A novel program of infiltrative control in astrocytomas: ADAM23 depletion promotes cell invasion by activating γ-Secretase and NOTCH1 signaling pathway

doi:10.21203/rs.3.rs-1538484/v1

Download PDF

Research Article

A novel program of infiltrative control in astrocytomas: ADAM23 depletion promotes cell invasion by activating γ-Secretase and NOTCH1 signaling pathway

https://doi.org/10.21203/rs.3.rs-1538484/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

Diffuse infiltration is a life-threatening growth pattern in malignant astrocytomas and a significant cause of therapy resistance. This pattern is determined by the balance between proliferative and invasive genetic programs and is mainly recapitulated by glioma stem cells (GSCs) during astrocytoma progression. Infiltration results in the GSCs spreading deeply into the surrounding brain tissue, fostering tumor recurrence and making complete surgical resection impossible. We need to thoroughly understand the mechanisms underlying diffuse infiltration to develop effective therapies.

Results

We found that ADAM23 is downregulated in astrocytoma compared with the normal brain. Based on the clinical data and preclinical models, ADAM23 downregulation is associated with a favorable prognosis in lower-grade astrocytoma patients and correlated with the emergence of a pro-invasive tumor phenotype. The expression of ADAM23 in astrocytoma cells is inversely correlated with increased γ-secretase complex activity, consequently contributing to NOTCH1 pathway activation in GSCs and amyloid-β (Aβ) deposition in the mouse brain. Finally, epigenetic inhibition or administration of γ-secretase inhibitors (GSIs) induced a significant inhibitory effect on the invasive programs, specifically in ADAM23low astrocytoma cells.

Conclusions

Overall, our findings reveal ADAM23 expression levels as a new prognostic biomarker and a predictive factor of GSI efficacy for diffuse infiltration in ADAM23 low-expressing astrocytomas.

ADAM23

glioblastoma

astrocytoma

invasion

Alzheimer

γ-secretase

Notch1

amyloid-β

PS1

Astrocytomas are the most common and deadly tumors of the central nervous system (CNS) (Ostrom et al., 2019). The World Health Organization (WHO) segregates astrocytomas into “circumscribed” (WHO grade 1), often exhibiting a solid morphology, as opposed to the inherently “diffuse” astrocytomas (WHO grades 2–4) (Louis et al., 2021). The gold-standard treatments — including, radical surgery, temozolomide, and radiotherapy — for diffuse astrocytomas do not guarantee tumor eradication and results in short survival gains (Oberheim Bush, Hervey-Jumper, and Berger, 2019; Furnari et al., 2007). One of the major causes of treatment failure is the infiltrative growth pattern of the glioma stem cells (GSCs) in the brain parenchyma. GSCs spread from the tumor core, hindering total tumor resection and fostering tumor recurrence (Vollmann-Zwerenz et al., 2020). Astrocytoma cells from the invasive front among distinct patients share conserved infiltrative molecular programs. These include mechanisms necessary for invasion and migration of the interstitial matrix that are associated with nervous system development and cell-cell adhesion (Darmanis et al., 2017). We need a detailed understanding of the mechanisms underlying the “diffuse invasion” of GSC in the brain parenchyma to develop effective therapies.

Genes identified to be responsible for inducing infiltrative growth in GSCs include NOTCH1, EGFR1, and SOX2 (Wang J, et al., 2019; Yi et al., 2019; Sun Z, et al., 2020; Ye et al., 2012). The Notch signaling pathway plays a fundamental role in CNS development and is highly active in GSCs (Bazzoni R et al., 2019). It is activated upon ligand (delta or jagged) binding in trans, followed by at least two sequential proteolytic cleavages. The first is mediated by members of the ADAM (A disintegrin and metalloproteinases) family of proteases with α-secretase activity, and the second is mediated by the γ-secretase complex, comprising presenilins 1 and 2 (PS1 and PS2), presenilin enhancer 2 (PEN-2), nicastrin, and anterior pharynx-defective 1 (APH-1) (LaVoie et al., 2003; Groot AJ, et al., 2012; Wolfe MS., 2019). The activity of α-secretase exerted by ADAM members, including ADAM 9, 10, 17, and 19, initiates Notch activation at the cell surface and is a rate-limiting step in the γ-secretase complex activity that releases the Notch intracellular domain (NICD) from the membrane, allowing nuclear translocation and transcription regulation (Wang et al., 2019). NICD is upregulated in human grade 4 astrocytomas compared with normal brain tissue (Wang et al., 2021). It functionally increases migration and invasion in astrocytoma cells by mediating the overexpression of Snail, Zeb1, and vimentin, together with activating AKT and the nuclear localization of β-catenin (Zhang, X., et al., 2011).

ADAM23 belongs to a subgroup of three catalytically inactive ADAM members (ADAM11, 22, and 23), known as “cerebral ADAMs” (Hsia et al., 2019). It is predominantly expressed in proliferative niches of the brain and plays a crucial role in neuronal development, myelination, and differentiation (Markus-Koch et al., 2017; Y. Wang, Sun, and Qiao, 2012). Several studies have shown that ADAM23 is downregulated in many types of human cancers, including breast (Costa et al., 2004; Verbisck et al., 2009; Fridrichova et al., 2015), pancreatic (Hagihara et al., 2004), gastric (Takada et al., 2005; Watanabe et al., 2009), head and neck (Calmon et al., 2007), colorectal (Choi et al., 2009), lung (Hu et al., 2011), bone (Conceição et al., 2015), ovarian (Ma et al., 2018), and brain (Kwon et al., 2017) cancers. Considering the large spectrum of human cancers associated with ADAM23 silencing and its major expression in the brain, it is likely that ADAM23 downregulation is associated with the activation of malignant programs in astrocytoma cells.

Here, we showed that the migratory and invasive behavior of astrocytoma cells is inhibited by ADAM23. The coding gene ADAM23 is downregulated in diffuse astrocytoma (grade 2–4) cells compared with the normal brain tissues. Moreover, depletion of endogenous ADAM23 in GSCs results in increased brain infiltration and prolonged survival in patients with astrocytoma and murine models owing to reduced intracranial pressure (cerebral herniation). We identified gene expression signatures associated with Alzheimer’s disease (AD) in ADAM23-depleted cells and investigated the molecular mechanisms by which ADAM23 regulates astrocytoma infiltration. We found that ADAM23 depletion increases γ-secretase complex activity, contributing to amyloid-β (Aβ) deposition in the mouse brain. Additionally, there was a concomitant increase in NICD cleavage and Notch signaling pathway activation. Interestingly, treatment with γ-secretase inhibitors (GSIs) or genetic knockdown of a catalytic member of the γ-secretase complex (PS1) reverts the invasive gains associated with ADAM23 depletion. Our data support that ADAM23 downregulation promotes diffuse astrocytoma invasiveness by increasing γ-secretase activity and Notch1 signaling pathway activation. Our results also suggest the possibility of using γ-secretase inhibitors as a new line of targeted therapeutic agents for diffuse astrocytomas.

