Targeting of glioma stem-like cells with a parthenolide derivative ACT001 through inhibition of AEBP1/PI3K/AKT signaling

Glioblastoma (GBM) is the most lethal primary brain tumor in adults with a median survival of around 15 months. A potential treatment strategy involves targeting glioma stem-like cells (GSCs) that are able to initiate, maintain, and repopulate the tumor mass. Here, we identify ACT001, a parthenolide derivative, targeting GSCs through regulation of adipocyte enhancer binding protein 1 (AEBP1) signaling. Methods: The effects of ACT001 on cell survival of normal human astrocytes (NHA) and patient-derived glioma stem-like cells (GSCs) were evaluated. RNA-Seq were performed to detect differentially expressed genes. ACT001 efficacy as a single agent or in combination with SHP-2 inhibitor SHP099 was assessed using a GSC orthotopic xenograft model. Results: GSCs exhibit high response to ACT001 in compared with normal human astrocytes. AEBP1 is a putative target of ACT001 by RNA-Seq analysis, which expression associates with prognosis of GBM patients. Knockdown of AEBP1 inhibits GSC proliferation and glioma sphere formation. Treatment with ACT001 or PI3K inhibitor or AEBP1 depletion would impair AKT phosphorylation and GSC proliferation, whereas constitutive AKT activation rescues ACT001 treatment or AEBP1 depletion-inhibited cell proliferation. Moreover, ACT001 blocks TGF-β-activated AEBP1/AKT signaling in GSCs. ACT001 exhibits antitumor activity either as a single agent or in combination with SHP099, which provides significant survival benefits for GSC tumor xenograft-bearing animals. Conclusions: Our data demonstrate AEBP1 as a new druggable target in GBM and ACT001 as a potential therapeutic option for improving the clinical treatment of GBM in combination with SHP099.


Introduction
Glioblastoma (GBM) is the most common and aggressive primary brain cancer with poor prognosis [1,2]. Typically, GBM patients underwent surgical resection, combined with postoperative radiotherapy and chemotherapy [3]. Chemotherapies including cisplatin and temozolomide (TMZ) are important clinical adjuvant therapy [4,5]. Given chemoresistance, GBM patients still have a poor median survival rate of 15 months. Glioma stem-like cells (GSCs) are recognized as a special population of GBM cells that contributes to tumorigenesis, radiochemoresistance, and recurrence [6][7][8][9]. Developing new therapeutic drugs combined with conventional therapy is strongly needed for GBM patients. ACT001 (also known as dimethylaminomicheliolide, DMAMCL) is developed from parthenolide, a sesquiterpene lactone by Accendatech Co., Ltd (Tianjin, China) [10]. ACT001 showed Ivyspring International Publisher anti-tumor functions in various cancers, including leukemia [11,12] and breast cancer [13]. ACT001 has been certified as an orphan drug by both the Food and Drug Administrations in the United States and in the European Union [10,11]. ACT001 was shown to selectively inhibit acute myelogenous leukemia stem and progenitor cells [11,12]. More important, ACT001 resulted in good control of GBM growth in China and Australia in Phase I clinical trials [12], and now it is currently undergoing Phase II clinical trials. ACT001 showed synergistic effects with combination of cisplatin by inhibition of PI3K/AKT signaling pathway in GBM [14]. However, the mechanisms by which ACT001 inhibits glioma progression is not clear, which limits the application of this new therapeutic target.
Adipocyte enhancer binding protein 1 (AEBP1) was at first discovered as a transcriptional repressor that binds to the AE-1 element of the ap-2 gene, coding for the fatty acid binding protein (FABP4) [15]. Then, AEBP1 was identified as an oncogenic transcription factor [16][17][18]. AEBP1 upregulation was observed in prostate cancer [16], which conferred to acquired resistance to BRAF (V600E) inhibition in melanoma. Through upregulation of sonic Hedgehog and NF-kappa B (NF-κB) pathway, AEBP1 functioned as a pro-inflammatory mediator and promoted mammary cell hyperplasia [17]. In gliomas, AEBP1 up-regulated PI3KCB transcription, leading to enhanced AKT phosphorylation and cell proliferation [18].
Here, we investigated whether AEBP1 is a new target of ACT001 in gliomas with activating AKT signaling. We determined the response of AEBP1 knockdown on ACT001 response in GSCs. Treatment with ACT001 either as a single agent or in combination with SHP099 was further performed using an orthotopic xenograft model of GSC.

