Analysis of genes involved in response to doxorubicin and a GD2 ganglioside-specific 14G2a monoclonal antibody in IMR-32 human neuroblastoma cells*

Neuroblastoma is the most common extra-cranial solid tumor of childhood and it is characterized by the pres- ence of a glycosphingolipid, GD2 ganglioside. Monoclonal antibodies targeting the antigen are currently tested in clinical trials. Additionally, several research groups reported results revealing that ganglioside-specific antibodies can affect cellular signaling and cause direct cytotoxicity against tumor cells. To shed more light on gene expression signatures of tumor cells, we used microarrays to analyze changes of transcriptome in IMR-32 human neuroblastoma cell cultures treated with doxoru- bicin (DOX) or a mouse monoclonal antibody binding to GD2 ganglioside 14G2a (mAb) for 24 h. The obtained re- sults highlight that disparate cellular pathways are regulated by doxorubicin and 14G2a. Next, we used RT-PCR to verify mRNA levels of selected DOX-responsive genes such as RPS27L, PPM1D, SESN1, CDKN1A, TNFSF10B, and 14G2a-responsive genes such as SVIL, JUN, RASSF6, TLX2, ID1 . Then, we applied western blot and analyzed levels of RPS27L, PPM1D, sestrin 1 proteins after DOX-treat-ment. Additionally, we aimed to measure effects of doxorubicin and topotecan (TPT) and 14G2a on expression of a novel human NDUFAF2 gene encoding for mimitin protein (MYC-induced mitochondrial protein) and cor- relate it with expression of the MYCN gene . We showed that expression of both genes was concomitantly de- creased in the 14G2a-treated IMR-32 cells after 24 h and 48 h. Our results extend knowledge on gene expression profiles after application of DOX and 14G2a in our mod- el and reveal promising candidates for further research aimed at finding novel anti-neuroblastoma targets. false discovery rate; GD2, GD2 ganglioside; GO, gene ontol- ogy; mAb, monoclonal antibody; RT-PCR, Reverse Transcription Polymerase Chain Reaction; qt-RT-PCR, quantitative Reverse Tran- scription Polymerase Chain Reaction; TPT, topotecan Also, we showed a correlation between changes of expression of a NDUFAF2 gene encoding for mimitin (MYC-induced mitochondrial protein) and the MYCN gene in the 14G2a-treated IMR-32 cells . The performed experiments allowed us to gain more insight into gene regulation in IMR-32 cells after treatment with 14G2a and DOX. Literature search on 14G2a-respon-sive genes allowed us to conclude that on one side the antibody affects pathways already established as impor- tant for neuroblastoma biology, on the other side novel gene candidates can be selected for further analyses.


INTRODUCTION
Neuroblastoma is the most common extra-cranial solid tumor diagnosed in childhood and the disease shows a remarkable biological heterogeneity (Øra & Eggert, 2011). Hence, a high rate of spontaneous regressions can be observed in infants with the 4S stage of the disease (Brodeur & Bagatell, 2014). On the contrary, an improvement of outcomes of patients with high risk neuroblastoma still remains a major challenge for pediatric on-cologists (Moreno et al., 2013). Neuroblastoma cells are marked with presence of GD2 ganglioside and tests of passive administration of monoclonal antibodies targeting the tumor-associated antigen are clinically advanced. Recently, Yu et al., reported that addition of a chimeric monoclonal antibody ch14.18 binding GD2 ganglioside, IL-2, and GM-CSF to 13-cis retinoic acid is effective in children with high risk neuroblastoma (Yu et al., 2010).
