PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
J Neuropathol Exp Neurol. Author manuscript; available in PMC Nov 25, 2013.
Published in final edited form as:
PMCID: PMC3839953
NIHMSID: NIHMS525738
Identification of Gene Markers Associated with Aggressive Meningioma by Filtering across Multiple Sets of Gene Expression Arrays
Jourdan E. Stuart, MA,1* Eriks A. Lusis, MD,2* Adrienne C. Scheck, PhD,3 Stephen W. Coons, MD,4 Anita Lal, PhD,5 Arie Perry, MD,6†† and David H. Gutmann, MD, PhD1
1Department of Neurology, Washington University School of Medicine, St. Louis, MO
2Department of Neurosurgery, Washington University School of Medicine, St. Louis, MO
3Department of Neuro-Oncology and Neurosurgery Research, Barrow Neurological Institute of SJHMC, Phoenix, AZ
4Division of Neuropathology, Barrow Neurological Institute of SJHMC, Phoenix, AZ
5Brain Tumor Research Center, Department of Neurosurgery, University of California, San Francisco, CA
6Division of Neuropathology, Washington University School of Medicine, St. Louis, MO
Correspondence and reprint requests to: David H. Gutmann, MD, PhD, Department of Neurology, Washington University School of Medicine, Campus Box 8111, 660 South Euclid Ave., St. Louis, MO 63110. Tel: (314) 362-7379; Fax: (314) 362-2388; gutmannd/at/neuro.wustl.edu
*These authors contributed equally to this work.
Current Address: Pathwork Diagnostics, Redwood City, CA
††Current Address: Department of Pathology, University of California San Francisco, San Francisco, CA
Meningiomasare common intracranial tumors but relatively little is known about the genetic events responsible for their clinical diversity. While recent genomic studies have provided clues, the genes identified often differ among publications. We used microarray expression profiling to identify genes that are differentially expressed, with at least a 4-fold change, between grade I and grade III meningiomas. We filtered this initial set of potential biomarkers through a second cohort of meningiomas and then verified the remaining genes by quantitative polymerase chain reaction followed by examination using a third microarray expression cohort. Using this approach, we identified 9 overexpressed (TPX2, RRM2, TOP2A, PI3, BIRC5, CDC2, NUSAP1, DLG7, SOX11) and 2 underexpressed (TIMP3, KCNMA1) genes in grade III vs. grade I meningiomas. As a further validation step, we analyzed these genes in a fourth cohort and found that patients with grade II meningiomas with high topoisomerase 2-α protein expression (greater than 5% labeling-index) had shorter times to death than patients with low expression. We believe that this multistep, multi-cohort approach provides a robust method for reducing false positives while generating a list of reproducible candidate genes that are associated with clinically aggressive meningioma and are suitable for analysis for their potential prognostic value.
Keywords: Gene expression, Immunohistochemistry, Malignancy, Meningioma, Microarray, Quantitative RT-PCR
Meningiomas are common tumors, accounting for over one quarter of all central nervous system tumors (1). Although most are benign, as many as 20% of these tumors display biologically aggressive features, leading to increased patient morbidity and mortality (2). In the 2007 World Health Organization (WHO) classification scheme, meningiomas are divided into 3 grades and 13 histologic variants (3). Whereas the current grading scheme has improved our ability to predict clinical behavior, significant variation exists within tumors of the same malignancy grade. Numerous studies have attempted to develop molecular markers that stratify meningiomas into clinically distinct subsets; however, the 2 best markers utilized clinically, progesterone receptor and Ki-67, have significant limitations. Progesterone receptor expression generally decreases with increasing meningioma grade, but does not predict reliably whether grossly resected tumors will recur (4). Similarly, Ki-67 is a marker of cell proliferation and is highly associated with tumor grade; increased labeling indices have been associated with decreased recurrence free survival, especially in tumors that exhibit borderline atypia (5). Once the malignancy grade and mitotic indices are taken into account, however, the Ki-67 labeling index is not independently predictive of recurrence-free survival.
With the advent of high throughput genetic analysis, multiple studies have employed microarray expression profiling to identify genes associated with aggressive clinical behavior. These studies have uncovered genes associated with poor clinical outcome but the lists differ significantly among the studies and the majority of the investigations have not examined these genes across multiple tumor sets (69). Using multiple sets of arrays, distinct microarray platforms, and different tumor cohorts, the clinical behavior of the much more commonly studied high-grade gliomas (i.e. anaplastic astrocytoma and glioblastoma) can be more reliably predicted using specific genetic signatures (10, 11). In this report, we employed a similar strategy to assess benign and malignant meningiomas.
