosis (no prior treatment). All of the other samples were from patients who had received some type of therapy.

Reverse transcriptase (RT) PCR of RNA samples was carried out with Superscript III (Invitrogen) using oligo(dT)₂₀ primers. Two microliters of RT reaction product was used for subsequent PCR (Taq DNA Polymerase, Invitrogen) consisting of 35 cycles with the following parameters: 94°C for 30 seconds, 60°C for 45 seconds, 72°C for 1 minute, followed by a final extension of 72°C for 10 minutes and storage at 4°C. Primers were designed to amplify a unique sequence of human NET (SLC6A2), each spanning intron–exon boundaries to ensure no genomic DNA was amplified. The PCR primers that were used for human: NET (GenBank accession number NM_001043) forward: ATTCTTCAAAGGCGTTGGCTAT and reverse: CAGGACACCACGCTCATAAA were designed to amplify a 262-bp fragment. Analysis of each PCR sample was then performed on 2% agarose gels containing 0.5 μg/mL ethidium bromide. Gels were visualized using a digital camera and image processing system (Kodak). RT-PCR analysis for glyceraldehyde 3-phosphate dehydrogenase (GAPDH) gene was used as an internal control for each sample.

For quantitative real-time (qRT) PCR analysis, 2 μg of total RNA isolated from NB cells and tumor samples were reverse transcribed using High Capacity cDNA Reverse Transcription Kit (Applied Biosytems) in a 20-μL volume reaction according to the manufacturer's protocol. The resulting cDNA was used as template for qRT-PCR using TaqMan Gene Expression Assays for human transporter NET (Assay ID: Hs00426573_m1*) and human GAPDH (Assay ID: Hs99999905_m1). qRT-PCR reaction was carried out in 96-well reaction plates in a volume of 10 μL using the TaqMan Fast Universal Master Mix (Applied Biosystems). Reactions were run on the Applied Biosystems 7500 Fast Real-Time PCR System with the following profile: 95°C for 20 seconds followed by 40 cycles of 95°C for 3 seconds and 60°C for 30 seconds. The relative expression of each mRNA was calculated by the comparative method (ΔΔCt method). Firstly, the ΔCt value for each sample was obtained by subtracting the Ct value of GAPDH mRNA from the Ct value of the NET mRNA. Then the ΔΔCt value for each sample was obtained by subtracting the ΔCt value of another endogenous control, PGK1, from the ΔCt value of each sample. The mRNA expression level of NET was reported as the percentage of the expression of PGK1 calculated using the arithmetic formula 2^−ΔΔCt (ABI PRISM 7700 Sequence Detection system User Bulletin No. 2, P/N 4303859) multiplied by 100.

The total protein extracts from transfected HEK293, NB cells and tumor samples were prepared using CellLytic M cell lysis buffer (Sigma) containing protease inhibitor cocktail at 4°C for 10 minutes. The cell homogenate was spun for 10 minutes at 14,000 rpm at 4°C and the protein concentration in the supernatant was determined by the BCA protein assay. Approximately 50 μg of the supernatant was resolved by SDS-PAGE and transferred onto polyvinylidene difluoride membrane (Bio-Rad). The blots were incubated for 1 hour in blocking buffer containing 5% nonfat dry milk in Tris-buffered saline (TBS) and then incubated overnight at 4°C with anti-human NET1 antibody diluted (1:750) in 2% milk in TBS buffer containing 0.1% Tween 20 (TBS-T). After washing with TBS-T, the blots were incubated for 1 hour at room temperature with the goat anti-rabbit IgG HRP-conjugated secondary antibody (1:2,500 dilution) in TBS-T buffer containing 5% milk. The blots were again washed thrice with TBS-T for 10 minutes and then developed with the ECL Western blotting detection reagent (GE Healthcare). All blots were reprobed for β-actin (1:10,000) as a loading control. Band intensities of scanned blots were quantified using ImageJ. The integrated intensity of a fixed area was measured, and background levels were subtracted.