Several research reports establish clearly that binding of ganglioside-specific antibodies can affect fate of cancer cells, e.g., lead to changes in cellular signaling and induce cell death that may encompass apoptosis, necrosis, and oncosis-like mechanisms (for review on the topic see Horwacik & Rokita, 2015). In 2009, our group showed that a mouse monoclonal antibody (mAb) 14G2a caused cell death with partial involvement of apoptosis in cultures of a human neuroblastoma cell line IMR-32, as evidenced by a rise in staining with Annexin V and propidium iodide and an increase in caspase 3 activity (Kowalczyk et al., 2009). Moreover, synergistic effects were observed when 14G2a was combined with standard chemotherapeutic drugs such as doxorubicin (DOX) that is an anthracycline antibiotic blocking activity of topoisomerase II, and topotecan (TPT) that is a camptothecan analogue and an inhibitor of topoisomerase I (Kowalczyk et al., 2009).
Biology of neuroblastoma is being intensively studied. MYCN amplification is an independent negative prognostic factor in neuroblastoma. Additionally, high expression levels of aurora A kinase were shown to indicate a shorter survival in neuroblastoma patients (Ramani et al., 2015). In 2013, our group showed that aurora A and MYCN oncogenic proteins are decreased upon the 14G2a-treatment of IMR-32, which correlated with an increase of a tumor suppressor protein P53. Moreover, the antibody enhanced effects of an aurora A inhibitor, MK-5108, and 13-cis retinoic acid in some MYCN-am-plified neuroblastoma cell lines (Horwacik et al., 2013). This fuels our efforts to characterize more deeply effects of the 14G2a mAb on survival of the tumor cells. Therefore, primarily we focused our experiments to decipher changes in expression of genes in IMR-32 cells treated with 14G2a using microarrays. Also, we aimed to compare the changes of transcriptome in 14G2atreated cells to doxorubicin-treated cells. Previously, the chemotherapeutic drug had been shown by us to act synergistically with 14G2a on IMR-32 cells (Kowalczyk et al., 2009). Then, based on the microarray results, we intended to analyze expression of selected 14G2a-and doxorubicin-responsive genes by RT-PCR and western blot techniques. Finally, we wanted to analyze changes in expression of NDUFAF2 upon addition of 14G2a, doxorubicin or topotecan to IMR-32 cells.
Here, we present result of microarray analyses of changes in transcriptome of IMR-32 cells treated with 14G2a or DOX for 24 h. Gene ontology (GO) analysis showed that genes regulated by DOX are involved in such cellular processes as induction and positive regulation of cell death (including apoptosis), DNA metabolic processes, negative regulation of cell proliferation, and response to UV. This contrasts with genes regulated by 14G2a which were assigned to a single group of genes involved in regulation of RNA metabolic processes. Although the results generated with microarrays had showed rather limited changes in mRNA profiles, they allowed us to select new gene candidates for further research. Thus, expression patterns of some selected genes: RPS27L, PPM1D, SESN1, CDKN1A, SVIL,JUN,RASSF6,TLX2,ID1 were analyzed in our model by application of RT-PCR and/or western blot techniques. Also, we showed a correlation between changes of expression of a NDUFAF2 gene encoding for mimitin (MYC-induced mitochondrial protein) and the MYCN gene in the 14G2a-treated IMR-32 cells. The performed experiments allowed us to gain more insight into gene regulation in IMR-32 cells after treatment with 14G2a and DOX. Literature search on 14G2a-responsive genes allowed us to conclude that on one side the antibody affects pathways already established as important for neuroblastoma biology, on the other side novel gene candidates can be selected for further analyses.