Clinical data and gene expression in normal brain and astrocytoma samples. 22 non-neoplastic brain (NNB) anonymized tissues from epilepsy patients and 143 WHO grades 2-4 astrocytomas were collected from patients undergoing treatment at the Hospital das Clinicas (HC) at the Faculdade de Medicina of the University of São Paulo (FM-USP, SP, Brasil). Specimens were obtained with appropriate consent from institutional review boards. ADAM23 expression levels were analyzed by qRT-PCR and normalized with several housekeeping genes (TBP, HPRT1, and GUSB). Details of patients, sample characteristics and preparation are described previously in Marie et al. (2016). Validation of ADAM23 expression was performed in a panel of 28 NNB, 147 astrocytoma grades 2-3, and 219 grade 4 diffuse astrocytoma samples from Rembrandt dataset (Rembrandt 2005). Clinical data on overall survival and RNA sequencing expression of 332 patients with grade II, III and IV astrocytomas were obtained from The Cancer Genome Atlas (TCGA) and viewed by the Gliovis portal (Bowman et al., 2017). Clinical data on overall survival and gene expression of ADAM23 by RNA sequencing of 215 patients with grades III and IV astrocytomas were obtained from the Chinese Glioma Glioma Atlas (CGGA) and viewed on the website http: //www.cgga.org.cn/

Cells, plasmids and reagents. Parental GBM6 and GBM39, GSC23 and GSC11, and HK296 and HK301 cells were provided by Frank Furnari (University of California, San Diego) and were cultured in DMEM/F12 medium supplemented with 2% B27 (GIBCO/Life Technologies) and 20 ng/ mL human recombinant EGF, 20 ng/mL bFGF, and 2 mg/mL Heparin. U87MG, U178MG and U343MG cell lines were maintained in complete medium (DMEM, 10% fetal bovine serum, FBS), and were obtained from ATCC. All cells were incubated at 37°C with 5% CO₂in a humidified incubator and were certified to be Mycoplasma free on a regular basis. The far-red fluorescent protein TurboFP635 (scientific name Katushka) was from Evrogen (pTurbo-FP635-C). PLKO.1-based ADAM23 and PS1 shRNA constructs were purchased from Sigma (Mission shRNA). U87 cells with the isopropyl-β-D-thiogalactoside (IPTG)-inducible ADAM23-shRNAs (pLKO_IPTG_3xLacO, Sigma Aldrich) (iU87) were transduced with viral particles at multiplicity of infection (MOI) of 2 and, after 24 h, they were selected with antibiotics. For inducible knockdown in vitro 1 mM IPTG was administered to the iU87 cells 96h before experiments. For ADAM23 rescue assays, IPTG was washed out with 1x PBS and iU87 cells were maintained for additional 10-20 days in complete medium. Cell line authentication testing was performed using GenePrint System (Promega). For γ-secretase pharmacological inhibition, cells were treated with 10 µM of GSI RO4929097 (Stem Cell Technologies) for 24h. Equivalent volume of diluent DMSO was added to control conditions. We inferred ADAM23 expression in normal tissues using data from GTEx portal (GTEx Consortium, 2013) (J. Lonsdale et al., 2013).

qRT-PCR Total RNA was extracted using Trizol reagent (Invitrogen), and cDNA was synthesized using using Superscript II reverse transcriptase (Invitrogen) as described. qPCR was performed using a 7300 Real-time PCR System and SYBR Green (Life Technologies). Gene expression levels were calculated as described previously (Costa et al., 2015).

Orthotopic transplantation and mass effect evaluation. Viable (100 – 10,000) GSC23 or U87MG cells or their derivatives stably transfected with TurboFP635 cells in 3 μl of DMEM/F12 without growth factors were stereotactically injected into the right striatum of 8-week-old female nude mice, as previously described (Ozawa and James, 2010). Tumor growth was assessed every two weeks by iVIS200 (iVIS Spectrum In Vivo Imaging System – PerkinElmer). Mice were sacrificed at the time of development of neurological symptoms or body weight loss > 20% by isoflurane inhalation (4-5 vol.%) followed by inhalation of CO₂. For mass effect analysis we evaluated the degree of deviation from the cerebral midline of the animals with the aid of a protractor. The “zero” grade of the protractor was positioned at the beginning of the midline. The deviation of the line average was determined by ascertaining the number of degrees deviated from the midline of a normal murine brain. Brains were then harvested followed by paraffin embedding.

Immunohistochemistry

Antigenic recovery of previously deparaffinized sections was performed by incubation with 10 mM citrate buffer (pH 6.0) at 95°C for 20 minutes. The detection of immunohistochemistry (IHC) human antigens were performed using the EXPOSE kit following the manufacturer's instructions (Abcam, Ab80436) and counterstaining with hematoxylin. Reactions were performed with the primary anti-Sox2 antibody (Cell Signaling, clone D6D9, Cat # 3579), amyloid-β, and ADAM23 (Sigma-Aldrich Co., HPA012130). Images were taken under a DS-Fi2 (Nikon) digital camera microscope coupled.

Quantitative analysis of in vivo cell invasion. To examine the invasion of GSC or U87MG cells into brain, we incubated paraffin-embedded sections with anti-human Sox2 antibodies (Dako, Carpinteria, CA, USA) overnight at 4°C, and then with mouse secondary antibodies (Santa Cruz Biotechnology) for 1 h at room temperature. For evaluating the ability of GSC to invade brain tissue we manually counted the number of TurboFP/Sox2+ invasive cells that visually had detached from the tumor bulk and migrated into the brain more than 100 μm from tumor border.

(TCIA-TCGA dataset) The Cancer Imaging Archive, TCIA, with gene expression data from The Cancer Genome Atlas, TCGA We identified 21 (10 ADAM23^hiand 11 ADAM23^lo) preoperative astrocytoma patients from TCGA whom had gene expression profiles and corresponding MR imaging available in the NCI's TCIA (http://cancerimagingarchive.net/). We used Fluid-Attenuated Inversion Recovery (FLAIR) sequences that reflects a mixture of edema and tumor infiltration and is routinely used to evaluate its extent. FLAIR sequences volumes and 3D reconstructions were acquired in 3D slicer software 4.10 version (Chaddad and Tanougast 2016).

F-actin labeling using phalloidin fluorescence staining. Actin filaments were stained with phalloidin conjugated with Alexa Fluor 488 (Cell Signaling) for 15 minutes at room temperature and then washed 3 times with PBS. Nuclei were stained with 4',6-Diamino-2-Phenylindole (DAPI, 0.5 μg/mL) for 2 min. Images were captured under EVOS FL-AUTO2 microscope (Thermo Fisher Scientific) equipped with the appropriate filters and aided by the “FibrilTool” plugin (Boudaoud et al. 2014) from ImageJ/FIJI software (Schindelin et al. 2012).

Time-lapse microscopy of cell migration. A total of 10,000 iU87MG cells treated with IPTG (1mM) for 72h were plated in a 24-well plate with or without 1 mg/mL of brain extracellular matrix hydrogel (MatriXpec Brain, TissueLabs) coating. After 24h of plating the cells, they were subjected to individual cell migration assay by real-time microscopy by monitoring on EVOS FL-AUTO 2 equipment (Thermo Fisher Scientific) equipped with a temperature and gas supply control. Post-processing of the images was performed using the ImageJ package Fiji with a cell tracking plug-in. For the assessment of the velocity and directionality of cell motility, the “Manual Tracking” and “Chemotaxis and Migration Tool” (Ibidi) plugins from ImageJ/FIJI software were used (Schindelin et al. 2012).

Transwell invasion assays. A total of 2 x 10⁵ cells were plated in the upper chamber of the matrigel-coated transwell and allowed to invade for 22 h at 37 °C. 600 μl of the DMEM/F12 + 10% FBS were used as chemo-attractant into the bottom of the lower chamber. At the end of the experiment, the top side of the filters were scraped with cotton swabs and migrating cells were fixed and stained with 4′,6-diamidino-2-phenylindole (Sigma-Aldrich Co.). Cells were counted at x100 magnification in 20 different optical fields/insert.

Cell invasion analysis in 3D purified brain matrix. We followed previously published method for tri-dimensional (3D) spheroids invasion assay (Costa & Camargo, 2018). Briefly, spheroids of 30,000 iU87MG cells treated with IPTG (1 mM) were embedded within 3D brain matrix (MatriXpec Brain, TissueLabs). The dynamic invasion of spheroids was monitored using the inverted microscope EVOS FL-AUTO2 (Thermo Fisher Scientific). Subsequently, the live images were analyzed with ImageJ/FIJI software.

