1. IntroductionThe Central Nervous System (CNS) tumors are the leading cause of cancer-related death in children [1] and exhibit molecular and genetic signatures different from those found in adult patients [2]. Pediatric high-grade glioma (pHGG) is a heterogeneous and aggressive primary brain tumor characterized by a rapid and infiltrative growth pattern [3,4]. Unfortunately, there is still no efficient treatment protocol for pHGG, which remains a devastating disease with significant morbidity and mortality [1,3,4]. The maximal safe surgical resection is recommended, and radiotherapy seems to be more effective for children over ten years old but not for younger patients [5]. Despite extensive research for novel treatment strategies, no new drug has dramatically increased the patient’s survival in the last decades, and the 5-year overall survival is nearly 17% [5,6]. HGGs are locally invasive and highly vascular tumors with extensive areas of necrosis and hypoxia. Hypoxia is a hallmark microenvironmental condition of HGG [7,8] that contributes the tumor development and growth [9]. In response to tumor growth, the microenvironment changes continuously and promotes specific conditions such as hypoxia and acidosis [10,11]. The tumor behavior is also affected by acidosis and, especially, by high lactate levels. In addition, studies described that acidosis promotes cell motility, migration, degradation, remodeling of the extra-cellular matrix (ECM,) and the radio and chemoresistance of tumor cells [10,11]. In this microenvironment, and under low oxygen concentrations, carbonic anhydrases (CAs) enzymes, are overexpressed and stimulate cell survival, invasion, and proliferation of tumor cells [12,13].Carbonic Anhydrases (CAs) are zinc-dependent metalloenzymes that catalyze the reversible hydration of carbon dioxide to bicarbonate, being involved in acid-base balance and other physiological processes [14]. The transmembrane isozymes CA9 and CA12 are overexpressed in many tumor types and their overexpression is correlated with poor prognosis in high-grade gliomas [11,15,16]. In addition, CAs have been described as potential anti-cancer therapeutic targets [17,18] and compounds derived of sulfonamide, such as Indisulam, have shown antitumor effects in vitro and in vivo [19,20].Indisulam/E7070 (N-(3-chloro-7-indolyl)-1,4-benzenedisulfonamide) is reported to be a potential inhibitor of CA9 and CA12 [21], which are overexpressed in diffuse gliomas from adult patients [20,22,23]. Studies have described that Indisulam targets multiple checkpoints of cell cycle phases, downregulates cyclins via p21/p53 dependent mechanisms, and promotes a significant impact on cellular metabolism [24,25,26]. Moreover, Indisulam can penetrate the blood-brain barrier [20,27] and may have promising anti-neoplastic effects in HGG. Despite this, the functions of CA9 and CA12 and the effects of Indisulam on pHGG are still unknown. In this study, we demonstrate that Indisulam reduces cellular growth and contributes to apoptosis via BCL and BAX imbalance. 2. Materials and Methods 2.1. Cell Culture

Pediatric HGG cell lines SF188 and KNS-42 were provided by Nada Jabado and Damien Faury, McGill University, Canada, and purchased from the Japanese Collection of Research Biosources Cell Bank—through the Rio de Janeiro Cell Bank, respectively. The cells were authenticated by short tandem repeat profiling, and mycoplasma testing was carried out. KNS-42 was cultured in MEM culture medium supplemented with 1% glutamine, 100 mg/mL streptomycin, 100 U/mL penicillin, and 5% fetal bovine serum (pH 7.2–7.4). SF188 was cultured in HAM F10 supplemented with 60 mg/L penicillin, 100 mg/mL streptomycin, and 10% (v/v) fetal bovine serum (Gibco, Life Technologies, Carlsbad, CA, USA).

2.2. Hypoxia Conditions—Treatment of Cells with Cobalt Chloride

To mimic hypoxic cultures, cells were cultured in the presence of Cobalt Chloride (CoCl2) (Mallinckrodt Chemicals, Dublin, Ireland). SF188 and KNS-42 cells were initially incubated with CoCl2 at increasing concentrations of 50, 100, and 150 μM for 24 h. For subsequent functional experiments, SF188 and KNS-42 were incubated with CoCl2 at 50 μM and 100 μM, respectively.