MATERIALS AND METHODS
Cell culture and treatment with doxorubicin and 14G2a mAb. Cell culture of IMR-32 (ATCC, USA, CCL-127), KELLY (Leibniz-Institut DSMZ -Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH, Germany, ACC 355), both MYCN-amplified human neuroblastoma cells, a MYCN-non-amplified SK-N-SH human neuroblastoma cells (ATCC, HTB-11), a hybridoma cell line producing the 14G2a mAb (kindly provided by Dr. R. Reisfeld, The Scripps Research Institute, La Jolla, CA, USA), purification of the 14G2a mAb, treatment of IMR-32 cells with doxorubicin, topotecan, and 14G2a were previously described in details (Horwacik et al., 2013;Kowalczyk et al., 2009). For most of the experiments IMR-32 cells were used. Briefly, IMR-32 cells were grown in EMEM containing 1% nonessential amino acid solution, 1 mM sodium pyruvate, 50 mg/ml of gentamicin (all from Sigma-Aldrich, Poland), supplemented with 10% fetal calf serum (Life Technologies, Poland) at 37ºC, and in an atmosphere containing 5% CO 2 . The cells were treated with doxorubicin, 14G2a mAb, or both agents for a given time (see the Results). Control cells were also included in experiments. The concentration of doxorubicin was 15 nM for microarray studies, 15 nM, 17 nM and 30 nM for RT-PCR and western blot studies (see the Results). The 14G2a mAb was used at concentration of 40 μg/ml for all results presented below. In some experiments cells were treated with 2 nM topotecan. The effects of the drugs on the IMR-32 cells were measured using cell viability tests, i.e., MTT tests (Sigma-Aldrich) or ATPlite Luminescence ATP Detection Assay System (Perkin Elmer, USA) according to instructions from manufacturers.
Microarray hybridization. Microarray hybridization was performed by Genbioinfo (Krakow, Poland). Total RNA was isolated using the TRIZOL reagent according to the manufacturer's instructions (Invitrogen, Carlsbad, CA). RNA concentration was measured using a ND-1000 Spectrometer (NanoDrop Technologies Inc., Montchanin, DE, USA). RNA quality was determined using an Agilent Bioanalyzer 2100 (Agilent, Palo Alto, CA, USA). A starting amount of 200 ng of high-quality total RNA was used to generate cDNA and cRNA with the Illumina TotalPrep RNA Amplification Kit (Illumina, San Diego, CA, USA). The procedure consisted of RT with an oligo(dT) primer bearing a T7 promoter using Array-Script. The obtained cDNA became a template for in vitro transcription with T7 RNA polymerase and biotin UTP, which generated multiple copies of biotinylated cRNA. The purity and concentration of the cRNA were determined using a NanoDrop ND-1000 Spectrometer. Quality cRNA was subsequently hybridized using a direct hybridization array kit (Illumina). Each cRNA sample (1.5 μg) was hybridized overnight using the Hu-manHT-12 BeadChip array (Illumina) in a multiple-step procedure; the chips were washed, dried, and scanned on the BeadArray Reader (Illumina). Raw microarray data were generated using BeadStudio v3.0 (Illumina). A total of 12 Illumina HumanHT-12v3 microarrays were used in the experiment.
Microarray data analysis. Microarray data analysis was performed by Genbioinfo (as shown on data presented on Fig. 1, Table 1-3). Microarray analysis and quality control of the arrays were performed using Bea-dArray R package v1.0.0. After background subtraction (using median background method), the data were normalized using quintile normalization and were subsequently log 2 -transformed. The obtained signal was taken as the measure of mRNA abundance derived from the level of gene (or particular transcriptional form) expression. The following criteria were applied for probe-set detection: at least three samples with signal intensity >7.4 (log2 data). This reduced the number on the analyzed probes to 19 642 (from total 49 895). Two-Way ANOVA was applied to calculate significance levels (P values) to discriminate differentially expressed genes between controls and the treated groups. The false discovery rate (FDR) was estimated using the Benjamini and Hochberg method. All statistical analyses were performed using R software version 2.10.1. Hierarchical clustering was performed with dChip software using Euclidean distance and average linkage method. The Database for Annotation, Visualization and Integrated Discovery (DAVID 2.1), a functional annotation analysis tool, was used to identify over-represented ontological groups among gene expression profiles and to group genes into functional classifications (with default options).
Statistical analyses. Data on graphs on Figs 3-5 are shown as means with SEM (standard errors of the means). Statistical significance between signals of controls and drug-treated samples were analyzed using pairwise t tests (Excel, Microsoft Office). For analysis of time-dependence of signals on the graphs presented on the Fig. 4 and the Fig. 5, we applied one-way ANOVA with repeats using R software (R version 2.15.1 Patched), *P<0.05, **P<0.01, ***P<0.001.