Tissue Specimens
Tissue specimens were collected from the Division of Neuropathology and the Siteman Cancer Center at Washington University School of Medicine, the Neurological Surgery Tissue Bank at UCSF, and the Barrow Neurological Institute. All tissue was obtained under protocols approved by the Human Studies Committee at all 3 institutions. Sections of snap-frozen tissue specimens from surgically resected meningiomas were first reviewed for specimen adequacy (at least 80% tumor). Subsequent 50-μm serial sections from each banked frozen specimen were then cut, placed immediately into Trizol reagent (Invitrogen, Carlsbad, CA), and homogenized for RNA preparation. Total RNA was isolated from Trizol homogenates using the manufacturer’s protocol. RNA was then further purified using RNeasy spin columns (Qiagen, Inc., Valencia, CA). All RNAs were quantified by UV absorbance at 260 and 280 nm and qualitatively assessed using an RNA Nano Assay and 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA). For validation studies using immunohistochemistry (IHC), formalin-fixed paraffin-embedded tissue was available from 4 of the same WHO grade I and 3 of the same WHO grade II meningiomas that were analyzed by expression profiling.
Expression Microarrays
Raw data (.CEL files) from 2 previously published studies were obtained from the authors and the Washington University Biomedical Informatics Core (12, 13). The first set consisted of 5 WHO grade I and 5 WHO grade III tumors processed on Affymetrix HG-U133A and HG-U133B microarrays (Affymetrix Inc., Santa Clara, CA) at Washington University. The second set included 8 grade I and 8 grade III samples processed on Affymetrix HG-U133 Plus 2.0 microarrays (Affymetrix) at UCSF. The Affymetrix HG-U133 Plus 2.0 microarrays contain all the probe sets found on the Affymetrix HG-U133A and HG-U133B microarrays as well as an additional 9,000 unique probe sets. It is possible that more data could have been generated if we had employed the newer microarray chips for both cohorts, however, because the second set contained all the probe sets found in the first set, it his highly unlikely that biomarker genes identified in the first set would not be identified in the second set. Signal intensity measures were generated from probe level data using the MAS 5.0 algorithm in Affymetrix Expression Console. Each array was scaled to a target intensity of 1500. Preprocessing included log2 transformation and removal of probes with all “absent” detection by absence/presence calls in DecisionSite (Spotfire, Somerville, MA). Significance Analysis of Microarrays (SAM) (Microsoft, Redmond, WA) 2-class unpaired analysis was used to calculate q-values and fold changes (FC) in expression levels between grade I and grade III meningiomas in Microsoft Excel, as previously described (14). Candidate genes were selected by filtering SAM results by FC >4 and q-value. To confirm the size of the samples needed in the second set for subsequent validation, we used the sample size calculator for microarrays (M.D. Anderson Cancer Center; http://bioinformatics.mdanderson.org/) to calculate the power of the gene expression array analysis.
A third set used for validation was provided by one of the authors (A.S.) (15, 16)and is available on the Gene Expression Omnibus as GEO4780 (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE4780). This dataset consisted of 5 U133A gene expression microarrays and 57 U133 Plus 2.0 arrays. We analyzed the 56 U133 Plus 2.0 arrays for which WHO tumor grades were available. The total number of tumors in this cohort included 33 grade I, 20 grade II, and 3 grade III meningiomas. Log2 data were imported into Microsoft Excel and the validated gene probe sets sorted. These were then imported into Graph Pad Prism® 5 (Graph Pad Software, La Jolla, CA) and histograms of relative gene expression vs. tumor grade were plotted. P values were calculated by the Mann Whitney U test; p values less than 0.05 were considered significant.
Patient/Tumor Cohorts
Demographic and clinical outcome data were available for most of the patients in the first 2 sets of expression data sets and from a separate cohort of grade II meningiomas that was utilized for IHC validation studies of candidate genes generated from expression profiling experiments (Tables 1, ,2).2). To exclude possible selection biases that could affect survival outcomes, these 3 cohorts were compared in terms of age, sex, tumor location, and extent of resection using the Fisher exact test. Tables 1 and and22 also identify which tissues were obtained from primary vs. recurrent tumor resections. Demographic data were not statistically different between the 3 independent patient populations (driver set vs. tester set, driver set vs. grade II meningiomas, and tester set vs. grade II meningiomas).