shRNA knockdown (KD) and infection
DNA and packaging plasmids were transduced into HEK293T cells using Lipofectamine 2000 reagent according to manufacturer's instruction (Invitrogen 52758). After 48-h transfection, the viral-containing supernatants were filtered and added into the culture media supplemented with 8 mg/ml polybrene. AEBP1 and AKT short hairpin RNAs (shRNAs) and control shRNAs were purchased from Shanghai GeneChem Company.

Cell proliferation and viability assays
Cells were plated in triplicate wells of a 96-well microplate (6000 cells/well). Cell proliferation analysis was performed using a WST-1 assay kit (BioVision Inc.). After 24 h, cells were treated with vehicle (ddH 2 O) or ACT001 from 0.1 to 100 μM for 72 h and cell viability was then analyzed using the WST-1 assay kit.
Half-maximal inhibitory concentration (IC 50 ) values were calculated from fitted concentration-response curves obtained from at least 3 independent experiments. IC 50 values were determined using GraphPad Prism 7 nonlinear regression curve fit.

Colony formation and limiting dilution glioma sphere-forming assays
Colony formation and limiting dilution assays were performed as we previously described [8,22]. In brief, dissociated cells from glioma spheres were seeded in 96-well plates containing GSC cell culture medium (20-200 cells per well). After 7 days, each well was examined for formation of tumor spheres. Stem cell frequency using extreme limiting dilution analysis was calculated as described in http://bioinf.wehi.edu.au/software/elda/.

Isobologram and combination index analysis
Cells were seeded in 96-well plates and treated with ACT001, SHP009 or combination for 72 h. Media were replaced with fresh media containing increasing concentrations of ACT001 (ranging from 0 μM to 64 μM) and SHP099 (ranging from 0 μM to 19.2 μM), with each concentration tested in triplicate wells. The concentrations used corresponded to 0.0625, 0.125, 0.25, 0.5, 1, 2, 4 times the IC50 of each agent. The concentration of the single drug that inhibits 50% of cell proliferation (IC50) was determined by fitting the dose-response curve utilizing the CompuSyn software. To calculate the Combination Index (CI), using the method of constant ratio drug combination proposed by Chou and Talalay [23,24] and described in result. The CI was calculated using the formula: C.I. = C A , X /IC X , A +C B , X /IC X , B , where C A , X and C B , X are the concentration of ACT001 and SHP009 used in combination to achieve 50% drug effect. IC X , A and IC X , B are the concentrations for single agents to achieve the same effect. A CI of less then, equal to, or more than 1 indicates synergic, additive or antagonistic effect, respectively.

Cell cycle analysis
After ACT001 treatment, culture supernatant and trypsinized cells were combined and centrifuged. After two washes in cold PBS, cells were fixed with 70% ethanol overnight at 4ºC and resuspended in 0.5 ml staining solution supplemented with 10 μl propidium iodine (PI) and 10 μl RNase A (YEASON). After incubation for 30 min at 37ºC in the dark, analysis was performed on a FACS ACCuri C6 (BD Biosciences).

Tumorigenicity Studies
All animal experiments were conducted in accordance with the animal use guidelines by the Shanghai Jiao Tong University Institutional Animal Care and Use Committee. Athymic nu/nu female mice aged 6 weeks (SLAC, Shanghai, P.R.C) were used. In total 2,000 GSCs were stereotactically implanted into the brain of the animals as previously described [20]. A scheme of treatment was described in the corresponding figure legend. Mice were euthanized when neuropathological symptoms developed.