Microarray analyses of IMR-32 cells treated with DOX and 14G2a
We performed research to extend our knowledge on mechanism involved in cell death upon binding of the 14G2a mAb to neuroblastoma cells. We used microarray technology to decipher changes in expression of genes in the model of MYCN-amplified IMR-32 cells treated with 14G2a mAb, and compared them to a standard chemotherapeutic drug, doxorubicin. 12 sets of Illumina HT-12v3 microarrays were used to analyze 3 independent experiments consisting of control cells, and cells treated for 24 h with 15 nM doxorubicin, 40 μg/ml of 14G2a, and both agents. Data analyses were performed on 19 642 probes from a total number of 49 895 probes. The list of probes, obtained with ANOVA at FDR<5%, showing response to doxorubicin contains 46 positions (see Table 1 for the list and Fig. 1A that shows a hierarchical clustering of the doxorubicin-responsive genes and particular transcriptional forms). The list of transcripts affected by 14G2a contains 28 positions (see Table 2 for the list and Fig. 1B that shows a hierarchical clustering of 14G2a-responive genes or particular transcriptional forms, ANOVA at FDR<5%). For the probes changed after treatment with DOX we measured increase of signal in the range from about 1.1-2.8 as compared to control cells. Particularly, an increase of ≥1.5 was measured for: CDKN1A, ACTA2, RN7SK, SESN1, TNFSF10B, RPS27L, BAX, PHLDA3, GADD45 and 2 sequences HS.545589 and HS.193406 (see Table 1). For the rest of probes changes between 1.5 and 1.1 were observed. Additionally, out of 46 probes only 9 were down-regulated I. Horwacik and others from about 0.96 to 0.8. The changes of expression of transcripts in the 14G2a-treated samples were even more restricted, when compared to control cells ( Table 2). None of the transcripts were increased more than 1.5fold. The PHLDA1 gene expression was changed 1.4fold, followed by JUN (1.3-fold). Some transcripts were decreased below 0.8-fold, including that of ID1, IGFBP4, MSX1, TLX2 and 3 sequences (C8ORF13, LOC389816, C9ORF7). Based on ontological classification analyses some genes were grouped according to their involvement into particular biological functions (see Table 3 for more details on gene ontology terms). Hence, some DOX-responsive genes were linked to: induction and positive regulation of cell death (including apoptosis), DNA metabolic processes, involvement in cell cycle, negative regulation of cell proliferation, and response to UV. On the contrary only one and unique gene ontology term was assigned to the 14G2a-responsive genes, i.e., regulation of RNA metabolic processes, as compared to the DOX-responsive genes. Based on the gathered data, we can conclude that although overall a restricted number of genes responded to the drugs tested in our model and changes of the most of the transcripts are limited, several interesting candidates can be selected for further analyses.

RT-PCR and western blot analyses of selected doxorubicin-responsive genes
In order to verify some changes of gene expression in the DOX-treated IMR-32 cells, we selected CDKN1A [cyclin-dependent kinase inhibitor 1A (p21, Cip1)], TNFRSF10B (tumor necrosis factor receptor superfamily, member 10B), RPS27L (ribosomal protein S27-like), SESN1 (sestrin 1), PPM1D (protein phosphatase, Mg 2+ /Mn 2+ dependent, 1D) as examples of the genes that were up-regulated more than 1.5-fold in the microarrays. We tested levels of their transcripts in IMR-32 cells treated with 15 nM and 30 nM DOX for 24 h using semiquantitative RT-PCR (as   . 2C). For the transcript of SESN1 an increase was clearly observed for all time points tested, while expression of the other two genes was consistently up-regulated from 24 h. Additionally, a western blot technique was applied to measure levels of proteins in the samples from IMR-32 cells treated with 15 nM and 30 nM DOX for 24 h, as compared to control cells. Signals from detection of GAPDH were used for protein normalization in the samples (Fig. 3). Only for RPS27L, we could observe increase in the protein levels for both concentration of DOX used (Fig. 3A). A level of PPM1D was increased only in samples treated with 15 nM DOX (Fig. 3C), while it was decreased along with sestrin 1 in samples treated with 30 nM DOX (Fig. 3B,C, pairwise t tests between controls and DOX-treated samples did not show statistical significance of the measured changes of levels of RPS27L, PPM1D, and sestrin 1). Based on the results of the experiments, we could confirm that DOX increased expression of transcripts of the five selected genes. However, additional experiments (e.g., performed for longer time points of incubation of cells with DOX) are necessary to establish effects of DOX on the expression of the genes on the protein levels.

