Targeting Several Biologically Reported Targets of Glioblastoma Multiforme by Assaying 2D and 3D Cultured Cells

Glioblastoma multiforme (GBM) is account for 70% of all primary malignancies of the central nervous system. The median survival of human patients after treatment is around 15 months. There are several biological targets which have been reported that can be pursued using ligands with varied structures to treat this disease. In our group, we have developed several ligands that target a wide range of proteins involved in anticancer effects, such as histone deacetylase (HDACs), G protein-coupled estrogen receptor 1 (GPER), estrogen receptor-beta (ERβ) and NADPH oxidase (NOX), that were screened on bidimensional (2D) and tridimensional (3D) GBM stem cells like (GSC). Our results show that some HDAC inhibitors show antiproliferative properties at 21–32 µM. These results suggest that in this 3D culture, HDACs could be the most relevant targets that are modulated to induce the antiproliferative effects that require in the future further experimental studies.


Introduction
Glioblastoma multiforme (GBM) is one of the most common glial tumors, presenting poor prognosis, and less than 5% survival over the 5 years after initial diagnosis (Liu et al. 2006;Kim et al. 2012;Singh et al. 2004). Multiple therapeutic strategies and biomarkers have risen over years, in this study we selected some of them in order to evaluate potential antimigratory or antiproliferative effect, leading potentially to interesting therapy for GBM during deeper investigations. It has been demonstrated that histone deacetylases (HDACs), (able to remove acetyl groups from histones protein on DNA) Lee et al. 2017 regulate the expression and activity of numerous proteins involved in cancer initiation, proliferation and progression and seem to be a potential target against GBM (Yelton and Ray 2018;Lee et al. 2017). Some HDAC inhibitors already developed exhibit hydroxamic acid groups that chelate pivotal Zn 2+ ions to the catalytic activity of the enzyme (Rajak et al. 2013), but due to a lack of specificity of those compounds, other chemical strategies have been employed to reach the nonpolar cavities of HDACs (Rajak et al. 2013;Wang 2009). G protein estrogen receptor (GPER) is implicated in diverse biological effects, including regulation of immune, endocrine, neuronal and cardiovascular functions. In cancer, GPER activation is already well described as regulating cell growth and proliferation in several cell line, migration and apoptotic cell death, but has not been studied in Glioma. Consequently, the use of GPER modulators to explore his role in GBM proliferation and migration appears as a great tool for in vivo studies, as well as possible target to treat GBM (Feldman and Limbird 2017). Estrogens have shown implication in development of gliomas, and estrogens appear having neuroprotective effects and improve glioblastoma outcome for women (Wan et al. 2018;Lan et al. 2017). Want et al. reported higher expression of ERs receptor in glioma cells in comparison with glial cells (Wan et al. 2018) in which and ERβ appears as a tumor suppressor in GBM (Liu et al. 2018). Consequently ERβ also appears as a promising target to treat GBM (Lan et al. 2017;Zhou et al. 2019).
Free radicals in GBM play an important role in cell proliferation being the NADPH oxidase (NOX) as putative targets for cancer therapy. NOX family is the main source of reactive oxygen species (ROS) (Meitzler et al. 2014) which could be an important mediator of neuroinflammation (Yang et al. 2020). The use of NOX inhibitors has been reported to inhibit invadopodium formation, the inflammatory response and cancer cell migration in vitro. Furthermore, NOXs as pharmacological targets appear as an interesting strategy for GBM treatment (Munson et al. 2013;Nayernia et al. 2014).
In this study, we explored a diverse set of ligands containing a wide range of functional groups as possible therapeutic agents to treat GBM. Then, 23 compounds were synthetized, including NADPH oxidase inhibitors (Ia-Ib), ER modulators (IIa-IId, IIIa-IIId), HDAC inhibitors (IVa-IVb, VIa-VIg) and GPER inhibitors (Va-Vd). Then, we performed a first preliminary screening in 2D and 3D cell culture to observe potential antiproliferative and/or antimigrating effect on Glioblastoma Stem Cell (GSC). This study is a first step leading to a selection on compound for future and deeper in vitro and in vivo investigations.

