Arsenic Trioxide exerts cytotoxic and radiosensitizing effects in pediatric Medulloblastoma cell lines of SHH Subgroup

We evaluated the potential effects of ATO in different pediatric SHH-MB cell lines (ONS-76: TP53-wild type; DAOY and UW402: TP53-mutated). MB cell lines molecular subgroup was confirmed and TP53 mutations were validated. Cell viability, clonogenicity and apoptosis were evaluated after ATO treatment at different concentrations (1–16 µM) alone or combined with irradiation doses (0.5, 1, 2 and 4 Gy). Rad51 and Ku86 proteins were evaluated by WB. ATO treatment reduced cell viability for all SHH-MB cell lines. Significant decrease of clonogenic capacity and higher apoptosis rates were also observed after ATO exposure, being cell death more pronounced (>70%) for the SHH-MB TP53-mutated. Combined treatment of ATO with irradiation also reduced colonies formation in UW402 tumor cells, which was independent of DNA damage repair proteins Rad51 and Ku86. In silico analyses suggested that a set of genes from cell cycle and p53 pathways are differentially expressed in SHH tumor subtypes, suggesting that cell lines may respond to therapies according to the gene expression profiles. Herein, we showed ATO cytotoxicity in pediatric SHH cell lines, with marked radiosensitizing effect for the MB-SHH TP53-mutated cells. These results highlight the potential of ATO, alone or in combination with radiotherapy, supporting further clinical investigations.


ATO controls cell viability, induces apoptosis and improves radiosensitivity in SHH-MB cells.
The MB molecular profile of the three MB cell lines models (DAOY, UW402 and ONS-76) was validated by TDLA, which confirmed the SHH molecular subgroup (Fig. 1A). Regarding the TP53 status, Sanger sequencing confirmed mutations in DAOY (TP53 -c.725G > T) and UW402 (TP53 -c.464C > A), while the ONS-76 cell line was shown to be SHH TP53 wild type ( Fig. 1B-D).
Treatment with ATO induced a significant reduction of cell viability in a dose-dependent manner for all three cell lines models ( Fig. 2A-C), being the UW402 the cell line more affected with lowest IC 50 values (Table 1). Also, non-neoplastic cells (MRC-5 cell line) were more resistant to ATO effect. While neoplastic cell lines presented a mean reduction of 81.8% in cell viability in the highest dose/time-point, MRC-5 decreased no more than 55.6% ( Supplementary Fig. S1). ATO also reduced cell colony formation at concentrations of 0.5, 1, 2 and 4 μM and increased apoptosis rates at 4 and 8 μM after 48 hours of treatment. The clonogenic effects were dose-dependent for all cell lines; however, DAOY showed to be the most sensible model either for apoptosis induction and colony capacity inhibition (Fig. 2G,H). In addition, clonogenic assays combining ATO with irradiation demonstrated that ATO was able to sensitize UW402 cell line (TP53 mutated) to irradiation, reducing clonogenic capacity from 1.7 to 3.4 times according to doses (0.5, 1, 2 and 4 Gy; p < 0.001), when compared to irradiation alone (Fig. 2D). DAOY (TP53 mutated) present a marginal radiosensitizing effect, with clonogenic capacity decreasing between 1.2 to 1.6 times (Fig. 2E). Interestingly, ONS-76 cell line (TP53 wild type) showed none radiosensitizing effect, as observed in Fig. 2F. The Supplementary Table S1 describes the relative clonogenic capacity reductions for all MB cell lines submitted to combined treatment.
In order to investigate the effects of ATO combined with irradiation, we evaluated the expression of key proteins involved in DNA repair of double-strand breaks caused by irradiation. However, the combination of ATO with irradiation did not change the expression of Rad51 (homologous recombination pathway -HR) or Ku86 heterodimer (non-homologous end-joining pathway -NHEJ) ( Supplementary Fig. S2). Additional In silico analyses were performed with data from a previous study on pediatric MB samples 2 . The results pointed a set of genes involved in the cell cycle, p53 pathways and chromosomal instability that presents specific expression patterns according to the SHH molecular subgroup (Supplementary Table S2). Thus, the expression profile of these genes, www.nature.com/scientificreports www.nature.com/scientificreports/ together with TP53 mutation status, showed it to be essential to address ATO's radiosensitizing effects in pediatric MB-SHH cells.