Qt-RT-PCR analyses of selected 14G2a-responsive genes
Previously, we reported that in IMR-32 cells treated with 14G2a both transcript and protein levels of the PHLDA1 gene (pleckstrin homology-like domain, family A, member 1) were statistically meaningfully increased (Horwacik et al., 2013). In order to verify some additional changes of transcripts in the 14G2a-treated IMR-32 cells, we selected 3 genes that were up-regulated, i.e., JUN (jun oncogene), SVIL (supervillin), RASSF6 (Ras association (RalGDS/AF-6) domain family member 6) and 2 genes that were down-regulated in the microarray analysis, i.e., ID1 (inhibitor of DNA binding 1, dominant negative helix-loop-helix protein) and TLX2 (thyroid adenoma associated). Also, we measured levels of the CDKN1A gene in the cells. Qt-RT-PCR was applied to quantify levels   of the transcripts in IMR-32 cells treated with 40 μg/ ml of 14G2a for 6, 16, 24, 48 h ( Fig. 4 A-F). The transcript of EF2 was used for signal normalization in the samples. The transcript of JUN was statistically significantly up-regulated in RNA samples of IMR-cells incubated with the 14G2a mAb for 6, 16, 48 h (P<0.05, for pairwise t tests between the control and treated samples). The transcripts of SVIL and RASSF6 were increased only for 48 h (P<0.05, for pairwise t tests between the control and treated samples), while for the time point of 6 h both transcripts were down-regulated (P<0.05, for pairwise t tests between the control and treated samples). The transcripts of ID1 and TLX2 were shown to be down-regulated for all 4 time points tested (for most of the time points evaluated P<0.05 was calculated using pairwise t tests between the control and treated samples). Additionally, we showed that the transcript of CDKN1A was statistically significantly decreased upon the treatment with the 14G2a mAb for 6 and 16 h. The data gathered allowed us to conclude that qt-RT-PCR results confirmed decrease of the transcripts for ID1 and TLX2 shown by the microarrays. The increase of the transcript of JUN was shown for 6, 16, 48 h, but not for 24 h as reported in the microarray experiments. Additionally, meaningful opposite effects, i.e., a decrease for the 6 h time point and an increase for the 48 h time point, were observed for the transcripts of SVIL and RASSF6. Finally, although direct comparison of changes in the transcript of CDKN1A were not performed for DOX-and 14G2a-treated cells contrasting effects can be observed ( Fig. 2 and Fig. 4), which require verification by side-by side sample analysis.

