Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2016 Sep 15.
Published in final edited form as: Cell Rep. 2016 Sep 13;16(11):2838–2845. doi: 10.1016/j.celrep.2016.08.040

Lineage-restricted OLIG2-RTK signaling governs the molecular subtype of glioma stem-like cells

Robert Kupp 1, Lior Shtayer 1,4, An-Chi Tien 2, Emily Szeto 1, Nader Sanai 1, David H Rowitch 2,3, Shwetal Mehta 1,*
PMCID: PMC5024710  NIHMSID: NIHMS811297  PMID: 27626655

SUMMARY

The bHLH transcription factor OLIG2 is a master regulator of oligodendroglial fate decisions and tumorigenic competence of glioma stem-like cells (GSCs). However, the molecular mechanisms underlying dysregulation of OLIG2 function during gliomagenesis remains poorly understood. Here, we show that OLIG2 modulates growth factor signaling in two distinct populations of GSCs, characterized by expression of either the EGFR or PDGFRα. Biochemical analyses of OLIG2 function in normal and malignant neural progenitors reveal a positive feedforward loop between OLIG2 and EGFR to sustain co-expression. Furthermore, loss of OLIG2 function results in mesenchymal transformation in PDGFRαHIGH GSCs, a phenomenon that appears to be circumscribed in EGFRHIGH GSCs. Exploitation of OLIG2’s dual and antithetical, pro-mitotic (EGFR-driven) and lineage-specifying (PDGFRα-driven) functions by glioma cells, appears to be critical for sustaining growth factor signaling and GSC molecular subtype.

Keywords: glioblastoma, cancer stem cell, growth factor, receptor tyrosine kinase, Olig2

INTRODUCTION

Glioblastoma multiforme (GBM) is the most common primary malignant brain tumor in adults, and despite profound molecular and genetic analyses, patient outcome following standard-of-care therapies remains dismal, with an average life expectancy of 12-15 months (Cloughesy et al., 2014). Intratumoral heterogeneity, diffuse infiltration throughout the brain parenchyma, and resistance to traditional therapies pose a major clinical challenge, which inevitably leads to tumor recurrence and the demise of the patient (Cloughesy et al., 2014). Recent work has highlighted the epigenetic and transcriptional networks of GSCs in human GBM (Suva et al., 2014), and their resemblance to niche-restricted progenitors of the adult mammalian brain (Alcantara Llaguno et al., 2015). Due to their unique ability to self-renew and differentiate into the neuro-glial lineages of the central nervous system (CNS), GSCs present a therapeutic burden by sustaining intratumoral heterogeneity and resistance to traditional therapies (Lathia et al., 2015).

In the developing CNS, the balance between progenitor proliferation and subtype specification is regulated by a diverse set of proteins and extracellular cues. At the apex of this process are transcription factors (TFs), such as the basic-Helix-Loop-Helix (bHLH) proteins, which harbor contradicting pro-neurogenic and anti-neurogenic functions (Imayoshi and Kageyama, 2014a). The dynamic expression of these proteins at various developmental stages reflects their bi-functional nature – where constitutive expression drives cell cycle exit and differentiation, and oscillatory expression sustains multipotency (Imayoshi and Kageyama, 2014b). During embryogenesis, the bHLH TF Olig2 is essential for expansion of the progenitor pools and for specification of the oligodendroglial lineage (Meijer et al., 2012). Universal expression of OLIG2 in almost all cases of diffuse pediatric and adult glioma suggests overlapping mechanisms between CNS development and gliomagenesis (Ligon et al., 2004; Otero et al., 2011). In this study, we set out to address the molecular mechanisms which sustain OLIG2 expression and its pro-mitotic functions in cycling neural progenitors. Cross-species investigative analysis suggests that OLIG2 forms a positive feedforward loop with the receptor tyrosine kinases (RTKs) EGFR and the lineage-restricted PDGFRα. We describe an OLIG2-dependent cell-autonomous nuclear mechanism for regulation of growth factor signaling in normal and malignant neural progenitors.

RESULTS

EGFR signaling sustains Olig2 protein in cycling neural progenitors

To address the mechanisms that sustain Olig2 expression in neural progenitors, we utilized the protein biosynthesis inhibitor cycloheximide (CHX) to determine the half-life of Olig2 protein. We observed a half-life between 4-8 hours for Olig2 protein in normal (wild-type; p53−/−; Cdkn2a−/−) and malignant (Cdkn2a−/−;EGFRvIII & p53−/−;EGFRvIII) neural progenitors, irrespective of the genetic background (Figure 1A). Importantly, Olig2 levels (mRNA and protein) were similar, with minimal deviation, across selected genotypes (Figure S1A). We observe no significant difference in the turnover time of either TPN (triple-phospho-null; Ser to Ala) or TPM (triple-phospho-mimetic; Ser to Glu/Asp) phospho-mutant of Olig2 (Figure 1B). When neural progenitors were plated in factor-free media (i.e., depleted of exogenous EGF/FGF2), phosphorylation of Olig2 at its amino-terminus (Ser10, 13, 14) rapidly declined (Figure S1B) (Sun et al., 2011) genetic background (Figure 1A).

