PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of oncotargetLink to Publisher's site
 
Oncotarget. 2017 January 3; 8(1): 191–202.
Published online 2016 November 12. doi:  10.18632/oncotarget.13329
PMCID: PMC5352106

IGFBP2 expression predicts IDH-mutant glioma patient survival

Abstract

Mutations of the isocitrate dehydrogenase (IDH) 1 and 2 genes occur in ~80% of lower-grade (WHO grade II and grade III) gliomas. Mutant IDH produces (R)-2-hydroxyglutarate, which induces DNA hypermethylation and presumably drives tumorigenesis. Interestingly, IDH mutations are associated with improved survival in glioma patients, but the underlying mechanism for the difference in survival remains unclear. Through comparative analyses of 286 cases of IDH-wildtype and IDH-mutant lower-grade glioma from a TCGA data set, we report that IDH-mutant gliomas have increased expression of tumor-suppressor genes (NF1, PTEN, and PIK3R1) and decreased expression of oncogenes(AKT2, ARAF, ERBB2, FGFR3, and PDGFRB) and glioma progression genes (FOXM1, IGFBP2, and WWTR1) compared with IDH-wildtype gliomas. Furthermore, each of these genes is prognostic in overall gliomas; however, within the IDH-mutant group, none remains prognostic except IGFBP2 (encodinginsulin-like growth factor binding protein 2). Through validation in an independent cohort, we show that patients with low IGFBP2 expressiondisplay a clear advantage in overall and disease-free survival, whereas those with high IGFBP2 expressionhave worse median survival than IDH-wildtype patients. These observations hold true across different histological and molecular subtypes of lower-grade glioma. We propose therefore that an unexpected biological consequence of IDH mutations in glioma is to ameliorate patient survival by promoting tumor-suppressor signaling while inhibiting that of oncogenes, particularly IGFBP2.

Keywords: DNA hypermethylation, glioma, IDH, IGFBP2, prognosis

INTRODUCTION

Malignant gliomas are the most common primary brain tumors in adults [1], yet recurrent genetic changes that drive pathogenesis of WHO grade II and grade III gliomas (referred to collectively hereafter as lower-grade glioma, LGG) remain to be identified [2, 3]. Single somatic mutations of the isocitrate dehydrogenase 1 (IDH1) gene, predominantly R132H, occurred in ~80% of LGGs as well as in secondary glioblastomas [4, 5]. IDH1 encodes the cytosolic isocitrate dehydrogenase 1, an enzyme that catalyzes the conversion of isocitrate to 2-oxoglutarate (aka α-ketoglutarate) concomitant with the production of reduced nicotinamide adenine dinucleotide phosphate. Tumors lacking mutations in IDH1 often have single mutations in IDH2, a mitochondrial gene in the citric acid cycle.

The most striking biochemical finding from IDH1 and IDH2 mutations is the acquired neomorphic activity to catalyze the reduction of 2-oxoglutarate to the (R)-enantiomer of 2-hydroxyglutarate [(R)2-HG] [6, 7]. (R)2-HG has been shown to be a competitive inhibitor of multiple 2-oxoglutarate-dependent dioxygenases, including histone demethylases and the TET family of 5-methylcytosine hydroxylases [8, 9]. In keeping with this, mutant IDH1 inhibits histone demethylation and induces DNA hypermethylation in cell culture and animal model, thereby blocking cell differentiation [1012]. Likewise, glioblastomas and LGGs harboring IDH mutations manifest a CpG island methylator phenotype [13, 14]. Although pharmacological targeting of mutant IDH1 or IDH2 induces tumor cell differentiation through reduction of (R)2-HG production [15, 16], the inhibitory effect on glioma growth remains less clear [16, 17], and there are no appreciable changes in genome-wide DNA methylation [16]. Therefore, the role of DNA hypermethylation and its target gene regulation remain obscure in glioma biology. An intriguing clinical feature of IDH-mutant glioma is the prolonged overall survival in comparison with the IDH-wildtype: ranging from a two- to threefold increase in patients with glioblastomas to a three- to fivefold increase in those with LGGs [2, 4, 5]. Whether DNA methylation-associated gene expression is a cause of the survival advantage has yet to be explored.

To address these questions, we performed comparative analyses of 286 cases of IDH-wildtype and IDH-mutant LGGs to identify differential expression of genes that not only account for prolonged survival but also correlate specifically with IDH mutations.

RESULTS

Increased expression of NF1 and PTEN in IDH-mutant gliomas correlates with improved patient survival

To provide an explanation for the improved survival in IDH-mutant patients relative to the wildtype patients, we first focused on glioma-relevant genes that possess tumor-suppressing activities, one of the key hallmarks of cancer [18, 19]. We compared the methylation and expression data between IDH-wildtype and IDH-mutant gliomas. Surprisingly, the IDH-mutant group exhibited notably decreased NF1 methylation concomitant with increased mRNA levels in reference to the wild type (Figures (Figures1A1A and and2A).2A). The mean mRNA and protein levels of NF2 and PTEN were also much higher than those in IDH-wildtype gliomas (Figure (Figure2B).2B). It should be noted that mutations of NF1 and PTEN were almost exclusive to IDH-wildtype glioma [2] (Table S1), indicating the functional importance of increased NF1 and PTEN expression in IDH-mutant glioma. Although IDH-mutant gliomas had lower RB1 expression along with higher methylation, the expression of RB pathway genes CDKN2A and CDKN2B was maintained (Figure S1A). Similarly, increased TP53 methylation in the IDH-mutant group did not seem to affect the gene expression, and, furthermore, genes that negatively regulate TP53 (i.e., MDM1, MDM2, and MDM4) were all expressed at significantly lower levels in IDH-mutant glioma than in IDH-wildtype glioma. Even though the roles of RB1 and TP53 seem less clear, the results nevertheless suggest that despite an overall increase in DNA methylation, IDH-mutant gliomas display increased expression of tumor-suppressor genes NF1, NF2, and PTEN.