Effects of doxorubicin, topotecan and the 14G2a mAb on expression of NDUFAF2 and MYCN
We have previously shown that the increase of transcript and protein of PHLDA1 in 14G2a-treated IMR-32 cells is correlated with a decrease of two oncogenic proteins i.e., aurora A (in whole cell extracts) and MYCN (in cytoplasmic and nuclear fractions) (Horwacik et al., 2013). Also, cytotoxicity induced with 14G2a in IMR-32 cells was accompanied by rise in activity of caspases 3 and 7 (Horwacik et al., 2013;Kowalczyk et al., 2009). Thus, we extended our research to correlate levels of MYCN and a novel NDUFAF2 [(NADH dehydrogenase (ubiquinone) complex I, assembly factor 2)] gene encoding for a mimitin protein (MYC-induced mitochondrial protein). We treated IMR-32 cells with two cytotoxic drugs, DOX and TPT, which are used for neuroblastoma treatment in clinics and were previously shown by us to act synergistically with 14G2a on IMR-32 cells (Kowalczyk et al., 2009). Again cells treated with the 14G2a mAb were included. We showed that MYCN and mimitin are expressed in IMR-32 and KELLY (both MYCN-amplified) neuroblastoma cell lines, contrasting to a SK-N-SH (a MYCN-non-amplified) neuroblastoma cell line, in which no MYCN was detected and lower intensity of the band of mimitin was measured (Fig.  5A). Also, in IMR-32 we confirmed that no MYC protein was detected (there was no MYC transcript as well, data not shown). To compare cytotoxic effects of DOX and TPT on expression of the genes, we selected concentration of 17 nM DOX and 2 nM TPT that caused reduction of cell viability to about 70% of control cells after 48 h (measured by a MTT test, data not shown). Western blot analysis of levels of mimitin and MYCN in DOX-and TPT-treated IMR-32 cells did not reveal their considerable changes (Fig. 5B, a result from a single experiment is shown). However, when we analyzed expression of the transcripts of NDUFAF2 and MYCN by qt-RT-PCR in the 14G2a-treated cells, we measured a statistically meaningful decrease in the levels of both transcripts (Fig. 5C, D). Furthermore, we used western blot to show that a decrease in the level of MYCN is correlated with a decrease of the level of mimitin in whole cell extracts from the 14G2a-treated IMR-32 cells at 24 and 48 h (P<0.05, for mimitin by a t test, controls vs. 14G2a-treated samples for 48 h, Fig. 5E, F).

DISCUSSION
Gene expression signatures of tumor cells are extensively studied to better understand differences between transformed and normal cells, but also in quest for diagnostic, or prognostic markers, and new targets for treatment of patients with cancer. Thus, we aimed to gain more insight into changes of transcriptome of IMR-32 cells treated with 14G2a and to compare the effects induced with the antibody and doxorubicin. In general, only minor changes in levels of the analyzed transcripts were observed between control, untreated cells, and 14G2a-treated cells in our whole-genome transcriptome studies, suggesting that in the next step we should focus on detailed proteome analyses. However, it should be stressed that the experiments allowed to find correlations between 14G2a and genes important for neuroblastoma biology. For instance, ID1 and TLX2 were downregulated in the arrays and in the following qt-RT-PCR analyses. TLX2 and PHOX2B are transcription factors important for differentiation of autonomic nerve cells (Borghini et al., 2006). Also, ID-1,-2, and -3 proteins are involved in keeping neuroblastoma cells in undifferentiated state as shown by Jögi and colleagues (Jögi et al., 2002).
Based on the GO analysis, the 2 aforementioned genes and 6 others (ASCL2, DHX9, CDKN1B, MSX1, ZNF224, JUN) are involved in regulation of RNA metabolic processes (GO: 0051252). Literature search allowed us to establish that ASCL2, a transcription factor protein involved in WNT signaling and miR-200s participate in regulation of transitions between mesenchymal and epithelial phenotypes (Tian et al., 2014). Also, MSX1, yet another transcription factor, was shown to be involved in Delta-Notch and WNT signaling in neuroblastoma (Revet et al., 2008;Revet et al., 2010). Recently, Chen and colleagues, showed that DHX9 (RNA helicase A) and its nuclear localization is involved in pro-apoptotic function of KIF1Bβ, a protein with tumor-suppressor properties that is encoded in a region 1p36.2 lost in a subset of neuroblastomas . Moreover, DHX9 was among ten genes with recurrent gene mutations that correlated with pathways deregulation and clinical outcomes among patients with lung adenocarcinoma who never-smoked (Sun et al., 2014). Finally, ZNF224 was implicated in apoptosis induced after DNA damage in leukemia cells (Montano et al., 2013).
