Figures
Abstract
Meningiomas are the most common primary brain tumors bearing in a minority of cases an aggressive phenotype. Although meningiomas are stratified according to their histology and clinical behavior, the underlying molecular genetics predicting aggressiveness are not thoroughly understood. We performed whole transcript expression profiling in 10 grade I and four grade II meningiomas, three of which invaded the brain. Microarray expression analysis identified deleted in colorectal cancer (DCC) as a differentially expressed gene (DEG) enabling us to cluster meningiomas into DCC low expression (3 grade I and 3 grade II tumors), DCC medium expression (2 grade I and 1 grade II tumors), and DCC high expression (5 grade I tumors) groups. Comparison between the DCC low expression and DCC high expression groups resulted in 416 DEGs (p-value < 0.05; fold change > 2). The most significantly downregulated genes in the DCC low expression group comprised DCC, phosphodiesterase 1C (PDE1C), calmodulin-dependent 70kDa olfactomedin 2 (OLFM2), glutathione S-transferase mu 5 (GSTM5), phosphotyrosine interaction domain containing 1 (PID1), sema domain, transmembrane domain (TM) and cytoplasmic domain, (semaphorin) 6D (SEMA6D), and indolethylamine N-methyltransferase (INMT). The most significantly upregulated genes comprised chromosome 5 open reading frame 63 (C5orf63), homeodomain interacting protein kinase 2 (HIPK2), and basic helix-loop-helix family, member e40 (BHLHE40). Biofunctional analysis identified as predicted top upstream regulators beta-estradiol, TGFB1, Tgf beta complex, LY294002, and dexamethasone and as predicted top regulator effectors NFkB, PIK3R1, and CREBBP. The microarray expression data served also for a comparison between meningiomas from female and male patients and for a comparison between brain invasive and non-invasive meningiomas resulting in a number of significant DEGs and related biofunctions. In conclusion, based on its expression levels, DCC may constitute a valid biomarker to identify those benign meningiomas at risk for progression.
Citation: Schulten H-J, Hussein D, Al-Adwani F, Karim S, Al-Maghrabi J, Al-Sharif M, et al. (2016) Microarray Expression Data Identify DCC as a Candidate Gene for Early Meningioma Progression. PLoS ONE 11(4): e0153681. https://doi.org/10.1371/journal.pone.0153681
Editor: Aamir Ahmad, University of South Alabama Mitchell Cancer Institute, UNITED STATES
Received: February 21, 2016; Accepted: April 1, 2016; Published: April 20, 2016
Copyright: © 2016 Schulten et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Microarray data have been deposited at the NCBI’s Gene Expression Omnibus under accession number GSE77259.
Funding: HJS (PI), and DH, SK, JM, AJ, SSB, MB (Co-Is) were supported by King Abdulaziz City for Science and Technology (KACST) grant AT-32-98, and DH (PI) and MB (Co-I) by KACST Technology Innovation Center in Personalized Medicine grant 13-CIPM-07.
Competing interests: The authors have declared that no competing interests exist.
Abbreviations: ANOVA, analysis of variance; BN, brain normal; DCC, deleted in colorectal cancer; DEG, differentially expressed gene; GBM, glioblastoma multiforme; GO, gene ontology; PCA, principal component analysis
Introduction
Meningiomas arise from arachnoidal cap cells of the arachnoidal membrane and are the most common intracranial neoplasms representing 20–35% of all primary brain tumors [1]. Meningiomas are histologically classified as benign (WHO grade I), atypical (WHO grade II), or anaplastic (WHO grade III) tumors comprising approximately 80%, 15%, and 5% of the cases, respectively. WHO grade I meningiomas are categorized into nine histological subtypes of which the meningothelial subtype is the most frequent one. Meningiomas are commonly cured by surgical treatment and for aggressive tumors additional treatment is advised. Besides existing histopathological criteria, predictive biomarkers for meningioma progression are not established yet. To improve prediction for the clinical behavior of meningiomas, a new scoring system for risk stratification had been proposed, based on age, WHO grade, cytogenetic profile, tumor size, and tumor location [2].
About half of the sporadic meningiomas harbor a mutation in the tumor suppressor gene NF2 encoding merlin that is a critical factor for regulation of contact-dependent inhibition of cell proliferation. The majority of grade I meningiomas exhibit a cytogenetically normal chromosome status whereas a minority is characterized by -22q and/or -1p [3]. In subsets of grade II meningiomas, additional chromosomal imbalances are common, including -6q, -14q, and -18q [3]. Chromosomal imbalances in grade III meningiomas can be more complex and may include -9p21 and -10. Besides the association of -22q with the impairment of NF2 function, impact of other chromosome arm losses are not thoroughly understood in the molecular etiology of meningiomas.
Deleted in Colorectal Cancer (DCC) is a member of the immunoglobulin superfamily and encodes a type I transmembrane receptor for netrin-1 and a number of other ligands including cerebellin precursor protein 4 (CBLN4) [4, 5]. The mature DCC protein contains an extracellular, a transmembrane, and a cytoplasmatic domain that contains a segment for mediating apoptotic signals. Crystal structure experiments indicated that DCC multimerizes via its extracellular FN type III domains 5 and 6 with netrin-1 [6]. Inhibition of netrin-1 induces multimerization of DCC or increases apoptosis mediated by a truncated DCC variant [7]. The netrin-1/DCC signaling complex has the capacity to activate a number of intracellular signal cascades that, depending on the cellular context, results for example in rearrangement of the cytoskeleton, vasculogenesis, or morphogenetic processes [8]. In the CNS, DCC is comparably highly expressed and a critical factor for axon guidance and neuronal migration. Recently, cell culture experiments demonstrated that netrin-1 and DCC are regulators of somatic cell reprogramming to pluripotency [9].
The well established tumor suppressor function of DCC is a result of its capacity to induce apoptosis as a dependence receptor for netrin-1 in case a ligand-receptor binding is interfered or the ligand is not present. This was demonstrated in vivo in netrin-1 knockout mice revealing that regulation of apoptosis is depending on the availability of netrin-1 for DCC [10, 11]. Furthermore, loss of DCC in mice bearing an inactivated TP53 gene promoted metastatic capacity [12] whereas induced DCC expression in glioblastoma cell lines inhibited spontaneous cell migration [13]. Consistent with its tumor suppressor function, downregulation of DCC expression has been demonstrated in a number of cancer types including colorectal [14], ovarian, pancreatic, hepatocellular carcinomas [15], and neuroblastoma [16]. In different histological types of gliomas, a number of mechanisms has been described that alter DCC expression on the transcriptional or protein level that in case of transcriptional downregulation or negative immunoreactivity is related to high grade and progressive tumors [17–19]. In our study, we employed whole transcript microarrays to identify molecular biomarkers for early meningioma progression. Limited numbers of samples, as in our study, have been successfully used in other microarray expression studies on meningiomas to detect tumor-related gene profiles [20, 21].
