Hyperactivation of the mTORC2 signaling pathway has been shown to contribute to the oncogenic properties of gliomas. Moreover, overexpression of the mTORC2 regulatory subunit Rictor has been associated with increased proliferation and invasive character of these tumor cells.
To determine whether Rictor overexpression was sufficient to induce glioma formation in mice, we inserted a Cre-lox-regulated human Rictor transgene into the murine ROSA26 locus. This floxed Rictor strain was crossed with mice expressing the Cre recombinase driven from the glial fibrillary acidic protein (GFAP) promoter whose expression is limited to the glial cell compartment. Double transgenic GFAP-Cre/RictorloxP/loxP mice developed multifocal infiltrating glioma containing elevated mTORC2 activity and typically involved the subventricular zone (SVZ) and lateral ventricle. Analysis of Rictor-dependent signaling in these tumors demonstrated that in addition to elevated mTORC2 activity, an mTORC2-independent marker of cortical actin network function, was also elevated. Upon histological examination of the neoplasms, many displayed oligodendroglioma-like phenotypes and expressed markers associated with oligodendroglial lineage tumors. To determine whether upstream oncogenic EGFRvIII signaling would alter tumor phenotypes observed in the GFAP-Cre/RictorloxP/loxP mice, transgenic GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP mice were generated. These mice developed mixed astrocytic-oligodendroglial tumors, however glioma formation was accelerated and correlated with increased mTORC2 activity. Additionally, the subventricular zone within the GFAP-Cre/RictorloxP/loxP mouse brain was markedly expanded, and a further proliferation within this compartment of the brain was observed in transgenic GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP mice.
These data collectively establish Rictor as a novel oncoprotein and support the role of dysregulated Rictor expression in gliomagenesis via mTOR-dependent and mTOR-independent mechanisms. Furthermore, oncogenic EGFRvIII signaling appears to potentiate the in vivo proliferative capacity of GFAP-Cre/RictorloxP/loxP gliomas.
Citation: Bashir T, Cloninger C, Artinian N, Anderson L, Bernath A, Holmes B, et al. (2012) Conditional Astroglial Rictor Overexpression Induces Malignant Glioma in Mice. PLoS ONE 7(10): e47741. https://doi.org/10.1371/journal.pone.0047741
Editor: Robert Oshima, Sanford Burnham Medical Research Institute, United States of America
Received: July 17, 2012; Accepted: September 14, 2012; Published: October 15, 2012
This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.
Funding: This work was supported, in part, by National Institutes of Health grant R01CA109312 and funds from the Veterans Administration. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Despite combined therapeutic modalities, including surgery, radiation and chemotherapy, the prognosis for patients with malignant glioma, such as glioblastoma multiforme (GBM) remains poor with a median survival of only twelve months . The appearance of de novo primary glioblastomas is associated with dysregulation of epidermal growth factor (EGFR) expression –. The most common EGFR mutation consists of an aberrantly spliced form which lacks exons 3 thru 6 (EGFRvIII) resulting in a constitutively activated receptor , . Phosphoinositol-3-kinase (PI3K) signaling is also hyperactivated in most GBMs in association with EGFR mutation and/or PTEN tumor suppressor protein loss , .
Downstream of both oncogenic EGFR and PI3K signaling, the mTOR serine/threonine kinase exists in at least two complexes (mTORC1 & 2) which relay signals to distinct effectors . PI3K mediated activation of mTORC1 via AKT, occurring due to inhibitory phosphorylation of the TSC1/TSC2 complex, links mTORC1 to PI3K signaling and the regulation of protein synthesis, cell size and proliferation . mTORC2 responds to growth factor receptor activation, including EGFRvIII , and regulates cell survival, metabolism and the cytoskeleton . mTORC1 specific components include Raptor and PRAS40, while the mTORC2 supracomplex uniquely contains Rictor, mSin1, Protor 1 and 2. Both mTORC1 and 2 contain mLST8, DEPTOR, the Tti1/Tel2 complex and the catalytic mTOR subunit .
Rictor, a 200 kD protein, was identified as a defining component of mTORC2 and does not exhibit significant sequence conservation between mammals and yeast, and lacks structural domains of known function . The relative amount of Rictor which complexes with mTOR varies by cell type and inversely correlates with Raptor expression , . Rictor binding to mTOR is unaffected by acute rapamycin exposure, however long term exposure results in reduced levels of complex abundance . Rictor is also known to bind other protein partners, including the unconventional myosin motor Myo1C, the integrin linked kinase (ILK), Cullin-1, PKCς, Hsp70, and has demonstrated roles in the regulation of the cytoskeleton , –.