Cell cycle analysis by flow cytometry A total of 250,000 iU87MG cells were treated with IPTG (1 mM) for 96h. Cells were then fixed with 70% ethanol, washed with PBS and stained with Propidium Iodide (PI) accordingly to the instructions of the FxCycle PI + RNAse reagent Staining Solution (Thermo Fisher). Cell-cycle analysis was performed with a FACSCalibur cytometer and indicated populations were quantified using the CellQuest software.

Clonogenic anchorage-independent assay. Single-cell suspensions of 1,000 viable U87MG cells/well were mixed to 300 μl of 0.5% low-melting agarose diluted in the complete medium and were plated in culture dishes coated with 5 mm of 1% agarose. Plates were incubated at 37 °C and 5% CO₂ for 4 weeks to allow colony formation. The number of colonies formed and the colony area (mm2) were determined under the light microscope using the ImageJ software (NIH, Bethesda, MD, USA).

Immunocytochemistry. U87 cells plated on coverslips or GSCs as neurospheres were fixed with 3.7% paraformaldehyde in PBS for 15 min and permeabilized with 0.3% Triton X-100 in PBS for 5 min. The cells were then blocked with 5% BSA and incubated with anti-Notch1 antibody (Cell Signaling, clone D1E11, Cat # 3608), anti-NICD antibody (Cell Signaling, Val1744, clone D3B8, Cat # 4147), anti-ADAM23 (Sigma-Aldrich Co., HPA012130), PSEN1 (Cell Signaling, clone D39D1, Cat # 5643), followed by incubation with Alexa-Fluor 488-conjugated secondary antibody. After several washes in PBS, nuclear labeling was performed with DAPI (0.5 μg/mL) for 2 min and cells were mounted on glass. The images were analyzed using the inverted microscope EVOS FL-AUTO2 (Thermo Fisher Scientific). For quantification of protein expression, the outline of each cell has been defined by the actin labeling (red fluorescence).

γ-secretase activity analysis. For determination of γ-secretase activity, cells were subjected to Amyloid-β ELISA immunoquantification (aa1-42, DAB142 R&D Systems). The assay was performed according to the instructions from the kit.

Degree of brain invasion by fluorescence molecular tomography (FMT). Intracranial fluorescence images obtained by the in vivo imaging systems (IVIS SpectrumCT) were used to indirectly track the invasion of GSC after ADAM23 depletion according to a protocol adapted from Benitez et al. (2018). Briefly, when tumors reached the same photon signal intensity (cellularity) (approximately 10⁹ photon counts/second) they were submitted to pixel distribution analysis using ImageJ/FIJI software to delimit the tumor area. The degree of invasion was inferred comparing the area of the tumors with the same cellularity.

RNAseq and GSEA analyses. Total RNA samples were extracted from GSC23 and U87MG cell variants using TRIzol. Total RNA was enriched for polyA+-containing RNA using oligo-dT dynabeads and then used as a template to cDNA synthesis. The cDNA was submitted to ultrasonic digestion, end repair, A-tailing reaction and adaptor ligation prior to size selection in a high-resolution agarose gel. Selected fragments ranging from 150 to 400 bp were amplified by PCR. Finally, library products were submitted to sequencing in the Illumina platform using the protocol “Hi-Seq Rapid Run Duo SR 100 Cycle”. Read sequences were aligned and mapped against the human genome using RNA Express Application (version 1.0.0) from BaseSpace. The list of genes with cpm > 5 (U87MG: 10,967 genes and GSC23: 11,591 genes) were used for the analyses. GSEA was performed using GSEA v2.0.12 with probes ranked by t test, and significance was determined by 1000 phenotype permutations, being considered the FDR < 0.05 (False Discovery ratio) to filter the groups significantly enriched (Subramanian et al., 2005). In this report we use lists of gene signatures (Gene sets) specific to neurodegenerative diseases and others identified throughout the report.

Statistical analyses. The results obtained were tested for normality using the Saphiro Wilk test. Parametric data (p> 0.05 in the Saphiro Wilk test) were expressed as mean ± standard deviation (SD) and analyzed using the unpaired t-test (for samples with different variances, Welch correction was used) or one-way ANOVA followed Bonferroni’s post-test. The overall survival data were plotted on a Kaplan-Meier curve and analyzed by Log-rank test. All data were analyzed using the statistical software GraphPad Prism version 6. The p-value was considered statistically significant when the risk α ≤ 0.05.

ADAM23 gene expression is downregulated in diffuse astrocytomas

ADAM23 transcripts are expressed in several normal tissues with particularly high levels in multiple regions of the CNS (Goldsmith et al., 2004) (Fig. S1). To evaluate ADAM23 gene expression in CNS tumors, we measured ADAM23 mRNA levels using qRT-PCR in the non-neoplastic brain (NNB) and 143 WHO grades 2-4 astrocytomas from the HC cohort. ADAM23 expression was significantly downregulated in diffuse astrocytomas compared with NNB (p < 0.001) (Fig. 1A). Downregulation of ADAM23 expression in CNS tumors was validated in an independent cohort (Rembrandt dataset). Independent of the tumor grade, all astrocytomas displayed a marked decrease in ADAM23 expression compared with NNB (p < 0.0001) (Fig. 1B).

ADAM23 depletion promotes brain infiltration

To assess the role of ADAM23 expression in astrocytomas, we examined ADAM23 levels in a panel of GSC and glioblastoma (GBM) cell lines using RT-qPCR. U87 and GSC23 cells amongst the glioma cells had the highest expression levels of ADAM23, as confirmed using immunofluorescence assays (Fig. 2A-B). GSC23 and U87 cells were transduced with two independent shRNAs against ADAM23 or a control hairpin against GFP (Fig. 2B-D) and grafted into nude mice via orthotopic transplantation. Qualitative blind microscopic examination by an expert neuropathologist revealed that GSC23 tumors were remarkably infiltrative, with GSC cells detected in both hemispheres with a partial recapitulation of key histological features of human GBM, includinghypercellularity and marked nuclear atypia, but excluding microvascular proliferation and necrosis (Fig. S2). In contrast, orthotopic xenografts of U87 cells that typically exhibit a non-infiltrative growth pattern in the brain parenchyma (de Vries et al., 2009; Miura et al., 2010) did not exhibit any infiltrative gains or significant changes in their histopathology, even after ADAM23 depletion (Fig. S3A-B).

Quantitative measurements of GSC23 infiltration showed a 3-fold increase in the number of disseminated GSCs in ADAM23^low tumors (GSC-A23^low) compared with the controls (GSC-A23^hi) (Fig. 2E). Additionally, in vivo biodistribution profiles generated using whole-brain fluorescence molecular tomography (FMT) analysis showed that GSC-A23^low tumors significantly spread over a larger area of the mouse brain compared with the GSC-A23^hi tumors (Fig. 2F-G).

The role of ADAM23 in the infiltrative behavior of diffuse astrocytomas was evaluated using 21 preoperative astrocytoma datasets included in the TCIA-TCGA repositories. Subgroup cases (ADAM23^hi and ADAM23^low) were compared to determine the fluid-attenuated inversion recovery (FLAIR) signal volume images that matched the level of peritumoral invasion/edema. Interestingly, these neuroimaging findings and ADAM23 expression were inversely correlated (Spearman correlation = −0.62, p = 0.003, Fig. 2H). We observed a 3.8-fold higher FLAIR volume in ADAM23^low lower astrocytomas (Fig. 2I). These results demonstrate that lower levels of ADAM23 facilitate the emergence of a brain infiltrative tumor phenotype in preclinical models and human patients.