Material and Methods
Tumor samples
The meningioma specimens studied were derived from patients who were treated surgically between April 2013 and January 2015 at the King Abdulaziz University Hospital, Jeddah. Written informed consent from the donors or the next of kin was obtained for the use of samples in research. The study was approved by the Research Ethics Committee of the King Abdulaziz University, Faculty of Medicine, #976–12. Histopathological diagnosis was performed by a team of pathologists (JM, AJ, and FG).
RNA and array processing
Isolation of total RNA and array sample processing were performed from in RNAlater (Qiagen, Hilden, Germany) preserved tumor specimens as described earlier [22]. In brief, the Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA) (11 cases) was employed to assess RNA integrity. The assigned RNA integrity number was for all samples > 5. The NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE) was utilized to determine RNA concentration. All RNA samples were processed using the Ambion WT Expression Kit (Life Technologies, Austin, TX), the GeneChip WT Terminal Labeling and Controls Kit (Affymetrix, Santa Clara, CA), and the Affymetrix GeneChip Hybridization, Wash and Stain Kit. The processed samples were hybridized to whole transcript Affymetrix Human Gene 1.0 ST GeneChip arrays that interrogate with a set of 764,885 probes 36,079 annotated reference sequences (NCBI build 36). The microarrays were scanned on a GeneChip Scanner 3000 7G and probe cell intensity data (CEL) files were generated using the GeneChip Command Console (AGCC) software. Microarray data have been deposited at the NCBI’s Gene Expression Omnibus under accession number GSE77259.
Gene and exon expression analysis
For gene and exon expression analysis, CEL files were imported into Partek Genomics Suite version 6.6 (Partek Inc., MO) using the default Robust Multi-array Average (RMA) settings for normalization. CEL files from three brain normal (BN) samples were imported from the Affymetrix Data Resource Center (http://www.affymetrix.com/support/mas/datasets.affx) and served as gene expression reference. Lists of differentially expressed genes (DEGs) were generated by applying analysis of variance (ANOVA) and using a p-value ˂ 0.05 and a fold change > 2. Where indicated, a p-value with a false discovery rate (FDR) (Step up method) < 0.05 and a fold change > 2.0 were employed. Quality of experiments was assessed on the basis of the QC metrics table and QC graphical report. Overall variance in gene expression between samples or groups of samples was assessed by principal component analysis (PCA). Average linkage hierarchical clustering was carried out using Spearman’s correlation as a similarity matrix. The Partek Gene Ontology (GO) enrichment tool was employed to group DEGs into functional categories that were ranked according to their p-values.
Functional network and pathway analysis
The Ingenuity Pathways Analysis software (IPA; build version 338830M) (Ingenuity Systems, Redwood City, CA) was employed to interpret biological significance of expression data and using the Ingenuity Knowledge Base as reference data set. Direct and indirect molecular relationships were included in analysis settings. Fisher`s exact test p-values indicate significance of relationships between the analyzed data set genes/molecules (both items were used interchangeably in the context of IPA) and the functional frameworks prebuilt or generated de novo by IPA. The Molecule Activity Predictor was utilized, as specified in the prediction legends of figures, to predict expression effects/coherence of expression effects of a molecule on other network molecules. Upstream regulators analysis was employed to explain how differences in target gene expression are regulated by upstream molecules and what kind of biological activities are involved. The overlap p-value is a statistical measure to indicate extent of overlap between the uploaded data set and genes that are known to be regulated by an upstream regulator. The regulator effects network analysis was employed to explain which regulators target DEGs from the uploaded data set and what kind of downstream effects, i.e. diseases and/or functions are connected. In how far a generated network is consistent with the Knowledge Base is scaled by a consistency score.
Semiquantitative RT-PCR
Expression of DCC was assessed by semiquantitative RT-PCR using housekeeping gene B2M as reference. Primer sequences used were DCC exon 28 forward, 5`TGAAGTGTCTGAGGAGAG-3`, DCC exon 29 reverse, 5`- GGTATGCTGCAAAGTTCC-3`, B2M forward, 5`- TCATCCAGCAGAGAATGG-3`, and B2M reverse, 5`- GAGATAGAAAGACCAGTCCT-3`. RT-PCR analysis was performed in 18 μL volumes containing each 2 μL buffer mix, 0.1% 2-mercaptoethanol, 0.0125% BSA, 3 mM MgCl2, 10 nmol of each dNTP, 5 pmol forward primers, 5 pmol reverse primers, 2.5 units GoTaq DNA Polymerase (Promega, Madison, WI), and 250 ng of second strand cDNA template that was generated using the Ambion WT Expression Kit. No cDNA template was included in negative control. The standard PCR protocol comprised an initial denaturation step at 95°C followed by 5 touch-down cycles with an annealing temperature decreasing 1°C per cycle from 65°C to 61°C. Subsequently, 24 cycles followed with 30 sec at 95°C, 30 sec at 60°C, and 30 sec at 72°C. The final step was performed for 10 min at 72°C. PCR products were resolved by electrophoresis on a 3% agarose gel. Product size for DCC was 187 bp and for B2M 156 bp. Gel band densities were measured using ImageJ 1.49v [23].
Results
The 14 analyzed meningiomas comprised 10 WHO grade I tumors and four WHO grade II tumors, one of which was a recurrence (Table 1). Four grade I tumors were classified according to their subtypes as meningothelial, three as transitional, two as fibroblastic, and one as psammomatous. Three of the four grade II tumors were brain invasive. Mean age of the study group was 54.1±10.1 years. Four patients were males and 10 were females.
Microarray expression analysis of DCC
We noticed DCC as a DEG when we applied ANOVA to compare each of the 14 meningiomas with the group of three BN samples to assess variation in expression levels among the tumors and between tumors and BN samples. We analyzed DCC in more detail based on its known critical function as a netrin-1 dependence receptor and its implications in various cellular key functions including axon guidance, cell migration, and pro-apoptotic activities; the latter known to be impaired in numerous cancer types. An exon expression plot and a corresponding heat map revealed that DCC exon expression levels varied widely between the samples (Fig 1). The resulting gene expression levels enabled us to categorize the meningiomas either as DCC low, DCC medium, or DCC high expression tumors (Table 1). A PCA scatter plot visualizes the relationship of expression of the different DCC groups and BN samples in a 3 dimensional space (Fig 2). ANOVA carried out between the DCC low expression and DCC high expression groups generated 416 DEGs. Unsupervised hierarchical cluster analysis on the 416 DEGs categorized the three DCC groups and BN samples into different branches (Fig 3). Detailed analysis of the 416 DEGs revealed 260 downregulated and 156 upregulated genes in the DCC low expression compared to the DCC high expression group (S1 Table). The most significantly downregulated genes in the DDC low expression group comprised, besides DCC, phosphodiesterase 1C (PDE1C), calmodulin-dependent 70kDa olfactomedin 2 (OLFM2), glutathione S-transferase mu 5 (GSTM5), phosphotyrosine interaction domain containing 1 (PID1), sema domain, transmembrane domain (TM) and cytoplasmic domain, (semaphorin) 6D (SEMA6D), and indolethylamine N-methyltransferase (INMT). The most significantly upregulated genes comprised chromosome 5 open reading frame 63 (C5orf63), homeodomain interacting protein kinase 2 (HIPK2), and basic helix-loop-helix family, member e40 (BHLHE40). Significance for these deregulated genes was based on a p-value with FDR < 0.05 and a fold change > 2. An RT-PCR analysis on the five DCC high expression and six DCC low expression meningiomas substantially supported the microarray expression data (S1 Fig).