Chemical Synthesis
All commercial grade reagents were used without further purification. Analytical thin-layer chromatography (TLC) was performed on silica gel 60 F254-coated aluminum sheets (0.25 mm thickness) with a fluorescent indicator. Visualization was accomplished with UV light (254 nm). Flash chromatography was performed using silica gel 60 (230-400 mesh). Melting points were determined in open capillary tubes with an IA 91,000 electrothermal melting point apparatus (Electrothermal, Bibby Scientific, Staffordshire, ST15 OSA, UK) and were uncorrected. 1 H and 13 C NMR spectra were recorded on either Bruker Avance III 750 or 400 MHz spectrometers or on a 500 MHz Varian System using deuterated dimethyl sulfoxide (DMSO-d 6 ) or chloroform (CDCl 3 ) as the solvent and TMS as the internal standard. Chemical shifts are given in ppm (δ) and are referenced to TMS as the internal standard. Coupling constant values are quoted in Hertz, with normal abbreviations (s, singlet; d, doublet; t, triplet; br, broad; m, multiplet). For complete data assignments, 2D HSQC and HMBC NMR spectra were used. IR spectra were obtained on a Spectrum 2000 Perkin Elmer FT-IR spectrometer. Only significant absorption bands are shown with absorption values expressed as wavenumbers (cm −1 ). Electrospray ionization (ESI) high-resolution mass spectrometry was performed with a Bruker micrQTOF-Q II instrument and an Agilent 6545 Q-TOF LC/MS. The chemical purity was determined by HPLC using an Agilent 1200 Infinity Series system.

Synthesis of Propanenitriles IIIa-IIId
NaBH 4 (1 equiv) was added to a 60-70 °C solution of acrylonitriles IIa-IId (1 equiv) in 10 mL of EtOH under a N 2 atmosphere. After stirring for 2.5 h, the reaction was cooled to room temperature and quenched with water. The reaction mixture was acidified with 6 M HCl, extracted with EtOAc (3 × 25 mL), washed with brine, dried over Na 2 SO 4 and concentrated in vacuo. The residue was purified by flash chromatography on silica (hexane/EtOAc 9:1) to give propionitriles IIIa-IIId (Scheme 4).

Synthesis of Valproic Acid Derivatives IVa-IVb as HDAC Inhibitors
Thionyl chloride SOCl 2 (3 mmol) was added dropwise to VPA (3.5 mmol) at 0 °C under a nitrogen atmosphere with constant stirring. The reaction mixture stirred for 1 h under the same conditions and then overnight at room temperature. After this time, Et 3 N (0.55 g, 5.45 mmol), the corresponding aniline (2.8 mmol) and 5 mL of hexane were added, and the reaction continued stirring for 3 h. At the end of the reaction time, 15 mL of EtOAc was added and followed by washing with a saturated NaHCO 3 solution (1 × 15 mL) and brine (1 × 15 mL), and the organic phase was dried over anhydrous Na 2 SO 4 (Scheme 5). The solvent was evaporated under reduced pressure, and the residue was purified by column flash chromatography using an eluent mixture of hexane/EtOAc (8:2).

Cell Culture
Isolation and culture of GBM cells using the classical nonadherent neurospheres (NS) method was performed according to a protocol used to derive NS from the adult human spinal cord as described by Dromard et al. (Dromard et al. 2008) and adapted by Guichet et al. for Gli4 cells (Guichet et al. 2013). In proliferation condition when GBM cells form NS, cells were cultivated in DMEM/ F12 medium supplemented with glucose, glutamine, insulin, N2, epidermal growth factor and fibroblast growth factor (proliferation medium). For migration, GSCs cell culture was realized in DMEM/F12 medium supplemented with glucose, glutamine, insulin, N2 and fetal bovine serum (0.5%). Prior to cell seeding, 2D or NF were either functionalized or not with poly-D-lysine added overnight and then the addition of LN (sigma L2020) (0.05 mg/mL) 1 h at 37 °C (Scheme 6).
For migration assay, same size GBM NS were performed by seeding 5000 cells in round bottom wells and remained in culture during 2 days until formation of single neurosphere due to sedimentation. Compounds were prepared in 100% DMSO and diluted in media to obtain in each condition a final percentage of 0.3% DMSO. Then GSCs NS were deposited on the top of NF + LN and let to migrate during 5 days in total, with 72 h in presence of either HDAC compounds or 0.3% DMSO alone for control.

Immunofluorescence
GSCs cultures were fixed by 4% PFA. Cell and brain sections were blocked and permeabilized using PBS-triton 0.5%-horse serum 5%. Fluorochrome-coupled phalloidin 546 was incubated 2 h at room temperature (dilution 1/500). Image acquisition was realized using Zeiss Axioimager Z1/Zen (with an apotome). Migration quantifications were done using ZEN 2012 software, and migration capacity was quantified by measuring an area of migration. To measure migration areas, we subtracted the area of the NS containing non-migrating cells from the total area where cells were detected.