Several reports have shown that the mTORC2 scaffolding protein Rictor is overexpressed in cancers including gliomas and has been shown to contribute to the oncogenic properties of these tumors, however, a causal role for Rictor in gliomagenesis has not been demonstrated , , –. We show here that glial fibrillary acidic protein Cre (GFAP-Cre) mediated conditional overexpression of Rictor is sufficient to induce intermediate-grade gliomas in mice. These oligodendroglial-like tumors exhibited elevated mTORC2 signaling in vivo and displayed increased growth, migratory capacity and invasiveness in vitro. Moreover, we demonstrate that cooperative signaling between GFAP-EGFRvIII and GFAP-Cre/RictorloxP/loxP mice leads to hyperactivated mTORC2 signaling and results in high-grade gliomas.
Generation of Rosa26hRictorloxP/wt and GFAP-Cre/Rictorloxp/loxp mice
To examine whether Rictor overexpression would induce tumor formation in vivo we generated a mouse model in which a myc-tagged human Rictor transgene was conditionally expressed in glial cells. Using homologous recombination in embryonic stem cells, the Rictor transgene was inserted into the ROSA26 locus downstream of a floxed PGK-neo cassette containing a strong transcriptional termination stop sequence (Figure 1A) . The transgene was not expressed unless Cre-mediated recombination resulted in removal of the cassette and allowed expression of Rictor. Southern blot and PCR analysis confirmed appropriate targeting of the transgene (Figure 1B) and Cre-dependent over-expression of the Rictor allele in heterozygous and homozygous animals was assessed by immunoblot (Figure 1C). Heterozygotes were intercrossed to obtain litters of pups with Mendelian distributions of the F2 genotypes. Glial cell restricted expression of the Cre recombinase was the result of matings to the GFAP-Cre mouse line developed in Albee Messing's laboratory . This line demonstrates widespread recombinase activity in neurons, glia and GFAP positive subventricular zone (SVZ) precursor cells in adult GFAP-Cre mouse brain , .
(A) Targeting strategy for Rictorloxp/+. A human myc-tagged Rictor sequence was cloned into the pBigT plasmid and subcloned into the targeting vector pROSA26PA. The predicted sizes of EcoRV (E) – digested DNA fragments are shown when hybridized to the indicated probe. (B). Southern blot of EcoRV-digested genomic DNA from +/+ or Rictorloxp/+mice hybridized to the probe shown in (A). In the lower panel, genomic PCR-based detection of the myc-Rictor transgene sequences is shown in +/+ and Rictorloxp/+ mice. (C) Immunoblot of brain extracts from GFAP-Cre, GFAP-Cre/Rictorloxp/+, or GFAP-Cre/Rictorloxp/loxp mice using antibodies against the myc-tagged Rictor demonstrating Cre-dependent expression of Rictor. Values in parentheses are fold increases in total Rictor expression relative to endogenous expression levels in GFAP-Cre mice as determined by densitometric analysis of an immunoblot using antibodies reactive to both human and murine Rictor.
Glioma formation and elevated mTORC2 activity in GFAP-Cre/RictorloxP/loxP mice
Offspring from both heterozygous (RictorloxP/+) and homozygous (RictorloxP/loxP) crosses to the GFAP-Cre line were born in the expected Mendelian ratios and appeared normal at birth. Upon examination of the brain at the time of euthanasia no macroscopic abnormalities were found, however, in GFAP-Cre/RictorloxP/loxP animals histological examination revealed bilateral, multifocal infiltrating glioma with nearly complete penetrance. In several cases, tumors involved the amygdalohippocampal region or surrounding cortex and were also found within close proximity to the subventricular zone (SVZ). The gliomas were characterized at low magnification by foci of hypercellularity on H & E stain (see Figure 2A). Infiltrating glioma cells exhibited oligodendroglial-like features with tumor cells displaying the typical fried egg appearance (Figure 2B). In addition, infiltrative tumors had regions of intratumoral hemorrage with chicken wire-like vascularity seen in human oligodendrogliomas (Figure 2C) , . The infiltrating growth pattern, degree of hypercellularity and atypism seen in these tumors were consistent with intermediate grade II–III gliomas based on WHO criteria . Many infiltrating cells were APC and GalC immunoreactive (Figure 2D&E, oligodendrocyte markers) staining positive for phospho-T346-NDRG1, phospho-S657-PKCα, phospho-S473-AKT, phospho-S32/36-IkBα, markers of mTORC2 activity (Figure 2F–H). Additionally, in a cell line (Ric0) derived from resected gliomas from newborn GFAP-Cre/RictorloxP/loxP mice expressed the Rictor transgene, stained positive for GalC and demonstrated hyperactivation of mTORC2 as compared to oligodendrocytes isolated from control littermates (Figure 2I–J).