Figure 1
Increased NF1 and PTEN expression in IDH-mutant glioma correlates with overall survival of glioma patients
Figure 2
Comparative analysis of DNA methylation and gene expression between IDH-wildtype and IDH-mutant gliomas

To assess the importance of these genes in LGG, we asked whether the expression levels correlate with patient survival, and if so, whether such correlation is relevant within the IDH-mutant group. IDH-mutant patients had a median overall survival of 87 months versus 20 months in IDH-wildtype patients (Table S2). By comparison, patients with high expression of NF1 and PTEN showed a similar survival advantage, whereas patients with low expression of these genes were still better off than those of IDH-wildtype (Figure (Figure1B;1B; Table S2). When analyzed within IDH-mutant gliomas, however, NF1 and PTEN expression did not confer a survival advantage (Figure S1B), suggesting that their correlation with improved survival is associated with IDH mutation-mediated increase in gene expression. Although no survival advantage was observed with respect to NF2 expression (data not shown), low expression of RB1, unexpectedly, showed a modest correlation with improved survival (Table S2). Additionally, none of the putative tumor-suppressor genes relevant to LGG, namely, ATRX, CIC, FUBP1, and NOTCH1 [2], showed correlations with overall survival (data not shown). Taken together, these results indicate that the increased NF1 and PTEN expression in IDH-mutant gliomas may be partly responsible for the improved patient survival compared with IDH-wildtype gliomas.

Repressed RTK-PI3K-AKT signaling in IDH-mutant glioma

Because sustaining proliferative signaling is another hallmark of cancer [18], we explored the receptor tyrosine kinase (RTK) signal pathways by analyzing genes known to promote the EGFR, FGFR, and PDGFR signaling commonly seen in malignant gliomas [2, 20]. We observed increased DNA methylation in EGFR, ERBB2, ERBB3, FGFR3, and PDGFRB in IDH-mutant gliomas as the mean methylation of each gene was statistically greater than that of IDH-wildtype (Figure (Figure2A;2A; Figure S2A). Concomitantly, the mean EGFR, ERBB2, FGFR3, and PDGFRB mRNA levels were statistically lower than those of IDH-wildtype gliomas, which also exhibited a marked increase in EGFR and ERBB2 protein abundance (Figure (Figure2C).2C). Consistent with the recent report of more chromosome gains in regions containing RTK pathways in IDH-wildtype gliomas [21], this finding suggests a weakened RTK signaling in IDH-mutant glioma compared with IDH-wildtype glioma, as suggested previously [2].

Despite increased DNA hypermethylation, the mean ERBB3 abundance and phosphorylation were statistically increased in IDH-mutant gliomas (Figure (Figure2C2C and and2E),2E), as shown previously [2]. Furthermore, the mean PDGFRA methylation was unexpectedly decreased in IDH-mutant glioma, but the mRNA levels had a modest, yet statistically insignificant, increase (Figure S2A), similar to that noted previously [22].

With respect to the downstream components of the phosphoinositol 3-kinase (PI3K)–AKT pathway, the mean mRNA levels of PIK3R1 (encoding the PI3K regulatory subunit 1) were much higher than those of IDH-wildtype, whereas the mean mRNA levels of AKT2 were statistically lower, consistent with the opposite directional changes in DNA methylation (Figure (Figure2A;2A; Figure S2B). Despite the increased PIK3CA (encoding the PI3K catalytic subunit α) and AKT3 expression in IDH-mutant glioma, the overall AKT activities were statistically lower than those of IDH-wildtype, as indicated by the mean AKT phosphorylation at Ser-473 and Thr-308 (Figure (Figure2D2D and and2E).2E). Taken together, these results indicate that IDH-mutant gliomas exhibit repressed RTK-PI3K-ATK signaling in addition to increased expression of tumor-suppressor genes.

Increased expression of PIK3R1 and ERBB3 and decreased expression of ERBB2, FGFR3, and AKT2 correlate with improved patient survival

To assess the functional relevance of the repressed RTK-PI3K-AKT signaling in patient survival, we focused on genes differentially expressed as described above. No statistical survival advantage was observed with EGFR nor with PDGFRA transcripts despite the distinct differences at mRNA and protein levels between IDH-wildtype and IDH-mutant gliomas (Figure S3A). Of note, high expression of PDGFRA seemed rather associated, if at all, with improved overall survival. However, low expression of ERBB2 correlated statistically with improved overall and disease-free survival (Figure (Figure3A;3A; Figure S4A). Likewise, low expression of FGFR3 and PDGFRB also correlated with overall but not disease-free survival. Unexpectedly, high ERBB3 abundance and Tyr-1289 phosphorylation correlated with improved overall and disease-free survival (Figure (Figure3B;3B; Figure S3B).

Figure 3
Evaluation of differentially expressed genes in the RTK-PI3K-AKT pathway with respect to glioma patient survival

High expression of PIK3R1 and low expression of AKT2 correlated statistically with improved overall and disease-free survival (Figure (Figure3C;3C; Figure S4B). By contrast, no survival benefits were observed with respect to PIK3CA expression and AKT phosphorylation at Ser-473 and Thr-308 (Figure (Figure3B;3B; Figure S3C). It is noteworthy that the median survival of IDH-wildtype patients remained the worst regardless of the expression levels of RTK-PI3K-AKT genes. Finally, within IDH-mutant gliomas, we found no specific predictive values for any of these genes (Figure S5). Therefore, we conclude that increased ERBB3 abundance and PIK3R1 expression and decreased ERBB2, FGFR3, and AKT2 expression may explain the improved LGG patient survival in IDH-mutant gliomas compared with IDH-wildtype gliomas.