The cytotoxicity of MIBG was measured by standard MTT assays in 96-well plates using NB cell lines. After seeding the cells at the desired density and incubating overnight, vehicle control [5% DMSO (dimethyl sulfoxide) in PBS] or vorinostat (5 μmol/L) was added to the culture medium. After 24 hours of treatment, the drug-containing medium was replaced with medium containing MIBG at various concentrations and the incubation was continued for additional 72 hours. At the end of the incubation, 20 μL of MTT stock solution (5 mg/mL) was added to each well. After additional incubation for 3 hours at 37°C, the MTT reaction medium was discarded and the purple-blue MTT formazan crystals were dissolved by the addition of 100 μL of 0.1N HCI in isopropanol. The optical density, which is a measure of the mitochondrial function of the viable cells, was read directly with a microplate reader (model versamax; Molecular Devices Co.) at 580 nm and a reference wavelength of 680 nm. Concentration response graphs were generated for each drug using GraphPad Prism software (GraphPad Software, Inc.). These graphs were analyzed using a curve fit for sigmoid dose–response, and IC₅₀ values were derived. Results are expressed as mean IC₅₀ values with the standard error of the mean.

Female CrTac:NCr-Foxn1nu athymic mice (4–5 weeks old) were purchased from Taconic and housed under aseptic conditions, which included filtered air, sterilized food, water, bedding, and cages. The experiments on mice were approved by the Institutional Animal Care and Use Committee of University of California at San Francisco. All experimental and animal handling procedures were in accordance with national ethical guidelines.

To study the effect of vorinostat treatment on NET expression in vivo, nude mice harboring xenografts of human NB1691 neuroblastoma cells expressing luciferase (NB1691-luc) were used. For preparation of NB xenografts, NB1691-luc cells were resuspended in serum free RPMI1640 media and 100 μL of this mixture containing 6 × 10⁶ cells were injected in mice by intravenous tail vein injection. Serial tumor measurements were obtained by weekly bioluminescence imaging after an intraperitoneal injection of d-luciferin (Xenogen) at 15 mg/mL in sterile PBS. Twenty-one days after injection of tumor cells, the animals were divided into vehicle control (5% DMSO in saline) and vorinostat-treated (150 mg/kg i.p.) groups (2 per group). Tumors were then harvested from each animal at various time points (1, 3, 6, and 18 hours) after initiation of treatment and lysed with a buffer containing (20 mmol/L Tris–HCl pH 7.5–8.0, 150 mmol/L NaCl, 0.5% sodium deoxycolate, 1% Triton X-100, 0.1 SDS, 1 mmol/L EDTA, 1 mmol/L phenylmethylsulphonyl fluoride, and protease inhibitor cocktail). qRT-PCR and Western blot analysis of these tissues was carried out as described under the Quantitative real-time PCR and Western blot analysis sections.

In order to study the effect of vorinostat administration on tissue expression of NET in healthy mice not bearing tumors, nude mice (5 per group) were injected with vorinostat (150 mg/kg i.p.) and various tissues were harvested 6 and 18 hours after drug treatment. Each of these tissues was then subjected to total RNA extraction and qRT-PCR analysis as described before.

Animals with NB xenografts (described above) were divided into 2 groups (3 per group) in the following manner: (i) control group: injected with vehicle (5% DMSO in saline i.p.) 6 hours prior to ¹²³I-MIBG (300 μCi/mouse) by intravenous tail vein injection; (ii) treatment group: injected with vorinostat (150 mg/kg i.p.) 6 hours before intravenous tail vein injection of ¹²³I-MIBG. Two hours after the treatment, the mice were euthanized. Samples of liver, kidney, spleen, heart, and tumor were excised, weighed, and their associated radioactivity was measured in an automated Wallac 1480 Wizard gamma counter using the program RiaCalc Wiz (Wallac Oy). Blood was also sampled and radioactivity content was normalized to its weight. The uptake of ¹²³I-MIBG was expressed as the fraction of the injected dose per gram of tissue. A correction was made for radioactive decay, which would take place since the time of injection.

Data were analyzed statistically by unpaired or paired Student t tests, as appropriate. Probability values lower than 0.05 were considered statistically significant.

Being an analog of monoamine neurotransmitter norepinephrine, the transport of MIBG by NET is well documented (11). We confirmed the uptake of MIBG by NET in our in vitro model using HEK293 cells stably transfected with human NET transporter (HEK-hNET). Due to its ease of handling and shorter half life, we used ¹²³I-labeled MIBG instead of ¹³¹I-labeled MIBG in our in vitro and in vivo studies (12). The cellular accumulation of ¹²³I-MIBG in HEK-hNET cells [0.484 ± 0.033 μCi/(mg protein min)] was 13.4-fold higher than in HEK-EV cells [0.036 ± 0.003 μCi/(mg protein min); P < 0.0001; Supporting Information; Supplementary Fig. S1]. Incubation of ¹²³I-MIBG with a NET inhibitor (desipramine, 10 μmol/L) significantly reduced the cellular accumulation of MIBG in hNET expressing cells [0.042 ± 0.0033 μCi/(mg protein min); P < 0.0001], with no effect in HEK-EV cells [0.030 ± 0.0026 μCi/(mg protein min); P > 0.05].