Table 1
Table 1
Clinical Features of the Profiled Meningiomas
Table 2
Table 2
Clinical Features of Analyzed Grade II Meningiomas
The histologic subtypes of the grade I meningiomas were similar in the 2 grade I cohorts. The “driver” set was composed of 1 meningothelial, 3 transitional and 1 fibroblastic meningiomas; the “tester” set was composed of 3 meningothelial, 4 transitional and 1 fibroblastic meningiomas. All of the grade III meningiomas were anaplastic tumors (Table 1). The cohort of grade II meningiomas used for the IHC validation studies were atypical meningiomas, except for 1 clear cell and 1 chordoid meningioma in the indolent group (Table 3). All meningiomas in the “driver” set had “absent” calls in at least 6 of the 8 represented probe sets representing the NF2 gene locus. In the “tester” set, 3 of the 8 WHO grade I and 7 of the 8 WHO III had “absent” calls at the NF2 gene locus. This was not statistically significant between the different tumor populations (Fisher exact test).
Table 3
Table 3
Histologic Characteristics of Grade II and WU Grade III Meningiomas
Quantitative Reverse Transcription-PCR (RT-qPCR)
Quantitative RT-qPCR was performed on 10 WHO grade I and 10 WHO grade III meningiomas. Aliquots of 2000 ng of total cellular RNA were combined with 2 μL 10X RTbuffer, 2 μL 5 mM dNTP mix, 1 μL Omniscript RT enzyme (Qiagen, Inc.), 0.5 μg oligo (dT)12–18 and 60 ng random hexamer primers, and 0.25 μL RNaseOUT (Invitrogen) to a volume of 20 μl in molecular-grade water. The mixture was incubated for 60 minutes at 37°C and then brought to a total volume of 100 μl with molecular-grade water.
Oligonucleotide primers were designed using the PrimerQuest tool (Integrated DNA Technologies, Coralville, IA). Each PCR reaction consisted of 50 ng cDNA, 100 nM forward primer, 100 nM reverse primer, and 10 μL 2X SYBR Green PCR master mix (Applied Biosystems, Foster City, CA) to a volume of 20 μL. Each sample was run in duplicate on the same 96-well plate and cycled at 95°C for 10 minutes, followed by 40 cycles of 95°C for 15 seconds and 60°C for 1 minute on an ABI SDS 7000 sequence detection system (Applied Biosystems). Fluorescent data were converted into cycle threshold (CT) measurements and thermal dissociation curves were reviewed for biphasic melting curves using SDS software. Any sample with ΔCT between duplicates ≥3 or dissociation curves varying in melting point from that of the positive control were excluded from further analysis. The positive control was pooled cDNA from 10 frozen meningiomas (5 grade I, 3 grade II, 2 grade III tumors) from the Siteman Cancer Center Tumor Bank. The ΔΔCT method was used to calculate relative expression relative to grade I samples using glyceraldehyde-3-phosphate dehydrogenase as the reference transcript (17). Relative expression between groups was compared by the Mann-Whitney U test with 2-tailed p < 0.05 considered significant.
For quality assurance, we ran 20 μl of each RT-qPCR sample with loading dye on a 1.5% agarose gel with 0.0003% ethidium bromide in 1X TAE buffer. We compared amplicon size to a 100 bp ladder (Lambda Biotech, St. Louis, MO). The product was run on the gel for 35 minutes at 100 V. The gel was visualized on a UVP BioDoc-It (UVP, LLC, Upland, CA).
Immunohistochemistry
IHC was conducted in whole tissue sections as previously published (18). The whole tissue sections included 4 WHO grade I (corresponding to driver set cohort), 40 WHO grade II (independent cohort), and 3 WHO grade III (from driver set cohort) meningiomas. Briefly, sections were deparaffinized in xylene, rehydrated in alcohol, and antigen retrieval was performed in boiling citrate solution (pH 6.0) for 15 minutes. Blocking included 0.3% H2O2 in methanol and the Avidin/Biotin Blocking kit (Vector Laboratories, Burlingame, CA), respectively. Slides were incubated with the mouse monoclonal antibodies to tissue inhibitor of metalloproteinase 3 (TIMP3) (Millipore, Billerica, MA: 1:2000). Baculoviral inhibitor of apoptosis repeat-containing 5 (BIRC5, also known as survivin) (Lifespan Biosciences, Seattle, WA; 1:100), and topoisomerase 2-α (TOP2A) (Lifespan Biosciences; 1:1000). Biotinylated anti-mouse secondary antibody (Vector Laboratories) was applied for 1 hour at the specified dilutions, followed by ABC reagent for 1 hour, and development in 3, 3′ diaminobenzidine (Vector) for 1 minute. Positive controls were used as specified by the manufacturer, and omission of the primary antibodies served as negative controls.