Suppression of ARAF in IDH-mutant glioma correlates with improved survival

Also implicated in LGG pathogenesis is the BRAF-MAP kinase pathway [2], but no significant changes in RAF1, BRAF, and MAPK1 methylation were observed in IDH-mutant glioma in comparison with IDH-wildtype (Figure S6A). Differential expression was found, however, with increased mean RAF1 transcripts and MAPK1 abundance and decreased mean ARAF transcripts in IDH-mutant gliomas; yet only low ARAF expression correlated with improved patient survival (Figure S6B).

IGFBP2 expression inversely correlates with overall survival of IDH-mutant glioma patients

In a continued effort to identify genes that are specifically associated with the survival of IDH-mutant gliomas, we examined IGFBP2, WWTR1, and YAP1, which are associated with glioma progression [2326]. As expected, IDH-mutant gliomas manifested increased methylation concomitant with decreased expression of IGFBP2, YAP1, and WWTR1 in comparison with IDH-wildtype (Figures (Figures2A2A and and4A).4A). The mean mRNA and protein levels of the three genes were statistically much lower in IDH-mutant gliomas. Patients with low expression of IGFBP2 and WWTR but not YAP1 had essentially the same median overall and disease-free survival as IDH-mutant patients (Figure (Figure4B4B and and4C;4C; Tables S2 and S3). Similar survival advantages were seen with analysis of IGFBP2 but not WWTR1 protein abundance (Figure S7A and S7B). It is noteworthy, however, that patients with high IGFBP2 mRNA levels were the only group that showed worse survival than IDH-wildtype patients. Furthermore, within the IDH-mutant group, IGFBP2 was the only gene that was still prognostic at the mRNA levels; the median overall survival of IDH-mutant gliomas with high IGFBP2 expression was ~25% worse than that of IDH-wildtype (Figure (Figure5A5A and and5B;5B; Table S2). Further analysis of IGFBP2 protein abundance confirmed the prognostic value in overall survival within the IDH-mutant group (Figure S7C).

Figure 4
Comparative and survival analyses of glioma progression genes between IDH-wildtype and IDH-mutant gliomas
Figure 5
IGFBP2 expression is a validated prognostic marker within IDH-mutant glioma

To test further the specificity of IGFBP2, we also analyzed MYC, FOXM1, and E2F2, which have been associated with malignant progression of IDH-mutant glioma [3]. Interestingly, IDH-mutant gliomas showed statistically lower FOXM1 but higher MYC mRNA levels, which nevertheless correlated with improved overall survival similar to that of IDH mutations (Figure (Figure2A;2A; Figure S8A), and yet the IDH-wildtype group remained the worst in survival regardless of the expression levels of these genes. Furthermore, within the IDH-mutant group, no survival benefits were observed with statistical significance (Figure S8B).

To validate our finding of IGFBP2 as a specific prognostic marker of IDH-mutant glioma, we used an independent cohort of gliomas ranging from WHO grade II to grade IV with IDH status determined by immunohistochemistry against IDH1R132H [27]. Again, not only were IGFBP2 mRNA levels statistically much lower in the IDH1R132H-positive than in the IDH1R132H-negative group, but low IGFBP2 mRNA levels also correlated with improved survival whereas high IGFBP2 mRNA levels exhibited worse survival than the IDH1R132H-negative group (Figure (Figure5C5C and and5D).5D). Importantly, the survival benefit of low IGFBP2 expression held true within the IDH1R132H-positive group (Figure (Figure5E).5E). Taken together, these results indicate that IDH-mutant glioma patients generally manifest low IGFBP2 expression, which is associated with improved survival independent of IDH status, whereas high IGFBP2 expression results in worse survival than in the IDH-wildtype group.

Across all histological types IGFBP2 expression inversely correlates with overall survival

LGGs have been classified histologically into oligodendroglioma, oligoastrocytoma, and astrocytoma subtypes [28], with oligodendroglioma demonstrating the best overall survival and astrocytoma having the worst [2]. To test the possibility that the above-described gene expression is associated with specific LGG histological types and/or grades, we evaluated the expression pattern of tumor-suppressor genes (NF1, PIK3R1, and PTEN), oncogenes (AKT2, ARAF, ERBB2, FGFR3, and PDGFRB), and progression genes (E2F2, FOXM1, IGFBP2, MYC, and WWTR1) that had been shown above to be correlative with patient survival. By using a heat map that corresponded to the histological types and IDH status (Figure (Figure6A),6A), we observed that irrespective of the histological types, 1) the IDH-mutant group expressed higher levels of tumor-suppressor genes than the IDH-wildtype group albeit with various degrees of statistical significance (Figure S9); 2) by contrast, the latter expressed higher levels of oncogenes, apparently with gene-specific variations among different histological types; and 3) the expression of progression genes was generally higher in the IDH-wildtype group, with the exception of MYC, where high expression correlated with improved survival in IDH-mutant patients (Figure S8). Altogether, these findings provide molecular explanations for the survival advantage of IDH-mutant patients in all histological types.

Figure 6
IGFBP2 is prognostic across all histological types of glioma

Furthermore, differential IGFBP2 expression between IDH-wildtype and IDH-mutant gliomas was seen with strong statistical significance across all three histological types, and increased IGFBP2 expression correlated with higher grade (Figure (Figure6B6B and and6C),6C), a finding that supports the general role of IGFBP2 in glioma progression. Consistently, differential IGFBP2 expression in each histological type was prognostic with strong statistical significance; decreased expression extended median survival by 24% in astrocytoma and 106% in oligoastrocytoma, whereas increased expression shortened median survival by >70% for astrocytoma and oligoastrocytoma and >80% for oligodendroglioma (Figure (Figure6D).6D). The latest classification of IDH-mutant glioma groups into two subtypes: one with 1p/19q co-deletion also harboring mutations in CIC, FUBP1, and NOTCH1, and the other without 1p/19q co-deletion but with extreme high frequencies of mutations in TP53 and ATRX [2,29]. When IDH-mutant gliomas were stratified accordingly with 1p/19q co-deletion, we found no statistical differences in IGFBP2 expression between the two subtypes even though sharp differences remained between IDH-wildtype group and each subtype (Figure (Figure6E).6E). Furthermore, no statistical differences in IGFBP2 expression were observed between subtypes with and without CIC, FUBP1, and/or NOTCH1 mutations, nor between those with and without TP53 and/or ATRX mutations, which further disputes the possibility of IGFBP2 association with specific glioma subtypes. Taken together, these results indicate that although IGFBP2 expression is low in IDH-mutant gliomas of all three histological types, the increased expression is a powerful, negative prognostic marker that predicts a median overall survival worse than that of IDH-wildtype group.