Statistical Analysis
All statistical analyses were performed using GraphPad Prism 7 software (GraphPad Software Inc.). A minimum of 3 independent biological replicates were analyzed using Student's two-tailed t-test or one-way ANOVA (Newman-Keuls post hoc test) as specified in the figure legends. Data are expressed as the mean ± SEM of 3 independent experiments. Survival analysis was performed using Kaplan-Meier analysis and the log-rank test. p < 0.05 was considered significant.

Results
GSCs are more responsive to ACT001 treatment ACT001 is an orally available chemotherapeutic agent with IC 50 of 27 μM and 21 μM against C6 and U87 glioma cells, respectively [10]. To assess the effects of ACT001 on glioma stem-like cells (GSCs), we performed cell viability analysis and found that compared to normal human astrocytes (NHA), patient-derived GSC 1123 and R39 cells were more responsive to ACT001 ( Figure 1A and 1B). ACT001 markedly inhibited glioma sphere formation in GSC 1123 and R39 cells ( Figure 1C). We also performed flow cytometric analyses of cell cycle and found that compared the vehicle, ACT001 did not impair the cell cycle in NHA, GSC 1123, or R39 cells (Supplementary Figure 1). This data indicate that ACT001 is more selective for GSCs than NHA.

ACT001 inhibits the expression of AEBP1, a prognostic factor of gliomas
To investigate the underlying mechanism by which GSCs are more responsive to ACT001 treatment, we employed RNA-Seq analysis in GSC 1123 cells treated with or without ACT001. Gene expression profile analysis identified 858 genes whose expression was significantly reduced by ACT001 treatment and 1075 genes whose expression was markedly increased by ACT001 treatment (false discovery rate < 0.05 and a folder change > 2) ( Figure  2A). AEBP1 is a top one in ACT001 down-regulated genes (Figure 2A). Given that AEBP1 is important for various cancers [16,17], including glioma [18], we further investigated if it is a target of ACT001 in GSCs. As shown in Figure 2B, the expression level of AEBP1 was significantly decreased by both 5 and 10 μM ACT001 treatment. This result was further validated by qRT-PCR ( Figure 2C) and Western blotting (WB) analyses ( Figure 2D) in GSC 1123 and R39 cells. This data suggests that AEBP1 is a putative target of ACT001 in GSCs.
We then assessed the expression levels of AEBP1 in normal brain tissues and clinical specimens of glioma patients. We downloaded the RAMBRANDT dataset (http://gliovis.bioinfo.cnio.es/) and found that compared to normal brain tissues, the expression levels of AEBP1 mRNA was significantly high in low grade tumors (Grade II and III) and GBM. Compared to low grade gliomas (Grade II and III), the expression levels of AEBP1 mRNA was markedly upregulated in GBM. This data suggests that the expression of AEBP1 is correlated with glioma progression.
Next, we determined the relationship of AEBP1 expression and glioma patient survival using the RAMBRANDT and the Cancer Genome Atlas (TCGA) datasets. As shown in Figure 2F, Kaplan-Meier survival analysis revealed a statistically significant worse prognosis for glioma patients with high AEBP1 (> median level) compared to those with low levels of AEBP1 (< median level) in the RAMBRANDT dataset. The median patient survival times of these patients were 12.8 and 37.4 months, respectively. Kaplan-Meier survival analysis of the TCGA low-grade glioma (TCGA-LGG) and TCGA GBM RNA-Seq datasets further indicated that AEBP1 is a prognostic factor for glioma patients (Figures 2G and  2H). These data demonstrate that ACT001 may inhibit glioma progression through targeting AEBP1, a prognostic factor for gliomas.