ADAM23 regulates astrocytoma cell migration and invasion in vitro

To better characterize the functional consequences of ADAM23 depletion, we used a Matrigel-based transwell invasion assay. A significant 16-fold increase in the invasion of 3D matrices was observed for GSC-A23^low cells compared to GSC-A23^hi cells (Fig. 3A).

We also tested the migration and invasion potential of the U87 cell line in which ADAM23-shRNAs were placed under the control of an IPTG-inducible promoter (iU87 cells) (Fig. 3B). IPTG-induced cells plated on exogenous brain extracellular matrix (B-ECM) acquired mesenchymal-like features, with highly dynamic filopodia protrusions and a marked increase in actin polymerization (Fig. 3C). Within 24 h of B-ECM, IPTG-induced cells exhibited a 25% global increase in migration speed (Fig. 3D), with up to 82% of cells switching to a faster migratory mode (>10 µm/h; Fig 3E), 13% increase in total distance, and 45% in directional movement (Fig. S4A-B) when compared with non-induced or control cells (i.e., U87 cells transduced with IPTG-inducible control-shRNAs). Complementarily, long-term exposure of iU87 cells to IPTG resulted in a significant increase their Matrigel invasion (Fig. 3F), a time-dependent decrease in cell proliferation (Fig. S4C), cell cycle arrest in the G0/G1 phase (Fig. S4D), and lower clonogenic efficiency than the controls (Fig. S4E).

Interestingly, when iU87 cells cultured as multicellular spheroids were embedded into 3D B-ECM, we observed a 48% increase in the brain matrix invasion speed after IPTG treatment (Fig. 3G). However, when IPTG was removed, ADAM23 expression levels were rescued (Fig. S4F), decreasing the cell invasion speed to levels similar to those observed in the control cells (Fig. 3G). These results indicated that the ADAM23-dependent invasive switch is reversible.

ADAM23 depletion promotes extension in overall survival (OS) associated with decreased brain herniation

Although no significant differences were observed in the tumorigenic potential or tumor latency (Fig. S5A-C), we observed a slower tumor growth rate (Fig. 4A) and an unexpected extension of OS in mice bearing GSC-A23^low tumors (Fig. 4B). To determine whether ADAM23 levels might be a biomarker of improved survival in human astrocytomas, we used the clinical and expression information from the TCGA and CGGA cohorts (Bowman et al., 2017). Patients were grouped according to their median ADAM23 expression into high and low groups. The OS was not different between patients with ADAM23^hi and ADAM23^low GBM (Fig. 4C, TCGA: HR = 1.26 [0.55–1.12], p = 0.188, and CGGA: HR = 0.95 [0.71–1.26], p = 0.71). Interestingly, subjects with ADAM23^hi low-grade astrocytomas had a worse prognosis than those with ADAM23^low ones (TCGA: HR = 0.49 [0.28–0.82], p = 0.008; CGGA: HR = 0.55 [0.32–0.92], p = 0.026) (Fig. 4D-E), successfully recapitulating the differences in OS observed in the GSC mouse model (Fig. 4B). The prognostic impact of ADAM23 expression level was independent of IDH status (Fig. S6A-B). Together, these data indicate that ADAM23 expression in lower-grade astrocytomas is associated with a poor prognosis. To further investigate the relationship between ADAM23 expression, GSC dissemination, and mouse OS, we used the midline shift (MLS) as a quantitative indicator of mass effect and herniation. GSC-ADAM23^low tumors exerted a significant 40% reduction in MLSs compared to GSC-ADAM23^hi tumors (Fig. 4F), although both conditions presented similar sizes.

ADAM23 depletion reprograms astrocytoma cells to a pro-infiltrative and Alzheimer’s disease-like (AD) gene expression signature

Next, gene expression profile analysis (RNA-seq) was performed to identify differentially expressed genes (DEGs) associated with ADAM23 status in GSC and U87 cells. We found that 5.2% (605/11,591) and 1.8% (197/10,967) of the total genes were DEGs in GSCs and U87 cells, respectively, after ADAM23 depletion (Fig. 5A, -0.75 > log2FC > 0.75, FDR < 0.05). Only seven DEGs, including ADAM23, were differentially expressed in both cell lines (NMB, GLTSCR2, MBNL3, CLIP2, TMEM134, and SLCO4A1) (Fig. S7A). A complete list of all the DEGs is provided in Supplementary Table S1.

Despite the low correspondence of DEGs in GSC and U87 cell lines, approximately 15% of DEGs in both models could be significantly integrated into a similar signaling pathway associated with AD (p = 6.38 x 10⁻¹⁹ for GSC and p = 8.57 x 10⁻⁹ for U87) (Tables S1, S2) (Blalock et al., 2004). These genes are enriched in transcription factors involved in cell proliferation control, matrix and extracellular adhesion (e.g., laminins and integrins), semaphorin pathway that inhibits axonal elongation, glial cell-mediated inflammation, and oxidative stress.

Gene set enrichment analysis (GSEA) was performed using several gene signatures to test the functional role of ADAM23 in silico. The GSC and U87 ADAM23^low cell transcriptional states significantly matched the biological processes associated with GBM invasion (Pang et al., 2019) (GSC NES = 3.20, FDR q-value < 0.05; U87 NES = 3.46, FDR q-value < 0.05) (Fig. 5B) and signaling pathways associated with AD (GSC NES = 2.13, FDR q-value < 0.05; U87 NES = 1.32, FDR q-value < 0.05) (Fig. 5C and Fig. S7B) (Blalock et al., 2004).

To extend these observations to patient data, gene co-expression analysis was used to identify regulatory networks associated with ADAM23 levels in grade 2-3 astrocytomas and GBM from two additional cohorts (TCGA and CGGA). Accordingly, gene ontology (GO) annotations of gene sets inversely correlated with ADAM23 mainly included terms related to tumor invasiveness and AD in both cohorts (Fig. 5D-E and Supplementary Tables S3-6). GSEA also identified a significant enrichment for genes associated with enhanced γ-secretase complex activity in ADAM23^low astrocytoma cells (Magold et al., 2009) (Fig. 5F, GSC NES = 1.81 and U87 NES = 1.69, FDR q-value < 0.05). Consistently, genes upregulated in response to pharmacological γ-secretase inhibition (Dohda et al., 2007) were significantly enriched in ADAM23^hi astrocytoma cells (Fig. 5G, GSC NES = −1.42 and U87 NES = −1.72, FDR q-value < 0.05).

These results suggest that, despite the poor overlap between the two sets of DEGs, reprogramming of cells toward a pro-infiltrative, AD-like phenotype is shared by both GSC and U87 models, and is also observed in two independent cohorts of grade 2-4 patients. This suggests that ADAM23-mediated alterations in gene expression are manifested at the level of functions rather than at the individual gene level.

ADAM23-depletion increased cleavage of amyloid precursor protein (APP) and proteolytic activation of NOTCH1

Given the persistent association between ADAM23^low status, transcriptional signatures of AD, and high γ-secretase activity in silico, we investigated the involvement of ADAM23 in the pathological hallmark of AD — amyloid-β (Aβ) deposition.

ADAM23^low tumors increase Aβ deposits by up to 55% in GSCs and 140% in U87 tumors compared with ADAM23^hi tumors (Fig. 6A). In line with these data, we observed a 50−60% increase in Aβ levels in ADAM23^low cells compared with ADAM23^hi cells supernatants (Fig. 6B). In both lineages, this increase was significantly reduced upon pre-treatment with a potent GSI (RO4929097) (Fig. 6B), suggesting an increase in the activity of γ-secretase following ADAM23 depletion.