Upper panel, DCC exons are interrogated with 29 probes. Lower panel, hierarchical cluster analysis based on the DCC expression values reveals two main branches, one of which contains the six DCC low expression meningiomas. Based on the expression values (Table 1) the meningiomas were grouped into DCC high, DCC medium, and DCC low expression samples.
Ellipsoids represent the 95% confidence interval and are a measure for the distance of relationships between samples of a group. Green, DCC low expression; purple, DCC medium expression; blue, DCC low expression; red, normal brain samples (BN).
BN samples are included in cluster analysis. A number of genes is represented by more than one transcript. Meningiomas are clustering into two main branches, one of which contains the DCC low expression samples and a DCC medium expression sample that was a brain invasive case. Color scheme bar indicates comparably higher and lower expression values in red and blue color, respectively. Color scheme for samples: yellow, DCC low expression; green, DCC medium expression; orange, DCC high expression; BN samples, red.
Biofunctional analysis of DCC low vs. DCC high expression
A GO enrichment analysis performed on the 416 DEGs displayed significantly overrepresented functional groups in the three categories cellular component, molecular functions, and biological process (S2 Fig). The two mostly enriched functions in the categories were extracellular matrix and extracellular region (cellular component), chemoattractant activity and molecular transducer activity (molecular functions), and biological adhesion and developmental process (biological process). The predicted top five upstream regulators for the 416 DEGs were beta-estradiol, TGFB1, TGF beta family, LY294002 which is an inhibitor of PI3Ks, and the anti-inflammatory and immunosuppressive steroid dexamethasone (S3 Fig). In the predicted top regulator effects network the upstream effectors were NFkB (complex), PIK3R1, and CREBBP that target a number of DEGs, i.e. MMP2, SERPINE2, DOK5, SLC2A5, FST, TGM2, NR4A3, TGFB3, BCL2, NCAM1, TLR2, AR, CTGF, VEGFA, CYR61, VCAM1, and GDF15 (Fig 4). Downstream network functions included adhesion of leukemia cell lines, differentiation of cells, sprouting, cell viability, and cell movement of phagocytes.
Upstream regulators NFkB, PIK3R1, and CREBBP target a number of DEGs including MMP2, SERPINE2, DOK5, SLC2A5, FST, TGM2, NR4A3, TGFB3, BCL2, NCAM1, TLR2, AR, CTGF, VEGFA, CYR61, VCAM1, and GDF15. Connected downstream functions are entitled adhesion of leukemia cell lines, differentiation of cells, sprouting (including cell morphological characteristics), cell viability, and cell movement of phagocytes.
Meningiomas from females vs. from males
To identify genes that are related to the overrepresentation of meningiomas in females, we compared the expression profiles of the 10 female with the four male cases resulting in a list of 43 DEGs (S2 Table). Among genes that were higher expressed in meningiomas from females were rhotekin 2 (RTKN2), neuritin 1 (NRN1), small nucleolar RNA, C/D box 114–26 (SNORD114-26), and leucine rich adaptor protein 1-like (LURAP1L). The gene list also contains 11 genes that are expressed from the Y chromosome. Autosomal genes that were comparably lower expressed in meningiomas from females included EPH receptor A3 (EPHA3), parvalbumin (PVALB), calbindin 2 (CALB2), and solute carrier family 38 member 3 (SLC38A3). Top upstream regulators comprised the calcineurin inhibitor tacrolimus, glutathione, ITPR, the resveratrol analogue (E)-2,3',4,5'-tetramethoxystilbene, and the zinc transporter SLC39A4 (Fig 5).
Target genes are APOD, KLRC4-KLRK1/KLRK1, MYH10, TNC, SLC7A11, CYP1B1, and NELL1. Upregulated and downregulated genes in red and blue color, respectively. Asterisk indicates a gene that is represented in the dataset by more than one transcript.
Brain invasive vs. non-invasive meningiomas
To identify an expression pattern that is related to meningioma invasion, we compared the three invasive meningiomas with the remaining 11 non-invasive meningiomas (Table 1). This comparison resulted in a set of 256 DEGs of which 94 were comparably down- and 162 upregulated in the invasive tumors (S3 Table). Among the most significantly upregulated genes in the invasive group were coiled-coil domain containing cell adhesion molecule L1-like (CHL1), RNA binding protein, fox-1 homolog (C. elegans) 3 (RBFOX3), peroxisomal biogenesis factor 5-like (PEX5L), and RAB3D, member RAS oncogene family (RAB3D), The most significantly downregulated genes comprised solute carrier family 2 (facilitated glucose transporter), member (SLC2A12), ABI family, member 3 (NESH) binding protein (ABI3BP), selectin P ligand (SELPLG), and InaD-like (Drosophila) (INADL). Top upstream regulators included heterogeneous nuclear ribonucleoprotein A2/B1 (HNRNPA2B1), class II trans-activator (CIITA), which regulates expression of MHC class II genes, the TP53 regulator structural maintenance of chromosomes protein 3 (SMC3), the cohesion component Rad21, and choline. In the predicted top regulator effects network the effector molecules were IFNG, IL1B, and TNF, that targeted a number of DEGs from the brain invasive vs. non-invasive comparison dataset (Fig 6). Downstream network functions were entitled, cell movement of myeloid cells, adhesion of blood cells, engulfment of cells, response of phagocytes, response of myeloid cells, binding of professional phagocytic cells, and recruitment of cells.
Effector molecules IFNG, IL1B, and TNF target a number of DEGs including OCLN, FLT1, CYBB, MCAM, RGS1, ITGA4, SPP1, FCGR1A, FCGR2A, TLR2, THBS1, C3, SELPLG, and FCGR3A/FCGR3B. Connected downstream functions are entitled, cell movement of myeloid cells, adhesion of blood cells, engulfment of cells, response of phagocytes, response of myeloid cells, binding of professional phagocytic cells, and recruitment of cells. Upregulated and downregulated genes in red and blue color, respectively. Asterisk indicates a gene that is represented in the dataset by more than one transcript.