Discussion
Currently, the ATO agent is used in the promyelocytic leukemia therapy 11 , but its clinical potential over other types of tumors continue to be underexplored. It has a well-recognized effect in suppressing the SHH pathway by inhibiting GLI proteins, a pathway related to human cancer pathogenesis 7,12,13 . In this study, we observed distinct responses to ATO according to the cell line and TP53 status. All SHH-MB models had significantly lower cell viability after ATO treatment in a dose-dependent manner. The cells bearing mutations on TP53 (mainly the DAOY cell line) presented higher sensitivity, showing increased apoptosis rates and reduced cell-colony formation. Previous studies also have shown ATO's efficiency against HGNET-BCO SHH+/GLI+ tumor resistant to  www.nature.com/scientificreports www.nature.com/scientificreports/ SMO inhibitors (e.g vismodegib) 14,15 . Similar results were observed in other in vitro models of pediatrics and CNS tumors (in e.g rhabdomyosarcoma, glioblastoma, leukemia and HGNET-BCOR tumor) 9,10,12,13,16 , but the exact connection between tumor response to ATO and TP53 mutational status has not been elucidated so far.
The combination of ATO with irradiation therapy decreased colony formation in both TP53 mutated SHH-MB cell lines, with a significant reduction in UW402 cells. A fair explanation for distinct ATO radiosensitizing effects in pediatric SHH-MB cells rely on molecular differences between the cell lines, which includes TP53 mutation sites 17,18 . Different point mutations in TP53 binding sites activate paradoxical cellular programs such as cell death by apoptosis or DNA damage response mechanisms 17,18 . Besides, additional experiments using non-malignant cells are essential to assess whether a combination of ATO with radiation therapy has a selective effect in cancer cells, a study in head and neck carcinoma corroborates with our findings 19 . The authors showed that the combination of ATO with irradiation was shown able to reduce colony formation in p53-deficient cells, which was associated with DNA damage, G2/M arrest, and apoptosis 19 . Moreover, the treatment (ATO + irradiation) was already explored in one cerebellar HGNET-BCO SHH+/GLI+ tumor, that achieved six-months tumor remission with therapy's proper safety and tolerability 12 . Kumthekar et al. (2017) 20 have confirmed the safety and tolerability of ATO administered at the dose of 0.15 mg/kg/day in combination with focal radiotherapy for infiltrating gliomas of children, which are essential aspects for ATO's applicability in humans.
Since significant radiosensitizing effects of ATO were found only in one SHH MB in vitro model, we hypothesize that alternatives molecular mechanisms are involved in the heterogeneity of SHH MB and thus promote distinct outcomes. We attempted to address this question performing In silico analyses using R2 genomic database visualization platform focused on MB SHH subtypes (SHH-alpha, beta, gamma, and delta). Using pathway enrichment analysis on 4 SHH MB subtypes, we found that these entities differ in signatures associated with cell cycle, P53 pathway and chromosomal instability (SHH-alpha, beta, gamma and delta). Similarly, Park and colleagues (2019) described genes from canonical pathways that are prognosis-related and differ according to SHH-MB subtypes (e.g P53 and cell cycle pathways), suggesting that these signatures potentially has an association with tumor aggressiveness and therapy response 21 . Although further analysis are needed, these findings may point towards to a signature set that predicts poor response to therapy within SHH subgroup. Moreover, the novel clinical subtype iSHH (infant patients), a standard to high-risk group, may also benefit from radiosensitizing compounds such as ATO) 14,15 considering that radiotherapy to treat MB in children below 3 to 5 years is still controversial 15,22 .
Taken together, we showed that ATO induces substantial cytotoxic effect, decreases the clonogenic capacity and induces apoptosis in MB-SHH cells. Interestingly, ATO induced significant radiosensitizing effect over one TP53 mutated cell line, with no association with Ku86 and Rad51 proteins. The gene expression profiles in MB-SHH samples suggested the cell cycle pathway as a vital prognosis marker, which may be related to the success or failure of combined therapies with irradiation. Therefore, our findings shed new light on a combined therapy of ATO with low-dose irradiation in MB-SHH to be exploited and evaluated by research protocols.

Molecular subgroup assignment of MB cell lines.
The total RNA from the three cell lines was isolated with TRIZOL LS reagent (Invitrogen, Carlsbad, CA, USA) and the cDNA was synthesized using the High Capacity kit (Applied Biosystems, Foster City, CA, USA) according to manufacturers' instructions. The molecular classification was performed by TaqMan Low-Density Array (TLDA PCR-Array). A detailed description of the TDLA method for MB classification can be found elsewhere 23 . Cell viability assay. Cell viability was assessed by the Resazurin reduction method. Cells were seeded in 96-well plates and maintained under standard culture conditions for 24 hours (2×10 3 cells/well). Cells were treated with different ATO concentrations (1 μM to 16 μM) and incubated for 24, 48, 72, 96 or 120 hours. Then, Resazurin solution (Sigma-Aldrich Co., Saint Louis, MO, EUA) was added and the plates were incubated for 4 hours under standard culture conditions. The absorbance at 570 nm wavelength with a reference wavelength of 595 nm was read using an iMax Microplate Reader (Bio-Rad, CA, USA). The IC 50 values were calculated using Calcusyn software (Biosoft, Ferguson, Missouri, US 24 . Three independent experiments were performed in triplicate. In silico analysis in R2 genomic database and visualization tool. Pathway enrichment was performed using R2: Genomics Analysis and Visualization Platform (http://r2.amc.nl) tool. The analysis held a microarray data set from 223 MB-SHH patients in out of 763 MB samples represented by GSE85217. We have focused the analysis on pre-selected SHH subtypes in R2 platform: alpha-beta-gamma-delta. Pathway enriched in each subtype was considered according to the Minimal t-test parameter as p < 0.01 with HugoOnce mode set as "yes". To visualize and compare specific gene signature from KEGG pathway from: cell cycle, P53 pathway and chromosomal instability among SHH MB subtypes, we have generated a Heatmap using draw Heatmap tool (Z-Score) with pre-set metrics as Pearson correlation distance and average-linkage algorithm.
Statistical analyses. Data were analyzed by one-way ANOVA followed by the Bonferroni post-test using the Statistical Package for the Social Sciences software (SPSS, Inc., Chicago, USA). Graphs were generated with GraphPad Prism 4.0 (GraphPad Software, San Diego, CA, USA). The level of significance was set at p < 0.05 in all analyses.
Ethics approval and consent to participate. This research was submitted to and approved by the HC/ FMRP-USP Research Ethics Committee and obtained exemption for not involving humans, only commercial cell lines.