As γ-secretase also cleaves other substrates, including Notch1, the levels of the Notch pathway downstream effector NICD1 were also evaluated. As predicted, we observed a strong increase of up to 20% and 190% in NICD1 levels in U87 and GSC ADAM23^low cells, respectively, compared with their corresponding controls and ADAM23^hi cells (Fig. 6C-D). To examine downstream NICD1 signaling, we assessed the expression of 27 genes that were downregulated in cells with constitutively active Notch1 (Vilimas et al., 2007). Of these genes, 60% (12/20) to 80% (15/19) were core-enriched in GSEA analyses of ADAM23^low cells (GSC NES = −1.78; U87 NES = −2.24, FDR q-value < 0.05) (Fig. 6E). Finally, analysis of the TCGA dataset revealed a positive correlation between ADAM23^lowastrocytomas and Notch1 signaling pathway genes (Fig. S8A-B) (NES = −1.49, FDR q-value < 0.05).

γ-secretase inhibition abrogates ADAM23-induced invasion

Transwell invasive assays with ADAM23^low and ADAM23^hi cells in the presence or absence of GSI were performed to determine whether astrocytoma cell invasion is dependent on γ-secretase activity. Interestingly, we observed inhibition of over 40% of invasive behavior, specifically on ADAM23^low cells, with no effect on the invasiveness of ADAM23^hicells, implying that only ADAM23^low-dependent increment of invasion requires γ-secretase activity (Fig. 7A). The same pattern was observed in the iU87 ADAM23-depleted cells (Fig. 7B).

The expression of presenilin-1 (PS1), the catalytic subunit of the γ-secretase complex, was ablated using two independent shRNAs against PS1 in cU87 and GSC ADAM23^low cells (ADAM23^low/PS1^low; Fig. S9A-B) or scrambled control hairpin. ADAM23^low/PS1^low cells secreted 60–70% less Aβ in the supernatant (Fig. 7C) and exhibited significant impairment of approximately 40–50% of the enhanced invasive phenotype compared with the GSC ADAM23^low control cells (Fig. 7D). Accordingly, we also observed a statistically significant decrease of 52% in the average migratory speed of U87 ADAM23^low/PS1^low cells compared with ADAM23^low/PS1^hi control cells (Fig. 7E).

Extending these observations in vivo, the infiltrative gain described for GSC ADAM23^low tumors (Fig. 2G) was no longer observed after silencing PS1 (Fig. 7F). In addition, no differences in tumor growth rates or mouse OS were observed between the ADAM23^low/PS1^low and ADAM23^low/PS1^hi control groups (Fig. S9C-E).

Taken together, these data suggest a mechanism where γ-secretase is activated and the Notch1 pathway is upregulated by the downregulation of ADAM23 in astrocytoma cells. Moreover, the migratory and invasive responses of astrocytoma cells lacking ADAM23 expression are more sensitive to γ-secretase inhibition.

We previously showed that ADAM23 depletion promotes the invasion of carcinoma cells (E.T. Costa et al., 2015). However, the role of ADAM23 in the infiltrative growth of gliomas has not yet been experimentally investigated. Initially, we observed that ADAM23 expression was downregulated in diffuse astrocytomas compared to normal brain samples. Next, we discovered a causal association between the infiltrative growth pattern of diffuse astrocytomas and ADAM23 expression. Our in vitro functional assays, histological findings of preclinical orthotopic mouse models, and radiogenomic analysis highlight a program of infiltrative growth control, in which ADAM23 gene downregulation induces a molecular reprogramming towards a pro-infiltrative gene signature. This leads to a dramatic increase in the number of brain-infiltrating GSCs and the reduction of the compression forces exerted by these tumors on the surrounding brain tissue — the so-called mass effect — widely recognized as an important factor in the death of brain tumor patients (Silbergeld et al., 1991).

Accordingly, ADAM23-dependent changes in infiltrative growth patterns lead to significantly increased OS across distinct cohorts of astrocytoma patients and in a preclinical model. Mice-bearing ADAM23^low infiltrative tumors become long-term survivors, with a 3-fold increase in the number of brain-infiltrating GSCs. Similarly, by combining MRI sequences with TCGA data, we observed a nearly 4-fold enhanced FLAIR volume in the ADAM23^low astrocytoma subgroup, reflecting increased peritumoral invasion.

Furthermore, the median OS of patients with ADAM23^low tumors was 93 months, with most patients (> 90%) dying within 5 years, compared with 44 months for patients bearing ADAM23^high tumors. These results demonstrate that FLAIR-volume data could be useful for infiltration features and that lower levels of ADAM23 facilitate the emergence of an invasive tumor phenotype. In line with our in vitro and xenotransplantation data, these findings suggested that a worse prognosis among astrocytoma patients (and in an established mouse model) with high ADAM23 expression could be due to divergent patterns of displacement of the surrounding brain tissue (mass effect).

Previous studies have demonstrated that a subgroup of astrocytoma patients with longer OS is associated with little mass effect and herniation and is significantly enriched for genes involved in invasive programs (Steed et al., 2018, Drumm et al., 2020). Thus, our initial findings provide sufficient evidence that ADAM23 levels are important mediators of the infiltrative growth pattern and a predictor of OS in astrocytoma patients, suggesting their importance as a prognostic biomarker. Additional studies on the clinical value of ADAM23 expression as a glioma biomarker will address this potential use.

ADAM23 is highly expressed in the brain and is a non-catalytic member of the ADAMs — a protein family that is traditionally associated with the process of ectodomain shedding of membrane-associated proteins (Goldsmith et al., 2004). Further appreciation of the catalytic activity of ADAMs has been gained from studies on regulated intramembrane proteolysis (RIP) by the PS-containing γ-secretase complex (Edwards et al., 2008; Le Gall et al., 2010). RIP acts on transmembrane substrates, releasing their cytoplasmic domain for trafficking to the nucleus to act as transcription factors. RIP dysregulation is found in human brain diseases such as gliomas and Alzheimer's disease (AD) (Haass and Selkoe, 2007; Murphy, 2008; Dries and Yu, 2009). Functionally, ADAM23 depletion in GSC and non-GSC models induces the characteristic transcriptional signatures associated with AD. This gene list comprises transcription factors involved in inhibiting cell proliferation (e.g., Rb, RBBP1, RBAK, PML, PRKR proteins); genes associated with the matrix, adhesion, and extracellular migration (e.g., laminins, tenascin, and integrins); semaphorin pathway genes that inhibit axonal elongation; genes associated with glial cell-mediated inflammatory processes; and oxidative stress (e.g., IFN-g, IL-18) (Blalock et al., 2004). Considering that the AD signature develops progressively with dysfunctional γ-secretase activity (Jurisch-Yaksi, Sannerud, and Annaert 2013; Viola and Klein 2015), which is altered in gliomas and is often implicated in cell motility through Notch1 pathway activation (Zhang et al., 2012; J. Wang et al., 2019; Kucheryavykh et al., 2019; Hiddingh et al., 2014; Gilbert et al., 2010, Bazzoni and Bentivegna, 2019), we examined the effects of ADAM23 on the γ-secretase activity of GSC and non-GSC models. Our results showed that ADAM23 downregulation increased γ-secretase activity, leading to elevated Aβ and NICD generation in vitro and in vivo. Although conflicting data exist and evidence for the efficacy of γ-secretase inhibitors in clinical trials is under debate, several classes of Notch inhibitors in human cancers have been developed, including GSI and α-secretase inhibitors that inhibit members of the ADAM family (Bazzoni and Bentivegna 2019). Notably, both GSIs, initially developed as AD therapy, and shRNA-mediated knockdown of PS1 reversed the invasive gains mediated by ADAM23 loss in vitro and in vivo. These findings raise important questions regarding the use of GSIs for targeting astrocytomas. For example, which specific astrocytoma subtypes exhibit Notch-dependent infiltrative growth? Recently, several studies have suggested that pharmacological inhibition of γ-secretase/Notch signaling dramatically increases irradiation-induced cell death and exerts specific effects on GSC subpopulations (Lin et al., 2010;Bazzoni and Bentivegna, 2019). More specifically, in our experimental and preclinical models, we showed that the efficacy of GSI was preferentially observed in ADAM23^low GSCs. This inhibition can be used in addition to conventional treatments, providing an opportunity for an “anti-invasive” therapy to reduce brain infiltration in ADAM23^low tumors.

