Cell lines and culture conditions

The neuroblastoma cell lines SK-N-AS and SH-SY5Y were maintained in RPMI-1640; and NB-1691 was maintained in DMEM, both supplemented with 10% FBS, 1% penicillin/streptomycin (P/S) and 1% L-glutamine (all from Quality Biological Inc., Gaithersburg, MD) at 37°C. To ensure consistency, a batch of cells was expanded, aliquoted and stored in liquid nitrogen prior to the screening. In each experiment, a vial of cells was defrosted and passaged 1:4 when 70% confluency was reached. Cells between passages 3 and 7 were used for all experiments.

Reagents

Topotecan hydrocholoride (Hycamtin; GlaxoSmithKline, Philadelphia, PA) and Bortezomib (Velcade; Millenium Pharmaceuticals, Cambridge, MA) were reconstituted and stored according to the manufacturers' instructions. NSC 676914 was obtained from the Developmental Therapeutics Program, Division of Cancer Treatment and Diagnostics, NCI/NIH.

High throughput siRNA screening

A set of synthetic siRNAs targeting 418 genes related to the apoptotic pathway (Qiagen Apoptosis Set V.1; Qiagen, Valencia, CA), with 2 siRNAs of different sequences per gene, was used for the first screen. For the second screen, 2 new siRNA pre-designed sequences were used (Qiagen). In the third confirmatory screen, one siRNA from each of the previous two screens was chosen. siRNAs were transfected at passage 4. Briefly, transfection reagent Dharmafect 1 (Dharmacon RNA Technologies, Lafayette, CO) was diluted in DCCR reagent medium at a ratio of 1:208 in volume. siRNA (20 nM) and 25 μL of the diluted transfection reagent were added to an individual well in a 96-well plate for complex formation with incubation for 20 min at RT. SK-N-AS cells were trypsinized, counted and resuspended in P/S free culture medium. 5000 cells were added to each individual well in 100 μL medium. The plate was incubated at RT for 30 min for cell attachment before being placed at 37°C for 24 h. Topotecan was then added to each well for additional 72 h incubation. Cell proliferation assay was performed at 96 h post siRNA transfection. The IC50 of topotecan for SK-N-AS cells at 24 h was 2 μM. In the 1st and 2nd screens, topotecan doses of 0, 1, 5 and 10 μM were used, whereas in the 3 rd screen, lower drug doses of topotecan at 0, 0.01, 0.1 and 1 μM were used to identify synergy. The criterion of hit selection for the enhancer genes was ≤0.8 fold cell growth compared to its own siRNA effect in the presence of topotecan. The final enhancer gene list was subjected to pathway analysis for identification of overrepresented genes within a target pathway. An inhibitor to the pathway was chosen and tested individually or in combination with topotecan in vitro and in vivo.

Cell proliferation assay

Cell proliferation was measured using Cell Titer Glo proliferation assay (Promega Corporation, Madison, WI) or a real-time cell sensing system (RT-CES; ACEA Biosciences, Inc. San Diego, CA) according to the manufacturer's instruction. We performed cell proliferation assays at a volume of 35 μL per well for 10 min and was measured at 562 nm on a Tecan plate reader (Tecan Inc., Durham, NC). Each treatment was performed in triplicates, averaged and normalized using untreated cells. For real-time cell electronic sensing assays, cells treated with drugs alone or in combinations or transfected with siRNA were added in triplicates to a 96-well plate device compatible with the real-time cell electronic sensing analyzer (RT-CES; ACEA Biosciences, Inc. San Diego, CA). Cell growth was monitored hourly for indicated durations via calculation of "cell index" (normalized impedance) for each well.

Quantitative RT-PCR

SK-N-AS cells were collected at 60 h post siRNA transfection. Total RNA was reverse transcribed using SuperScript II reverse transcriptase system (Life Technologies, Foster City, CA). The target regions were then pre-amplified using a standard PCR for 10 cycles. The pre-amplified cDNA was quantified by Taqman gene expression assay (Life Technologies) using Fluidigm digital array (South San Francisco, CA) according to the manufacturer's protocol. Fold expression was calculated using a comparative threshold cycle method (2^{-Δ ΔCT}) [16].