(A) Low magnification of H & E-stained multifocal gliomas. (B) H & E-stained oligodendroglial tumor cells. (C) Representative gliomas displaying chicken wire-like vascularity observed in oligodendrogliomas. (D&E) Oligodendroglial tumor cells immunoreactive for APC and GalC, respectively. (F–H) Sections were stained with antibodies to phospho-T346-NDRG1, phospho-S657-PKCα and phospho-S473-AKT, respectively and demonstrated increased staining in oligodendroglial cells. In A–H, scale bar; 20 µm. (I) Increased mTORC2 signaling determined by immunoblot of Ric0 oligodendroglial cell line established from a newborn GFAP-Cre/Rictorloxp/loxp glioma. The Rictor transgene was detected using antisera specific for human Rictor. (J) GalC immunoreactive Ric0 cells demonstrating oligodendroglial lineage. Scale bar; 20 µm.
mTORC2-independent functions of Rictor are increased in GFAP-Cre/RictorloxP/loxP mice
Several reports have implicated Rictor in functions distinct from its role within mTORC2 , , , . Rictor has been shown to directly interact with the unconventional myosin Myo1c to form a complex which regulates the cortical actin cytoskeletal network . Tyrosine 118 phosphorylation of paxillin is associated with the regulation of dynamic cortical actin events and requires the Myo1c-Rictor complex , . Moreover, this phosphorylation event occurs independent of mTOR or PI3K signaling . To address whether gliomas from GFAP-Cre/RictorloxP/loxP mice contained elevated levels of phospho-Y118-paxillin sections were stained via immunohistochemistry using phospho-specific antibodies. As shown in Figure 3A, phospho-Y118-paxillin positive cells were detected in many oligodendroglial cells suggesting that Rictor overexpression may indeed promote Myo1C-Rictor/paxillin signaling. Similarly, in Ric0 cells, phospho-Y118-paxillin levels were elevated compared to purified oligodendrocytes from control littermate mice (Figure 3B, left panel). We were also able to detect the endogenous Myo1C-Rictor complex via coimmunoprecipitation in Ric0 cells. As shown in the right panel of Figure 3B, Myo1C antibodies coimmunoprecipitated myc-tagged Rictor, as well as, in the reciprocal coimmunoprecipitation reaction using anti-myc to detect Myo1C. These data taken together strongly suggest that in gliomas from GFAP-Cre/RictorloxP/loxP mice Rictor participates in mTORC2-independent activation of cortical actin dynamics.
(A) Elevated phospho-Y118-paxillin staining in GFAP-Cre/Rictorloxp/loxp gliomas. Left panel, no primary antibody; right panel, anti-phospho-Y118-paxillin antibody. Scale bar; 20 µm. (B) Elevated phospho-Y118-paxillin levels in Ric0 cells as determined by immunoblot analysis (left panel); co-immunoprecipitation of the myc-tagged Rictor-Myo1C complex in Ric0 cells (right panel). Whole cell lysate was immunoprecipitated with the indicated antibodies and immunoprecipitates were immunoblotted for either Myo1C or the myc-tag.