Originating from the sympathetic nervous system, neuroblastomas retain the ability of neuronal cells to uptake NE through the neuronal NE transporter. Hence the baseline expression level of NET transcripts in 7 NB cell lines and 13 NB tumor samples was examined by qRT-PCR analysis. All the NB cell lines tested showed detectable and variable expression of NET mRNA. The highest expression of NET was observed in LAN5, SK-N-SH, and NB1691 cell lines and was in the range of 23% to 45% of the expression of PGK1 (3-phosphoglycerate kinase), an abundantly expressed endogenous control (Fig. 1A). NET mRNA was detected in 12 out of 13 NB tumor samples across a range of expression levels. NET transcripts were expressed at more than 40% of the endogenous control PGK1 in 9 out of the 13 tumor samples (Fig. 1B). The comparison of NET transcripts with the endogenous control signifies the high abundance and significance of NET protein to NB cell types.

To determine the effect of vorinostat treatment on the expression of NET, we cultured 2 NB cell lines in the presence of either DMSO (vehicle control) or vorinostat (0.05, 0.10, 0.25, 0.50, 0.75, 1.00, 2.5, and 5.00 μmol/L) for 24 hours. The cell lines selected for this purpose (SH-SY-5Y and Kelly) were based on their moderate NET expression and to allow evaluation of MYCN amplified (Kelly) and MYCN nonamplified (SH-SY-5Y) status of these cell lines. The doses of vorinostat were based on clinically relevant plasma concentrations (13, 14). RT-PCR analysis of mRNA isolated from vorinostat-treated cells demonstrated qualitatively a dose-dependent increase in the expression of NET transcripts in both of the cell lines (Fig. 2A).

Using qRT-PCR analysis the increase in NET expression after vorinostat treatment observed earlier was quantified (Fig. 2B). Even the lowest concentration of vorinostat (0.025 μmol/L) resulted in a 2-fold increase in NET expression in Kelly cells over the nontreated cells and a maximum increase of 6.7-fold was achieved at 5 μmol/L dose of vorinostat. A dose-dependent increase in NET mRNA expression was also observed in SH-SY-5Y cells after vorinostat treatment with maximum attainable increase of 5.6-fold after exposure to 5 μmol/L concentration of vorinostat. The increase in the expression of NET mRNA observed by RT-PCR and qRT-PCR data, however, does not necessarily prove the presence of functional NET transporter protein. These studies were therefore supplemented with Western blotting and functional characterization of elevated NET protein after vorinostat treatment in NB cell lines.

Increases in NET mRNA expression in NB cells after vorinostat treatment were examined for correlations with NET protein levels. NET protein concentrations in Kelly and SH-SY-5Y cells exposed to 1 and 5 μmol/L vorinostat for 24 hours were analyzed by Western blot (Fig. 2C). The increase in NET mRNA levels observed in both the NB cell lines was in agreement with the increase in NET protein synthesis. Kelly and SH-SY-5Y cells treated with vorinostat (5 μmol/L) showed approximately 5.3- and 6.9-fold increases in NET protein compared to untreated cells, respectively.

We next performed a functional characterization of NET following vorinostat treatment. The cellular accumulation of a protypical NET substrate, norepinephrine (NE), was determined in the presence and absence of a range of vorinostat concentrations in 2 NB cell lines (Fig. 3A). The cellular accumulation rate of ³H-NE (15 nmol/L) in NB cells treated for 24 hours at the highest vorinostat concentration (5 μmol/L) was 2.5- to 3.4-fold higher than the corresponding untreated control cells. Specifically, the uptake of ³H-NE in Kelly cells in the presence and absence of vorinostat (5 μmol/L) treatment was 0.252 ± 0.017 nmol/(mg protein.min) and 0.009 ± 0.006 nmol/(mg protein.min), respectively (P < 0.001). Similarly, the highest concentration of vorinostat resulted in an accumulation rate of 0.178 ± 0.011 nmol/(mg protein.min) for ³H-NE in SH-SY-5Y cells, which was significantly greater than 0.052 ± 0.004 nmol/(mg protein.min) in untreated cells (P < 0.001). Both cell lines exhibited a dose-dependent increase in NE uptake across the range of evaluated vorinostat concentrations (0–5 μmol/L).