2.3. Drug and Treatments

Indisulam (Medkoo Bioscience, Morrisville, NC, USA) was dissolved as stock at 100 mM in DMSO and stored at −20 °C until use according to manufacturer recommendations.

2.4. Cell Viability Assay

A total of 2 × 103 cells were seeded in quadruplicate in 96-well plates and maintained under standard culture conditions, under hypoxia for 24 h. The cells were treated with Indisulam at 2 to 256 μM and incubated for 24, 48, and 72 h under hypoxia. At each treatment interval, a resazurin (Sigma-Aldrich Co., St. Louis, MO, USA) solution was added to the plates (10% of the initial volume). The plates were incubated for 4 h and read at 570 nm (iMax Microplate Reader (Bio-Rad, Hercules, CA, USA). These data were used to obtain the IC50 value for KNS-42 and SF188, defined as the concentrations necessary for 50% cell viability reduction, using the Calcusyn software (Biosoft, Ferguson, MO, USA). Three independent experiments were performed in quadruplicate. The results are expressed as mean plus standard deviation and were compared between treatments and control.

2.5. Clonogenic Survival To carry out the clonogenic assay, 500 cells were seeded in triplicate in six-well plates. After 24 h incubation with CoCl2, the cells were treated with Indisulam (2 to 256 µM) for 48 h. Thus, these tests allowed us to identify the lowest dose of Indisulam required to inhibit clonogenicity. The cells were washed with PBS, and a culture medium without the drug was added to permit colony growth for approximately 9–11 days until the colonies were visible but not confluent. Colonies were fixed in methanol and stained with 1% Giemsa. Colonies of at least 50 cells were counted with a magnifying glass. The plating efficacy (PE) represents the percentage of cells seeded that grow into colonies under a specific culture condition of a given cell line. It was calculated as described by Franken et al. [28]. Three independent experiments were performed. 2.6. Apoptosis Assay

The assay for cell death detection was carried out by labeling apoptotic cells with annexin V fluorescein isothiocyanate (BD Biosciences, San Jose, CA, USA) and necrotic cells with propidium iodide (PI). To perform the experiments, 4 × 104 KNS-42 or SF188 cells were incubated with CoCl2 (100 and 50 μL, respectively) for 24 h. After cells were treated with the different concentrations of Indisulam (8, 32, 128, and 256 µM) for 48 h, they were trypsinized, washed with ice-cold PBS, and resuspended in 1X annexin binding buffer (BD Biosciences, San Jose, CA, USA). The cells were labeled with annexin V and propidium iodide (PI) solution and analyzed using a BD FACSCalibur™ flow cytometer (BD Biosciences, San Jose, CA, USA). The values represent the mean and standard deviation of three independent experiments performed in triplicate.