Cooperative signaling between EGFRvIII and mTORC2 in mice results in accelerated glioma development
To determine whether their existed a potential interaction between oncogenic EGFRvIII and Rictor proteins, we crossed GFAP-Cre/RictorloxP/loxP mice with a previously characterized transgenic strain which expresses the constitutively active EGFRvIII allele driven from the GFAP promoter . In addition to genotyping to confirm the presence of the transgenes in these mice, derived oligodendroglial tumor cultures from these transgenic mice brain (R0E3 cells) showed expression of EGFRvIII protein by immunoblot analysis. Moreover, immunoblot analysis confirmed Y1068-EGFRvIII phosphorylation, as well as specific myc-Rictor transgene expression only in GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP mice (see Supplementary Figure 2 and Figures 5A&B). As shown in Figure 4A, GFAP-EGFRvIII mice (hemizygous and homozygous) were healthy and did not demonstrate any clinical abnormalities as previously described . In homozygous GFAP-Cre/RictorloxP/loxP mice 50% of animals developed tumors consistent with Grade I–II histopathology by 12–14 weeks and with 75% of animals succumbing to gliomas by 22 weeks. Transgenic GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP mice exhibited a marked reduction in overall survival with 50% of animals developing gliomas by 4–6 weeks and all of the mice succumbing by 18 weeks. The histopathology of the tumors from GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP mice was more advanced and consistent with higher-grade mixed astrocytic-oligodendroglial lineage (Grade II–III, Figure 4B&C) as compared to tumors from the GFAP-Cre/RictorloxP/loxP mice (Figure 2A&B). GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP glial tumors were strongly APC and GalC positive by immunohistochemical staining (Figure 4D&E) and to a lesser extent GFAP positive consistent with a mixed astrocytic-oligodendroglial tumor. These tumors also displayed elevated mTORC2 signaling and stained positive for phospho-T346-NDRG1, phospho-S657-PKCα, phospho-S473-AKT shown in Figures 4G–I, respectively. Moreover, as shown in Figure 4J, quantification of the immunohistochemical staining results comparing relative mTORC2 marker levels in tumors from GFAP-EGFRvIII; GFAP-Cre/RictorloxP/lox mice to those from GFAP-Cre/RictorloxP/loxP mice demonstrated significantly greater numbers of immuno-positive cells for all three markers in GFAP-EGFRvIII; GFAP-Cre/RictorloxP/lox tumors. We also examined the in vitro signaling and oncogenic properties of the derived tumors cell lines from both GFAP-Cre/RictorloxP/loxP and GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP glial tumors. As shown in Figure 5B, while the total amount of myc-Rictor expressed from the transgene was comparable in the Ric0 and R0E3 cells, the mTORC2 markers phospho-T346-NDRG1, phospho-S657-PKCα, phospho-S473-AKT, phospho-S32/36-IkBα demonstrated elevated levels in R0E3 cells. Total IkBa levels were also undetectable in Ric0 or R0E3 cells consistent with its observed phosphorylation status. We also examined the relative growth, migration and invasive characteristics of the Ric0 and R0E3 lines in vitro. As shown in Figure 5C, R0E3 cells grew approximately twice as fast as Ric0 cells as determined by XTT proliferation assays. To assess whether cell migration was also enhanced in R0E3 cells we compared the ability of cells to traverse either a vitronectin-coated or fibronectin-coated Boyden chamber utilizing bovine serum albumin (BSA)-coated chambers as a control. We additionally tested whether R0E3 cells showed increased invasive properties as determined by their ability to invade Matrigel. As shown in Figures 5D&E, R0E3 cells derived from GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP glial tumors demonstrated increased motility as compared to Ric0, GFAP-Cre/RictorloxP/loxP -derived glial tumors. To demonstrate that both the Ric0 and R0E3 were neoplastic and not dysplastic cells, we also determined whether they would form tumors in xenografted SCID mice. Tumors developed with both lines in recipient mice within 2–3 weeks (data not shown) with histological characteristics similar to the original gliomas from the respective transgenic mice (see supplementary Figure 3). These data taken together strongly suggest a cooperative interaction between oncogenic EGFRvIII and Rictor in vivo to promote gliomagenesis via increases in mTORC2 activity.
In (A), wild-type mice had 100% tumor-free survival in these experiments (not shown), as did GFAP-EGFRvIII transgenic mice. Twenty-six transgene-positive mice were followed for twenty-two weeks (EGFRvIII). Twenty-nine transgene-positive GFAP-Cre/RictorloxP/loxP mice were followed for twenty-two weeks (Rictor). Transgene-positive GFAP-EGFRvIII and GFAP-Cre/RictorloxP/loxP mice were crossed and a cohort of twenty-four animals were followed for eighteen weeks (EGFRvIII x Rictor). (B&C) Increased grade of oligodendroglioma in GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP mice. H & E sections of a typical tumor displayed increased regions of necrosis and cells with pyknotic nuclei. Increased endothelial proliferation was also observed consistent with anaplastic oligodendroglioma in humans (grade 3 WHO). (D–F) APC, GalC and GFAP immunoreactive cells from representative oligodendroglial tumors. (G–I) mTORC2 markers displayed elevated immunohistochemical staining in sections of gliomas from GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP mice. In B–I, scale bar; 20 µm. (J) Quantification of mTORC2 signaling markers in gliomas from GFAP-Cre/RictorloxP/loxP and GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP mice. A representative count in triplicate out of three individual counts with similar results is shown with + S.D. *, P<0.05.