In conclusion, we demonstrated that ADAM23 regulates the activity of the PS1-containing γ-secretase complex. Moreover, low expression levels of ADAM23 are good prognostic markers in diffuse (grades 2–3) astrocytomas and can help identify a subclass of patients that might benefit from the use of GSI. Although shRNA-mediated γ-secretase inactivation was insufficient to inhibit astrocytoma growth or increase mouse OS, this approach might be promising for inhibiting the migratory and invasive phenotypes of GSCs, considering that infiltrative growth in the brain parenchyma is a major factor in therapeutic failure, precluding complete tumor removal by surgery.

A disintegrin and metalloproteinases (ADAM); A disintegrin and metalloproteinases 23 (ADAM23); Alzheimer’s disease (AD); Amyloid-β (Aβ); Anterior pharynx-defective 1 (APH-1); Brain extracelular matrices (B-ECM); Central nervous system (CNS); Differentially expressed genes (DEGs); Epidermal growth factor receptor 1 (EGFR1); Fluid-attenuated inversion recovery (FLAIR); Fluorescence molecular tomography (FMT); Gene set enrichment analysis (GSEA); Glioblastoma (GBM); Glioma stem cells (GSCs); Interferon gamma (IFN-g); Interleukin-18 (IL-18); Isopropyl β- d-1-thiogalactopyranoside (IPTG); Non-neoplastic brain (NNB); Notch intracellular domain (NICD); Notch Receptor 1 (NOTCH1); Overall Survival (OS); Presenilin enhancer 2 (PEN-2); Presenilins 1 and 2 (PS1 and PS2); Protein kinase B (PKB or AKT); Regulated intramembrane proteolysis (RIP); SRY-Box Transcription Factor 2 (SOX2); World Health Organization (WHO); Zinc finger E-box binding homeobox 1 (Zeb1); Zinc finger protein SNAI1 (Snail); γ-secretase inhibitors (GSIs).

Ethical Approval. All animal protocols and procedures were performed in accordance with ethics guidelines and were approved by the Animal Use Ethics Committee of FM-USP (Protocol: 529/1252). All procedures involving human samples were approved and conducted in accordance with the Ethic Committee of School of Medicine of the University of São Paulo (Protocol# 0263/07).

Consent for publication Not applicable

Availability of supporting data The accession code for the RNA-seq data deposited to ENA is PRJEB52539. Additional data used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests The authors have no conflict of interest to declare.

Funding This work was supported by grants from FAPESP (2014/04945-2, 2016/06857-9, 2016/07463-4, 2017/13622-0), Hospital Sírio-Libanês, and LICR.

Authors' contributions E.H.F.J. performed the cell lines generation, all in vitro and in vivo analytical experiments, samples processing, transcriptome analysis, patients’ cohort analyses, statistical analyses, data interpretation, and wrote the manuscript. G.F.B contributed in cell lines generation, gene expression in HC cohort, performed the RNA sequencing, pre-processed RNA sequencing data and aided to project conception. P.F.A. helped with RNA sequencing experiments and analyses. L.T.I. aided in cell culture experiments and theoretical discussions. N.C.S performed the histopathological staining and the in vivo single cell invasion analysis. A.R., C.Z., and P.A.F.G. helped with RNA sequencing analyses. S.K.N.M. and S.M.O. provided RNA samples from HC Cohort. T.G.S. and R.C. managed the animals’ protocols. C.L.P.L. performed histopathological analyses. F.B.F. donated the cells lines, aided in obtaining funding for this project and provided advice. A.A.C aided in obtaining funding for this project and provided advice. E.T.C. provided funding for the project, was responsible for the conception of the work, contributed to data interpretation, project design, and manuscript writing and edition. All authors contributed to revising the manuscript.

Acknowledgements We are grateful to Dr Luis Felipe Campesato and Dr Bruna Renata Silva Correa for scientific discussions and Mara de Souza Junqueira and Dr. Renata de Freitas Saito for technical assistance.

Authors' information ¹Molecular Oncology Center, Hospital Sírio-Libanês, São Paulo, SP, Brazil; ²Ludwig Institute for Cancer Research (LICR), University of California, San Diego, CA, USA; ³Department of Neurology, Laboratory of Molecular and Cellular Biology, LIM15, School of Medicine, University of São Paulo, São Paulo, Brazil; ⁴Centro Internacional de Pesquisa, A.C. Camargo Cancer Center, Fundação Antônio Prudente, São Paulo, SP, Brazil; ⁵Laboratório de Oncologia Experimental, Centro de Investigação Translacional em Oncologia, Instituto do Câncer do Estado de São Paulo, São Paulo, SP, Brazil; ⁶Departamento de Patologia, Santa Casa de São Paulo, São Paulo, Brazil.

Bazzoni, R., and Bentivegna, A. (2019). Role of Notch Signaling Pathway in Glioblastoma Pathogenesis. Cancers (Basel) 11.

Benitez, Jorge A et al. “Fluorescence Molecular Tomography for In Vivo Imaging of Glioblastoma Xenografts.” Journal of visualized experiments : JoVE ,134 57448. 26 Apr. 2018, doi:10.3791/57448

Blalock, E.M., Geddes, J.W., Chen, K.C., Porter, N.M., Markesbery, W.R., and Landfield, P.W. (2004). Incipient Alzheimer's disease: microarray correlation analyses reveal major transcriptional and tumor suppressor responses. Proc Natl Acad Sci U S A 101, 2173-2178.

Bowman, R.L., Wang, Q., Carro, A., Verhaak, R.G., and Squatrito, M. (2017). GlioVis data portal for visualization and analysis of brain tumor expression datasets. Neuro Oncol 19, 139-141.

Calmon, M.F., Colombo, J., Carvalho, F., Souza, F.P., Filho, J.F., Fukuyama, E.E., Camargo, A.A., Caballero, O.L., Tajara, E.H., Cordeiro, J.A., and Rahal, P. (2007). Methylation profile of genes CDKN2A (p14 and p16), DAPK1, CDH1, and ADAM23 in head and neck cancer. Cancer Genet Cytogenet 173, 31-37.

Choi, J.S., Kim, K.H., Jeon, Y.K., Kim, S.H., Jang, S.G., Ku, J.L., and Park, J.G. (2009). Promoter hypermethylation of the ADAM23 gene in colorectal cancer cell lines and cancer tissues. Int J Cancer 124, 1258-1262.

Conceição, A.L., Babeto, E., Candido, N.M., Franco, F.C., De Campos Zuccari, D.A., Bonilha, J.L., Cordeiro, J.A., Calmon, M.F., and Rahal, P. (2015). Differential Expression of ADAM23, CDKN2A (P16), MMP14 and VIM Associated with Giant Cell Tumor of Bone. J Cancer 6, 593-603.

Costa, E.T., Barnabé, G.F., Li, M., Dias, A.A., Machado, T.R., Asprino, P.F., Cavalher, F.P., Ferreira, E.N., Del Mar Inda, M., Nagai, M.H., Malnic, B., Duarte, M.L., Leite, K.R., De Barros, A.C., Carraro, D.M., Chammas, R., Armelin, H.A., Cavenee, W., Furnari, F., and Camargo, A.A. (2015). Intratumoral heterogeneity of ADAM23 promotes tumor growth and metastasis through LGI4 and nitric oxide signals. Oncogene 34, 1270-1279.