Antiproliferative Activity: MTT Assay
The GBM stem cells are primary cell culture coming from patient, gently donated by Luc Bauchet from the Hospital Scheme 5 Synthesis of valproic acid derivatives (IVa-IVb)

Scheme 6 Synthesis of GPER ligands (Va-Vd)
Guy de Chauliac and isolated by the Professor Jean-Philippe Hugnot (Guichet et al. 2013;Saleh et al. 2019). Gli4 were seeded into 96-well plates at 3000 cells per well in a 2D system and 15,000 cells in a 3D system, 2D culture were culture with laminin (2D + LN) or without laminin (2D-LN), while 3D systems were also culture with laminin (3D + LN) as single cells or without laminin (3D-LN) for collective migrations. Once the cells were adherent, different concentrations of the compound were added at the µM range (Table 1) to each well at a final DMSO concentration of 0.3%, and Phostine (PST3) was tested as positive control at 1 µM in 0.3% DMSO. The MTT assay was carried out as previously described 72 h after the addition of compounds (Mosmann 1983). Active compounds were tested in four independent experiments, and each experiment was carried out in triplicate for each condition.

Statistical Analyses
The experiments were carried out at least in triplicate, depending on the number of independent variables. Student's or ANOVA statistical tests were applied using Serf software (bram.org/serf/CellsAndMaps.php) and GraphPad. EC 50 and Ki values were calculated using the Hill equation from the dose − log response curves.