Two genes, SVIL and RASSF6 showed a decrease in expression for the 6 h time point and an increase of expression for the 48 h time point in samples of 14G2a-treated IMR-32 cell, thus stressing the need for detailed analyses of the 14G2a-responsive genes yielded by microarrays. They are necessary to verify the results generated with the whole genome analysis approach. Additionally, analysis of both mRNA and protein levels in samples collected for additional time points or other 14G2a-sensitive neuroblastoma cell lines are warranted to elucidate roles of the selected genes in our model. Of interest to us is a report on a role of RASSF6 as a tumor suppressor and its function in P21 accumulation, induction of G1 cell cycle arrest, and apoptosis in clear cell renal cell carcinoma (Liang et al., 2014).
Previously we reported that the PHLDA1 transcript and protein levels were up-regulated in our model. Also, the P53 protein was up-regulated in the IMR-32 cells treated with 14G2a from 2 h and its nuclear accumulation was shown from 6 h (Horwacik et al., 2013). Both proteins are involved in interactions with aurora A (Johnson et al., 2011). The total level of the oncogenic protein and the extent of its phosphorylation at Thr288 were decreased in the 14G2a-treated IMR-32 and LA-N-1 cells (Horwacik et al., 2013). In addition, in the microarrays data the transcript of SCN3B, a P53 inducible gene, (Adachi et al., 2004) was down-regulated in the 14G2a-treated cells. Again, this highlights the need for analysis of samples collected at additional time points. Also, the transcript of FKBP4 (FKBP54) encoding a chaperone protein (Shim et al., 2009) was decreased in the microarrays. This adds to our previous findings that HSP40 and HSP70 proteins were decreased in IMR-32 cells treated with 14G2a (Horwacik et al., 2013). Thus, we can conclude that the data from the microarray and the qt-RT-PCR analyses of transcriptome of IMR-32 cells after 14G2a extend our knowledge on mechanism of action of the antibody on the cell line.
Yet another transcription factor, MYCN, was decreased in cytoplasmic and nuclear extracts of IMR-32 cells after 14G2a treatment (Horwacik et al., 2013). Here, we add data from qt-RT-PCR showing that down-regulation of MYCN mRNA correlates with decreased levels of the protein in whole cell extracts of IMR-32 after the mAb addition. Interestingly, mRNA and protein levels of NDUFAF2 are also down-regulated in IMR-32 cells treated with 14G2a. The novel NDUFAF2 gene encoding for mimitin was shown to be directly stimulated by MYC and involved in proliferation of esophageal squamous cell carcinoma (Tsuneoka et al., 2005). Its functions are still elucidated. The mimitin protein was showed to act as a molecular chaperone for mitochondrial complex I assembly (Ogilvie et al., 2005). On the other hand, in 2013 Schlehe and colleagues showed that the protein deficiency reduced complex-1 activity, increased levels of oxidative stress and deletion of mitochondrial DNA, but was dispensable for maturation of the complex (Schlehe et al., 2013). Moreover, over-expression of mimitin was reported to indirectly correlate with reduction of activity of caspase 3/7 in HepG2 cells (Wegrzyn et al., 2009). In our model, the observed reduction of mimitin protein correlates with our previous report on a rise in caspase 3/7 activity at 48 h and 72 h of IMR-32 treatment with 14G2a (Horwacik et al., 2013). This may indicate a role of mimitin in observed apoptosis and encourages further research on the involvement of the protein in our model. Finally, as no MYC is expressed in MYCN-amplified IMR-32 neuroblastoma cells and the levels of both MYCN and mimitin are reduced in the 14G2a-treated IMR-32 cells, it might be of interest to further investigate whether the transcription factor is directly involved in regulation of expression of the NDUFAF2 gene. Recently, Murphy and colleagues showed that in neuroblastoma cell lines MYCN binds more selectively to the CATGTG motif than to the classic E-box CACGTG motif. Also, other motifs, i.e., CATTTG, CATCTG, CAACTG were bound by MYCN in MYCN-amplified cells. (Murphy et al., 2009). On that note, possible role of MYCN in the NDUFAF2 gene expression could be investigated.
To summarize, based on the reported data from our experiments we gained more insight into effects of 14G2a on IMR-32 neuroblastoma cells and selected new and interesting genes for further investigation.