2.7. mRNA Extraction and RT-qPCRTo evaluate the effect of Indisulam on CA9 and CA12 gene expression (refs. Hs00154208 and Hs01080909), the cells were exposed to CoCl2 for 24 h and treated with Indisulam (169 µM) for 48 h. mRNA was extracted using the Trizol® reagent (Invitrogen Inc., Carsdab, CA, USA) according to the manufacturer’s instructions. cDNA was synthesized using High Capacity® kit (Applied Biosystems, Foster City, CA, USA). qRT-PCR was performed using the QuantStudio 12k Flex device (Applied Biosystems, Foster City, CA, USA). Samples treated with vehicle were used as controls. TBP (TATA binding protein, ref. 4326322E) and HPRT (hypoxanthine guanine phosphoribosyl transferase, ref. 4326321E) genes were used as internal mRNA controls. The relative quantification of gene expression was determined using the 2−ΔΔCT method [29]. 2.8. Protein Extraction and Western BlottingTo evaluate the effect of Indisulam on protein expression of HIF1α, CA9, CA12, BAX, and BCL-2, the cells were exposed to CoCl2 for 24 h, subjected to treatment with the drug (169 µM) for 48 h, and then collected. Protein extraction was performed using RIPA lysis buffer (Sigma Aldrich Co., Saint Louis, MO, USA) with protease and phosphatase inhibitors according to manufacturer’s guidance. Protein concentration was determined by Bradford’s method [30]. Subsequently, 40 µg of the samples were separated by SDS-PAGE electrophoresis. The proteins were transferred to nitrocellulose membranes and then incubated in a 5% non-fat milk solution in 0.1% TBST for 2 h at room temperature. Following this, the membranes were incubated overnight with specific primary antibodies for each protein, diluted according to the manufacturers’ instructions (Table 1). The levels of the GAPDH protein and β-actin were used as an endogenous control in all experiments. The membranes were then washed with TBST and incubated with specific peroxidase-conjugated secondary antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA) for 1 h and submitted to another wash cycle. The protein bands were visualized using the ECL™ Prime Western Blotting Detection Reagent (GE Healthcare, Amersham, UK) and ChemiDoc system (Bio-Rad Laboratories, Hercules, CA, USA). The membranes were exposed for 10 to 500 s and then analyzed. The relative quantification of protein expression was determined using ImageJ software (National Institutes of Health, Bethesda, MD, USA). 2.9. Statistical Analysis

Results are presented as mean with standard deviation. Significant differences were tested using Student’s t test for gene and protein expression, while one-way analysis of variance (ANOVA) with a post-hoc Bonferroni test was carried out for the remaining experiments. All analyses were performed using SPSS 20.0 software (SPSS, Chicago, IL, USA) or Prism GraphPad6.0 software (Graph Pad, San Diego, CA, USA) with level of significance p < 0.05.

4. DiscussionPediatric high-grade gliomas represent a group of tumors with wide heterogeneity, as demonstrated by identifying subgroups with different genetic, epigenetic, and morphological characteristics [31]. Based on their molecular analysis, the WHO divides pediatric-type diffuse high-grade gliomas into four subgroups: Diffuse midline glioma H3 K27-altered, Diffuse hemispheric glioma H3 G34-mutant, Diffuse pediatric-type high-grade glioma H3-wildtype and IDH-wildtype and Infant-type hemispheric glioma [32]. Using information from the methylation profile of these tumors, as well as tumor location and altered oncogenic pathways, the German Cancer Research Center establishes six different subtypes: K27, G34, IDH, RTK-I, Mesenchymal and PXA-like [33]. Regarding the cells selected to realize this study, KNS-42 presents mutation in histone H3.3 G34V and activation of NMYC and IGFR1 pathways [34]. Among the pHGGs, histone H3 mutation is present in 40% of the total, being related to a less aggressive subtype with a better prognosis [35]. Already, SF188 has overexpression of PDGFRA and amplified MYC [34], the latter being present in only 8 to 9% of cases of pHGGs, being related to a poor prognosis [35]. However, targeted therapies for each subgroup of pediatric CNS tumors have not yet been established. Therefore, clinical treatments are generally based on traditional protocols, which lack long-term efficacy because they do not take into account the distinct characteristics of pHGG subgroups or their differences with adult tumors [36,37]. In addition to tumor heterogeneity, the presence of diffuse and ill-defined tumor borders [37], as well as hypoxic regions [12], reduce the efficacy of surgical resection [3,38] and adjuvant chemo- and radiotherapy [39]. In particular, the relationship between hypoxia and radioresistance of cancer has been demonstrated since the 1950s [40,41]. Under the hypoxic microenvironment of the tumor, CA9 and CA12 have an important role in the pH regulation and therefore contribute to treatment resistance and tumor progression [14,15,16,19,42]. Indeed, prior studies from our group have demonstrated that the inhibition of CA9 and CA12 can block cell cycle progression and sensitize adult GBM cells to the treatment of radio- (in vitro) and chemotherapy (in vitro and in vivo) [20]. However, to the best of our knowledge, the effect of anti-CA treatment on pediatric brain tumors has not been studied.