DISCUSSION

The suggestion that IGFBP2 is involved in glioma progression dates back to 1999 when IGFBP2 overexpression was discovered in glioblastoma [30]. Further studies demonstrated the important role of IGFBP2 in glioma progression [23, 24] and indicated that IGFBP2 expression was a poor prognostic marker in a mixed cohort ranging from grade II glioma to glioblastoma [31] as well as in glioblastoma alone [32]. Interestingly, our analysis of 135 cases of TCGA glioblastoma data also showed a 2-month statistical advantage in the median overall survival with low IGFBP2 mRNA levels (data not shown). However, this advantage reduced by half and lost statistical significance when the only 7 IDH-mutant cases were subtracted from the analysis, further supporting an intimate relationship between IDH mutations and IGFBP2 in glioblastoma prognosis.

To our knowledge, IGFBP2 is the first gene that serves as a prognostic marker in IDH-mutant glioma patients; in particular, increased IGFBP2 expression in IDH-mutant glioma patients is prognostic of poor outcomes worse than those of IDH-wildtype glioma patients. Although many genes identified in this study may serve as prognostic markers of lower-grade gliomas, only IGFBP2 remains prognostic within IDH-mutant glioma patients. Low IGFBP2 expression seems attributable to global DNA hypermethylation [33], but the regulatory mechanism seems complex with the possible involvement of copy number (data not shown); thus, further investigations are warranted to understand how IGFBP2 becomes deregulated to drive glioma progression.

The key hallmarks of cancer involve sustaining proliferative, replicative, and angiogenic signaling while evading tumor-suppressive and cell-death signaling [18]. We demonstrate here the strong statistical correlations of IDH-mutant gliomas with increased expression of mostly intact tumor-suppressor genes (NF1, PIK3R1, and PTEN), and decreased expression of oncogenic genes (ERBB2, FGFR3, PDGFRB, AKT2, and ARAF) and progression genes (IGFBP2, FOXM1, and WWTR1). Furthermore, these changes in gene expression are associated statistically with improved patient survival. Additionally, we observed decreased replicative potential and reduced apoptotic and hypoxic/angiogenic signaling (data not shown) in IDH-mutant gliomas. Therefore, these results not only provide molecular explanation to the mystery of prolonged survival of IDH-mutant glioma patients but also prompt us to revisit the functional role of IDH mutations in glioma.

Acquisition of somatic mutations in the genome is an evolutionary process of cancer [34]. Whereas mutations that are causally implicated in oncogenesis are defined as driver mutations, the rest are grouped as passenger mutations for the lack of clear contributions to cancer development. Although IDH mutations are believed to drive glioma initiation by producing (R)2-HG, which results in DNA hypermethylation and metabolic reprograming [3537], expression of mutant Idh1 in the mouse brain has yet to induce glioma formation despite robust production of (R)2-HG [11, 38]. To the contrary, our results show strong correlation between increased DNA methylation and suppression of oncogenic signaling, suggesting that the effects of DNA hypermethylation in IDH-mutant glioma are unexpectedly anti-oncogenic, which is consistent with a newly identified subtype of IDH-mutant glioma featuring DNA demethylation associated with poor survival [39]. Our hypothesis is also supported by the strong correlations between IDH mutations and increased tumor-suppressor gene expression that correlates with patient survival. Likewise, targets of epigenetic silencing identified in IDH1-mutant gliomas, including glycolytic genes [40], retinol binding protein 1 [41], and microRNA miR148a [42], were all associated with reduced cell proliferation and improved survival. Furthermore, introduction of mutant IDH1 in glioblastoma cells and transformed astrocytes inhibited cell proliferation, reduced tumor growth, and improved mouse survival [4345] (Tiburcio et al., manuscript in preparation). Thus, taken together, the evidence indicates that IDH mutations in glioma may represent a different category, i.e., beneficial mutations, resulting from an aberrant cellular response that counteracts glioma progression, even though the mechanism by which transformed cells acquire this type of mutations necessitates further investigation.

A counterargument to the theory of IDH mutation-associated survival advantage is that IDH-wildtype glioma is simply a different type of disease with more genetic abnormalities and therefore more malignant traits than IDH-mutant glioma. Given the fact that IDH mutations are early events (“founder mutation”) in LGG development [46], it stands to reason that the early gain of IDH mutations prevents glioma from acquiring additional malignant traits seen in IDH-wildtype glioma, whereas functional loss of IDH mutations, such as the loss of DNA hypermethylation in the newly identified IDH-mutant glioma subtype [39], results in unfavorable clinical outcome. In fact, the propensity of IDH-mutant tumor cells to lose the mutant allele and (R)2-HG production has been observed in culture [40] even though whether this occurs in vivo merits further investigation.