Discussion
By utilizing whole transcript oligonucleotide arrays, we identified DCC as a candidate gene for tumor progression in grade I and II meningiomas using an initial multiple group comparison that was performed on each meningioma against the BN samples which served as control set. Subsequent grouping of the meningiomas according to their DCC expression levels resulted in a set of 416 DEGs between the DCC low expression and DCC high expression groups and the former one included the more aggressive tumors with the exception of one brain invasive meningioma that was categorized as a DCC medium expression tumor. DCC has not been so far identified in other microarray expression studies as a gene specifically associated with meningioma progression. In our opinion, this is based on the fact that DCC expression is apparently not a suitable discriminator between clinical benign and more aggressive meningiomas as DCC low expression was already observed in our study in those clinical benign meningiomas resembling with their expression profiles more aggressive tumors. As meningiomas are slow growing tumors, cell culture experiments on clinical meningioma samples with WHO grade I but DCC low expression profile could probably rule out the capacity of these tumors to progress into more aggressive cancer cells.
Comparison with external expression data
A systematic review on 13 microarray studies on meningiomas reported that, due to the large number of markers present on the microarrays, a comparison between the studies to identify common gene signatures with implication in the tumor biology of meningioma is difficult [24]. Alternatively, this observed discrepancy between the results can be attributed to the fact that molecular and histopathological classification of meningiomas diverge and a number of microarray expression studies reported that unsupervised cluster analysis of meningiomas grouped the samples not according to histopathological classification schemes [25, 26]. A meta-analysis on seven microarray expression studies, representing 10 to 68 samples each, identified a set of 49 genes associated with progression and/or recurrence [3, 20, 27]. The majority of these genes were only shared between any two studies incorporated in the meta-analysis. In this regard, it is notable that from this 49er gene set, 14 genes (28.5%), namely CKS2, CYR61, FGF7, FHL1, FOLR1, KIF23, LTBP2, MYLK, MYO5B, PHLDA1, PPAP2B, SERPINF1, SFRP4, and SVEP1 were shared with our set of 416 DEGs. In contrast, comparison of the 49er gene set with the 249 DEGs from the comparison between grade I and grade II meningiomas of our study (compilable from GSE77259 data set) resulted only in four shared genes.
Lower expressed in DCC low expression meningiomas
PDE1C encodes a calmodulin dependent phosphodiesterase. This gene has a glioblastoma-derived neural stem cell signature that is significantly associated with poor prognosis of glioblastoma patients [28]. Glioblastoma multiforme (GBM) cell culture experiments demonstrated that PDE1C is a promoting factor for cell proliferation, migration and invasion which were inhibited by PDE1C silencing; however, a metadata analysis revealed that only 5.3% (32/596 cases) of primary GBM overexpress PDE1C [29]. OLFM2 encodes a secreted glycoprotein of the member of the olfactomedin domain-containing proteins and is known to interact with NgR1 that modulates the functions of the NgR1 complex in axonal growth [30]. The function of OLFM2 in cancer is not thoroughly elucidated. Among other genes, GSTM5 had been identified with a hypermethylated promoter in GBM and accordingly was downregulated compared to control brain tissue [31]. In a combined cytogenetic and expression study, GSTM5 was among those genes that were downregulated in meningiomas with complex karyotypes [32]. PID1 was originally identified as an obese associated gene that, when overexpressed in 3T3-L1 adipocytes, impaired insulin-stimulated glucose uptake and phosphorylation of Akt and Irs1 [33]. Furthermore, cell culture experiments in medulloblastomas and gliomas revealed that cisplatin, etoposide, and vincristine induced transcriptional upregulation of PID1 [34]. Of notice, higher PID1 expression was positive associated with favorable prognosis of glioma patients. SEMA6D encodes a member of the semaphorin containing proteins and is expressed in different splice variants in a number of tissues including embryonic and adult brain where it is involved in axon guidance [35]. A metadata analysis on invasive breast carcinomas found a positive association between higher SEMA6D expression and favorable prognosis [36]. INMT had been identified as the top hub gene among 8000 genes which were analyzed by the TCNG database consortium on a set of 85 meningiomas [3, 37]. In prostate and lung cancer, INMT expression has been found to be downregulated [38].
Higher expressed in DCC low expression meningiomas
C5orf63 encodes a glutaredoxin-like protein that is mainly localized to mitochondria. Predicted interacting proteins include DENN/MADD domain containing 5B (DENND5B) and vesicular, overexpressed in cancer, prosurvival protein 1 (VOPP1) [39]. Its molecular functions in cancer are virtually unknown. HIPK2 encodes a nuclear serine/threonine kinase that is a positive regulator of TP53 with a relation to tumor growth suppression and induction of apoptosis [40]. HIPK2 amplification had been detected in pilocytic astrocytomas and HIPK2 overexpression in U87 human glioma cells resulted in enhanced cell growth [41]. BHLHE40, also known as DEC1, encodes a hypoxia inducible protein that is involved in diverse developmental and differentiation processes. A microarray expression analysis in mesenchymal gliomas found that expression of BHLHE40 and a number of other transcriptional regulators of mesenchymal transition including CEBPB, CEBPD, and STAT3 correlates with the extent of necrosis [42]. In breast cancer, expression of BHLHE40 correlated with invasiveness and its gene silencing in breast cancer cell lines resulted in reduced invasiveness [43].
Higher expressed in meningiomas from females
RTKN2 encodes a Rho-GTPase effector. Induced expression of RTKN2 in HEK cells resulted in NF-KappaB dependent resistance to intrinsic apoptotic signals and RTKN2 gene silencing in CD4+ lymphocytes lead to decreased BCL-2 expression [44]. Neuritin 1 is a neurotrophic factor involved in regulating synaptic plasticity and neuronal migration and has been identified as a downregulated gene in low grade meningiomas compared to normal meninges [45]. In contrast, a molecular genetic study in astrocytomas demonstrated that overexpression of neuritin 1 correlates with proliferation, apoptosis, and angiogenesis [46]. SNORD114-26 is expressed from the 14q(II) snoRNA cluster that is highly expressed in brain and uterus. Elevated SNORD114-26 expression was identified in a microarray study in dental follicle tissue in comparison to the periodontal ligament [47]. Of notice, besides SNORD114-26, the two small nucleolar RNAs SNORD113-4 and SNORD114-3 were also significantly higher expressed in meningiomas from females. Based on sequence similarities, LURAP1L is predicted to function as an adaptor implicated in regulating cell motility [48]. In osteoblastic cell line models, impairment of the estrogen receptor signalling pathway resulted in downregulation of LURAP1L [49]. In prostate cancer cell line PC-3, gene silencing of Toll-like receptor-9 (TLR9), reversed migration and invasiveness and resulted in downregulation of a number of genes including LURAP1L [50].
Lower expressed in meningiomas from females
If transcription of the Y chromosome specific genes represents only a sex biased expression remains to be elucidated as for instance, for DDX3Y, EIF1AY, KDM5D, NLGN4Y, RPS4Y1, USP9Y, UTY, and ZFY a sex-biased gene level expression in one or more CNS regions has been reported [51]. EPHA3 encodes a protein tyrosine kinase and was found to be highly expressed in the comparable aggressive mesenchymal subtype of GBMs [52]. Furthermore, in vitro experiments indicated that the receptor has the capacity to maintain tumor cells in a dedifferentiated state and its tumorigenic capacity was further demonstrated in orthotopic xenograft experiments. PARVALB and CALB2 are critical modulators of intracellular calcium dynamics in neurons. In a microarray expression study on dorsal hippocampi of young female mice, administration of 17β-estradiol resulted in a number of DEGs including Calb2 that was downregulated upon 17β-estradiol treatment [53]. SLC38A3 is an electroneutral, bidirectional glutamine transporter and is enriched in perisynaptic astroglial cell membranes. Strong immunoreactivity was revealed for SLC38A3 in high grade gliomas compared to metastases to the brain and control brain tissue [54].