Costa ET, Camargo AA. Beyond the Proteolytic Activity: Examining the Functional Relevance of the Ancillary Domains Using Tri-Dimensional (3D) Spheroid Invasion Assay. Methods Mol Biol. 2018;1731:155-168. doi: 10.1007/978-1-4939-7595-2_15. PMID: 29318552.

Costa, F.F., Verbisck, N.V., Salim, A.C., Ierardi, D.F., Pires, L.C., Sasahara, R.M., Sogayar, M.C., Zanata, S.M., Mackay, A., O'hare, M., Soares, F., Simpson, A.J., and Camargo, A.A. (2004). Epigenetic silencing of the adhesion molecule ADAM23 is highly frequent in breast tumors. Oncogene 23, 1481-1488.

Darmanis, S., Sloan, S.A., Croote, D., Mignardi, M., Chernikova, S., Samghababi, P., Zhang, Y., Neff, N., Kowarsky, M., Caneda, C., Li, G., Chang, S.D., Connolly, I.D., Li, Y., Barres, B.A., Gephart, M.H., and Quake, S.R. (2017). Single-Cell RNA-Seq Analysis of Infiltrating Neoplastic Cells at the Migrating Front of Human Glioblastoma. Cell Rep 21, 1399-1410.

De Vries, N.A., Beijnen, J.H., and Van Tellingen, O. (2009). High-grade glioma mouse models and their applicability for preclinical testing. Cancer Treat Rev 35, 714-723.

Dohda, T., Maljukova, A., Liu, L., Heyman, M., Grandér, D., Brodin, D., Sangfelt, O., and Lendahl, U. (2007). Notch signaling induces SKP2 expression and promotes reduction of p27Kip1 in T-cell acute lymphoblastic leukemia cell lines. Exp Cell Res 313, 3141-3152.

Dries, D.R., and Yu, G. (2009). Rip exposed: how ectodomain shedding regulates the proteolytic processing of transmembrane substrates. Proc Natl Acad Sci U S A 106, 14737-14738.

Drumm, M.R., Dixit, K.S., Grimm, S., Kumthekar, P., Lukas, R.V., Raizer, J.J., Stupp, R., Chheda, M.G., Kam, K.L., Mccord, M., Sachdev, S., Kruser, T., Steffens, A., Javier, R., Mccortney, K., and Horbinski, C. (2020). Extensive brainstem infiltration, not mass effect, is a common feature of end-stage cerebral glioblastomas. Neuro Oncol 22, 470-479.

Edwards, D.R., Handsley, M.M., and Pennington, C.J. (2008). The ADAM metalloproteinases. Mol Aspects Med 29, 258-289.

Fridrichova, I., Smolkova, B., Kajabova, V., Zmetakova, I., Krivulcik, T., Mego, M., Cierna, Z., Karaba, M., Benca, J., Pindak, D., Bohac, M., Repiska, V., and Danihel, L. (2015). CXCL12 and ADAM23 hypermethylation are associated with advanced breast cancers. Transl Res 165, 717-730.

Furnari, F.B., Fenton, T., Bachoo, R.M., Mukasa, A., Stommel, J.M., Stegh, A., Hahn, W.C., Ligon, K.L., Louis, D.N., Brennan, C., Chin, L., Depinho, R.A., and Cavenee, W.K. (2007). Malignant astrocytic glioma: genetics, biology, and paths to treatment. Genes Dev 21, 2683-2710.

Gilbert, C.A., Daou, M.C., Moser, R.P., and Ross, A.H. (2010). Gamma-secretase inhibitors enhance temozolomide treatment of human gliomas by inhibiting neurosphere repopulation and xenograft recurrence. Cancer Res 70, 6870-6879.

Goldsmith, A.P., Gossage, S.J., and Ffrench-Constant, C. (2004). ADAM23 is a cell-surface glycoprotein expressed by central nervous system neurons. J Neurosci Res 78, 647-658.

Groot, A.J., Habets, R., Yahyanejad, S., Hodin, C.M., Reiss, K., Saftig, P., Theys, J., and Vooijs, M. (2014). Regulated proteolysis of NOTCH2 and NOTCH3 receptors by ADAM10 and presenilins. Mol Cell Biol 34, 2822-2832.

Gusev, Y., Bhuvaneshwar, K., Song, L., Zenklusen, J.-C., Fine, H., and Madhavan, S. (2018). The REMBRANDT study, a large collection of genomic data from brain cancer patients. Scientific Data 5, 180158.

Haass, C., and Selkoe, D.J. (2007). Soluble protein oligomers in neurodegeneration: lessons from the Alzheimer's amyloid β-peptide. Nature Reviews Molecular Cell Biology 8, 101-112.

Hagihara, A., Miyamoto, K., Furuta, J., Hiraoka, N., Wakazono, K., Seki, S., Fukushima, S., Tsao, M.S., Sugimura, T., and Ushijima, T. (2004). Identification of 27 5' CpG islands aberrantly methylated and 13 genes silenced in human pancreatic cancers. Oncogene 23, 8705-8710.

Hiddingh, L., Tannous, B.A., Teng, J., Tops, B., Jeuken, J., Hulleman, E., Boots-Sprenger, S.H., Vandertop, W.P., Noske, D.P., Kaspers, G.J., Wesseling, P., and Wurdinger, T. (2014). EFEMP1 induces γ-secretase/Notch-mediated temozolomide resistance in glioblastoma. Oncotarget 5, 363-374.

Hsia, H.E., Tüshaus, J., Brummer, T., Zheng, Y., Scilabra, S.D., and Lichtenthaler, S.F. (2019). Functions of 'A disintegrin and metalloproteases (ADAMs)' in the mammalian nervous system. Cell Mol Life Sci 76, 3055-3081.

Hu, C., Lv, H., Pan, G., Cao, H., Deng, Z., Wen, J., and Zhou, J. (2011). The expression of ADAM23 and its correlation with promoter methylation in non-small-cell lung carcinoma. Int J Exp Pathol 92, 333-339.

Jurisch-Yaksi, N., Sannerud, R., and Annaert, W. (2013). A fast growing spectrum of biological functions of γ-secretase in development and disease. Biochim Biophys Acta 1828, 2815-2827.

Kucheryavykh, L.Y., Ortiz-Rivera, J., Kucheryavykh, Y.V., Zayas-Santiago, A., Diaz-Garcia, A., and Inyushin, M.Y. (2019). Accumulation of Innate Amyloid Beta Peptide in Glioblastoma Tumors. Int J Mol Sci 20.

Kwon, N.S., Kim, D.S., and Yun, H.Y. (2017). Leucine-rich glioma inactivated 3: integrative analyses support its prognostic role in glioma. Onco Targets Ther 10, 2721-2728.

Lavoie, M.J., and Selkoe, D.J. (2003). The Notch ligands, Jagged and Delta, are sequentially processed by alpha-secretase and presenilin/gamma-secretase and release signaling fragments. J Biol Chem 278, 34427-34437.

Le Gall, S.M., Maretzky, T., Issuree, P.D., Niu, X.D., Reiss, K., Saftig, P., Khokha, R., Lundell, D., and Blobel, C.P. (2010). ADAM17 is regulated by a rapid and reversible mechanism that controls access to its catalytic site. J Cell Sci 123, 3913-3922.

Lin, J., Zhang, X.M., Yang, J.C., Ye, Y.B., and Luo, S.Q. (2010). γ-secretase inhibitor-I enhances radiosensitivity of glioblastoma cell lines by depleting CD133+ tumor cells. Arch Med Res 41, 519-529.