(A) Immunoblot analysis for myc-tagged Rictor in oligodendrocytes obtained from control littermates, GFAP-EGFRvIII, or GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP mice. (B) Cell line established from GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP oligodendroglioma (R0E3) displays elevated mTORC2 signaling in comparison to Ric0 cells. Lysates of normal mature oligodendrocytes (NMO), Ric0 and R0E3 cells were immunoblotted for the indicated proteins. (C) XTT proliferation assay of established normal mature oligodendrocytes (squares), transgenic Ric0 (circles) and R0E3 (triangles) cells. (D) Motility of NMO, Ric0 and R0E3 cells in Boyden chamber assays migrating towards, BSA (white bars), vitronectin (grey bars) or fibronectin (black bars), (*, P<0.05) (E) Invasive potential of NMO, Ric0 and R0E3 cells in matrigel invasion assay (*, P<0.05). Counts were normalized to the viable cell count to exclude any effects of increased cell number on transmigration. The data shown in (D&E) are the mean + S.D. of three independent experiments.
GFAP-Cre/RictorloxP/loxP and GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP mice display marked subventricular zone expansion
There are examples demonstrating the ability of several mutated receptor tyrosine kinases, as well as tumor suppressor loss, to converge on the mTOR pathway to enhance gliomagenesis in vivo –. Moreover, recent studies demonstrate that in the Nf1-loss model of glioma in mice, mTORC2 activity promotes neural stem cell proliferation and expansion in a manner dependent on Rictor expression . The most extensively characterized neural stem cell compartment resides in the subventricular zone (SVZ) –, thus we determined whether this region displayed alterations in the transgenic mice. In both GFAP-Cre/RictorloxP/loxP and GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP mice the subventricular zone was markedly thickened, with mean thickness increases of approximately 2- and 3-fold, respectively, as compared to age-matched control transgene-negative littermates (Figure 6A&B). Most of the expanded cells were morphologically similar with oval to angular nuclei, containing little cytoplasm and were not immunoreactive for GFAP, however were TuJ1 positive (early neuronal differentiation marker) (Figure 6C, center panel). Apoptotic bodies were observable, as well as mitotic figures (Figure 6C, left panel). We also noted a second cell type that was larger, contained increased cytoplasm and many of these cells stained immuno-positive for GFAP (Figure 6C, right panel). Cells within the expanded SVZ stained positive for Ki-67 and phospho-S473-AKT in both GFAP-Cre/RictorloxP/loxP and GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP mice (Figure 6D) and a greater number of SVZ phospho-S473-AKT immunopositive cells were observed in the GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP mice compared with age-matched GFAP-Cre/RictorloxP/loxP mice (Figure 6E). These data suggest that in vivo expression of Rictor alone is sufficient to result in marked expansion of the SVZ and correlates with elevated mTORC2 activity in these cells. Furthermore, cooperative interactions between Rictor and oncogenic EGFRvIII signaling can potentiate SVZ cell expansion consistent with increased numbers of cells containing activated mTORC2.
(A) Parasagittal sections of the SVZ from control, GFAP-Cre/RictorloxP/loxP (Rictor) and GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP (Rictor x EGFRvIII) mice. Arrows show the SVZ/rostral migratory stream in mature control mouse brain; ST, striatum; CC, corpus callosum; V; ventricle. Scale bar, 100 µm. (B) The SVZ was measured at its single widest point (left panel, representative measurement; arrow) and quantified. Five mice of each strain were examined. (C) Expanded SVZ in GFAP-Cre/RictorloxP/loxP mice are composed of distinct cell populations. Representative adjacent sections stained with H&E, immunohistochemically with anti-TuJ1 and anti-GFAP antibodies as indicated. Similar results were observed in SVZ sections in GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP mice. (D) Representative SVZ sections stained immunohistochemically with anti-Ki-67 and anti-phospho-S473-AKT antibodies from GFAP-Cre/RictorloxP/loxP (Rictor) and GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP (Rictor x EGFRvIII) mice as indicated. In B-D, scale bar; 20 µm. (E) Quantification of phospho-S473-AKT immunopositive cell numbers in the SVZ of control, GFAP-Cre/RictorloxP/loxP (Rictor) and GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP (Rictor x EGFRvIII) mice. A representative count in triplicate out of three individual counts with similar results is shown with + S.D.