Pathway analysis

The pathway analysis was performed using MetaCore (http://www.genego.com/metacore.php, GeneGo Inc., St Joseph, MI). MetaCore is an integrated software suite for functional analysis of experimental data and it contains curated protein interaction networks on the basis of manually curated database of human protein-protein, protein-DNA, protein-RNA and protein-compound interactions. Metacore uses a hypergeometric model to determine the significance of enrichment. The enhancer genes from our experiment and the genes in the GeneGo maps from the MetaCore database were used to identify the enriched GeneGo pathway maps.

Gene Set Enrichment Analysis

To investigate gene set enrichment, GSEA (http://www.broad.mit.edu/gsea/) was performed for genes, ranked by log₂ ratio of gene expressions between topotecan-treated (1 μM and 10 μM) and untreated control SK-N-AS cells, using a weighted Kolmogorov-Smirnov-like statistics [17]. The gene set of NF-κB target genes used in GSEA were downloaded from http://bioinfo.lifl.fr/NF-KB.

Western blotting

Cells were seeded on a 100 mm² culture dish for overnight and were treated with topotecan alone or with bortezomib. For topotecan alone effect, SK-N-AS were treated with 5 μM of topotecan for 0, 3, 6, or 24 h at 37°C. For the combined effect of bortezomib and topotecan, cells were treated with bortezomib at 0, 1 and 10 nM for 24 h, followed by the addition of topotecan at 5 μM for 6 h. Total cell lysate was collected with RIPA buffer containing 1% phosphatase inhibitor and 1% protease inhibitor (all from Thermo Fisher Scientific, Rockford, IL). Protein concentration was measured by BCA Protein Assay Kit (Thermo Fisher Scientific, Waltham, MA) and protein lysates (20 ug per lane) were resolved on 4-12% TRIS-gradient gel (Invitrogen Life Technologies, Carlsbad, CA) and were transferred to nitrocellulose membranes by iBlot blotting system (Invitrogen Life Technologies). The membranes were blocked with 5% non-fat dry milk in PBS with 0.1% Tween 20 (PBST) for 1 h at RT, followed by incubation with mouse monoclonal antibodies against total or phosphorylated IκB-α (Ser32/36), rabbit monoclonal antibodies against total or phosphorylated p65/RelA or rabbit monoclonal antibody against NFKB1 (all from Cell Signaling Technology, Danvers, MA) at 4°C overnight. Peroxidase-conjugated goat anti-mouse or anti-rabbit antibody was used as secondary antibodies (Santa Cruz Biotechnology, Santa Cruz, CA) for 1 h incubation at RT. Immunoreactive bands were visualized by ECL analysis system (GE Healthcare, Piscataway, NJ) and enhanced chemiluminescence. For loading control, the membranes were washed with Restore Western blot stripping buffer (Thermo Fisher Scientific) and probed with goat anti-human actin HRP conjugated antibody (Santa Cruz Biotechnology).

In vitro drug combination

Neuroblastoma cells were trypsinized, counted and resuspended in P/S free culture medium. 5000 cells per well in 100 μL medium were seeded in 96-well white plates for overnight. Topotecan and NSC 676914 or bortezomib were added individually or in combination at various doses and the plates were incubated at 37°C for 24 h and 48 h respectively. Cell proliferation assay was performed as described above. Combination index was calculated using CompuSyn software (ComboSyn Inc., Paramus, NJ). Briefly, the combination index theorem was used to quantify synergy or antagonism for two drugs by the formula C.I. = (D)₁/(D_x)₁ + (D)₂/(D_x)₂, where D1 and D2 are drug 1 and drug 2, and × is growth inhibition by X% [18].

Apoptosis assay

The caspase-3 activity was measured using the PE Active Caspase-3 Apoptosis kit (BD Pharmingen, San Diego, CA). Briefly, SK-N-AS cells (untreated or treated with topotecan and bortezomib alone or in combination for 24 h) were trypsinized, fixed, and stained with PE rabbit anti-active caspase-3 antibody. Fluorescence intensity was measured by FACS Calibur and data were analyzed using CellQuest software (BD Biosciences, Franklin Lakes, NJ).