Figure 1.

Figure 1

Open in a new tab

EGFR signaling regulates Olig2 protein turnover in cycling neural progenitors. A. Murine neural progenitors were challenged with Cycloheximide (25 μg/mL CHX). Lysates were collected at indicated time points and immunoblotted with antibodies directed against Olig2 and vinculin. B. Olig2-null cells expressing either phospho-mutants (TPN or TPM) were treated with CHX and lysates were collected at indicated time points. Immunoblotting was performed with antibodies directed against Olig2 and vinculin. C. Neural progenitors were treated with CHX alone or in combination with EGFRi (10 μM Erlotinib). Immunoblotting was performed similarly to above. Quantifications shown on right. Error bar indicates mean ± SEM (n ≥ 3). ****p<0.0001. See also Figure S1.

Previous reports have implicated mitogenic signaling as important regulator of Olig2+ cells during CNS development and gliomagenesis (Gonzalez-Perez et al., 2009; Persson et al., 2010). To address the direct role of active EGF signaling in regulating Olig2 protein expression, we utilized a known EGFR inhibitor, Erlotinib (henceforth referred to as EGFRi). Treatment of neural progenitors expressing Olig2 (Olig2−/−;mOlig2) with EGFRi in the presence of CHX over a 24 hour time course resulted in rapid depletion of the Olig2 protein (T1/2 = 1 hr) (Figure 1C). Consistent with this observation, treatment of murine neural progenitors with EGFRi resulted in depletion of the Olig2 protein, irrespective of the genetic background and phosphorylation status (Figure S1C and S1D). These data suggests that EGFR signaling mediates the stability of Olig2 in neural progenitors and is independent of amino-terminal phosphorylation and genetic background.

Phosphorylation of Olig2 alters expression of Growth Factor Receptor Tyrosine Kinases

We and others have shown the Egfr is a direct genetic target of Olig2 in neural progenitors in ChIP-seq and RNA-seq studies (Mateo et al., 2015; Meijer et al., 2014). Qualitative analysis of Olig2 binding sites reveals its localization to intronic and distal upstream regions of the Egfr loci, marked by deposition of H3K4me1 and H3K27ac, indicating these sites as putative enhancers (Figure S2A). To validate these ChIP-seq findings on whether Egfr is a direct genetic target of Olig2, we analyzed Olig2-null neural progenitors expressing control GFP (KO), wild-type (WT), TPN, or TPM forms of murine Olig2. Immunoblot analysis suggests that germline ablation (KO) or loss of amino-terminal phosphorylation (TPN) of Olig2 reduces the expression of Egfr in normal (WT) and malignant (Cdkn2a−/−;EGFRvIII) neural progenitors (Figure 2A). We have previously shown that the phosphorylation state of Olig2 alters the mRNA levels of various RTKs (Meijer et al., 2014); however, these cell lines were engineered to express either phospho-mutant under a constitutive viral promoter against an Olig2-null genetic background. We extend these observations to knock-in (KI) neural progenitors, wherein the endogenous Olig2 locus was replaced with either the TPN or TPM coding sequences (Olig2TPN/TPN, Olig2TPM/TPM). Consistent with the over-expression system, we observed a significant up-regulation of the RTKs Egfr, Fgfr3 and Pdgfrα mRNA and repression of the cell-cycle inhibitor Cdkn1a and astrocytic marker Gfap mRNA in TPM versus TPN-expressing cells (Figure S2B).

Figure 2.

Figure 2

Open in a new tab

Olig2 regulates expression of Egfr in murine neural progenitors. A. Olig2-null cells expressing control GFP, wild type (WT), or phospho-mutants (TPN or TPM) of Olig2 were assayed for Egfr and Olig2 protein levels. Quantifications of Egfr (and EGFRvIII) protein levels shown on right. B. Knock-in neural progenitors expressing the TPN or TPM mutant of Olig2 (Olig2TPN/TPN and Olig2TPM/TPM) expressing either control GFP or dominant-negative p53 (p53DD). Protein levels of Egfr, Olig2, and p53 were assayed as above. C. Comparison of intracranial xenografts derived from murine glioma cells expressing TPN or TPM mutants of Olig2 (V5-tagged). Sections were stained with Egfr (green) and V5 (red) antibody to visualize tumor cells (scale bar = 100 μM). D. Olig2-null cells expressing control GFP, wild type (WT), or FLAG/HA-tagged DNA-binding mutant (Δ) of Olig2 were assayed for Egfr and Olig2 protein levels .(n ≥ 3). See also Figure S2.