Higher expressed in invasive meningiomas
CHL1 encodes a neural cell adhesion factor and has been identified in a microarray expression analysis as a considerably upregulated gene in schwannomas compared to control nerve cell samples [55]. Interestingly, besides its putative tumor suppressor functions in early tumor development, repression of CHL1 has been shown in a cancer profiling array study to be associated with local tumor invasion in ovarian, colon, and breast cancer [56]. RBFOX3 encodes a neuron-specific RNA binding protein that is functionally involved in miRNA biogenesis and neuronal developmental differentiation [57, 58]. PEX5L is expressed in the brain and its underexpression has been found in vestibular schwannomas associated with hearing loss [59]. Rab3D is a secretory small GTPase and in vitro and in vivo experiments established its critical function in cancer metastatic processes [60].
Lower expressed in invasive meningiomas
The glucose transporter member SLC2A12, alias GLUT12, has been recognized as a major glucose transporter in cancer and therefore considered as a possible therapeutic cancer target [61]. ABI3BP was among other genes found to be upregulated in serum-differentiated GBM cells upon SOX2 knockdown [62]. In thyroid cancer cells, reexpression of ABI3BP lead to reversion of the tumorigenic capacity of cells [63]. SELPLG encodes a selectin ligand and has essential functions in leukocyte trafficking during inflammation. Decreased SELPLG expression has been identified in recurrent compared to primary gliomas [64]. INADL, also known as PATJ, is a member of the cell polarity genes which have been recognized as critical factors in cancer progression [65].
In summary, this microarray expression study identified DCC as a candidate gene for early meningioma progression which could provide an explanation for the observed discrepancies between histopathological and molecular classification of meningiomas.
Supporting Information
S1 Fig. Semiquantitative RT-PCR on the five DCC high expression meningiomas, Jed13_MN, Jed26_MN, Jed40_MN, Jed43_MN, and Jed70_MN and the six DCC low expression meningiomas, Jed18_MN, Jed34_MN, Jed36_MN, Jed64_MN, Jed49_MN, and Jed58_MN.
DCC primers located in exons 28 and 29 generated a PCR product of 187 bp and B2M primers generated a PCR product of 156 bp. Relative band densities of DCC compared to B2M varied in the DCC high expression group between 0.77 and 0.93 and in the DCC low expression group between < 0.01 and 0.09.
https://doi.org/10.1371/journal.pone.0153681.s001
(TIF)
S2 Fig. GO enrichment analysis for the 416 DEGs from the comparison DCC low vs. DCC high expression meningiomas.
A, most prevalent in the cellular component domain were the categories extracellular matrix and extracellular region. B, in the molecular function domain, the most predominant categories were chemoattractant activity and molecular transducer activity. C, in the biological process domain, the most significant categories were biological adhesion and developmental process. The functional categories were scored by their p-values.
https://doi.org/10.1371/journal.pone.0153681.s002
(TIF)
S3 Fig. Merged network of the top five upstream regulators beta-estradiol, TGFB1, Tgf beta complex, LY294002, and dexamethasone that target a number of DEGs from the DCC low vs. DCC high expression comparison.
Upregulated genes include, ACKR3, ACTG2, ADAMTS5, APOD, BAMBI, BHLHE40, BMP5, BMP6, CKS2, COCH, DMTN, DSG2, FANCE, FGFR2, FOSL2, GDF15, GPRC5A, GRB10, HEY1, HIST2H2BE, ITGA2, KCNK2, LOC102724428/SIK1, MUC4, NR4A3, PCDH9, PHLDA1, PMEPA1, PODXL, RPRM, SSTR2, TMEM2, TP53I11, VEGFA, and ZNF385D. Downregulated genes include, ADAM12, ANGPT1, AQP1, AR, ARID5B, BCL2, CA12, CACNA1D, COL1A1, COL1A2, COL8A1, CTGF, CYP1B1, CYR61, DDIT4, ELN, ERBB4, FBLN1, FGF7, FHL1, FLRT2, FST, GAB1, GPR65, GSTM3, HLA-DQB1, HLA-DRB1, HOXC5, HOXC6, IL33, ITGA11, ITGA4, LIPC, LOX, LTBP1, LTBP2, MATN2, MFGE8, MME, MMP2, MSR1, MT1M, MXRA5, MYLK, NAV3, NCAM1, NNMT, NOV, OLFM2, PAPPA, PIK3R1, PLCB1, PLPP3, PLSCR4, PROM1, PTGER2, PTPN20, PXDN, PYGL, RASSF2, RBP4, ROR1, SCG2, SEMA3F, SERPINE2, SERPINF1, SLC23A2, ST8SIA1, SVEP1, TGFB3, TGFBI, TGM2, THBS1, TLR2, TPD52L1, TRIM2, TSPAN7, and VCAM1.
https://doi.org/10.1371/journal.pone.0153681.s003
(TIF)
S1 Table. Genes differentially expressed between DCC low and DCC high expression meningiomas.
https://doi.org/10.1371/journal.pone.0153681.s004
(XLSX)
S2 Table. Genes differentially expressed between meningiomas from female and male patients.
https://doi.org/10.1371/journal.pone.0153681.s005
(XLSX)
S3 Table. Genes differentially expressed between brain invasive and non-invasive meningiomas.
https://doi.org/10.1371/journal.pone.0153681.s006
(XLSX)
Acknowledgments
We thank Alaa Alamdi, Reem Alotibi, Nadia Bagtian, Ohoud Subhi, Manal Shabaat, Ishaq Khan, Mona Al-Omary, and Shireen Hussain for excellent technical assistance.
Author Contributions
Conceived and designed the experiments: HJS DH FA SK JM MS AJ FG SSB MB AC MQ. Performed the experiments: HJS DH FA AJ FG. Analyzed the data: HJS SK JM MS AJ FG SSB MB AC MQ. Contributed reagents/materials/analysis tools: HJS SK JM AJ FG SSB MB AC MQ. Wrote the paper: HJS DH AC MQ.
References
- 1. Shibuya M. Pathology and Molecular Genetics of Meningioma: Recent Advances. Neurologia medico-chirurgica. 2015;55 Suppl 1:14–27. Epub 2015/08/04. pmid:26236799.
- 2. Domingues PH, Sousa P, Otero A, Goncalves JM, Ruiz L, de Oliveira C, et al. Proposal for a new risk stratification classification for meningioma based on patient age, WHO tumor grade, size, localization, and karyotype. Neuro-oncology. 2014;16(5):735–47. Epub 2014/02/19. pmid:24536048; PubMed Central PMCID: PMCPMC3984558.