In vivo xenograft model

All animal experiments have been reviewed and approved by the NIH Animal Care and User Committees. A minimal residual disease xenograft model in mice bearing neuroblastoma was established in 8-10 week-old female SCID Beige mice (Charles River Laboratories, Fredrick, MD). Briefly, five million SK-N-AS cells expressing luciferase (gift from Dr. Bryan Clary, Duke University Medical Center) were injected intravenously via the lateral tail vein into the mice. Tumors were allowed to grow for 7 d, and then mice were randomly assigned to cohorts treated with topotecan and bortezomib administered individually or in combination, or with saline solution (control mice). Bortezomib (0.6 mg/kg) was injected intraperitoneally three times a week for two weeks, rested for two weeks and repeated with another course of treatment. Topotecan (0.5 mg/kg) was injected intraperitoneally five times a week for two weeks, rested for two weeks, followed by another course of treatment. Body weight and general wellness of the mice were monitored, and tumor size was monitored by Xenogen IVIS 100 imaging system (Caliper Life Sciences Inc., Hopkinton, MA). The in vivo xenograft experiment was repeated, and results from two independent experiments were combined (n = 43).

Statistical analysis

Non-parametric Mann-Whitney test was used to compare among various groups in cell growth assay. For relative luciferase intensity results from two independent in vivo experiments, we normalized the log2 transformed intensities from each experiment using median-centered method and then combined the results. T-test was used to compare the difference of two groups.

Identification and validation of enhancer genes

Inhibition of NF-κB pathway enhanced topotecan-mediated growth inhibition in neuroblastoma cells

We next explored if there were any signaling pathways associated with these nine hits using MetaCore (see Methods). The top of the list was anti-apoptotic TNFs/NF-κB/IAP pathway (p = 6.03E-06; Table 2). Out of the total 27 curated genes in this pathway, five of them were among the nine hits, including BIRC4, NFKB1, RIPK1, TNFRSF25 and TNFRSF8. The knockdown on these five target genes by siRNAs was confirmed using Taqman assays (Figure 1A). Therefore we hypothesized that the NF-κB pathway might be activated as a response to topotecan in SK-N-AS cells. To test this hypothesis, we performed Gene Set Enrichment Analysis (GSEA) [17] on microarray gene expression data of SK-N-AS cells treated with topotecan to determine if there was enrichment for NF-κB target genes (http://bioinfo.lifl.fr/NF-KB/). We found that these NF-κB target genes were significantly up-regulated in the leading edge subset genes in the GSEA analysis in the topotecan-treated cells (p < 0.0001; Figure 1B). Because activation of the NF-κB pathway results in degradation of IκB-α (through phosphorylation) and increase of p65/RelA phosphorylation, we further examined the effect of topotecan on the NF-κB pathway using Western blotting for these two key components of the pathway. Indeed, the Western blotting showed decreased amount of total IκB-α and increased phosphorylation of p65/RelA after topotecan treatment, confirming activation of NF-κB by topotecan (Figure 1C). These results suggested that the NF-κB pathway was activated by topotecan possibly as a protective mechanism against topoisomerase-I inhibition resulting in single stranded DNA breaks. To verify if the up-regulation of NF-κB pathway was protective from topotecan-mediated growth inhibition, we specifically knocked down NFKB1, another key component in the pathway, by siRNA. Knocking down NFKB1 resulted in further inhibition of cell proliferation in topotecan-treated cells (p < 0.01, Figure 1D), and a Western blot confirmed the effective knockdown of NFKB1 by the siRNA at the protein level (Figure 1E). Taken together, our results indicated that combination of topotecan with a NF-κB inhibitor might be synergistic, and this synergistic effect provided a rational combination therapy against neuroblastoma.