ACT001 targets AEBP1 to decrease AKT phosphorylation and GSC proliferation
ACT001 inhibits glioma cell proliferation by inhibiting PI3K/AKT activation [14] and AEBP1 activates PI3K/AKT signaling through upregulation of PI3KCB transcription [18]. Thus, we investigated whether AEBP1 mediates ACT001 response through PI3K/AKT signaling. We first treated GSC 1123 and R39 cells with ACT001 and detected AEBP1 expression and AKT phosphorylation (p-AKT). As shown in Figure 3A, consistent with the above, ACT001 treatment decreased AEBP1 protein expression. ACT001 treatment also significantly attenuated p-AKT in GSC 1123 and R39 cells ( Figure  3A). Then, AEBP1 was knocked down by two different shRNAs ( Figure 3B). Compared with the shRNA control, AEBP1 knockdown (KD) markedly decreased p-AKT ( Figure 3B) and cell proliferation ( Figure 3C) in GSC 1123 and R39 cells. However, AEBP1 KD in combination with ACT001 treatment did not further reduced p-AKT and cell proliferation compared to ACT001 treatment ( Figures 3B and 3C). Additionally, ectopic expression of AEBP1 increased cell proliferation and attenuated ACT001-inhibited cell proliferation in GSC 1123 and R39 cells ( Figures  3D and 3E). These data suggest that ACT001 targets AEBP1 to decrease p-AKT and GSC proliferation.

ACT001 inhibits GSC proliferation through AEBP1/PI3K/AKT signaling
To further investigate whether ACT001 inhibits GSC proliferation through AEBP1-activated AKT, we treated GSC 1123 cells with PI3K inhibitor LY294002 with or without ACT001. As shown in Figures 4A and  4B, compared to the vehicle control, LY294002 treatment inhibited p-AKT and cell proliferation but not AEBP1 expression. ACT001 treatment in combination with LY294002 further decreased p-AKT and cell proliferation compared with the single treatment of ACT001 or LY294002 (Figures 4A and  4B).
Next, we determined the effects of AKT activation on AEBP1-mediated ACT001 response. As shown in Figures 4D and 4E, compared with the vector control, ectopic expression of a constitutively activated (CA) AKT (Myr-AKT) mutant did affect AEBP1 expression decreased by ACT001 treatment or/and AEBP1 KD in GSC 1123/shC or 1123/shAEBP1 cells. However, overexpression of the CA AKT mutant rescued ACT001 treatment or/and AEBP1 KD-decreased p-AKT ( Figure 4D) and cell proliferation ( Figure 4E). These data indicate that ACT001 inhibits GSC proliferation through AEBP1-activated PI3K/AKT signaling.