Several studies have described the overexpression of Rictor in a variety of cancers including gliomas , –, as well as, its ability to function as a haploinsufficiency gene protecting Pten+/− heterozygous mice from prostate cancer . Our previous study demonstrated that Rictor overexpression promoted the oncogenic properties of GBM lines and as a prelude to formal transgenic studies we also examined whether overexpression of Rictor was sufficient to induce carcinogenesis in the classic NIH3T3-focus forming assay (see supplementary Figure 1A). Rictor overexpression induced focus formation and additionally cooperated with activated H-rasV12 to increase transformed focus formation in Ink4a/Arf-deficient MEFs (supplementary Figure 1B). These observations provided the impetus for us to examine whether Rictor overexpression would induce gliomagenesis in the mouse. In this report, we describe the generation and characterization of transgenic mice with conditional overexpression of Rictor in astroglial cells resulting in multifocal-intermediate and low-grade glioma. Moreover, we demonstrate that mice derived from crosses between these overexpressors and those carrying a GFAP-driven constitutively active EGFR variant transgene harbor more aggressive, higher-grade gliomas consistent with hyperactivated mTORC2 signaling.
GFAP-Cre/RictorloxP/loxP mice developed gliomas reminiscent of human oligodendroglioma staining immunopositive for APC and GalC. We also observed regions of satellitosis and more rarely, subpial accumulation of tumor cells often observed in oligodendrogliomas. Tumor cells from this strain exhibited immunoreactivity to known mTORC2 substrates demonstrating increased mTORC2 activity as a direct result of Rictor overexpression. These data are consistent with the notion that increased Rictor levels are sufficient to lead to elevated nucleation of mTORC2 resulting in a larger pool of supracomplexes which are competent to transduce signal. mTOR immunoprecipitations performed on Ric0 cells support this hypothesis (supplementary figure 3B). Additionally, we observed increased mTORC2-independent signaling effects of Rictor overexpression in gliomas from GFAP-Cre/RictorloxP/loxP mice and in cell lines established from these tumors (see Figure 3). These data suggest that in addition to stimulating mTORC2, Rictor may also impinge on other signaling pathways. One such pathway may be downstream activation of the Rho family of GTPases . The S. cerevisiae homolog of Rictor, AVO3 is known to contain a RasGEFN domain . This domain exists N-terminal to the catalytic GDP/GTP exchange domain in some guanine nucleotide exchange factors such as the Ras-like small GTPases . Thus, it is possible that Rictor may mediate activation of GTPase signaling.
We observed significant expansion of the SVZ in GFAP-Cre/RictorloxP/loxP mice and this was even more pronounced in the GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP transgenic strain. Our data suggested that both GFAP and TuJ1-expressing cell populations within the SVZ from both transgenic strains were present in this compartment and likely participate in the observed proliferation. Current data suggests that type B cells within the SVZ (GFAP positive cells) are the actual primary progenitors within this compartment which give rise to actively proliferating transit amplifying, which give rise to immature neuroblasts (TuJ1 positive) . The expansion of the SVZ in our transgenic strains is similar to those observed in transgenic mice expressing GFAP-Cre+/K-rasG12D as well as in wild-type mice receiving intraventricular infusions of EGF or PDGF –. Our data suggest that elevated Rictor levels alone or in cooperation with upstream EGFRvIII overexpression leads to proliferation in the SVZ and may constitute an originating stem cell in these transgenic strains, however we cannot rule out contributions from other germinal regions of the brain.
Our results are also consistent with the observation that AKT activity is critical initiator of gliomagenesis. Previous reports have demonstrated that pharmacological inhibition of the PI3K/AKT axis reduces the number of GFAP-expressing cells in the olfactory bulb . Additionally, abrogating PI3K/AKT signaling alters the subcellular localization of Olig2 which is required for astrocyte differentiation . The mTOR complex containing Rictor positively regulates AKT activity via phosphorylation on serine 473 and recent data suggests that variable Rictor expression levels may also regulate brain region specific stem cell proliferation and gliomagenesis in a manner dependent on selective AKT hyperactivity .
In summary, our results demonstrate that Rictor overexpression alone is sufficient to produce gliomas which resemble human oligodendroglial tumors in the mouse. Moreover, oncogenic EGFRvIII signaling cooperates with Rictor overexpression to induce more aggressive mixed astrocytic-oligodendroglial tumors. GFAP-Cre/RictorloxP/loxP tumors harbor elevated mTORC2 activity which is further induced in tumors from GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP mice. Finally, hyperactive mTORC2 signaling in cells of the SVZ results in marked expansion of this stem cell compartment suggesting that these cells may constitute the glioma cell of origin in these models. These models may provide future insights into the mechanistic basis of gliomagenesis and resistance to chemotherapeutic intervention.