- 3. Lee Y, Liu J, Patel S, Cloughesy T, Lai A, Farooqi H, et al. Genomic landscape of meningiomas. Brain Pathol. 2010;20(4):751–62. Epub 2009/12/18. pmid:20015288; PubMed Central PMCID: PMCPmc3167483.
- 4. Haddick PC, Tom I, Luis E, Quinones G, Wranik BJ, Ramani SR, et al. Defining the ligand specificity of the deleted in colorectal cancer (DCC) receptor. PLoS One. 2014;9(1):e84823. Epub 2014/01/09. pmid:24400119; PubMed Central PMCID: PMCPmc3882260.
- 5. Wei P, Pattarini R, Rong Y, Guo H, Bansal PK, Kusnoor SV, et al. The Cbln family of proteins interact with multiple signaling pathways. J Neurochem. 2012;121(5):717–29. Epub 2012/01/10. pmid:22220752; PubMed Central PMCID: PMCPmc3342465.
- 6. Finci LI, Kruger N, Sun X, Zhang J, Chegkazi M, Wu Y, et al. The crystal structure of netrin-1 in complex with DCC reveals the bifunctionality of netrin-1 as a guidance cue. Neuron. 2014;83(4):839–49. Epub 2014/08/16. pmid:25123307; PubMed Central PMCID: PMCPMC4412161.
- 7. Mille F, Llambi F, Guix C, Delloye-Bourgeois C, Guenebeaud C, Castro-Obregon S, et al. Interfering with multimerization of netrin-1 receptors triggers tumor cell death. Cell death and differentiation. 2009;16(10):1344–51. Epub 2009/06/23. pmid:19543238; PubMed Central PMCID: PMCPMC2841642.
- 8. Cirulli V, Yebra M. Netrins: beyond the brain. Nature reviews Molecular cell biology. 2007;8(4):296–306. pmid:17356579.
- 9. Ozmadenci D, Feraud O, Markossian S, Kress E, Ducarouge B, Gibert B, et al. Netrin-1 regulates somatic cell reprogramming and pluripotency maintenance. Nature communications. 2015;6:7398. Epub 2015/07/15. pmid:26154507.
- 10. Castets M, Broutier L, Molin Y, Brevet M, Chazot G, Gadot N, et al. DCC constrains tumour progression via its dependence receptor activity. Nature. 2012;482(7386):534–7. Epub 2011/12/14. pmid:22158121.
- 11. Mehlen P, Rabizadeh S, Snipas SJ, Assa-Munt N, Salvesen GS, Bredesen DE. The DCC gene product induces apoptosis by a mechanism requiring receptor proteolysis. Nature. 1998;395(6704):801–4. Epub 1998/10/31. pmid:9796814.
- 12. Krimpenfort P, Song JY, Proost N, Zevenhoven J, Jonkers J, Berns A. Deleted in colorectal carcinoma suppresses metastasis in p53-deficient mammary tumours. Nature. 2012;482(7386):538–41. pmid:22358843.
- 13. Jarjour AA, Durko M, Luk TL, Marcal N, Shekarabi M, Kennedy TE. Autocrine netrin function inhibits glioma cell motility and promotes focal adhesion formation. PloS one. 2011;6(9):e25408. Epub 2011/10/08. pmid:21980448; PubMed Central PMCID: PMCPMC3182204.
- 14. Hibi K, Mizukami H, Shirahata A, Goto T, Sakata M, Sanada Y. Aberrant methylation of the netrin-1 receptor genes UNC5C and DCC detected in advanced colorectal cancer. World journal of surgery. 2009;33(5):1053–7. Epub 2009/02/27. pmid:19242752.
- 15. Yan W, Han P, Zhou Z, Tu W, Liao J, Li P, et al. Netrin-1 induces epithelial-mesenchymal transition and promotes hepatocellular carcinoma invasiveness. Digestive diseases and sciences. 2014;59(6):1213–21. pmid:24442237.
- 16. Reyes-Mugica M, Lin P, Yokota J, Reale MA. Status of deleted in colorectal cancer gene expression correlates with neuroblastoma metastasis. Laboratory investigation; a journal of technical methods and pathology. 1998;78(6):669–75. Epub 1998/06/30. pmid:9645757.
- 17. Reyes-Mugica M, Rieger-Christ K, Ohgaki H, Ekstrand BC, Helie M, Kleinman G, et al. Loss of DCC expression and glioma progression. Cancer research. 1997;57(3):382–6. Epub 1997/02/01. pmid:9012460.
- 18. Ekstrand BC, Mansfield TA, Bigner SH, Fearon ER. DCC expression is altered by multiple mechanisms in brain tumours. Oncogene. 1995;11(11):2393–402. Epub 1995/12/07. pmid:8570191.
- 19. Hara A, Saegusa M, Mikami T, Okayasu I. Loss of DCC expression in astrocytomas: relation to p53 abnormalities, cell kinetics, and survival. Journal of clinical pathology. 2001;54(11):860–5. Epub 2001/10/31. pmid:11684721; PubMed Central PMCID: PMCPMC1731311.
- 20. Lusis EA, Watson MA, Chicoine MR, Lyman M, Roerig P, Reifenberger G, et al. Integrative genomic analysis identifies NDRG2 as a candidate tumor suppressor gene frequently inactivated in clinically aggressive meningioma. Cancer Res. 2005;65(16):7121–6. Epub 2005/08/17. pmid:16103061.
- 21. Wang X, Gong Y, Wang D, Xie Q, Zheng M, Zhou Y, et al. Analysis of gene expression profiling in meningioma: deregulated signaling pathways associated with meningioma and EGFL6 overexpression in benign meningioma tissue and serum. PLoS One. 2012;7(12):e52707. Epub 2013/01/04. pmid:23285163; PubMed Central PMCID: PMCPmc3532066.
- 22. Schulten HJ, Alotibi R, Al-Ahmadi A, Ata M, Karim S, Huwait E, et al. Effect of BRAF mutational status on expression profiles in conventional papillary thyroid carcinomas. BMC Genomics. 2015;16 Suppl 1:S6. Epub 2015/04/30. pmid:25922907; PubMed Central PMCID: PMCPmc4315163.
- 23. Schneider CA, Rasband WS, Eliceiri KW. NIH Image to ImageJ: 25 years of image analysis. Nat Methods. 2012;9(7):671–5. Epub 2012/08/30. pmid:22930834.
- 24. Aarhus M, Lund-Johansen M, Knappskog PM. Gene expression profiling of meningiomas: current status after a decade of microarray-based transcriptomic studies. Acta neurochirurgica. 2011;153(3):447–56. Epub 2011/01/15. pmid:21234620; PubMed Central PMCID: PMCPMC3040823.
- 25. Carvalho LH, Smirnov I, Baia GS, Modrusan Z, Smith JS, Jun P, et al. Molecular signatures define two main classes of meningiomas. Molecular cancer. 2007;6:64. Epub 2007/10/17. pmid:17937814; PubMed Central PMCID: PMCPmc2173907.