ACT001 inhibits GSC tumorigenicity in combination with SHP-2 inhibitor SHP099
SHP099 is a highly potent and orally available inhibitor of SHP-2 and inhibits multiple cancer progression [33][34][35]. Our recent study demonstrated that SHP099 exhibits antitumor activity either as a single agent or in combination with TMZ and provides significant survival benefits for GBM tumor xenograft-bearing animals [20]. We first test if the superior effect of ACT001 plus SHP099 by the synergistic effect of the two compounds addition. We used the Chou-Talalay method for drug combination studies [23] and determined the dose-response curves of ACT001 and SHP099 individually and in combination at a constant potency ratio. For each of the GSC cell lines, we observed the single-agent activity of ACT001 and SHP099 and a remarkably potent synergistic inhibition of cell viability after combined treatment. The calculated combination index (CI) values for GSC 1123 and R39 cells at doses effective in reducing cell viability by 90% were 0.48, 0.21, respectively ( Figure 6A). The CI value of ACT001 in combination with SHP099 was less than 1 (< 1), indicating that ACT001 has a synergistic effect with SHP099. We further assessed the activities of AEBP1/AKT axis in response to ACT001 and SHP099 alone or in combination in GSC 1123 and R39 cells. As shown in Figures 6B and 6C, compared to the vehicle control, SHP099 decreased p-AKT and cell proliferation but not AEBP1 expression. Combined treatment with ACT001 and SHP099 significantly decreased p-AKT and cell proliferation compared to single ACT001 or SHP099 treatment ( Figure 6A-6C), suggesting that ACT001 in combination with SHP099 decreases GSC proliferation in vitro.
Next, to examine the in vivo antitumor effect of single ACT001 treatment or in combination with SHP099, we employed an orthotopic xenograft model in immunodeficient mice. Oral gavage of ACT001 (100 mg/kg) with or without intraperitoneal injection of SHP099 (25 mg/kg) was initiated after randomization at Day 3 after GSC 1123 implantation ( Figure 7A). Compared to the vehicle control, ACT001 treatment markedly decreased tumor growth (Figures 7B-7E) and extended the survival GBM-bearing animals ( Figure 7F). Additionally, IHC analysis showed that ACT001 significantly reduced p-AKT and the expression of AEBP1, cell proliferation marker Ki-67, and GSC marker hNestin (human Nestin) in GSC 1123 xenograft tumors ( Figure 7G). ACT001 in combination with SHP099 further reduced tumor burden ( Figures  7B-7E), p-AKT, the expression of AEBP1, Ki-67, hNestin ( Figure 7G), and extended survival to a greater extent than ACT001 treatment alone ( Figure  7F), suggesting that the combination of ACT001 and SHP099 synergizes to extend the survival in our GSC xenograft model ( Figure 7H).   . (H) A working model of ACT001-targeted TGF-β/AEBP1/AKT signaling pathway in GSC tumor growth and self-renewal. Data represent the mean ± SEM of replicates from three independent experiments. *p < 0.05, **p < 0.01, **p < 0.001, by two-tailed Student's t-test or log rank analysis.

Discussion
Despite ACT001 had exhibited good potential for GBM treatment in clinical trials [12], the mechanism by which ACT001 inhibits tumor growth is still unclear. In this study, we conducted an extensive preclinical in vitro and in vivo analysis of ACT001 inhibition of multiple GSC cells and one GSC xenograft model. GSCs were more responsive to ACT001 compared with NHA in vitro. ACT001 inhibited GSC proliferation and glioma sphere formation through targeting AEBP1/PI3K/AKT signaling and prolonged the survival of animals as a single agent or in combination with SHP099 in an orthotopic GSC xenograft.
Biological analyses have established that ACT001 inhibits various tumor growth. However, the mechanism by which ACT001 inhibits cell proliferation remains unclear. ACT001 functioned as a PAI-1 inhibitor and exerted synergistic effects in combination with cisplatin through inhibiting PI3K/AKT pathway in glioma [14]. ACT001 also impaired NF-κB activation in breast cancer and glioma cells [36,37]. In this study, we identified AEBP1 as a new target of ACT001. ACT001 treatment decreased AEBP1 expression and AEBP1-activated PI3K/AKT signaling in GSCs. Knockdown of AEBP1 attenuated GSC response to ACT001. PI3K inhibitor treatment sensitized GSCs to ACT001, whereas overexpression of a constitutively activated AKT mutant reduced GSC response to ACT001. ACT001 treatment also blocked TGF-β -induced AEBP1/PI3K/AKT signaling.
Taken together, this study identifies AEBP1 as a new target of ACT001 and a putative prognostic factor in glioma. The newly elucidated roles of AEBP1 in GSC proliferation and glioma sphere formation provide a strong rationale for targeting this molecule in clinical treatment of human gliomas. ACT001 as a single agent or in combination with SHP-2 inhibitor SHP099 provides significant survival benefits to GBM tumor xenograft-bearing animals, which also provide a strong rationale for validation of combination of ACT001 and SHP099 as a therapeutic approach against GBM.