We have previously reported that Olig2 phosphorylation is dispensable for progenitor proliferation and tumor initiation when p53 is genetically inactivated. However, the rate of tumor formation is significantly increased in the presence of Olig2 (i.e., Olig2cre/+ vs. Olig2cre/cre) in a murine model of malignant glioma harboring deletion of the p53 locus (p53flox/flox;EGFRvIII) (Mehta et al., 2011). To assess if Olig2-mediated regulation of Egfr is dependent on the tumor suppressor protein p53, we attenuated p53 function in phospho-mutant KI cells. As shown in Figure 2B, ectopic expression of dominant-negative p53 (p53DD) in neural progenitors expressing TPN mutant, was insufficient to induce Egfr expression suggesting that Olig2-mediated regulation of Egfr is independent of p53 status.

To confirm phospho-dependent up-regulation of Egfr in vivo, we orthotopically transplanted murine glioma cells (Cdkn2a −/−;EGFRvIII) expressing either of the Olig2 phosphomutants into the brains of immunocompromised mice. This murine glioma model re-capitulates the hallmark genetic lesions of the ‘Classical’ GBM molecular subtype: constitutive EGFR signaling (EGFRvIII), homozygous deletion of the Cdkn2a locus, and an intact p53 gene (Verhaak et al., 2010). Olig2 function is required to initiate tumors in this glioma model and amino-terminal phosphorylation regulates its tumorigenic competence (Ligon et al., 2007; Sun et al., 2011). Consistent with our in vitro findings, we observed stronger immunoreactivity of Egfr protein in TPM-expressing tumors compared to TPN-expressing tumors (Figure 2C).

To confirm if Olig2 DNA-binding is required to promote expression of Egfr, we analyzed Olig2-null cells expressing control GFP (KO), WT-Olig2, or a DNA-binding mutant (ΔOlig2) that harbors a point mutation (N114H) in the basic region which abrogates canonical E-Box recognition (Meijer et al., 2014). As seen in Figure 2D, germline ablation of Olig2 or loss of DNA-binding function led to diminished expression of Egfr in wild-type murine neural progenitors. Finally, we tested whether ΔOlig2 can antagonize endogenous Olig2 protein function. Ectopic expression of ΔOlig2 in malignant neural progenitors (Cdkn2a−/−; EGFRvIII and p53−/−; EGFRvIII) led to substantial repression of Egfr protein levels (Figure S2C). These findings suggest that Olig2 promotes Egfr expression through activity of the amino-terminal phosphorylation motif in a DNA-binding dependent manner.

OLIG2 regulates expression of EGFR and PDGFRα in hGSCs

To address whether the cross-regulatory loop between OLIG2 and RTKs persists in human GSCs (hGSCs), we isolated and enriched for neurosphere-forming cells from patient-derived GBM tissues (Table S1). Immunoblot analysis showed that hGSCs with higher levels of pOLIG2 shared expression of the EGFR (BT112, BT286, GB82, BT145), with varying levels of the PDGFRα. In contrast, pOLIG2 levels are largely diminished in EGFRLOW (BT187, BT142, GB3, GB16, GB80) hGSCs, which often displayed high expression of PDGFRα, reminiscent of the Proneural GBM subtype (Figure 3A). To simplify the nomenclature, we have termed hGSCs displaying EGFRHIGH expression as Classical (CLS) and EGFRLOW/PDGFRαHIGH expression as Proneural (PN). To confirm a role for active EGFR signaling in sustaining OLIG2 protein in hGSCs (BT145, BT112), we treated the cells with EGFRi + CHX and observed a rapid depletion of the OLIG2 protein, compared to CHX alone (Figure S3A). These findings suggest that, similar to murine neural progenitors, EGFR signaling sustains the OLIG2 protein in CLS hGSCs.

Figure 3.

Figure 3

Open in a new tab

OLIG2 regulates expression of RTKs in hGSCs. A. Lysates derived from hGSCs were immunoblotted with antibodies to pOLIG2, OLIG2, EGFR, PDGFRα, and Vinculin. B. hGSCs (BT145, BT112, GB3, GB16) were transduced with shNT (non-target control) or shOLIG2. Lysates were collected 48 hours post-infection, and immunoblotted for EGFR, PDGFRα, OLIG2 and Vinculin. C. hGSCs expressing GFP or ΔOlig2 (FLAG-tag) and the levels of EGFR and PDGFRα were assayed similarly as above. D. hGSCs were transduced with retrovirus expressing GFP or WT-Olig2 (V5-tag) and the levels of EGFR and PDGFRα were assayed using standard immunoblotting techniques. E. Lysates from hGSCs expressing GFP or Olig2 phospho-mutants (TPN, TPM) were collected to assay protein levels of EGFR and PDGFRα, ectopic (V5-tag) and endogenous OLIG2. (n ≥ 3). *denotes 90 kDa EGFR band, uniquely expressed in GB16. See also Figure S3.