Similarly, vorinostat pretreatment (0–5 μmol/L) caused a dose-dependent increase in ¹²³I-MIBG uptake in the NB cells. In the absence of vorinostat treatment, the intracellular concentration of ¹²³I-MIBG after its exposure for 3 minutes to Kelly cells was 0.025 ± 0.001 μCi/(mg protein.min). This intracellular concentration of ¹²³I-MIBG increased 3.9-fold after pretreatment with 5 μmol/L vorinostat [0.097 ± 0.018 μCi/(mg protein.min); P < 0.001]. Similarly, a 2.6-fold enhancement in ¹²³I-MIBG uptake was observed in SH-SY-5Y cells pretreated with vorinostat [5 μmol/L; 0.062 ± 0.003 μCi/(mg protein.min)] compared to untreated cells [0.024 ± 0.001 μCi/(mg protein.min); P < 0.001; Fig. 3B]. Furthermore, the increase in cellular uptake of NE and MIBG observed after vorinostat pretreatment was inhibited by desipramine (10 μmol/L), a NET inhibitor (data not shown). When these uptake experiments were performed under various preincubation periods or vorinostat (2, 4, 6, 12, and 24 hours), the maximum changes in the uptake were observed at the 6-hour time point with a negligible further increase at 24 hours (Supporting Information; Supplementary Fig. S4), in agreement with maximum increase in NET mRNA and protein observed at that time point (Supporting Information; Supplementary Fig. S3). The vorinostat pretreatment period of 24 hours was selected to capture maximal changes in the uptake of NET substrates and for simplicity of experimentation.

To obtain a correlation between vorinostat-induced increases in MIBG uptake in NB cell lines via NET upregulation and the functional consequences of this enhanced uptake on cell growth, we examined the cytotoxicity of MIBG in the presence and absence of vorinostat. The results of standard MTT assays indicate that MIBG in combination with vorinostat is prominently more toxic against all NB cell lines tested when compared to MIBG alone (Table 1). The IC₅₀ value of MIBG in the presence of vorinostat is 1.88 to 3.57 (depending on the NB cell line) order of magnitude lower than its IC₅₀ as a single agent. Changes in cell proliferation following exposure to 5 μmol/L vorinostat were less than 10% for 36-hour exposure periods and less than 20% for 48 hours.

The next aim for this study was to obtain in vivo proof-of-principle of the observed vorinostat-induced NET upregulation and its functional consequences. Neuroblastoma xenografts created by intravenous tail vein injection of NB1691-luc cells were used for these experiments. After allowing tumors to develop for 21 days, mice were injected with 150 mg/kg of vorinostat or vehicle control intraperitoneally. Tumors harvested at various time points after vorinostat administration were analyzed by Western blotting (Fig. 4A). A 2.1- and 2.5-fold increase in NET protein level compared to the untreated control mice was noticeable at the 1- and 3-hour time points, respectively, in the vorinostat-treated mice. The tumors at 6- and 18-hour time points exhibited 3.8- and 4.2-fold increases in NET protein with vorinostat treatment over the untreated control mice, respectively (Supporting Information; Supplementary Fig. S5A). Similarly, the corresponding changes in NET mRNA expression levels were observed in these samples (Supporting Information; Supplementary Fig. S5B). The rapid tumor progression observed in this metastatic mouse model and negligible differences in NET expression enhancement prompted us to use the 6-hour time point for biodistribution studies. No change in NET protein expression level was observed in mice treated with the vehicle control over the entire period of the experiment.

The effect of vorinostat on NET transcription was studied in different tissues of healthy mice at 6 and 18 hours following intraperitoneal vorinostat administration (Fig. 4B). The greatest relative increases in NET mRNA expression levels were in the brain (7.82-fold over untreated control) and in the liver (7.69-fold over untreated control) after 6 hours. At the 18-hour time point, increases in NET mRNA levels were also observed in the adrenal gland and skin (3.64- and 2.25-fold over untreated controls, respectively).