In the present study, we have analyzed whether the treatment of pHGG with Indisulam has inhibitory effects on CA isozymes and cell viability, as previously shown with adult GBM cells. To induce hypoxia, we selected CoCl2, a chemical hypoxia model that increases HIF1α and upregulated genes that encode proteins related to pH regulation such as CA9 and CA12. Then, our results have shown that a hypoxia-mimic is sufficient to increase the expression of HIF1α, CA9, and CA12, suggesting that this is an adequate condition to investigate Indisulam’s effects on pHGG cells.

Carbonic anhydrases, controlled by oxygen levels via the HIF1α [43], are overexpressed in several solid tumors, with a fundamental role in tumor pH homeostasis and very low, or negligible expression, in normal tissues [14]. The CoCl2 increases HIF1α in dose (25 µM to 200 µM) and time-dependent, according to the cell type selected, and generally, the stabilization is observed for 2 h, with a maximum of 12–48 h [44,45,46,47,48,49]. A similar effect of HIF expression was observed in breast cancer and HGG cells exposed to hypoxia conditions [34,44]. For the U87 GBM cell, in the treatment with CoCl2 100 mM, the expression of HIF1α was time-dependent. In addition, doses ≤200 µM CoCl2, and 24 h of treatment, did not increase or decrease cell viability of glioblastoma and colon cancer; however, the viability decreased after 48 h of treatment [45,47]. In this way, recent publications have demonstrated that the hypoxia exposition for 72 h to 168 h significantly decreased the tumor cell viability [34,46,47]. According to Rana et al., the cytotoxic effect of CoCl2 is due to the inhibition of DNA repair pathways and accumulation of ROS (Reactive Oxygen Species), with consequent cell death [44]. In vitro, this cytotoxicity appears to be dose-dependent, as observed in breast cancer cells [44,49] and the KNS-42 cell line in the present study. Furthermore, using 50 and 100 µM of CoCl2 promoted an increase in the number of both cell lines selected to develop this study (SF188 and KNS-42). Thus, as the observed effect did not decrease but increased cell proliferation for the selected CoCl2 doses, the analysis of the Indisulam effect in the present work was not impaired. High expression of CAs, associated with high proliferation and expansion of tumor cells that deprive the tissues of oxygen, has been described in brain tumors and functionally linked to the malignant behavior of these cancer cells [50]. A metabolic adaptation of GBM to these oxygen concentrations is also present through the Warburg effect, leading to lactate production as a metabolic by-product and increasing resistance to radiotherapy [50]. Furthermore, malignant cells, especially tumor stem cells, can endure hypoxic conditions by changing their expression of genes involved in cell proliferation, metabolism, apoptosis, and angiogenesis, thus resulting in tumor expansion, treatment resistance, and metabolic adaptation [51]. The high expression of CA9 and CA12 in a hypoxic microenvironment is currently considered a potential prognostic biomarker and therapeutic target [10]. CAs have restricted expression in non-neoplastic tissues, suggesting that it may be possible to inhibit their pro-tumoral effects in the microenvironmental of the tumor, with minimal toxicity on normal tissues [52]. Accordingly, this study aimed to use Indisulam to inhibit the upregulation of CA9 and CA12 observed in hypoxic pHGG cells. Interestingly, we observed inhibition of mRNA and protein expression for CA9 but not for CA12. Similar results have been reported in adult GBM cells [20]. According to Chiche et al., 2009 [53], this effect might represent a compensatory feedback mechanism for tumor cells to maintain an intracellular alkaline pH.Overall, our results suggest that Indisulam triggered an antitumor effect through the reduction of cell proliferation and clonogenicity of pHGG cells. The antitumor activity of Indisulam has been demonstrated in other tumor types [20,21,25,54,55]. This effect may be related to the ability of Indisulam to inhibit cell cycle progression through G1/S and G2/S phase disarray and decreased expression of cyclins A, B1, and E, and CDK2/CDK4 and altered glucose metabolism [56,57]. In turn, the antiproliferative effect of Indisulam may be caused by dysregulation of cellular pH