Louis, D.N., Perry, A., Wesseling, P., Brat, D.J., Cree, I.A., Figarella-Branger, D., Hawkins, C., Ng, H.K., Pfister, S.M., Reifenberger, G., Soffietti, R., Von Deimling, A., and Ellison, D.W. (2021). The 2021 WHO Classification of Tumors of the Central Nervous System: a summary. Neuro Oncol 23, 1231-1251.

Ma, R., Tang, Z., Sun, K., Ye, X., Cheng, H., Chang, X., and Cui, H. (2018). Low levels of ADAM23 expression in epithelial ovarian cancer are associated with poor survival. Pathol Res Pract.

Magold, A.I., Cacquevel, M., and Fraering, P.C. (2009). Gene expression profiling in cells with enhanced gamma-secretase activity. PLoS One 4, e6952.

Marie, S.K.N., Oba-Shinjo, S.M., da Silva, R. et al. Stathmin involvement in the maternal embryonic leucine zipper kinase pathway in glioblastoma. Proteome Sci 14, 6 (2016). https://doi.org/10.1186/s12953-016-0094-9

Markus-Koch, A., Schmitt, O., Seemann, S., Lukas, J., Koczan, D., Ernst, M., Fuellen, G., Wree, A., Rolfs, A., and Luo, J. (2017). ADAM23 promotes neuronal differentiation of human neural progenitor cells. Cell Mol Biol Lett 22, 16.

Miura, F.K., Alves, M.J., Rocha, M.C., Da Silva, R., Oba-Shinjo, S.M., and Marie, S.K. (2010). Xenograft transplantation of human malignant astrocytoma cells into immunodeficient rats: an experimental model of glioblastoma. Clinics (Sao Paulo) 65, 305-309.

Murphy, G. (2008). The ADAMs: signalling scissors in the tumour microenvironment. Nat Rev Cancer 8, 929-941.

Oberheim Bush, N.A., Hervey-Jumper, S.L., and Berger, M.S. (2019). Management of Glioblastoma, Present and Future. World Neurosurg 131, 328-338.

Ostrom, Q.T., Cioffi, G., Gittleman, H., Patil, N., Waite, K., Kruchko, C., and Barnholtz-Sloan, J.S. (2019). CBTRUS Statistical Report: Primary Brain and Other Central Nervous System Tumors Diagnosed in the United States in 2012-2016. Neuro Oncol 21, v1-v100.

Pang, L., Hu, J., Li, F., Yuan, H., Yan, M., Liao, G., Xu, L., Pang, B., Ping, Y., Xiao, Y., and Li, X. (2019). Discovering Rare Genes Contributing to Cancer Stemness and Invasive Potential by GBM Single-Cell Transcriptional Analysis. Cancers (Basel) 11.

REMBRANDT. (2005). http://rembrandt.nci.nih.gov. In, (National Cancer Institute).

Silbergeld, D.L., Rostomily, R.C., and Alvord, E.C. (1991). The cause of death in patients with glioblastoma is multifactorial: clinical factors and autopsy findings in 117 cases of supratentorial glioblastoma in adults. J Neurooncol 10, 179-185.

Steed, T.C., Treiber, J.M., Brandel, M.G., Patel, K.S., Dale, A.M., Carter, B.S., and Chen, C.C. (2018). Quantification of glioblastoma mass effect by lateral ventricle displacement. Sci Rep 8, 2827.

Sun, Z., Wang, L., Zhou, Y., Dong, L., Ma, W., Lv, L., Zhang, J., and Wang, X. (2020). Glioblastoma Stem Cell-Derived Exosomes Enhance Stemness and Tumorigenicity of Glioma Cells by Transferring Notch1 Protein. Cell Mol Neurobiol 40, 767-784.

Takada, H., Imoto, I., Tsuda, H., Nakanishi, Y., Ichikura, T., Mochizuki, H., Mitsufuji, S., Hosoda, F., Hirohashi, S., Ohki, M., and Inazawa, J. (2005). ADAM23, a possible tumor suppressor gene, is frequently silenced in gastric cancers by homozygous deletion or aberrant promoter hypermethylation. Oncogene 24, 8051-8060.

Verbisck, N.V., Costa, E.T., Costa, F.F., Cavalher, F.P., Costa, M.D., Muras, A., Paixão, V.A., Moura, R., Granato, M.F., Ierardi, D.F., Machado, T., Melo, F., Ribeiro, K.B., Cunha, I.W., Lima, V.C., Maciel, M.O.S., Carvalho, A.L., Soares, F.F., Zanata, S., Sogayar, M.C., Chammas, R., and Camargo, A.A. (2009). ADAM23 negatively modulates alpha(v)beta(3) integrin activation during metastasis. Cancer Res 69, 5546-5552.

Vilimas, T., Mascarenhas, J., Palomero, T., Mandal, M., Buonamici, S., Meng, F., Thompson, B., Spaulding, C., Macaroun, S., Alegre, M.L., Kee, B.L., Ferrando, A., Miele, L., and Aifantis, I. (2007). Targeting the NF-kappaB signaling pathway in Notch1-induced T-cell leukemia. Nat Med 13, 70-77.

Viola, K.L., and Klein, W.L. (2015). Amyloid β oligomers in Alzheimer's disease pathogenesis, treatment, and diagnosis. Acta Neuropathol 129, 183-206.

Vollmann-Zwerenz, A., Leidgens, V., Feliciello, G., Klein, C.A., and Hau, P. (2020). Tumor Cell Invasion in Glioblastoma. Int J Mol Sci 21.

Wang, J., Xu, S.L., Duan, J.J., Yi, L., Guo, Y.F., Shi, Y., Li, L., Yang, Z.Y., Liao, X.M., Cai, J., Zhang, Y.Q., Xiao, H.L., Yin, L., Wu, H., Zhang, J.N., Lv, S.Q., Yang, Q.K., Yang, X.J., Jiang, T., Zhang, X., Bian, X.W., and Yu, S.C. (2019). Invasion of white matter tracts by glioma stem cells is regulated by a NOTCH1-SOX2 positive-feedback loop. Nat Neurosci 22, 91-105.

Wang, Y., Sun, Y., and Qiao, S. (2012). ADAM23 knockdown promotes neuronal differentiation of P19 embryonal carcinoma cells by up-regulating P27KIP1 expression. Cell Biol Int 36, 1275-1279.

Watanabe, Y., Kim, H.S., Castoro, R.J., Chung, W., Estecio, M.R., Kondo, K., Guo, Y., Ahmed, S.S., Toyota, M., Itoh, F., Suk, K.T., Cho, M.Y., Shen, L., Jelinek, J., and Issa, J.P. (2009). Sensitive and specific detection of early gastric cancer with DNA methylation analysis of gastric washes. Gastroenterology 136, 2149-2158.

Wolfe, M.S. (2019). Structure and Function of the γ-Secretase Complex. Biochemistry 58, 2953-2966.

Yi, L., Zhou, X., Li, T., Liu, P., Hai, L., Tong, L., Ma, H., Tao, Z., Xie, Y., Zhang, C., Yu, S., and Yang, X. (2019). Notch1 signaling pathway promotes invasion, self-renewal and growth of glioma initiating cells via modulating chemokine system CXCL12/CXCR4. J Exp Clin Cancer Res 38, 339.

Zhang, X., Chen, T., Zhang, J., Mao, Q., Li, S., Xiong, W., Qiu, Y., Xie, Q., and Ge, J. (2012). Notch1 promotes glioma cell migration and invasion by stimulating β-catenin and NF-κB signaling via AKT activation. Cancer Sci 103, 181-190.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

A novel program of infiltrative control in astrocytomas: ADAM23 depletion promotes cell invasion by activating γ-Secretase and NOTCH1 signaling pathway

Status:

Version 1

Abstract

Background

Results

Conclusions

Figures

Background

Methods

Results

Discussion

Conclusions

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1