The expression of OLIG2 protein in both PN and CLS hGSCs led us to inquire if these developmentally-regulated pro-neural (lineage-specifying) and anti-neural (proliferative) functions of OLIG2 were co-opted in hGSCs. After acute silencing of OLIG2 in both CLS (BT145, BT112) and PN (GB3, GB16) hGSCs, we observed a significant down-regulation of both RTKs, independent of the molecular subtype (Figure 3B). We confirmed specificity and effects of shRNA-mediated knockdown by utilizing three alternate hairpins targeting the human OLIG2 ORF (Figure S3B). In contrast to recent findings (Lu et al., 2016), we did not observe a parallel up-regulation of EGFR protein following acute silencing of OLIG2 in PDGFRαHIGH hGSCs (Figure 3B and S3C). Furthermore, analysis of published OLIG2/H3K27ac ChIP-seq data in hGSCs (Suva et al., 2016), confirms OLIG2 binding at both distal and intronic enhancers at both the EGFR and PDGFRA loci (data not shown). We also observed a significant down-regulation of both RTK transcripts (EGFR and PDGFRA) following acute OLIG2 silencing (Figure S3D).

We next addressed whether ΔOlig2 can assert itself in a dominant-negative manner in hGSCs. We assessed protein levels of EGFR and PDGFRα from two CLS (BT145, BT112) and two PN (GB3, GB16) lines where we have chronically over-expressed ΔOlig2. We observed a substantial down-regulation of the RTKs in all hGSCs surveyed (Figure 3C and S3E), indicating an OLIG2 loss-of-function phenotype mediated by the DNA-binding mutant (Δ). qRTPCR analysis confirmed the down-regulation of RTK transcripts (i.e., EGFR, PDGFRA) in presence of ΔOlig2 (Figure S3F). To further confirm OLIG2-mediated regulation of RTKs, we ectopically introduced wild-type Olig2 or control GFP into either the BT145CLS or GB16PN hGSC line. Surprisingly, following over-expression of WT-Olig2 we observed an induction of PDGFRα protein in our CLS/EGFRHIGH line (BT145) and induction of EGFR protein in our PN/PDGFRαHIGH line (GB16) (Figure 3D). Finally, we asked if the phosphorylation status of OLIG2 is critical for sustaining the expression of RTKs in human GSCs. We utilized retroviral expression vectors to introduce control GFP or the phospho-mutants (TPN, TPM) in either the BT145CLS or GB16PN hGSC lines. We observed a reduction in the protein levels of both the EGFR and PDGFRα (albeit more modest) protein following over-expression of TPN-Olig2, but not TPM-Olig2 (Figure 3D). These data suggest that OLIG2 regulates the expression of both the EGFR and PDGFRα in hGSCs.

OLIG2 is the Nuclear Gateway for Proneural-Mesenchymal Transition

We next asked if loss of OLIG2 leads to a shift in GBM molecular subtype. To investigate the functional outcomes of OLIG2 knockdown in hGSCs, we utilized shRNA to ablate the OLIG2 protein in both CLS/EGFRHIGH and PN/PDGFRαHIGH hGSCs. We observed a robust induction of the astrocytic marker GFAP in our CLS lines (BT145, GB71, GB82), whereas the mesenchymal marker CD44 is induced in PN lines (GB3, GB16, GB80) (Figure 4A and S4A). However, this ‘mutually-exclusive’ up-regulation of the GFAP and CD44 proteins does not fully extend to all hGSC lines examined (e.g., GB71CLS): lines which express appreciable GFAP protein prior to OLIG2 knockdown, appear to up-regulate both GFAP and CD44 (Figure S4A). Consistent with a loss of ‘stemness’, we observed a substantial down-regulation of the glioma master transcription factor SOX2 following OLIG2 knockdown in all CLS/PN lines examined (Figure S4A). These data suggest that acute loss of OLIG2 collapses the ‘stem’ state in hGSCs, with a subsequent shift to a more differentiated state.

Figure 4.

Figure 4

Open in a new tab

OLIG2 is the Nuclear Gateway for Proneural-Mesenchymal Transition. A. hGSCs (BT145, GB3) were transduced with shNT (non-target control) or shOLIG2. Lysates were collected 48 hours post-transduction, and immunoblotted for the indicated proteins. B-C. Lysates from hGSCs expressing control GFP, WT-Olig2 (V5-tag), or ΔOlig2 (FLAG-tag) were collected and immunoblotted for CD44, GFAP, and ectopic Olig2. Note: blots presented here were reprobed from Figure 3C-D. D. qRT-PCR analysis of hGSCs (BT145, GB16) expressing control GFP or ΔOlig2 for indicated subclass genes. HPRT and 18s were utilized as housekeeping genes. (n ≥ 3). E. Proposed model of OLIG2-mediated regulation of growth factor signaling. See also Figure S4.