Genego Maps	Number of genes in the map	Number of overlapped enhancer genes	p-value
Anti-apoptotic TNFs/NF-κB/IAP pathway	27	5	6.03E-06
Apoptotic TNF-family pathways	41	4	8.67E-04
HTR1A signaling pathway	38	3	8.21E-03
APRIL and BAFF signaling pathway	38	3	8.21E-03

Synergistic growth inhibition in three neuroblastoma cell lines treated with topotecan and NF-κB inhibitors

In order to examine if the cell growth inhibition in neuroblastoma mediated by topotecan can be synergized with NF-κB pathway inhibition, we tested a specific NF-κB inhibitor, NSC 676914 [19] and a FDA approved proteasome inhibitor which also inhibits the NF-κB pathway, bortezomib, in combination with topotecan. We tested these combinations in SK-N-AS and two additional neuroblastoma cell lines NB-1691 (MYCN amplified) and SH-SY5Y (MYCN non-amplified) to ensure that the synergistic effect was not specific to any cell line. We first measured the IC50 of topotecan and NSC676914 in these cells at 24 hours, and they were 2 μM and 7 μM for SK-N-AS; 550 nM and 4 μM for NB-1691; 200 nM and 2 μM for SH-SY5Y respectively. Sub-IC50 doses of topotecan and NSC 676914 were then used in combination in all three cell lines and the combination index was calculated for each dose combination [18]. All the combination treatment data points in normalized isobolograms were away from the diagonal additive line and towards the origin indicating that low doses of NSC676914 and topotecan act synergistically in all three cell lines (Figure 2A). To facilitate future translation of our findings to the clinic, we tested bortezomib, an FDA-approved drug known to inhibit NF-κB as one of its mechanisms, in the subsequent in vitro and in vivo studies. In addition, this drug has been reported to inhibit all of the top four pathways identified in our synthetic lethal screen (Table 2) [20–22]. Therefore, we predicted bortezomib would have synergistic effects in combination with topotecan in neuroblastoma cells. To test this hypothesis, we first measured the IC50 of bortezomib for these three cell lines at 24 hours as 7 nM, 3.5 nM and 3.5 nM for SK-N-AS, NB-1691 and SH-SY5Y respectively. As expected, we found from normalized isobolograms that sub-IC50 doses of topotecan and bortezomib acted synergistically to inhibit cell growth as the data points in normalized isobolograms were also away from the diagonal additive line (Figure 2B). This synergistic inhibitory effect between topotecan and bortezomib was confirmed in an independent experiment using electronic cell sensing to monitor cell growth (P < 0.01) (Figure 2C). Furthermore, we were able to detect an increase in apoptosis with the combination of topotecan and bortezomib as evidenced by elevated caspase-3 activity (Figure 2D). To confirm that bortezomib enhances topotecan-induced growth inhibition through NF-κB pathway, Western blotting was performed to show that SK-N-AS cells treated with both bortezomib and topotecan inhibited the degradation of total IκB-α protein, suggesting that the NF-κB was indeed inactivated (Figure 2E). Taken together, these data demonstrated that topotecan and NF-κB inhibitors could synergistically inhibit growth through inhibition of the NF-κB pathway and induction of apoptosis in neuroblastoma cells.

Delayed tumor progression in human neuroblastoma xenograft treated with topotecan and bortezomib

In order to investigate the synergistic effect of topotecan and bortezomib in vivo, we tested the drug combination in SCID beige mice bearing human neuroblastoma xenografts. SK-N-AS cells were intravenously injected into the mice and monitored by Xenogen. Treatments started seven days after injection, and our treatment scheme encompassed only two courses of drug administration allowing us to focus on the immediate effect of the drug combination on tumor formation and progression. One week after the first course of treatment, tumors were observed in mice treated with control and bortezomib alone. By the end of the second course of treatment, these mice became morbid and were euthanized (data not shown). We also found that the mice treated with combination of topotecan and bortezomib showed a delay in tumor progression when compared to mice treated with topotecan alone (Figure 3A). Results from two independent in vivo experiments showed that the relative luciferase signals from tumor cells were significantly decreased in mice treated with combination of topotecan and bortezomib (p = 0.01; Figure 3B). Therefore, we concluded that combination therapy of topotecan and bortezomib caused a significant reduction of tumor growth compared to individual drugs alone, which confirmed our hypothesis that bortezomib potentiated the effects of topotecan, via targeting the NF-κB pathway as one mechanism together with the topoisomerase-1 inhibition.