The functional effect of increased NET protein level after vorinostat treatment was next evaluated in terms of changes in MIBG in vivo biodistribution. Figure 5 presents the biodistribution of ¹²³I-MIBG 2 hours after intravenous administration in nude mice bearing NB1691-luc human neuroblastoma xenografts. Half of these xenografts received pretreatment with intraperitoneal vorinostat (150 mg/kg) 6 hours before MIBG administration. Differences in tissue accumulation due to vorinostat administration reaching statistical significance were observed only in the liver and in tumors. Without vorinostat treatment, the concentration of MIBG in the liver was 0.025 ± 0.003 μCi/(mg tissue-dose injected) whereas with vorinostat pretreatment, the total concentration of MIBG in the liver increased to 0.049 ± 0.006 μCi/(mg tissue-dose injected) (P < 0.05; Fig. 5). The greatest increases in MIBG accumulation due to vorinostat treatment were observed in tumors, which were 2.8-fold higher than the corresponding control group [with vorinostat pretreatment: 0.062 ± 0.011 μCi/(mg tissue-dose injected) vs. untreated: 0.022 ± 0.003 μCi/(mg tissue-dose injected); P < 0.05; Fig. 5]. Although there was a trend toward increase in MIBG concentration after vorinostat pretreatment in other tissues (e.g., kidney, spleen, heart), the difference in concentrations between the vorinostat-treated and -untreated group did not reach statistical significance. A trend was also observed toward reduced MIBG accumulation in the blood in the vorinostat pretreated group, but the difference did not reach statistical significance [with vorinostat pretreatment: 0.004 ± 0.001 μCi/(mg tissue-dose injected) vs. no vorinostat pretreatment: 0.007 ± 0.001 μCi/(mg tissue-dose injected); P > 0.05; Fig. 5]. Because the compound does not cross the BBB, MIBG brain accumulation was below the detection limit of the assay (15).

We have demonstrated that exposure to vorinostat at clinically relevant concentrations enhances NET mRNA and protein expression in neuroblastoma cell lines and tumor xenografts. This enhanced NET expression translates into increases in MIBG uptake into model systems, suggesting that the newly synthesized NET protein is functional. While increased MIBG uptake was observed in nontumor tissues, the greatest vorinostat-induced increase in uptake was in neuroblastoma tumors. This suggests a possible favorable therapeutic index for ¹³¹I-MIBG when administered together with vorinostat. We hypothesize that combination of vorinostat and ¹³¹I-MIBG will augment intracellular accumulation of ¹³¹I-MIBG in patients with neuroblastoma. This hypothesis forms the basis for an ongoing phase 1 clinical trial of this combination in children and young adults with relapsed or refractory neuroblastoma (NANT N2007-03).

Results from our laboratory and others clearly demonstrate that NET is expressed at higher than endogenous levels in almost all NB tumors evaluated (Fig. 1). MIBG, a metabolically stable NE analog, accumulates in neuroendocrine tissues via NET (Supporting Information; Supplementary Fig. S1) and MIBG labeled with either ¹³¹I or ¹²³I is used for the scintigraphic visualization (diagnosis) and therapy of pheochromocytomas and neuroblastomas (16–18). The polar guanidinium group borne by MIBG precludes its movement across the BBB into the brain (15), making MIBG highly appropriate for our aim of achieving a high therapeutic index.

The experiments in this study were designed to evaluate the hypothesis that chemically induced upregulation of NET would enhance partitioning of MIBG into NB tumors. The chemical induction of NET expression was inhibition of HDAC through administration of vorinostat. RT-PCR, qRT-PCR, and Western blot analysis of NB cells treated with increasing concentrations of vorinostat revealed dose-dependent increases in NET transcription and translation (Fig. 2). The mechanism through which vorinostat upregulates NET expression remains unclear. Vorinostat chelates an essential Zn²⁺ ion, deactivating a host of HDAC isoenzymes. HDAC inhibition causes chromatin opening, making it more accessible to transcription factors and thereby leading to increased expression of cellular proteins. However, due to the intricate secondary interrelationships between cellular transcription factors and cellular proteins, the net result may not necessarily be upregulation of a particular protein. For example, HDAC inhibition may also promote increased acetylation of transcription factors, thereby increasing their activity, e.g., in the case of the tumor suppressor p53, HDAC inhibition increases its activity, which in turn may lead to cell cycle arrest and even apoptosis (19). The pathway through which vorinostat upregulates NET may therefore involve several levels of interplay between transcription factors and chromatin opening. Vorinostat is already approved for the treatment of cutaneous T cell lymphoma and has shown moderate activity as a single agent against neuroblastoma cell lines in preclinical studies (20). Furthermore a large body of preclinical data also indicates that HDAC inhibitors including vorinostat can act as radiosensitizers across a range of human cancers (21–24). Combinations of ¹³¹I-MIBG with radiosensitizing agents like topotecan (25) and irinotecan (NANT2004-06) have been successfully evaluated and shown to be tolerable in clinical trials. These results prompted us to evaluate vorinostat/¹³¹I-MIBG combination for NB. Vorinostat pretreatment indeed caused increased accumulation of MIBG in NB cells (Fig. 3). The cytotoxicity of MIBG also was vastly enhanced when NB cells were pretreated with vorinostat as opposed to MIBG alone (Table 1). These data conclusively link the increased cellular accumulation of MIBG in the presence of vorinostat with a corresponding increase in its toxicity.