Materials and Methods
Vector Construction and ES cell transgenics
The human myc-tagged Rictor was obtained from Dr. David Sabatini (MIT, Cambridge, MA) and cloned into the shuttling vector pBigT and subsequently subcloned into ROSA26PA (kind gift of Dr. Frank Costantini, Columbia University, NY) . 129ES cells were electroporated and grown without feeders under selection in 300 mg/ml G418 for 7 days. 129ES cells (Dr. Hong Wu, UCLA) were derived from 129/Sv mice using standard methods . Clones were isolated and screened by genomic PCR and Southern hybridization. The 5′ UTR probe used detects an 11 kb wild-type band and a 3.8 kb targeted band.
Mice and Genotyping
Rosa26hRictorloxP/wt mice were generated following a protocol similar to that described previously . Mice were bred on a mixed 129/Sv-FVB/N genetic background. GFAP-EGFRvIII mice, originally on an outbred ICR background were backcrossed at least eight generations into FVB/N inbred mice. Genotyping of the GFAP-Cre+ [FVB-Tg(GFAP-Cre)25Mes/J; The Jackson Laboratory] and GFAP-EGFRvIII mice was performed as previously described , . Genotyping of Rosa26hRictorloxP/wt mice was as described for Gt(ROSA)26Sortm1Sor with minor modification . The primers used were as follows: RF216, 5′- GTGGTGGTGGGTCGACGATGG -3′; RR39, 5′-AAGCGCTCGTAGCCCTGCTG -3′. All protocols conformed to the guidelines established by the Association for the Assessment and Accreditation of Laboratory Animal Care and were approved by the UCLA-VA Animal Care and Use Committees under protocol numbers 0010-11036 and 0009-09029.
Histology and Immunohistochemistry
Tissue for histological analysis was immersion-fixed in buffered formalin, paraffin-embedded, sectioned and stained with H&E according to established protocols. Immunohistochemical analysis was performed according to standard procedures with antigen retrieval performed as previously described . Samples were blocked in either 20% goat serum or 10% rabbit serum in PBS, followed by overnight incubation in primary antibody at 4°C. The primary antibodies used were as follows: rabbit anti-GFAP (Dako, Carpinteria, CA), rabbit anti-GalC (Proteintech, Chicago, IL), Anti-APC (Ab-7, EMD Millipore, Billerica, MA), phospho-T346- NDRG1, phospho-S657-PCKα, phospho-S473-AKT (Cell Signaling, Danvers, MA), rabbit anti-phospho-Y118-paxillin (GenWay, San Diego, CA), Ki-67 (Vector Laboratories, Burlingame, CA), TUJ1 (Covance, Vienna, VA). Quantification of immunopositive cell numbers was performed on three representative sections from three mice for each genotype.
Establishment of primary cells from transgenic mice and growth as xenografts
Oligodendrocytes were purified from postnatal mice as described . Lines were maintained in DMEM supplemented with 10% FCS (Omega Scientific, Tarzana, CA). 1×106 cells were injected in the flanks of SCID (JAX Laboratories, Bar Harbor, ME) mice and growth assessed by direct caliper measurements.
Total protein from whole brain homogenates or cultured lines were extracted and subjected to SDS-PAGE, followed by transfer to PVDF membranes, blocked with 2.5% nonfat milk and incubated with primary antibody overnight at 4°C. The following primary antibodies were used in addition to those described above: Myc tag clone 9E10 (Millipore), Rictor, mTOR (Bethyl Laboratories, Montgomery, TX), phospho-S32/36-IkBα, IkBα, paxillin, phospho-Y1068-EGFR, EGFR (Cell Signaling), EGFR/EGFRvIII cocktail antibody (Upstate), anti-Myo1C (kind gift from Dr. Peter Gillespie, Vollum Institute, Oregon Health & Science University). Detection was performed using Amersham ECL (GE Healthcare, Piscataway, NJ). Immunoprecipitations were performed as previously described , utilizing normal mouse IgG (Millipore) as a non-specific control as indicated.
Cell proliferation, clonogenic, cell migration and colony-forming assays
XTT assays (Roche, Indianapolis, IN) were used to assess cell proliferation in vitro. Cells were plated into 96-well plates at 1,000 per well and incubated for various time points. Cell numbers were subsequently determined by the addition of 2,3-bis[2-methoxy-4-nitro-5-sulfophenyl]-2H-tetrazolium-5-carboxanilide inner salt and assaying relative absorbance at 490 nm as described by the manufacturer. NIH3T3 and Ink4a/Arf-null MEF focus-forming assays were performed by plating 1×103 cells per well in 24-well plates in a volume of 400 µL using a two-layered soft agar system as previously described . The H-rasV12, Rictor and Myc expression vectors utilized in these assays have been previously described , , . Cell migration assays were conducted using precoated modified Boyden chambers from Chemicon as recommended by the manufacturer and as previously described . Invasion assays through Matrigel were performed by loading 2×105 cells in the top well of Boyden chambers that contained growth factor-reduced Matrigel over a polyethylene terephthalate membrane with 8-mm pores (BD Biosciences, Franklin Lakes, NJ). Cells were allowed to invade and Matrigel removed prior to fixation and staining. Cells adhering to the bottom surface of the membrane were counted.