This study has also prompted us to reassess therapeutic targets of malignant glioma. Although pharmacological targeting of IDH mutations induces tumor cell differentiation by reducing (R)2-HG production [15,16], the inhibitory effect on glioma growth remains less clear [16,17]. Additionally, a recent study suggests that IDH1-mutant inhibitors may alter oxidative stress responses in glioma patients and therefore diminish the therapeutic efficacy of irradiation [47]. Our results not only suggest that targeting of mutant IDH in glioma patients could be counterproductive but also raise concerns about the selection of molecular targets for malignant glioma treatment. For instance, owing to the lack of correlation between overall survival and expression of EGFR or PDGFRA, targeting the corresponding pathway may not be effective, even in tumors overexpressing EGFR or PDGFRA. By contrast, targeting ERBB3 signaling could be counterproductive given the statistical association of ERBB3 abundance and activity with improved survival. In sum, deep understanding of the signaling pathway critical for patient survival may serve as the foundation for improving therapeutic efficacy of malignant glioma.

MATERIALS AND METHODS

Glioma data sets

Genomic data and patient survival data from the Brain Lower Grade Glioma (TCGA, Provisional) data set with 286 cases of sequenced tumors were acquired from cBioPortal (http://www.cbioportal.org/) [48,49]. They include mutation data from whole-exome sequencing, mRNA expression z-scores (RNA Seq V2 RSEM), which compared with the gene distribution of each gene in tumors that are diploid for this gene, DNA methylation (HM450) beta-values, and protein expression and phosphorylation results by reverse-phase protein array. The genomic data along with glioma grades and histologic types were matched with patient survival data according to the case identification numbers. There were 53 cases of IDH-wildtype gliomas and 233 cases with either IDH1 or IDH2 mutations, which were grouped into a single group, IDH-mutant. The GSE16011 cohort of gliomas ranging from WHO grade II to grade IV was used for validation [27], wherein the IDH status was determined by immunohistochemistry against IDH1R132H.

Comparative analysis and generation of heat maps

We used the GraphPad Prism software to perform column analyses where unpaired t-tests were used to compare the differences between IDH-wildtype gliomas and IDH-mutant gliomas with respect to methylation, mRNA, protein, and phosphorylation of individual genes. The results were presented in box-and-whisker plots. Two-tailed p values were used for statistical significance (*, p<0.05; **, p<0.01; ***, p<0.001; and ****, p<0.0001). Heat maps of DNA methylation and gene expression were generated using the mean of the genomic data and plotted accordingly.

Survival study

Log-rank (Mantel-Cox) tests were performed to compare gene expression z-scores below zero and above zero in relation to overall survival or disease-free survival data. Median overall survival and disease-free survival are listed in Table S2 and Table S3 along with log-rank hazard ratio and 95% confidence interval of ratio. Kaplan-Meier survival curves were plotted with censored points marked. For overall survival analysis, there were 53 IDH-wildtype cases and 229 IDH-mutant cases available. For disease-free survival analysis, 42 IDH-wildtype and 218 IDH-mutant cases were available.

SUPPLEMENTARY MATERIAL

Acknowledgments

The authors wish to thank Kristin Kraus for editorial assistance.

Abbreviations

IDH
isocitrate dehydrogenase
IDHm
IDH-mutant
IGFBP2
insulin-like growth factor binding protein 2
LGG
lower-grade glioma
PI3K
phosphoinositol 3-kinase
RTK
receptor tyrosine kinase
TCGA
The Cancer Genome Atlas
WHO
World Health Organization.

Footnotes

CONFLICTS OF INTEREST

The authors declare no conflicts of interest.

FUNDING

This work was supported in part by the Huntsman Cancer Foundation and the National Cancer Institute of the National Institutes of Health under Award Number P30CA042014.

Contributed by

Authors' contributions

LEH conceived and designed the study and wrote the manuscript. LEH and AL acquired and analyzed the data. AL, HC, RLJ, DWF, and WTC reviewed the manuscript.

Editorial note

This paper has been accepted based in part on peer-review conducted by another journal and the authors' response and revisions as well as expedited peer-review in Oncotarget.