To further assess OLIG2-dependent regulation of GFAP and CD44, we utilized hGSC lines (BT145CLS and GB16PN) to ectopically express WT-Olig2 or the DNA-binding mutant ΔOlig2. We observed repression of both the CD44 and GFAP proteins in lines where we have over-expressed WT-Olig2 (Figure 4B). In striking contrast, chronic over-expression of dominant-negative Olig2 (ΔOlig2) resulted in up-regulation of the CD44 protein (but not GFAP) - suggesting Mesenchymal (but not Astroglial) conversion (Figure 4C). To assess if loss of OLIG2 function triggers a Proneural-Mesenchymal Transition (PMT), we performed expression analysis for GBM molecular subtype specific genes in four independent hGSCs lines expressing ΔOlig2 (BT145, BT112, GB3, GB16). We observed a catastrophic loss of Proneural class genes and upregulation of Mesenchymal class genes in our ΔOlig2-expressing hGSCs compared to GFP-expressing controls, independent of the original molecular subtype (i.e., CLS vs. PN) (Figure 4D and S4B). Conversely, over-expression of WT-Olig2 results in up-regulation of Proneural class genes and down-regulation of Mesenchymal class genes (Figure S4C). We further extend these “PMT” expression shift observations to murine glioma lines of defined genetic backgrounds (Cdkn2a−/−;EGFRvIII and p53−/−;EGFRvIII) expressing ΔOlig2 (Figure S4D). Thus, our data support a model in which OLIG2 serves as a critical mediator of growth factor signaling (i.e., RTK expression) and GSC identity (Figure 4E).

DISCUSSION

A key determinant of intratumoral heterogeneity in malignant gliomas is likely to be glioma stem-like cells, which promote tumor growth, recurrence and therapeutic resistance (Lathia et al., 2015). Identification of proteins that drive GSC biology is of special interest in neuro-oncology. One such protein that is expressed in GSCs, glial-restricted progenitors and their respective progeny is the bHLH TF Olig2 (Ligon et al., 2007). During the early stages of CNS development, Olig2 serves to prevent premature cell cycle exit in developmentally-uncommitted neural progenitors, and at later time points, promote differentiation into oligodendrocytes (Meijer et al., 2012). Previous work has coupled the pro-mitotic functions of Olig2 to phosphorylation of a cluster of serines at its amino-terminus (Ser10, 13, 14), also known as the phospho-serine-motif. Mutation of this motif into a “null” state (Ser to Ala) impedes neural progenitor proliferation and the tumorigenic capacity of murine GSCs (Sun et al., 2011). Further biochemical analyses of these mutants suggest little to no alteration in genomic targeting, but rather a difference in intranuclear localization and co-regulator association (Meijer et al., 2014). Here, we investigated the mechanisms that sustain OLIG2 expression and its proliferative functions in neural progenitors, and assessed whether these signals are co-opted in malignant gliomas. We have identified an “OLIG2-RTK” signaling axis, which functions in a feedforward manner to sustain co-expression. This loop appears to be conserved in both normal and malignant neural progenitors, highlighting a previously unappreciated relationship between OLIG2 and EGFR signaling (Aguirre et al., 2007; Persson et al., 2010).

Constitutive growth factor signaling is a hallmark of diffuse gliomas, propagated by activating genetic lesions in RTKs, most commonly within the EGFR and PDGFRA locus (Verhaak et al., 2010), which appear to be mutually exclusive oncogenic events (Snuderl et al., 2011). Mutations at the EGFR locus occur in the Classical subtype and mutations in the PDGFRA locus occur primarily in the Proneural subtype (Verhaak et al., 2010). While these two molecular subtypes appear similar on a histopathological level, their expression profiles suggest that they might have disparate cellular origins. Recent elegant mouse modeling studies (Alcantara Llaguno et al., 2015) suggests that both multipotent neural stem cells (NSCs) and lineage-restricted oligodendrocyte precursor cells (OPCs) are capable of giving rise to anatomically and molecularly unique types of GBM when transformed with identical mutations (i.e., Tp53, Pten, Nf1 deletion) using cell specific Cre-drivers (NestincreER for NSCs, Ascl1creER for NSC/OPCs, Cspg4creER for OPCs). These authors observed expression of Olig2 in both Type I (‘NSC’-like) and Type II (‘OPC-like’) tumors in their genetically engineered mouse models. In our patient-derived GSCs we observe universal expression of OLIG2, and two distinct states of phosphorylated OLIG2 (low vs. high), which appear to correlate with dominant expression of the RTKs EGFR or PDGFRα (CLS vs. PN) (Figure 3A). As the developmental functions of Olig2 amino-terminal phosphorylation are presumably complete following the formation of mature white matter – its presence in PDGFRα-expressing (PN) hGSCs presents a paradox. Is the phospho-serine-motif performing an identical function in these hGSCs (compared to their CLS counterparts)? Our biochemical analyses of OLIG2 function in both hGSC populations (CLS and PN) suggest a requirement for intact phosphorylation to sustain expression of the EGFR, while having only a modest impact on PDGFRα expression (Figure 3E). This finding is consistent with previous observations that indicated amino-terminal phosphorylation is not critically required for oligodendroglial lineage specification (Sun et al., 2011). However, expression of both RTKs is repressed upon introduction of a DNA-binding mutant (ΔOlig2), which appears to function in a dominant-negative manner (Figure 3C and 4D). This suggests that while phosphoserine-motif functions might persist in PN hGSCs, the larger pool of un-phosphorylated OLIG2 may be performing a secondary function.