Scoring was recorded as the TOP2A labeling index using a cutoff of 5%, as previously described (19). The scores for TIMP3 and BIRC5 were recorded as the percentage of immunopositive tumor cells (0 = none, 1 = <10%, 2 = 10% –50%, 3 = 50%–90%, and 4 = >90%) and the extent of staining intensity (0 = negative, 1 = weak, 2 = moderate, and 3 = strong). These measurements were combined into a multiplied score, as described (20). Slides were independently scored by 2 of the authors in a blinded fashion and any discrepant scores were re-examined to arrive at consensus scores. All statistics and Kaplan Meier curves were calculated and generated using Graph Pad Prism® 5 (Graph Pad Software).
We examined expression patterns of WHO grade I and III meningiomas using multiple previously generated microarray expression data sets (12, 13). Initially, using supervised hierarchical clustering, the differential expression patterns of all genes in the first set of meningiomas revealed distinct genetic signature differences between these 2 histologic tumor grades (Fig. 1). To narrow the list of genes, we applied a filtering algorithm using an independently generated dataset (Fig. 2). In the “driver” set using a FC ≥4 and a false discovery rate (FDR) q-value) < 25%, SAM identified 28 genes that were overexpressed and 20 genes that were underexpressed in grade III vs. grade I meningiomas. No differentially expressed genes were identified on the U133B arrays. To confirm the size of the samples needed in the “tester” set for subsequent validation, we used the sample size calculator for microarrays on 48 genes from the SAM analysis generated in set 1 using a false discovery rate q < 0.25 and a FC of 4. We calculated that 8 samples per group would provide 80% power to validate these 48 genes with no greater than 3 false positives.
Figure 1
Figure 1
A representative supervised hierarchical clustering map revealing genes that are differentially expressed in World Health Organization (WHO) grade I vs. WHO grade III meningiomas (p < 0.05).
Figure 2
Figure 2
Filtering algorithm used to generate Table 4. FC = fold change, q = false discovery rate q value.
We then filtered our candidate gene list on an independent “tester” second set. Using a FC ≥4 and q < 5%, 14 of the 28 overexpressed and 9 of the 20 underexpressed candidate genes from set 1 were validated, resulting in a total of 23 candidate genes (Table 4). This analysis additionally identified 177 overexpressed and 144 underexpressed genes not identified in the first set.
Table 4
Table 4
Candidate Genes Identified by Expression Microarray Analyses
To confirm the differential expression of the 23 candidate genes, we next performed RT-qPCR using 10 grade I and 10 grade III frozen meningioma samples, a third of which overlapped with the “tester” set because we did not have any more available frozen samples. By ΔΔCT analysis, the expression of 11 of the 23 candidate genes was validated (Fig. 3; Table 5): TPX2, RRM2, TOP2A, PI3, BIRC5, CDC2, NUSAP1, DLG7, and SOX11 were overexpressed in grade III vs. grade I meningiomas, whereas TIMP3 and KCNMA1 were underexpressed in grade III vs. their benign counterparts (Fig. 4).
Figure 3
Figure 3
Filtering algorithm used to generate Table 5. FC = fold change, q = false discovery rate q value. PCR = real time reverse transcribed RNA PCR.
Table 5
Table 5
Candidate Genes with Expression Validated by qRT-PCR
Figure 4
Figure 4
Differential RNA expression in human meningiomas as measured by real-time reverse transcribed RNA PCR. (a–k) Relative gene expression was increased in grade III meningiomas compared to grade I meningiomas (9 genes; a–i) and decreased in (more ...)
Of the remaining candidate genes that did not validate, 8 genes had relative expression patterns that positively trended with microarray results, but failed to reach statistical significance using only 10 tumors of each malignancy grade: AURKA(p = 0.49), FANCI (p = 0.11), GPX3 (p = 0.054), HMMR (p = 0.13), KIF20A (p = 0.23), NEK2 (p = 0.23), SSPN (p = 0.97), and TCEAL2 (p = 0.054). Based on the microarray data, 4 candidate genes showed expression patterns opposite to those expected, but none of these reached statistical significance. These included LMOD1 (p = 0.55), NPY1R (p = 0.46), P2RY14 (p = 0.21), and PID1 (p = 0.25).
As a separate validation step, we obtained meningioma expression data from a third independent microarray consisting of 33 WHO grade I, 20 WHO grade II, and 3 WHO grade III meningiomas from the Barrow Neurological Institute. Ten of the 11 genes that validated by RT-qPCR (TPX2, RRM2, TOP2A, PI3, BIRC5, CDC2, NUSAP1, SOX11, TIMP3, and KCNMA1 )were found on the Affymetrix HG-U133 plus 2 GeneChip. Similar patterns of over- and underexpression were observed in this dataset. Statistically significant differences were found for 9 of the 10 genes (Mann-Whitney U test p value ≤0.05) (Fig. 5). Within the group of WHO grade I meningiomas there was some variability in the expression of specific genes, but the variability was random and was not in the same cases that proved to be outliers for all of the genes evaluated.