after inhibition of CA9 [53,58,59,60,61], making the tumor more sensitive to external agents [11,62,63,64,65]. Inhibition of CAs by Indisulam has been shown to downregulate genes involved in chemoresistance and reduce proliferation capacity and invasion of cancer cells [21]. Then, a parallel mechanism by which hypoxia promotes pHGG chemoresistance is through inhibiting pro-apoptotic pathways. Hypoxia is considered a tumor microenvironment factor that favors the resistance to therapy. For instance, under hypoxic conditions, the proapoptotic protein appears to be modified and unable to interfere with pro-survival factors; therefore, an increase in the expression of antiapoptotic proteins has been reported [47,48]. For breast cancer, hypoxia upregulated proapoptotic protein expression [44]. In this way, the family of apoptosis-related proteins BCL-2 (B-cell lymphoma 2) inhibits apoptosis, and BAX (BCL-2-associated X protein) promotes a proapoptotic effect [47]. However, the criterion for programmed cell death is the imbalance in the ratio of BCL-2/BAX expression.Our results indicate that Indisulam, inhibiting CA9, promotes apoptosis in pHGG cells in hypoxia. The BAX/BCL-2 ratio increase in the Indisulam-treated group makes evident the imbalance in the pro-survival machinery and triggering of the apoptotic process [66,67]. In this sense, the drastic reduction of the anti-apoptotic factor BCL-2, more markedly than the proapoptotic factor BAX, leads to cell death through the intrinsic pathway of apoptosis, which is controlled by members of the BCL-2 family [67,68]. The CA9 inhibition-induced apoptosis by other agents has also been described in colorectal [69], cervical [70,71] and breast cancer [72]. Inhibition of CA9 decreases intracellular pH, causing DNA damage and ROS accumulation that activate apoptosis [73]. Cianchi et al. have also found that the inhibition of CA9 leads to the phosphorylation of p38/MAPK as well as the synthesis of peroxynitrite [73]. In addition, the p38/MAPK factor is described as a mediator of apoptosis [74], possibly relating this pathway to apoptotic induction promoted by the inhibition of CAs. Recently, it was observed that the induction of apoptosis after inhibition of CA9 and CA12 is mediated by the activation of caspase 7, with an increase in ROS and a decrease in PARP-1 levels [75,76]. Moreover, Indisulam promotes the overexpression of p53 and p21, leading to apoptosis [25,77]. The pro-apoptotic stimulus is one of the most efficient effects in non-surgical treatments [78], and the fact that Indisulam presents this activity in pHGG cells highlights its potential in anti-cancer therapeutic strategies. The current standard-of-care chemotherapy to treat high-grade glioma, including pHGG, is Temozolomide (TMZ). However, the treatment is not curative, even with a combination of surgery, radiotherapy, and adjuvant chemotherapy, with relapse being a frequent outcome [79]. The best-known resistance mechanism to TMZ is effectuated by O6-Methylguanine-DNA methyltransferase (MGMT), which corrects DNA damage caused by TMZ and reduces the effectiveness of the treatment [80]. This resistance mechanism is probably also associated with the poor response to TMZ in pHGG [81,82]. The combination of TMZ with carbonic anhydrase inhibitors [52], including Indisulam [20], results in increased TMZ efficacy and indicates a promising strategy toward more effective treatments. In addition, Indisulam has already been tested in combination with another chemo- and radiotherapeutic agents in vitro, in vivo, and clinical studies, showing promising results as an adjuvant strategy [20,21,83,84,85]. Despite the limitation of working with established cell lines and in vitro models, several studies and different research groups with the purpose of better understanding the tumor biology of pediatric glioblastoma have already used the SF188 and KNS-42 cell lines, being two reliable models of investigation of this tumor subtype. Although in vitro tumor models with cell culture have provided essential tools for cancer research and serve as low-cost models for drug therapy investigation, serial passage of cell lines can further cause genotypic and phenotypic variation over an extended period. Therefore, further assays with different tumor models are needed to validate the results found in this study.