We have previously reported that the pro-mitotic functions of Olig2 indirectly promote radio-resistance through antagonization of p53 transactivation (Mehta et al., 2011). Recent work (Bhat et al., 2013; Carro et al., 2010) has proposed a phenomenon termed “Proneural-Mesenchymal Transition” as mediating radio-resistance and aggressive recurrence in malignant gliomas. In this model of progressive resistance, glioma cells acquire Mesenchymal traits and lose Proneural gene expression. In this study, we attempted to reconcile these results through cross-species investigative analysis of OLIG2 molecular function in normal and malignant neural progenitors. We observe that CLS hGSCs resist Mesenchymal transformation and shift towards the astroglial lineage (i.e., GFAP vs. CD44 up-regulation) following ablation of OLIG2. In contrast, PN hGSCs strongly up-regulate the CD44 protein (Figure 4A and S4A). Notably, Mesenchymal transformation does not appear to be entirely exclusive to astrocytic differentiation, as we observe a parallel up-regulation of GFAP in some lines following OLIG2 loss (Figure S4A). This does not however exclude fate conversion –as during development the CD44 protein has been shown to label astroglial-restricted progenitors which do not yet express GFAP (Liu et al., 2004).

Thus, OLIG2 which is uniquely CNS-restricted amongst a quartet of glioma master TFs (Suva et al., 2014), appears to function as a “nuclear gateway” for Proneural gene and RTK expression in hGSCs (Figure 4D and 4E). Similar molecular subtype conversion findings were recently presented (Lu et al., 2016), however, these authors observed a “Proneural-Classical” subtype shift following Olig2 deletion. These authors employed a murine glioma model which initiates tumors in a very specific cell type (white matter OPCs) bearing mutations unique to the Proneural subtype (Tp53;Pten deletion; Pdgfb over-expression), excluding the possibility of transformation in more naïve cells (i.e., neural stem/progenitor cells) and Olig2 functions therein (i.e., sustaining Egfr expression; Figure 2). Similar to our findings, Lu et al also observed a global collapse of the Proneural gene expression program following chronic Olig2 deletion, consistent with a loss of OPC character. Unlike our observed PMT, these authors observed an up-regulation of the Egfr and Classical subtype genes. However, it is important to note that the EGFR can be expressed in a plethora of normal and neoplastic cells, many of which do not express OLIG2 (e.g., astrocytes, immortalized glioma cell lines) and lack cancer stem cell identity (Suva et al., 2014). Thus, this observation by Lu and colleagues of Egfr up-regulation following chronic Olig2 deletion is possibly a result of astroglial restriction (and dissimilar to our acute silencing experiments). Importantly, a functional role for Egfr in oligodendroglial lineage cells during CNS development and gliomagenesis has been well documented (Aguirre et al., 2007; Persson et al., 2010), and a role for Olig2 in repression of the Egfr locus is difficult to reconcile with this pre-existing data (as postulated in Lu et al). Consistent with our murine data and unlike Lu et al study, we observed a decrease in EGFR protein after loss of OLIG2 function in hGSCs. It is possible that differences in other genomic alterations apart from PDGFRα amplifications also play a role in subtype specification (Ozawa et al., 2014). Further analysis of the micro-environmental cues and intracellular mediators of this proliferation-differentiation (EGFR-PDGFRα) switch will provide additional mechanistic insight into the process of subtype conversion and may present novel therapies for patients with malignant gliomas (Singh et al., 2016). We therein propose a model where OLIG2 resides at the intersection of two cross regulatory loops in normal and malignant neural progenitors, forming a compulsory link between cell cycle regulation and sub-type specification, dictated through spatiotemporal protein modifications which promote self-renewal or fate choice.

METHODS

Animal procedures, tissue harvest and cell culture

Animal husbandry was performed according to BNI guidelines under IACUC-approved protocols. Derivation and culture of neural progenitor cell lines used have been described previously ( Mehta et al., 2011). Generation of murine Olig2 mutant constructs have been previously described (Meijer et al., 2014).

Immunoblotting and Quantitative PCR (qRT-PCR)

Immunoblotting was performed using standard protocols previously described (Meijer et al., 2014). Full list of antibodies used is provided in the supplemental information. RNA and cDNAwas prepared using previously described methods (Meijer et al., 2014).

Statistical analysis

For each experiment, data were collected from at least three biological repeats and analyzed by 1way or 2way ANOVA with Bonferroni multiple comparisons test as indicated. SD, standard deviation; SEM, standard error of the mean; *, p<0.05; **, P<0.01; ***, P<0.001

Supplementary Material

HIGHLIGHTS.