REFERENCES

1. Ostrom QT, Gittleman H, Liao P, Rouse C, Chen Y, Dowling J, Wolinsky Y, Kruchko C, Barnholtz-Sloan J. CBTRUS statistical report: primary brain and central nervous system tumors diagnosed in the United States in 2007-2011. Neuro-oncology. 2014;16(Suppl 4):iv1–63. doi: 10.1093/neuonc/nou223. [PMC free article] [PubMed] [Cross Ref]
2. The Cancer Genome Atlas Research Network Comprehensive, integrative genomic analysis of diffuse lower-grade gliomas. N Engl J Med. 2015;372:2481–98. doi: 10.1056/NEJMoa1402121. [PMC free article] [PubMed] [Cross Ref]
3. Bai H, Harmancı AS, Erson-Omay EZ, Li J, Coşkun S, Simon M, Krischek B, Özduman K, Omay SB, Sorensen EA, Turcan S, Bakırcığlu M, Carrión-Grant G, et al. Integrated genomic characterization of IDH1-mutant glioma malignant progression. Nat Genet. 2016;48:59–66. doi: 10.1038/ng.3457. [PMC free article] [PubMed] [Cross Ref]
4. Parsons DW, Jones S, Zhang X, Lin JC-H, Leary RJ, Angenendt P, Mankoo P, Carter H, Siu I-M, Gallia GL, Olivi A, McLendon R, Rasheed BA, et al. An integrated genomic analysis of human glioblastoma multiforme. Science. 2008;321:1807–12. doi: 10.1126/science.1164382. [PMC free article] [PubMed] [Cross Ref]
5. Yan H, Parsons DW, Jin G, McLendon R, Rasheed BA, Yuan W, Kos I, Batinic-Haberle I, Jones S, Riggins GJ, Friedman H, Friedman A, Reardon D, et al. IDH1 and IDH2 mutations in gliomas. N Engl J Med. 2009;360:765–73. doi: 10.1056/NEJMoa0808710. [PMC free article] [PubMed] [Cross Ref]
6. Dang L, White DW, Gross S, Bennett BD, Bittinger MA, Driggers EM, Fantin VR, Jang HG, Jin S, Keenan MC, Marks KM, Prins RM, Ward PS, et al. Cancer-associated IDH1 mutations produce 2-hydroxyglutarate. Nature. 2009;462:739–44. doi: 10.1038/nature08617. [PMC free article] [PubMed] [Cross Ref]
7. Ward PS, Patel J, Wise DR, Abdel-Wahab O, Bennett BD, Coller HA, Cross JR, Fantin VR, Hedvat CV, Perl AE, Rabinowitz JD, Carroll M, Su SM, et al. The common feature of leukemia-associated IDH1 and IDH2 mutations is a neomorphic enzyme activity converting alpha-ketoglutarate to 2-hydroxyglutarate. Cancer Cell. 2010;17:225–34. doi: 10.1016/j.ccr.2010.01.020. [PMC free article] [PubMed] [Cross Ref]
8. Xu W, Yang H, Liu Y, Yang Y, Wang P, Kim S-H, Ito S, Yang C, Wang P, Xiao M-T, Liu L-X, Jiang W-Q, Liu J, et al. Oncometabolite 2-hydroxyglutarate is a competitive inhibitor of α-ketoglutarate-dependent dioxygenases. Cancer Cell. 2011;19:17–30. doi: 10.1016/j.ccr.2010.12.014. [PMC free article] [PubMed] [Cross Ref]
9. Loenarz C, Schofield CJ. Physiological and biochemical aspects of hydroxylations and demethylations catalyzed by human 2-oxoglutarate oxygenases. Trends Biochem Sci. 2011;36:7–18. doi: 10.1016/j.tibs.2010.07.002. [PubMed] [Cross Ref]
10. Lu C, Ward PS, Kapoor GS, Rohle D, Turcan S, Abdel-Wahab O, Edwards CR, Khanin R, Figueroa ME, Melnick A, Wellen KE, O'Rourke DM, Berger SL, et al. IDH mutation impairs histone demethylation and results in a block to cell differentiation. Nature. 2012;483:474–8. doi: 10.1038/nature10860. [PMC free article] [PubMed] [Cross Ref]
11. Sasaki M, Knobbe CB, Munger JC, Lind EF, Brenner D, Brüstle A, Harris IS, Holmes R, Wakeham A, Haight J, You-Ten A, Li WY, Schalm S, et al. IDH1(R132H) mutation increases murine haematopoietic progenitors and alters epigenetics. Nature. 2012;488:656–9. doi: 10.1038/nature11323. [PMC free article] [PubMed] [Cross Ref]
12. Turcan S, Rohle D, Goenka A, Walsh LA, Fang F, Yilmaz E, Campos C, Fabius AWM, Lu C, Ward PS, Thompson CB, Kaufman A, Guryanova O, et al. IDH1 mutation is sufficient to establish the glioma hypermethylator phenotype. Nature. 2012;483:479–83. doi: 10.1038/nature10866. [PMC free article] [PubMed] [Cross Ref]
13. Verhaak RGW, Hoadley KA, Purdom E, Wang V, Qi Y, Wilkerson MD, Miller CR, Ding L, Golub T, Mesirov JP, Alexe G, Lawrence M, O'Kelly M, et al. Integrated genomic analysis identifies clinically relevant subtypes of glioblastoma characterized by abnormalities in PDGFRA, IDH1, EGFR, and NF1. Cancer Cell. 2010;17:98–110. doi: 10.1016/j.ccr.2009.12.020. [PMC free article] [PubMed] [Cross Ref]
14. Noushmehr H, Weisenberger DJ, Diefes K, Phillips HS, Pujara K, Berman BP, Pan F, Pelloski CE, Sulman EP, Bhat KP, Verhaak RGW, Hoadley KA, Hayes DN, et al. Identification of a CpG island methylator phenotype that defines a distinct subgroup of glioma. Cancer Cell. 2010;17:510–22. doi: 10.1016/j.ccr.2010.03.017. [PMC free article] [PubMed] [Cross Ref]
15. Wang F, Travins J, Delabarre B, Penard-Lacronique V, Schalm S, Hansen E, Straley K, Kernytsky A, Liu W, Gliser C, Yang H, Gross S, Artin E, et al. Targeted inhibition of mutant IDH2 in leukemia cells induces cellular differentiation. Science. 2013;340:622–6. doi: 10.1126/science.1234769. [PubMed] [Cross Ref]
16. Rohle D, Popovici-Muller J, Palaskas N, Turcan S, Grommes C, Campos C, Tsoi J, Clark O, Oldrini B, Komisopoulou E, Kunii K, Pedraza A, Schalm S, et al. An inhibitor of mutant IDH1 delays growth and promotes differentiation of glioma cells. Science. 2013;340:626–30. doi: 10.1126/science.1236062. [PMC free article] [PubMed] [Cross Ref]