Results and Discussion
Cancer cells are presenting numerous deregulation favoring the overexpression of several proteins which could be a target for drug design as GPER1 (Martínez-Muñoz et al. 2018;Zacarias-Lara et al. 2018) leading to cell proliferation and migration. GBM is a disease with a poor survival rate and limited therapeutic options. Therefore, a multitargeted approach for drug screening is necessary in order to explore several pharmacological options for GBM treatment (Shea et al. 2016). Traditional 2D cell cultures (monolayers) are widely used to assay antiproliferative activity as one of the first approaches to decide if a compound should be further investigated as an anticancer drug (Shan et al. 2018;Verjans et al. 2018). However, it is well admitted that the use of this strategy is the cause of the failure to discover new anticancer drugs because 2D cell cultures not mimic the in vivo 3D fibrillar microenvironment, cell-cell interactions and cell-extracellular matrix interactions (Lv et al. 2017). Therefore, 3D cultures may be a powerful tool for drug discovery, offering a more realistic approach for drug screening (Lv et al. 2017). Our group has previously focused work on drug design employing in silico tools (Zacarias-Lara et al. 2019;Sixto-Lopez et al. 2020). In this drug screening preliminary study, compounds are chemically synthesized Scheme 7 General solid-phase synthesis of hydroxamic acid derivatives VIa-VIg (Sixto-Lopez et al. 2020) targeting the 23 candidates potentially able to impair proliferative and migrating behavior of GSCs (Zacarias-Lara et al. 2019;Sixto-Lopez et al. 2020). (Rajak et al. 2013;Wang 2009;Lan et al. 2017;Batistatou et al. 2004;Wan et al. 2018;Liu et al. 2018;Zhou et al. 2019;Martinez-Archundia et al. 2018;Feldman and Limbird 2017;Meitzler et al. 2014;Munson et al. 2013;Nayernia et al. 2014). Those targeted proteins are tackling several pathways, including HDACs leading to epigenetic modulation, ROS production and GPER and ER modulators associated in multiple cancer with cell proliferation and metastasis. First, to evaluate the potential antiproliferative effect of the compounds, we performed MTT test on GSC in 2D and 3D with or without LN by adding the different compounds during 72 h. During this first screening, the following compounds (from the Table 1) appeared unefficient on the GSC for the: NADPH oxidase inhibitors Ia-Ib, estrogen receptor modulators IId and IIIa-IIId, valproic acid derivatives IVa-IVb, GPER ligands Va-Vd and two hydroxamic acid derivatives VIa and VIg (Table 1). Those compounds didn't present any cytotoxic or antiproliferative potential. Nevertheless, five HDACis hydroxamic acid derivatives (VIb-VIf) exhibited equal antiproliferative efficiency on GSC either in 2D or in 3D. The assays were performed in presence and absence of laminin, in order to recapitulate collective and individual migration mode (Saleh et al. 2019). The antiproliferative effect of the compounds (VIb-VIf) targeting HDACs didn't present any specificity against single or collective migratory cells (Table 2) (Fig. 1d). ERβ, here targeted by the compounds IIa, has been described in other studies as inhibiting the proliferation of glioma cell lines (Sareddy et al. 2016). However, the estrogen modulator IIa seems to present antiproliferative properties in 2D, but not in 3D, validating the 3D model choice as stringent drug screening model for candidate selection.
The HDAC antiproliferative effects are also already well described Li et al. 2015). For instance, HDAC8 inhibition by a specific inhibitor produces DNA damage, cell cycle arrest and decreases cell viability and those effects were associated with the inhibition of the O 6 -methyl-guanine DNA methyltransferase (Santos- Barriopedro et al. 2019). In its turn, HDAC6 inhibition decreases GBM cell growth due to of SMAD2 phosphorylation inhibition and down expression regulation of p21 (Li et al. 2015).
HDACi (VIb-VIf) were previously designed and synthesized depicting antiproliferative properties against a panel of cancer cell line such as hepatocellular carcinoma (HepG2), pancreatic cancer (MIA PaCa-2), breast cancer (MCF-7 and HCC1954), renal cancer (RCC4-VHL and RCC4-VA) and neuroblastoma (SH-SY5Y). According to molecular docking and MD simulations studies, these compounds might block Table 1 Percent inhibition of cell proliferation of the tested compounds obtained in 2D and 3D GBM cell cultures with or without laminin *These compounds were selected for a second test. In the table, the indicated value represents the average of the first and second experiments 2D-LN = 2D cell culture without laminin; 2D + LN = 2D cell culture with laminin; 3D-LN = 3D cell culture without laminin; 3D + LN = 3D cell culture with laminin  the catalytic cavity of HDAC1, HDAC6 and HDAC8, which could be due to their aromatic portion that is able to interact with residues surrounding the catalytic tunnel (Sixto-Lopez et al. 2020). Therefore, due to the previous promissory activity against several cancer cell lines, these HDACic were tested against GBM cell lines in order to explore the antiproliferative and antimigratory activity against this disease. GSCs during migration start differentiation process, hence decreasing proliferation capacity. As HDACs compounds are described as targeting proliferation, also following the proliferation assay for the compounds (VIb-VIf), we decided to evaluate HDACs inhibitor effects on migration ( Fig. 1a, b) and then potential impairment of cell viability. We deposited same size neurosphere (NS) (5000 GSCs) on those 3D matrices with LN, leave them to migrate during 5 days and applied treatment during 72 h and quantified the NS size and the migration area around the NS after subtracting the NS area from the total migration (Fig. 1c). The HDACs inhibitor compounds (VIb-VIf) appear as not decreasing the migrating capacity of the GSC comparing to the control condition. We can hypothesize that HDACs hydroxamic acid derivatives are not cytotoxic but impair proliferation capacity without impairing migration in 3D fibrillar matrix. Also, the compounds VIe and VIf show Fig. 1 a 5000 cells NS deposited on 3D + LN after 5 days of migration and percent survival of cell proliferation of the tested compounds obtained from the 2D and 3D GBM cell cultures with (2D + LN or 3D + LN) or without (2D-LN or 3D-LN) laminin. VIf, VIc, VIe, VId, VIb and IIa only NS size decrease after 72 h treatment illustrating the antiproliferative effect of the HDACs, independently of a potential antimigration effect. Following those preliminary experiment and for upcoming studies, we have, on first hand, to evaluate EC50 of the HDACs hydroxamic acid derivatives and, on the other hand, evaluate in vivo efficiency of those compounds, potential cytotoxicity and side effects. In fact, HDACs, are well described as having great therapeutic potential but also pleiotropic effects at the cellular and systemic levels and seems to present ability to modulate the immune system affecting many cellular functions. Indeed, their use is recently extended far beyond cancer treatment in terms of their therapeutic capacity (Hull et al. 2016). Consequently, it seems essential to perform in vivo trials as next step for those HDACs inhibitor potential evaluation and ultimate determine their action mechanism.

Conclusion
In this study, we are evaluating the efficiency of 24 compounds targeting some deregulated proteins described as improving cancer proliferation (including HDACs, GPER and ERβ modulators, NADPH oxidase inhibitor), therefore, impairing patient survival. Among those 24 compounds, according to proliferation and migration assay results in 2D and 3D, we selected HDACs inhibitor compounds (VIb-VIf) as possible candidate for in vivo trial in order to evaluate their antitumoral efficiency and potential toxicity.