- 26. Perez-Magan E, Rodriguez de Lope A, Ribalta T, Ruano Y, Campos-Martin Y, Perez-Bautista G, et al. Differential expression profiling analyses identifies downregulation of 1p, 6q, and 14q genes and overexpression of 6p histone cluster 1 genes as markers of recurrence in meningiomas. Neuro-oncology. 2010;12(12):1278–90. Epub 2010/08/06. pmid:20685720; PubMed Central PMCID: PMCPmc3018937.
- 27. Perez-Magan E, Campos-Martin Y, Mur P, Fiano C, Ribalta T, Garcia JF, et al. Genetic alterations associated with progression and recurrence in meningiomas. Journal of neuropathology and experimental neurology. 2012;71(10):882–93. Epub 2012/09/12. pmid:22964784.
- 28. Engstrom PG, Tommei D, Stricker SH, Ender C, Pollard SM, Bertone P. Digital transcriptome profiling of normal and glioblastoma-derived neural stem cells identifies genes associated with patient survival. Genome medicine. 2012;4(10):76. Epub 2012/10/11. pmid:23046790; PubMed Central PMCID: PMCPMC3556652.
- 29. Rowther FB, Wei W, Dawson TP, Ashton K, Singh A, Madiesse-Timchou MP, et al. Cyclic nucleotide phosphodiesterase-1C (PDE1C) drives cell proliferation, migration and invasion in glioblastoma multiforme cells in vitro. Molecular carcinogenesis. 2015. Epub 2015/01/27. pmid:25620587.
- 30. Nakaya N, Sultana A, Lee HS, Tomarev SI. Olfactomedin 1 interacts with the Nogo A receptor complex to regulate axon growth. The Journal of biological chemistry. 2012;287(44):37171–84. Epub 2012/08/28. pmid:22923615; PubMed Central PMCID: PMCPmc3481317.
- 31. Etcheverry A, Aubry M, de Tayrac M, Vauleon E, Boniface R, Guenot F, et al. DNA methylation in glioblastoma: impact on gene expression and clinical outcome. BMC genomics. 2010;11:701. Epub 2010/12/16. pmid:21156036; PubMed Central PMCID: PMCPmc3018478.
- 32. Domingues PH, Teodosio C, Otero A, Sousa P, Ortiz J, Macias Mdel C, et al. Association between inflammatory infiltrates and isolated monosomy 22/del(22q) in meningiomas. PloS one. 2013;8(10):e74798. Epub 2013/10/08. pmid:24098347; PubMed Central PMCID: PMCPmc3788099.
- 33. Zhang CM, Chen XH, Wang B, Liu F, Chi X, Tong ML, et al. Over-expression of NYGGF4 inhibits glucose transport in 3T3-L1 adipocytes via attenuated phosphorylation of IRS-1 and Akt. Acta pharmacologica Sinica. 2009;30(1):120–4. Epub 2008/12/17. pmid:19079291; PubMed Central PMCID: PMCPmc4006536.
- 34. Erdreich-Epstein A, Xu J, Ren X. PID1 IS A NOVEL SENSITIZER OF BRAIN TUMOR CELLS TO CHEMOTHERAPY. Neuro-oncology. 2014;16(Suppl 3):iii26. PubMed Central PMCID: PMCPmc4144555.
- 35. Qu X, Wei H, Zhai Y, Que H, Chen Q, Tang F, et al. Identification, characterization, and functional study of the two novel human members of the semaphorin gene family. The Journal of biological chemistry. 2002;277(38):35574–85. Epub 2002/07/12. pmid:12110693.
- 36. Chen D, Li Y, Wang L, Jiao K. SEMA6D Expression and Patient Survival in Breast Invasive Carcinoma. International journal of breast cancer. 2015;2015:539721. Epub 2015/05/15. pmid:25973277; PubMed Central PMCID: PMCPmc4417987.
- 37. Tamada Y, Imoto S, Araki H, Nagasaki M, Print C, Charnock-Jones DS, et al. Estimating genome-wide gene networks using nonparametric Bayesian network models on massively parallel computers. IEEE/ACM transactions on computational biology and bioinformatics / IEEE, ACM. 2011;8(3):683–97. Epub 2010/08/18. pmid:20714027.
- 38. Chu UB, Vorperian SK, Satyshur K, Eickstaedt K, Cozzi NV, Mavlyutov T, et al. Noncompetitive inhibition of indolethylamine-N-methyltransferase by N,N-dimethyltryptamine and N,N-dimethylaminopropyltryptamine. Biochemistry. 2014;53(18):2956–65. Epub 2014/04/16. pmid:24730580; PubMed Central PMCID: PMCPmc4025572.
- 39. Jensen LJ, Kuhn M, Stark M, Chaffron S, Creevey C, Muller J, et al. STRING 8—a global view on proteins and their functional interactions in 630 organisms. Nucleic acids research. 2009;37(Database issue):D412–6. Epub 2008/10/23. pmid:18940858; PubMed Central PMCID: PMCPmc2686466.
- 40. D'Orazi G, Rinaldo C, Soddu S. Updates on HIPK2: a resourceful oncosuppressor for clearing cancer. Journal of experimental & clinical cancer research: CR. 2012;31:63. Epub 2012/08/15. pmid:22889244; PubMed Central PMCID: PMCPmc3432601.
- 41. Deshmukh H, Yeh TH, Yu J, Sharma MK, Perry A, Leonard JR, et al. High-resolution, dual-platform aCGH analysis reveals frequent HIPK2 amplification and increased expression in pilocytic astrocytomas. Oncogene. 2008;27(34):4745–51. Epub 2008/04/15. pmid:18408760.
- 42. Cooper LA, Gutman DA, Chisolm C, Appin C, Kong J, Rong Y, et al. The tumor microenvironment strongly impacts master transcriptional regulators and gene expression class of glioblastoma. The American journal of pathology. 2012;180(5):2108–19. Epub 2012/03/24. pmid:22440258; PubMed Central PMCID: PMCPmc3354586.
- 43. Liu Y, Miao Y, Wang J, Lin X, Wang L, Xu HT, et al. DEC1 is positively associated with the malignant phenotype of invasive breast cancers and negatively correlated with the expression of claudin-1. International journal of molecular medicine. 2013;31(4):855–60. Epub 2013/02/22. pmid:23426649.
- 44. Collier FM, Loving A, Baker AJ, McLeod J, Walder K, Kirkland MA. RTKN2 Induces NF-KappaB Dependent Resistance to Intrinsic Apoptosis in HEK Cells and Regulates BCL-2 Genes in Human CD4(+) Lymphocytes. Journal of cell death. 2009;2:9–23. Epub 2009/01/01. pmid:26124677; PubMed Central PMCID: PMCPmc4474337.