  • OLIG2 forms a feedforward loop with mitogenic signaling in neural progenitors

  • Inhibition of EGFR results in depletion of OLIG2 protein

  • Phosphorylation state of OLIG2 alters expression of RTKs

  • Loss of OLIG2 function in GSCs results in mesenchymal transformation

ACKNOWLEDGEMENTS

Patient-derived glioma cells were provided by the Biobank Core Facility @ St. Joseph’s Hospital and Barrow Neurological Institute and the Living Tissue Bank at Dana-Farber Cancer Institute. The Biobank is funded by the Arizona Biomedical Research Commission and the Barrow Neurological Foundation. We thank Drs. John Alberta and Charles Stiles (DFCI) for unrestricted use of Olig2/phospho-Olig2 antibodies and helpful discussions. We thank Dr. Shiv Singh, Dr. Nadine Bakkar, Ashley Boehringer, and members of the Mehta Lab for technical assistance and manuscript comments. This work was supported by grants to S.M. from Barrow Neurological Foundation and NIH (R01 NS088648A); N.S. is supported by NIH R01 NS082745 and R01 CA175391.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

SUPPLEMENTAL INFORMATION.

Supplemental information includes extended experimental procedures, four figures and one table, and can be found with this article online.

AUTHOR CONTRIBUTIONS.

R.K. and S.M. conceived and designed the experiments. R.K. performed all experiments except for the following: L.S. performed IF staining, A-C.T. generated phospho-mutant knock-in neural progenitors, and E.S. generated patient-derived lines. A-C.T., N.S., and D.H.R. contributed unpublished reagents and materials. R.K. and S.M. analyzed data. R.K. and S.M. wrote the manuscript.