17. Turcan S, Fabius AWM, Borodovsky A, Pedraza A, Brennan C, Huse J, Viale A, Riggins GJ, Chan TA. Efficient induction of differentiation and growth inhibition in IDH1 mutant glioma cells by the DNMT inhibitor decitabine. Oncotarget. 2013;4:1729–36. doi: 10.18632/oncotarget.1412. [PMC free article] [PubMed] [Cross Ref]
18. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell. 2000;100:57–70. doi: 10.1016/S0092-8674(00)81683-9. [PubMed] [Cross Ref]
19. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell. 2011;144:646–74. doi: 10.1016/j.cell.2011.02.013. [PubMed] [Cross Ref]
20. Cancer Genome Atlas Research Network Comprehensive genomic characterization defines human glioblastoma genes and core pathways. Nature. 2008;455:1061–8. doi: 10.1038/nature07385. [PMC free article] [PubMed] [Cross Ref]
21. Cohen A, Sato M, Aldape K, Mason CC, Alfaro-Munoz K, Heathcock L, South ST, Abegglen LM, Schiffman JD, Colman H. DNA copy number analysis of Grade II-III and Grade IV gliomas reveals differences in molecular ontogeny including chromothripsis associated with IDH mutation status. Acta Neuropathol Commun. 2015;3:34. doi: 10.1186/s40478-015-0213-3. [PMC free article] [PubMed] [Cross Ref]
22. Flavahan WA, Drier Y, Liau BB, Gillespie SM, Venteicher AS, Stemmer-Rachamimov AO, Suvà ML, Bernstein BE. Insulator dysfunction and oncogene activation in IDH mutant gliomas. Nature. 2015;529:110–4. doi: 10.1038/nature16490. [PMC free article] [PubMed] [Cross Ref]
23. Dunlap SM, Celestino J, Wang H, Jiang R, Holland EC, Fuller GN, Zhang W. Insulin-like growth factor binding protein 2 promotes glioma development and progression. Proc Natl Acad Sci USA. 2007;104:11736–41. doi: 10.1073/pnas.0703145104. [PubMed] [Cross Ref]
24. Moore LM, Holmes KM, Smith SM, Wu Y, Tchougounova E, Uhrbom L, Sawaya R, Bruner JM, Fuller GN, Zhang W. IGFBP2 is a candidate biomarker for Ink4a-Arf status and a therapeutic target for high-grade gliomas. Proc Natl Acad Sci USA. National Acad Sciences. 2009;106:16675–9. doi: 10.1073/pnas.0900807106. [PubMed] [Cross Ref]
25. Orr BA, Bai H, Odia Y, Jain D, Anders RA, Eberhart CG. Yes-associated protein 1 is widely expressed in human brain tumors and promotes glioblastoma growth. J Neuropathol Exp Neurol. 2011;70:568–77. doi: 10.1097/NEN.0b013e31821ff8d8. [PMC free article] [PubMed] [Cross Ref]
26. Bhat KPL, Salazar KL, Balasubramaniyan V, Wani K, Heathcock L, Hollingsworth F, James JD, Gumin J, Diefes KL, Kim SH, Turski A, Azodi Y, Yang Y, et al. The transcriptional coactivator TAZ regulates mesenchymal differentiation in malignant glioma. Genes Dev. 2011;25:2594–609. doi: 10.1101/gad.176800.111. [PubMed] [Cross Ref]
27. Gravendeel LAM, Kouwenhoven MCM, Gevaert O, de Rooi JJ, Stubbs AP, Duijm JE, Daemen A, Bleeker FE, Bralten LBC, Kloosterhof NK, De Moor B, Eilers PHC, van der Spek PJ, et al. Intrinsic gene expression profiles of gliomas are a better predictor of survival than histology. American Association for Cancer Res. 2009;69:9065–72. doi: 10.1158/0008-5472.CAN-09-2307. [PubMed] [Cross Ref]
28. Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, Burger PC, Jouvet A, Scheithauer BW, Kleihues P. The 2007 WHO classification of tumours of the central nervous system. Acta Neuropathol. 2007;114:97–109. doi: 10.1007/s00401-007-0243-4. [PMC free article] [PubMed] [Cross Ref]
29. Louis DN, Perry A, Reifenberger G, Deimling von A, Figarella-Branger D, Cavenee WK, Ohgaki H, Wiestler OD, Kleihues P, Ellison DW. The 2016 World Health Organization Classification of Tumors of the Central Nervous System: a summary. Acta Neuropathol. 2016;131:803–20. doi: 10.1007/s00401-016-1545-1. [PubMed] [Cross Ref]
30. Fuller GN, Rhee CH, Hess KR, Caskey LS, Wang R, Bruner JM, Yung WK, Zhang W. Reactivation of insulin-like growth factor binding protein 2 expression in glioblastoma multiforme: a revelation by parallel gene expression profiling. 1999;59:4228–32. [PubMed]
31. Holmes KM, Annala M, Chua CYX, Dunlap SM, Liu Y, Hugen N, Moore LM, Cogdell D, Hu L, Nykter M, Hess K, Fuller GN, Zhang W. Insulin-like growth factor-binding protein 2-driven glioma progression is prevented by blocking a clinically significant integrin, integrin-linked kinase, and NF-κB network. Proc Natl Acad Sci USA. 2012;109:3475–80. doi: 10.1073/pnas.1120375109. [PubMed] [Cross Ref]
32. Brennan CW, Verhaak RGW, McKenna A, Campos B, Noushmehr H, Salama SR, Zheng S, Chakravarty D, Sanborn JZ, Berman SH, Beroukhim R, Bernard B, Wu C-J, et al. The somatic genomic landscape of glioblastoma. Cell. 2013;155:462–77. doi: 10.1016/j.cell.2013.09.034. [PMC free article] [PubMed] [Cross Ref]
33. Zheng S, Houseman EA, Morrison Z, Wrensch MR, Patoka JS, Ramos C, Haas-Kogan DA, McBride S, Marsit CJ, Christensen BC, Nelson HH, Stokoe D, Wiemels JL, et al. DNA hypermethylation profiles associated with glioma subtypes and EZH2 and IGFBP2 mRNA expression. Neuro-oncology. 2011;13:280–9. doi: 10.1093/neuonc/noq190. [PMC free article] [PubMed] [Cross Ref]