Figure 5
Figure 5
Histograms of relative gene expression in a series of 33 World Health Organization (WHO) grade I, 20 WHO grade II, and 3 WHO grade III meningiomas. (a–j) Relative increased gene expression (a–h) or decreased gene expression (i, j) was (more ...)
To examine whether these filtered and validated genes could be validated at the protein level, IHC was performed on 3 of the proteins for which suitable antibodies were commercially available (BIRC5, TOP2A, and TIMP3). For this analysis, whole section paraffin-embedded tumor specimens including 4 grade I and 3 grade III meningiomas that were represented in first “driver” set of meningiomas. We added 40 WHO grade II tumors with similar demographics (Table 2). These 40 tumors were selected as groups of 21 “indolent” and 19 “aggressively” behaving WHO grade II tumors to determine whether the protein biomarkers could stratify them on the basis of clinical behavior. The predicted expression patterns of the 3 proteins were confirmed in the grade I and grade III meningiomas (i.e. the same patterns as encountered in expression profiling experiments). Patients with WHO grade II meningiomas and high TOP2A labeling indices (>5%) had shorter times to death (TTD) (p = 0.0308; Fig. 6), and a trend towards shorter times to recurrence (TTR) (p = 0.1062). TIMP3 and BIRC5 protein expression showed similar trends, but these did not reach statistical significance when immunopositive score cut-offs of 4% and 3%, respectively, were used (20) (TIMP3; TTD p = 0.1693, TTR p = 0.1215; BIRC5 TTD p = 0.1705). Representative photomicrographs of TOP2A IHC are illustrated in Figure 7.
Figure 6
Figure 6
Survival rates of patients with World Health Organization (WHO) grade II meningioma stratified by topoisomerase 2-α (TOP2A) protein expression. Patients in this group withhigh TOP2A expression (>5%)had shorter time to death (TTD) than (more ...)
Figure 7
Figure 7
Photomicrograph showing (A) low and (B) high levels of nuclear topoisomerase 2-α immunoreactivity, magnification: 200x.
The potential biomarkers for meningiomas identified using microarray gene expression approaches to date differ significantly from report to report(69, 12, 13). This likely reflects the relatively small numbers of tumors studied, differences in bioinformatics strategies, and other technical artifacts that are difficult to control when many transcripts are assessed at the same time. To minimize the risk of false positive discovery, a number of strategies are typically employed, including statistical tests such as multiple test correction, validation by use of several different platforms (quantitative PCR), and the inclusion of multiple independent tumor cohorts.
In the current study, we applied all of these strategies, filtering genes across 2 independent sets of gene expression microarrays from 2 separate institutions, validating common genes by RT-qPCR, and confirming our findings using a third set of meningioma microarray gene expression data. This stringent approach generated smaller lists of candidate genes that are more likely to be biologically relevant. At the same time we recognize that other biologically relevant markers might have been discarded in the process. By requiring each candidate to be identified in every cohort examined, we almost certainly have missed some of the genes that may only be implicated in smaller subsets of malignant meningiomas. For example, we eliminated a previously identified gene (NDRG2) reported by our group in clinically aggressive meningiomas (13). Nonetheless, NDRG2 likely is biologically important because it 1) was previously validated at the protein level, 2) is linked to a cytogenetically relevant site of deletion (chromosome 14q), and 3) was found to have its second allele frequently inactivated by promoter region methylation. Similarly, in the current study, we also identified genes in only 1 cohort that could not be validated in a second cohort, thereby further underscoring the need for multiple independent validation sets. These validation “failures” likely result from the small number of samples in each cohort and the required use of different q value cut-offs when comparing across independently generated microarray data sets. In this regard, the smaller the number of samples in any group, the larger the change in differential gene expression is required to reach statistical significance. Despite these problems inherent in the high stringency filtering approach, it was a useful strategy for identifying the most common and reproducible progression-associated genetic alterations.
The purpose of this study was to identify proteins whose differential expression correlated with aggressive meningioma behavior and poor patient outcome. It was not our intention to replace well-validated markers (e.g. Ki-67) in the routine evaluation of meningiomas, but to demonstrate that the use of an algorithm over multiple independent sets of meningiomas facilitates the identification of other protein biomarkers useful for stratifying meningiomas in clinically relevant subgroups. Ki-67 has proven to be an excellent biomarker for predicting clinical behavior in univariate analyses but it has some limitations. One of us (A.P.) specifically examined this question in a large set of 425 meningiomas using univariate analysis. In that study, Ki-67 was predictive of recurrence-free survival whereas on multivariate analysis after malignancy grade and mitotic index were taken into account, Ki-67 was not independently predictive (5). Similarly, in the present study Ki-67 exhibited the highest p value in the grade II meningiomas on Kaplan Meier survival curves (p = 0.5754, p = 0.2554; TTD, TTR, respectively). Thus, although Ki-67 is strongly correlated with tumor grade in general, it does not distinguish between clinically aggressive and indolent atypical meningiomas.