REFERENCES

  1. Aguirre A, Dupree JL, Mangin JM, Gallo V. A functional role for EGFR signaling in myelination and remyelination. Nat Neurosci. 2007;10:990–1002. doi: 10.1038/nn1938. [DOI] [PubMed] [Google Scholar]
  2. Alcantara Llaguno SR, Wang Z, Sun D, Chen J, Xu J, Kim E, Hatanpaa KJ, Raisanen JM, Burns DK, Johnson JE, et al. Adult Lineage-Restricted CNS Progenitors Specify Distinct Glioblastoma Subtypes. Cancer Cell. 2015;28:429–440. doi: 10.1016/j.ccell.2015.09.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bhat KP, Balasubramaniyan V, Vaillant B, Ezhilarasan R, Hummelink K, Hollingsworth F, Wani K, Heathcock L, James JD, Goodman LD, et al. Mesenchymal differentiation mediated by NF-kappaB promotes radiation resistance in glioblastoma. Cancer Cell. 2013;24:331–346. doi: 10.1016/j.ccr.2013.08.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Carro MS, Lim WK, Alvarez MJ, Bollo RJ, Zhao X, Snyder EY, Sulman EP, Anne SL, Doetsch F, Colman H, et al. The transcriptional network for mesenchymal transformation of brain tumours. Nature. 2010;463:318–325. doi: 10.1038/nature08712. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Cloughesy TF, Cavenee WK, Mischel PS. Glioblastoma: from molecular pathology to targeted treatment. Annu Rev Pathol. 2014;9:1–25. doi: 10.1146/annurev-pathol-011110-130324. [DOI] [PubMed] [Google Scholar]
  6. Gonzalez-Perez O, Romero-Rodriguez R, Soriano-Navarro M, Garcia-Verdugo JM, Alvarez-Buylla A. Epidermal growth factor induces the progeny of subventricular zone type B cells to migrate and differentiate into oligodendrocytes. Stem Cells. 2009;27:2032–2043. doi: 10.1002/stem.119. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Imayoshi I, Kageyama R. bHLH factors in self-renewal, multipotency, and fate choice of neural progenitor cells. Neuron. 2014a;82:9–23. doi: 10.1016/j.neuron.2014.03.018. [DOI] [PubMed] [Google Scholar]
  8. Imayoshi I, Kageyama R. Oscillatory control of bHLH factors in neural progenitors. Trends Neurosci. 2014b;37:531–538. doi: 10.1016/j.tins.2014.07.006. [DOI] [PubMed] [Google Scholar]
  9. Lathia JD, Mack SC, Mulkearns-Hubert EE, Valentim CL, Rich JN. Cancer stem cells in glioblastoma. Genes Dev. 2015;29:1203–1217. doi: 10.1101/gad.261982.115. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Ligon KL, Alberta JA, Kho AT, Weiss J, Kwaan MR, Nutt CL, Louis DN, Stiles CD, Rowitch DH. The oligodendroglial lineage marker OLIG2 is universally expressed in diffuse gliomas. J Neuropathol Exp Neurol. 2004;63:499–509. doi: 10.1093/jnen/63.5.499. [DOI] [PubMed] [Google Scholar]
  11. Ligon KL, Huillard E, Mehta S, Kesari S, Liu H, Alberta JA, Bachoo RM, Kane M, Louis DN, Depinho RA, et al. Olig2-regulated lineage-restricted pathway controls replication competence in neural stem cells and malignant glioma. Neuron. 2007;53:503–517. doi: 10.1016/j.neuron.2007.01.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Liu Y, Han SS, Wu Y, Tuohy TM, Xue H, Cai J, Back SA, Sherman LS, Fischer I, Rao MS. CD44 expression identifies astrocyte-restricted precursor cells. Dev Biol. 2004;276:31–46. doi: 10.1016/j.ydbio.2004.08.018. [DOI] [PubMed] [Google Scholar]
  13. Lu F, Chen Y, Zhao C, Wang H, He D, Xu L, Wang J, He X, Deng Y, Lu EE, et al. Olig2-Dependent Reciprocal Shift in PDGF and EGF Receptor Signaling Regulates Tumor Phenotype and Mitotic Growth in Malignant Glioma. Cancer Cell. 2016;29:669–683. doi: 10.1016/j.ccell.2016.03.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Mateo JL, van den Berg DL, Haeussler M, Drechsel D, Gaber ZB, Castro DS, Robson P, Crawford GE, Flicek P, Ettwiller L, et al. Characterization of the neural stem cell gene regulatory network identifies OLIG2 as a multifunctional regulator of self-renewal. Genome Res. 2015;25:41–56. doi: 10.1101/gr.173435.114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Mehta S, Huillard E, Kesari S, Maire CL, Golebiowski D, Harrington EP, Alberta JA, Kane MF, Theisen M, Ligon KL, et al. The central nervous system-restricted transcription factor Olig2 opposes p53 responses to genotoxic damage in neural progenitors and malignant glioma. Cancer Cell. 2011;19:359–371. doi: 10.1016/j.ccr.2011.01.035. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Meijer DH, Kane MF, Mehta S, Liu H, Harrington E, Taylor CM, Stiles CD, Rowitch DH. Separated at birth? The functional and molecular divergence of OLIG1 and OLIG2. Nat Rev Neurosci. 2012;13:819–831. doi: 10.1038/nrn3386. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Meijer DH, Sun Y, Liu T, Kane MF, Alberta JA, Adelmant G, Kupp R, Marto JA, Rowitch DH, Nakatani Y, et al. An amino terminal phosphorylation motif regulates intranuclear compartmentalization of Olig2 in neural progenitor cells. J Neurosci. 2014;34:8507–8518. doi: 10.1523/JNEUROSCI.0309-14.2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Otero JJ, Rowitch D, Vandenberg S. OLIG2 is differentially expressed in pediatric astrocytic and in ependymal neoplasms. J Neurooncol. 2011;104:423–438. doi: 10.1007/s11060-010-0509-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Ozawa T, Riester M, Cheng YK, Huse JT, Squatrito M, Helmy K, Charles N, Michor F, Holland EC. Most human non-GCIMP glioblastoma subtypes evolve from a common proneural-like precursor glioma. Cancer Cell. 2014;26:288–300. doi: 10.1016/j.ccr.2014.06.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Persson AI, Petritsch C, Swartling FJ, Itsara M, Sim FJ, Auvergne R, Goldenberg DD, Vandenberg SR, Nguyen KN, Yakovenko S, et al. Non-stem cell origin for oligodendroglioma. Cancer Cell. 2010;18:669–682. doi: 10.1016/j.ccr.2010.10.033. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Singh SK, Fiorelli R, Kupp R, Rajan S, Szeto E, Lo Cascio C, Maire CL, Sun Y, Alberta JA, Eschbacher JM, et al. Post-translational Modifications of OLIG2 Regulate Glioma Invasion through the TGF-beta Pathway. Cell Rep. 2016 doi: 10.1016/j.celrep.2016.06.045. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Snuderl M, Fazlollahi L, Le LP, Nitta M, Zhelyazkova BH, Davidson CJ, Akhavanfard S, Cahill DP, Aldape KD, Betensky RA, et al. Mosaic amplification of multiple receptor tyrosine kinase genes in glioblastoma. Cancer Cell. 2011;20:810–817. doi: 10.1016/j.ccr.2011.11.005. [DOI] [PubMed] [Google Scholar]
  23. Sun Y, Meijer DH, Alberta JA, Mehta S, Kane MF, Tien AC, Fu H, Petryniak MA, Potter GB, Liu Z, et al. Phosphorylation state of Olig2 regulates proliferation of neural progenitors. Neuron. 2011;69:906–917. doi: 10.1016/j.neuron.2011.02.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Suva ML, Rheinbay E, Gillespie SM, Patel AP, Wakimoto H, Rabkin SD, Riggi N, Chi AS, Cahill DP, Nahed BV, et al. Reconstructing and reprogramming the tumor-propagating potential of glioblastoma stem-like cells. Cell. 2014;157:580–594. doi: 10.1016/j.cell.2014.02.030. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Verhaak RG, Hoadley KA, Purdom E, Wang V, Qi Y, Wilkerson MD, Miller CR, Ding L, Golub T, Mesirov JP, 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. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

NIHMS811297-supplement.pdf (829.9KB, pdf)

RESOURCES