- 45. Torres-Martin M, Lassaletta L, Isla A, De Campos JM, Pinto GR, Burbano RR, et al. Global expression profile in low grade meningiomas and schwannomas shows upregulation of PDGFD, CDH1 and SLIT2 compared to their healthy tissue. Oncology reports. 2014;32(6):2327–34. pmid:25333347; PubMed Central PMCID: PMCPMC4240498.
- 46. Zhang L, Zhao Y, Wang CG, Fei Z, Wang Y, Li L, et al. Neuritin expression and its relation with proliferation, apoptosis, and angiogenesis in human astrocytoma. Medical oncology. 2011;28(3):907–12. Epub 2010/04/21. pmid:20405246.
- 47. Lee HS, Lee J, Kim SO, Song JS, Lee JH, Lee SI, et al. Comparative gene-expression analysis of the dental follicle and periodontal ligament in humans. PloS one. 2013;8(12):e84201. Epub 2014/01/01. pmid:24376796; PubMed Central PMCID: PMCPmc3871683.
- 48. Lee IC, Leung T, Tan I. Adaptor protein LRAP25 mediates myotonic dystrophy kinase-related Cdc42-binding kinase (MRCK) regulation of LIMK1 protein in lamellipodial F-actin dynamics. The Journal of biological chemistry. 2014;289(39):26989–7003. Epub 2014/08/12. pmid:25107909; PubMed Central PMCID: PMCPmc4175338.
- 49. Roforth MM, Atkinson EJ, Levin ER, Khosla S, Monroe DG. Dissection of estrogen receptor alpha signaling pathways in osteoblasts using RNA-sequencing. PLoS One. 2014;9(4):e95987. Epub 2014/04/30. pmid:24776842; PubMed Central PMCID: PMCPMC4002480.
- 50. Luo Y, Jiang QW, Wu JY, Qiu JG, Zhang WJ, Mei XL, et al. Regulation of migration and invasion by Toll-like receptor-9 signaling network in prostate cancer. Oncotarget. 2015;6(26):22564–74. Epub 2015/06/19. pmid:26087186; PubMed Central PMCID: PMCPmc4673182.
- 51. Trabzuni D, Ramasamy A, Imran S, Walker R, Smith C, Weale ME, et al. Widespread sex differences in gene expression and splicing in the adult human brain. Nature communications. 2013;4:2771. Epub 2013/11/23. pmid:24264146; PubMed Central PMCID: PMCPmc3868224.
- 52. Day BW, Stringer BW, Al-Ejeh F, Ting MJ, Wilson J, Ensbey KS, et al. EphA3 maintains tumorigenicity and is a therapeutic target in glioblastoma multiforme. Cancer cell. 2013;23(2):238–48. Epub 2013/02/16. pmid:23410976.
- 53. Pechenino AS, Frick KM. THE EFFECTS OF ACUTE 17β-ESTRADIOL TREATMENT ON GENE EXPRESSION IN THE YOUNG FEMALE MOUSE HIPPOCAMPUS. Neurobiology of learning and memory. 2009;91(3):315–22. pmid:18938255; PubMed Central PMCID: PMCPmc2674265.
- 54. Sidoryk M, Matyja E, Dybel A, Zielinska M, Bogucki J, Jaskolski DJ, et al. Increased expression of a glutamine transporter SNAT3 is a marker of malignant gliomas. Neuroreport. 2004;15(4):575–8. Epub 2004/04/20. pmid:15094455.
- 55. Torres-Martin M, Lassaletta L, San-Roman-Montero J, De Campos JM, Isla A, Gavilan J, et al. Microarray analysis of gene expression in vestibular schwannomas reveals SPP1/MET signaling pathway and androgen receptor deregulation. Int J Oncol. 2013;42(3):848–62. Epub 2013/01/29. pmid:23354516; PubMed Central PMCID: PMCPmc3597452.
- 56. Senchenko VN, Krasnov GS, Dmitriev AA, Kudryavtseva AV, Anedchenko EA, Braga EA, et al. Differential expression of CHL1 gene during development of major human cancers. PLoS One. 2011;6(3):e15612. pmid:21408220; PubMed Central PMCID: PMC3049765.
- 57. Kim KK, Yang Y, Zhu J, Adelstein RS, Kawamoto S. Rbfox3 Controls the Biogenesis of a Subset of MicroRNAs. Nature structural & molecular biology. 2014;21(10):901–10. pmid:25240799; PubMed Central PMCID: PMCPmc4189996.
- 58. Kim KK, Nam J, Mukouyama Y, Kawamoto S. Rbfox3-regulated alternative splicing of Numb promotes neuronal differentiation during development. The Journal of cell biology. 2013;200(4):443–58. pmid:23420872; PubMed Central PMCID: PMCPmc3575530.
- 59. Celis-Aguilar E, Lassaletta L, Torres-Martin M, Rodrigues FY, Nistal M, Castresana JS, et al. The molecular biology of vestibular schwannomas and its association with hearing loss: a review. Genetics research international. 2012;2012:856157. Epub 2012/05/09. pmid:22567403; PubMed Central PMCID: PMCPmc3335540.
- 60. Yang J, Liu W, Lu X, Fu Y, Li L, Luo Y. High expression of small GTPase Rab3D promotes cancer progression and metastasis. Oncotarget. 2015;6(13):11125–38. Epub 2015/04/01. pmid:25823663; PubMed Central PMCID: PMCPmc4484444.
- 61. Pujol-Gimenez J, de Heredia FP, Idoate MA, Airley R, Lostao MP, Evans AR. Could GLUT12 be a Potential Therapeutic Target in Cancer Treatment? A Preliminary Report. Journal of Cancer. 2015;6(2):139–43. Epub 2015/01/07. pmid:25561978; PubMed Central PMCID: PMCPmc4280396.
- 62. Berezovsky AD, Poisson LM, Cherba D, Webb CP, Transou AD, Lemke NW, et al. Sox2 promotes malignancy in glioblastoma by regulating plasticity and astrocytic differentiation. Neoplasia. 2014;16(3):193–206,.e19–25. Epub 2014/04/15. pmid:24726753; PubMed Central PMCID: PMCPmc4094829.
- 63. Latini FR, Hemerly JP, Oler G, Riggins GJ, Cerutti JM. Re-expression of ABI3-binding protein suppresses thyroid tumor growth by promoting senescence and inhibiting invasion. Endocr Relat Cancer. 2008;15(3):787–99. pmid:18559958; PubMed Central PMCID: PMC2742300.
- 64. van den Boom J, Wolter M, Kuick R, Misek DE, Youkilis AS, Wechsler DS, et al. Characterization of gene expression profiles associated with glioma progression using oligonucleotide-based microarray analysis and real-time reverse transcription-polymerase chain reaction. The American journal of pathology. 2003;163(3):1033–43. Epub 2003/08/26. pmid:12937144; PubMed Central PMCID: PMCPmc1868272.
- 65. Shin K, Wang Q, Margolis B. PATJ regulates directional migration of mammalian epithelial cells. EMBO reports. 2007;8(2):158–64. Epub 2007/01/20. pmid:17235357; PubMed Central PMCID: PMCPmc1796763.