Of the 11 genes identified in our study, 4 of them (BIRC5, TOP2a, TIMP3 and KCNMA1) were previously implicated in meningioma tumor progression, thereby strengthening their association with aggressive meningioma behavior. BIRC5 (survivin) is an anti-apoptotic protein that is highly expressed in meningiomas (18, 2124). BIRC5 is located on 17q, a known region of chromosomal gain in grade III meningiomas. Another gene located on 17q is TOP2A. TOP2A controls DNA folding states during transcription and 2 previous studies have shown that TOP2A expression is an independent predictor of meningioma recurrence (19, 25). A third study did not find such an association in benign and atypical meningiomas (26). Our finding that TOP2A protein expression was associated with poor overall survival in grade II meningiomas underscores the utility of this specific biomarker. TIMP3 is located at 22q12.3, the same chromosomal arm as NF2 (22q12.2). This reg ion is commonly lost in all grades of meningiomas. TIMP3 is a member of a family of proteins that regulate cell motility and invasion by regulating metalloproteinase function. We found that TIMP3 was underexpressed in grade III meningiomas and has been associated with poor survival in both esophageal squamous cell carcinoma (27)and meningioma (20). In addition, 2 recent reports found reduced TIMP3 expression in tumors with more complex karyotypes (8) or clinically aggressive behavior (9). In the latter study, BIRC5 expression was also found to be increased in grade III meningiomas. Finally, we also identified KCNMA1 as a potential meningioma biomarker. Although we did not analyze KCNMA1 expression at the protein level, expression of this potassium large-conductance calcium-activated channel protein was also reduced in high-grade meningiomas in previous studies (8, 9).
Other genes identified in this study have not previously been linked to meningioma: RRM2 overexpression has been linked to angiogenesis, tumor invasion, cell proliferation, and drug resistance in a variety of different cancers (2830). CDC2, a cyclin-dependent kinase, is important for regulating cell proliferation (34) and TPX2, a microtubule-associated protein, is important in normal mitotic spindle formation (31). Although its role in meningioma tumor progression has not been explored to date, TPX2 overexpression is associated with increased proliferation in malignant salivary gland tumors and with decreased 5 y survival rates in patients with squamous cell carcinoma of the lung (32, 33). PI3, also known as elafin, is an elastase-specific inhibitor with anti-microbial, anti-inflammatory, and immunomodulatory functions (34); high PI3 levels have recently been correlated with poor survival in glioblastoma (35). DLG7 overexpression results in increased cell growth and its expression is upregulated in liver and colon carcinomas (36). Increased NUSAP1 expression is associated with melanoma progression (37). SOX11, a sex-determining region-box protein, is a transcription factor that functions in the developing nervous system and has been reported to be overexpressed in medulloblastoma and malignant glioma (3841).
Using IHC on select grade I and grade III meningiomas from the original “driver set”, we confirmed the predicted differential protein expression of TOP2A, TIMP3, and BIRC5. In particular, our data suggest that TOP2A may be a useful ancillary marker since there is a commercial antibody to it. Thus, our data suggest that IHC may help distinguish clinically aggressive from more indolent WHO grade II meningiomas.
One of the unique features of the approach we employed is the application of a filtering algorithm using a second independently generated microarray dataset to reduce the numbers of false positives. This resulted in the identification of a robust set of 11 genes that were validated by RT-qPCR on a third set of meningiomas and by microarray analysis on a fourth set of tumors. This approach increases the predictive value of the candidate genetic biomarkers discovered. Future studies employing larger and prospectively collected cohorts of tumors will be required to determine whether these differentially expressed genes have prognostic value in the management of meningioma.
Acknowledgments
We acknowledge support from the Vincent Buono Research Fund (to D.H.G.) and the Arizona Biomedical Research Commission (to A.C.S). J.E.S. was supported on an institutional TL1 Predoctoral Grant. E.A.L. received supported from a Departmental T32 training grant. A.L. received support from a SPORE developmental program grant (P50CA097257). Funds were also provided from a Washington University Division of Anatomic and Molecular Pathology grant.