NB tumors are biologically heterogeneous, and may in theory respond dissimilarly to vorinostat exposure. A clinically utilized indicator of NB prognosis is MYCN gene amplification and is an indication of advanced disease (26, 27). In order to address some of this heterogeneity, we evaluated 2 cell lines that differ in their MYCN status: the Kelly cell line is MYCN amplified while the SH-SY-5Y cell line is not. Throughout the experiments, vorinostat-induced NET upregulation was independent of the MYCN status of the cell line. It should, however, be noted that HDAC inhibitors such as valproic acid have been previously shown to downregulate MYCN (28). Indeed, vorinostat was found to downregulate MYCN (data not shown) suggesting that any underlying correlation between NET upregulation and MYCN status may be masked by the effect of vorinostat on MYCN.

The mapping of vorinostat-induced NET overexpression in vivo demonstrated increase in NET mRNA transcription in the brain, liver, adrenal gland, and skin for a healthy mouse (Fig 4B). This effect was prominent even 6 hours after vorinostat administration. Similar increases in NET protein levels was observed in NB tumors at 6-hour time-point (Fig. 4A) which correlated with the significantly increased MIBG uptake in the tumors of vorinostat-treated mice over untreated controls (Fig. 5). Although MIBG uptake also increased in the livers of vorinostat-treated mice, the animal model used in our study contained high degrees of tumor metastasis. The increased MIBG uptake in the liver may possibly have occurred due to the contribution of NB cells metastatic to the liver. ¹³¹I-MIBG is associated with a very low-incidence of hepatic toxicity (29). The current findings nevertheless raise the possibility of increased hepatic toxicity when ¹³¹I-MIBG is combined with vorinostat. Clinical trials of the combination will need to monitor for this possibility.

The main dose limiting toxicity of ¹³¹I-MIBG is hematologic toxicity. We observed a trend toward decreased plasma concentrations of MIBG in mice treated with vorinostat (Fig. 5). While we were not able to evaluate bone marrow levels of MIBG, it is possible that the combination of vorinostat and ¹³¹I-MIBG may result in less hematologic toxicity through 2 mechanisms. First, if the combination results in increased ¹³¹I-MIBG into tumor cells, then a lower dose of ¹³¹I-MIBG may be as efficacious as ¹³¹I-MIBG given as a single agent at a higher dose. In addition, if bone marrow uptake of ¹³¹I-MIBG is reduced by vorinostat treatment though preferential channeling of ¹³¹I-MIBG into tumors, then hematologic toxicity may also be reduced.

In spite of many combination therapies tried in the treatment of neuroblastoma, the goal of a very high therapeutic index with sparing of the normal and the developing tissues remains elusive. The strategy designed and developed in this study is perhaps one of the few that are derived specifically from addressing each of the unique aspects of neuroblastoma that differentiate them from other tumors. While the enhancement of MIBG uptake by vorinostat comes about from a biochemical change induced by vorinostat, the sparing of CNS results from biodistribution aspects of MIBG. Through selective partitioning of MIBG into NB tumors by vorinostat administration, it may be possible to achieve desired efficacy from a lower dose of MIBG, potentially mitigating the problematic hematologic toxicity often observed with MIBG therapy.

No potential conflicts of interest were disclosed.

This work was supported by NIH grants GM61390, GM36780, PO1 81403 (K.K. Matthay and W.A. Weiss), R01CA102321 (W.A. Weiss) and KL2 RR024130; by a grant (17RT-0126) from the Tobacco-Related Disease Research Program (TRDRP); and by the Alex Lemonade, Dougherty and the Campini Foundation.

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