34. Stratton MR, Campbell PJ, Futreal PA. The cancer genome. Nature. 2009;458:719–24. doi: 10.1038/nature07943. [PMC free article] [PubMed] [Cross Ref]
35. Venneti S, Garimella MT, Sullivan LM, Martinez D, Huse JT, Heguy A, Santi M, Thompson CB, Judkins AR. Evaluation of histone 3 lysine 27 trimethylation (H3K27me3) and enhancer of Zest 2 (EZH2) in pediatric glial and glioneuronal tumors shows decreased H3K27me3 in H3F3A K27M mutant glioblastomas. Brain Pathol. 2013;23:558–64. doi: 10.1111/bpa.12042. [PMC free article] [PubMed] [Cross Ref]
36. Cairns RA, Mak TW. Oncogenic isocitrate dehydrogenase mutations: mechanisms, models, and clinical opportunities. Cancer Discov. 2013;3:730–41. doi: 10.1158/2159-8290.CD-13-0083. [PubMed] [Cross Ref]
37. Zhang C, Moore LM, Li X, Yung WKA, Zhang W. IDH1/2 mutations target a key hallmark of cancer by deregulating cellular metabolism in glioma. Neuro-oncology. 2013;15:1114–26. doi: 10.1093/neuonc/not087. [PMC free article] [PubMed] [Cross Ref]
38. Sasaki M, Knobbe CB, Itsumi M, Elia AJ, Harris IS, Chio IIC, Cairns RA, McCracken S, Wakeham A, Haight J, Ten AY, Snow B, Ueda T, et al. D-2-hydroxyglutarate produced by mutant IDH1 perturbs collagen maturation and basement membrane function. Genes Dev. 2012;26:2038–49. doi: 10.1101/gad.198200.112. [PubMed] [Cross Ref]
39. Ceccarelli M, Barthel FP, Malta TM, Sabedot TS, Salama SR, Murray BA, Morozova O, Newton Y, Radenbaugh A, Pagnotta SM, Anjum S, Wang J, Manyam G, et al. Molecular profiling reveals biologically discrete subsets and pathways of progression in diffuse glioma. Cell. 2016;164:550–63. doi: 10.1016/j.cell.2015.12.028. [PMC free article] [PubMed] [Cross Ref]
40. Chesnelong C, Chaumeil MM, Blough MD, Al-Najjar M, Stechishin OD, Chan JA, Pieper RO, Ronen SM, Weiss S, Luchman HA, Cairncross JG. Lactate dehydrogenase A silencing in IDH mutant gliomas. Neuro-oncology. 2014;16:686–95. doi: 10.1093/neuonc/not243. [PMC free article] [PubMed] [Cross Ref]
41. Chou AP, Chowdhury R, Li S, Chen W, Kim AJ, Piccioni DE, Selfridge JM, Mody RR, Chang S, Lalezari S, Lin J, Sanchez DE, Wilson RW, et al. Identification of retinol binding protein 1 promoter hypermethylation in isocitrate dehydrogenase 1 and 2 mutant gliomas. J Natl Cancer Inst. 2012;104:1458–69. doi: 10.1093/jnci/djs357. [PMC free article] [PubMed] [Cross Ref]
42. Li S, Chowdhury R, Liu F, Chou AP, Li T, Mody RR, Lou JJ, Chen W, Reiss J, Soto H, Prins R, Liau LM, Mischel PS, et al. Tumor-suppressive miR148a is silenced by CpG island hypermethylation in IDH1-mutant gliomas. Clin Cancer Res. 2014;20:5808–22. doi: 10.1158/1078-0432.CCR-14-0234. [PMC free article] [PubMed] [Cross Ref]
43. Bralten LBC, Kloosterhof NK, Balvers R, Sacchetti A, Lapre L, Lamfers M, Leenstra S, de Jonge H, Kros JM, Jansen EEW, Struys EA, Jakobs C, Salomons GS, et al. IDH1 R132H decreases proliferation of glioma cell lines in vitro and in vivo. Ann Neurol. 2011;69:455–63. doi: 10.1002/ana.22390. [PubMed] [Cross Ref]
44. Pellegatta S, Valletta L, Corbetta C, Patanè M, Zucca I, Riccardi Sirtori F, Bruzzone MG, Fogliatto G, Isacchi A, Pollo B, Finocchiaro G. Effective immuno-targeting of the IDH1 mutation R132H in a murine model of intracranial glioma. Acta Neuropathol Commun. 2015;3:4. doi: 10.1186/s40478-014-0180-0. [PMC free article] [PubMed] [Cross Ref]
45. Chen R, Nishimura MC, Kharbanda S, Peale F, Deng Y, Daemen A, Forrest WF, Kwong M, Hedehus M, Hatzivassiliou G, Friedman LS, Phillips HS. Hominoid-specific enzyme GLUD2 promotes growth of IDH1R132H glioma. Proc Natl Acad Sci USA. 2014;111:14217–22. doi: 10.1073/pnas.1409653111. [PubMed] [Cross Ref]
46. Watanabe T, Nobusawa S, Kleihues P, Ohgaki H. IDH1 mutations are early events in the development of astrocytomas and oligodendrogliomas. Am J Pathol. 2009;174:1149–53. doi: 10.2353/ajpath.2009.080958. [PubMed] [Cross Ref]
47. Molenaar RJ, Botman D, Smits MA, Hira VV, van Lith SA, Stap J, Henneman P, Khurshed M, Lenting K, Mul AN, Dimitrakopoulou D, van Drunen CM, Hoebe RA, et al. Radioprotection of IDH1-mutated cancer cells by the IDH1-mutant inhibitor. AGI-5198. 2015;75:4790–802. doi: 10.1158/0008-5472.CAN-14-3603. [PubMed] [Cross Ref]
48. Gao J, Aksoy BA, Dogrusoz U, Dresdner G, Gross B, Sumer SO, Sun Y, Jacobsen A, Sinha R, Larsson E, Cerami E, Sander C, Schultz N. Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci Signal. 2013;6:pl1. doi: 10.1126/scisignal.2004088. [PMC free article] [PubMed] [Cross Ref]
49. Cerami E, Gao J, Dogrusoz U, Gross BE, Sumer SO, Aksoy BA, Jacobsen A, Byrne CJ, Heuer ML, Larsson E, Antipin Y, Reva B, Goldberg AP, et al. The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. Cancer Discov. 2012;2:401–4. doi: 10.1158/2159-8290.CD-12-0095. [PMC free article] [PubMed] [Cross Ref]

Articles from Oncotarget are provided here courtesy of Impact Journals, LLC