We thank Dr. Mark Watson and Dr. Jacqueline Payton (Department of Pathology and Immunology, Division of Laboratory and Genomic Medicine Washington University School of Medicine, St. Louis, MO) for their advice during the analysis of these data sets. We thank Dr. Jingqin (Rosy) Luo (Department of Biostatistics, Washington University School of Medicine, St. Louis, MO) for her advice in the analysis of the expression microarrays. We also thank William P. Hendricks (Department of Neuro-Oncology Research, Barrow Neurological Institute, Phoenix, AZ) for his help with the Barrow Neurological Institute sample set.
1. CBTRUS. Statistical Report: Primary Brain Tumors in the United States, 1995–1999. Published by the Central Brain Tumor Registry of the United States; 2002.
2. Perry A, Scheithauer BW, Stafford SL, et al. “Malignancy” in meningiomas: a clinicopathologic study of 116 patients, with grading implications. Cancer. 1999;85:2046–56. [PubMed]
3. Perry ALD, Scheithauer BW, Budka H, Von Deimling A. WHO classification of tumours of the central nervous system. IARC; Lyon: 2007. Meningiomas.
4. Roser F, Nakamura M, Bellinzona M, et al. The prognostic value of progesterone receptor status in meningiomas. J Clin Pathol. 2004;57:1033–7. [PMC free article] [PubMed]
5. Perry A, Stafford SL, Scheithauer BW, et al. The prognostic significance of MIB -1, p53, and DNA flow cytometry in completely resected primary meningiomas. Cancer. 1998;82:2262–9. [PubMed]
6. Martinez-Glez V, Franco-Hernandez C, Rey JA. Microarray gene expression profiling in meningiomas and schwannomas. Curr Med Chem. 2008;15:826–33. [PubMed]
7. Watson MA, Gutmann DH, Peterson K, et al. Molecular characterization of human meningiomas by gene expression profiling using high-density oligonucleotide microarrays. Am J Pathol. 2002;161:665–72. [PubMed]
8. Tabernero MD, Maillo A, Gil-Bellosta CJ, et al. Gene expression profiles of meningiomas are associated with tumor cytogenetics and patient outcome. Brain Pathol. 2009;19:409–20. [PubMed]
9. Fevre-Montange M, Champier J, Durand A, et al. Microarray gene expression profiling in meningiomas: differential expression according to grade or histopathological subtype. Int J Oncol. 2009;35:1395–407. [PubMed]
10. Nutt CL, Mani DR, Betensky RA, et al. Gene expression-based classification of malignant gliomas correlates better with survival than histological classification. Cancer Res. 2003;63:1602–7. [PubMed]
11. Freije WA, Castro-Vargas FE, Fang Z, et al. Gene expression profiling of gliomas strongly predicts survival. Cancer Res. 2004;64:6503–10. [PubMed]
12. Carvalho LH, Smirnov I, Baia GS, et al. Molecular signatures define two main classes of meningiomas. Mol Cancer. 2007;6:64. [PMC free article] [PubMed]
13. Lusis EA, Watson MA, Chicoine MR, et al. Integrative genomic analysis identifies NDRG2 as a candidate tumor suppressor gene frequently inactivated in clinically aggressive meningioma. Cancer Res. 2005;65:7121–6. [PubMed]
14. Tusher VG, Tibshirani R, Chu G. Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci U S A. 2001;98:5116–21. [PubMed]
15. Hendricks W, Coons S, Preul M, et al. Global expression profiles predict meningioma recurrence and reveal novel pathways leading to aggressive phenotypes: a new clinical model of meningioma aggression. AACR Meeting Abstracts. 2005;1:1149.
16. Hendricks WP, Pfisterer W, Nieman R, et al. Characterization of molecular, genetic, and biochemical profiles associated with meningioma progression and recurrence. Neuro Oncol. 2004;6:337.
17. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods. 2001;25:402–8. [PubMed]
18. Lusis EA, Chicoine MR, Perry A. High throughput screening of meningioma biomarkers using a tissue microarray. J Neurooncol. 2005;73:219–23. [PubMed]
19. Korshunov A, Shishkina L, Golanov A. DNA topoisomerase II-alpha and cyclin A immunoexpression in meningiomas and its prognostic significance: an analysis of 263 cases. Arch Pathol Lab Med. 2002;126:1079–86. [PubMed]
20. Barski D, Wolter M, Reifenberger G, et al. Hypermethylation and transcriptional downregulation of the TIMP3 gene is associated with allelic loss on 22q12. 3 and malignancy in meningiomas. Brain Pathol. 2010;20:623–31. [PubMed]
21. Al-Khalaf HH, Lach B, Allam A, et al. Expression of survivin and p16(INK4a)/Cdk6/pRB proteins and induction of apoptosis in response to radiation and cisplatin in meningioma cells. Brain Res. 2008;1188:25–34. [PubMed]
22. Das A, Tan WL, Smith DR. Expression of the inhibitor of apoptosis protein survivin in benign meningiomas. Cancer Lett. 2003;193:217–23. [PubMed]
23. Sasaki T, Lopes MB, Hankins GR, et al. Expression of survivin, an inhibitor of apoptosis protein, in tumors of the nervous system. Acta Neuropathol. 2002;104:105–9. [PubMed]
24. Katoh M, Wilmotte R, Belkouch MC, et al. Survivin in brain tumors: an attractive target for immunotherapy. J Neurooncol. 2003;64:71–6. [PubMed]
25. Korshunov A, Shishkina L, Golanov A. Immunohistochemical analysis of p16INK4a, p14ARF, p18INK4c, p21CIP1, p27KIP1 and p73 expression in 271 meningiomas correlation with tumor grade and clinical outcome. Int J Cancer. 2003;104:728–34. [PubMed]
26. Konstantinidou AE, Patsouris E, Korkolopoulou P, et al. DNA topoisomerase IIalpha expression correlates with cell proliferation but not with recurrence in intracranial meningiomas. Histopathology. 2001;39:402–8. [PubMed]
27. Gu P, Xing X, Tanzer M, et al. Frequent loss of TIMP-3 expression in progression of esophageal and gastric adenocarcinomas. Neoplasia. 2008;10:563–72. [PMC free article] [PubMed]
28. Zhang K, Hu S, Wu J, et al. Overexpression of RRM2 decreases thrombspondin-1 and increases VEGF production in human cancer cells in vitro and in vivo: implication of RRM2 in angiogenesis. Mol Cancer. 2009;8:11. [PMC free article] [PubMed]
29. Zhou B, Yen Y. Characterization of the human ribonucleotide reductase M2 subunit gene; genomic structure and promoter analyses. Cytogenet Cell Genet. 2001;95:52–9. [PubMed]
30. Duxbury MS, Whang EE. RRM2 induces NF-kappaB-dependent MMP-9 activation and enhances cellular invasiveness. Biochem Biophys Res Commun. 2007;354:190–6. [PubMed]
31. Gruss OJ, Wittmann M, Yokoyama H, et al. Chromosome-induced microtubule assembly mediated by TPX2 is required for spindle formation in HeLa cells. Nat Cell Biol. 2002;4:871–9. [PubMed]
32. Shigeishi H, Ohta K, Hiraoka M, et al. Expression of TPX2 in salivary gland carcinomas. Oncol Rep. 2009;21:341–4. [PubMed]
33. Ma Y, Lin D, Sun W, et al. Expression of targeting protein for xklp2 associated with both malignant transformation of respiratory epithelium and progression of squamous cell lung cancer. Clin Cancer Res. 2006;12:1121–7. [PubMed]
34. Moreau T, Baranger K, Dade S, et al. Multifaceted roles of human elafin and secretory leukocyte proteinase inhibitor (SLPI), two serine protease inhibitors of the chelonianin family. Biochimie. 2008;90:284–95. [PubMed]
35. Saidi A, Javerzat S, Bellahcene A, et al. Experimental anti -angiogenesis causes upregulation of genes associated with poor survival in glioblastoma. Int J Cancer. 2008;122:2187–98. [PubMed]
36. Gudmundsson KO, Thorsteinsson L, Sigurjonsson OE, et al. Gene expression analysis of hematopoietic progenitor cells identifies Dlg7 as a potential stem cell gene. Stem Cells. 2007;25:1498–506. [PubMed]
37. Ryu B, Kim DS, Deluca AM, et al. Comprehensive expression profiling of tumor cell lines identifies molecular signatures of melanoma progression. PLoS One. 2007;2:e594. [PMC free article] [PubMed]
38. Hyodo-Miura J, Urushiyama S, Nagai S, et al. Involvement of NLK and Sox11 in neural induction in Xenopus development. Genes Cells. 2002;7:487–96. [PubMed]
39. Wurm A, Sock E, Fuchshofer R, et al. Anterior segment dysgenesis in the eyes of mice deficient for the high-mobility-group transcription factor Sox11. Exp Eye Res. 2008;86:895–907. [PubMed]
40. Lee CJ, Appleby VJ, Orme AT, et al. Differential expression of SOX4 and SOX11 in medulloblastoma. J Neurooncol. 2002;57:201–14. [PubMed]
41. Weigle B, Ebner R, Temme A, et al. Highly specific overexpression of the transcription factor SOX11 in human malignant gliomas. Oncol Rep. 2005;13:139–44. [PubMed]