Parasagittal sections from GFAP-Cre/RictorloxP/loxP and GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP mice, as well as, age matched littermate controls were aligned according to anatomical position based on hippocampal configuration and size. In this orientation a horizontal segment of the SVZ is displayed subjacent to the corpus callosum and an oblique segment connecting to a vertical section. The width of the SVZ at its widest point in the vertical section was measured perpendicular to the ependymal layer at x400 using a Nikon Optiphot-2® microscope equipped with a Optronics Microfire® digital camera. Optronics PictureFrame® imaging processing suite was used containing a calibrated scale bar for measurements. Mean + S.D. are shown and relative significant differences were determined using Student's t test.
Rictor overexpression results in increased carcinogenesis. (A) NIH3T3 focus forming assay in which cells were transfected with empty expression vector (vec), Rictor or c-myc containing expression vector DNA. (B) Cooperative increase in foci formation between Rictor and H-rasV12 in Ink4a/Arf-deficient MEFs. Data shown are mean + S.D. of three independent experiments.
(A) PCR amplification from tail clipped DNA showing the appropriate size bands for both transgenes of interest (GFAP-EGFRvIII and GFAP-Cre/RictorloxP/loxP) in GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP transgenic mice. (B) Immunoblot analysis comparing EGFRvIII-Y1068-phosphorylation, Rictor and Myo1C expression in NMO, Ric0 and R0E3 cells.
(A) H & E stained sections of xenografted Ric0 and R0E3 tumors from SCID mice. 1×106 cells were injected into the flanks of recipient mice and tumors harvested and sectioned for histological analysis. Arrows show oligodendroglial morphology of transplanted glioma cells derived from GFAP-Cre/RictorloxP/loxP and GFAP-EGFRvIII; GFAP-Cre/RictorloxP/loxP transgenic mice. Scale bar, 20 µm. (B) Cell extracts from NMO or Ric0 cells were immunoprecipitated with anti-mTOR antibodies and precipitates were immunobloted for the indicated proteins. Rictor demonstrates increased association with mTOR in Ric0 cells as compared to NMOs.
We thank Rong Qian and Drs. Frank Costantini, Peter Gillespie, Meisheng Jiang and Xin Liu for reagents and assistance with ES cell transgenics. We are also grateful to Dr. Denson Fujikawa for assistance with tissue processing, embedding and sectioning. We acknowledge Lindsay Cooper and Jonathan Li for excellent technical assistance. Dr. Abhijit Guha succumbed to leukemia November 8, 2011 and this work is dedicated to his memory.
Conceived and designed the experiments: TB AG JG. Performed the experiments: TB CC NA LA AB BH ABS. Analyzed the data: TB RNN JG. Contributed reagents/materials/analysis tools: NS AG. Wrote the paper: TB RNN JG.
- 1. Stupp R, Mason WP, van den Bent MJ, Weller M, Fisher B, et al. (2005) Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med 352: 987–996.
- 2. Hurtt MR, Moossy J, Donovan-Peluso M, Locker J (1992) Amplification of epidermal growth factor receptor gene in gliomas: histopathology and prognosis. J Neuropathol Exp Neurol 51: 84–90.
- 3. Jaros E, Perry RH, Adam L, Kelly PJ, Crawford PJ, et al. (1992) Prognostic implications of p53 protein, epidermal growth factor receptor, and Ki-67 labelling in brain tumours. Br J Cancer 66: 373–385.
- 4. Schlegel J, Stumm G, Brandle K, Merdes A, Mechtersheimer G, et al. (1994) Amplification and differential expression of members of the erbB-gene family in human glioblastoma. J Neurooncol 22: 201–207.
- 5. Heimberger AB, Hlatky R, Suki D, Yang D, Weinberg J, et al. (2005) Prognostic effect of epidermal growth factor receptor and EGFRvIII in glioblastoma multiforme patients. Clin Cancer Res 11: 1462–1466.
- 6. Shinojima N, Tada K, Shiraishi S, Kamiryo T, Kochi M, et al. (2003) Prognostic value of epidermal growth factor receptor in patients with glioblastoma multiforme. Cancer Res 63: 6962–6970.
- 7. Yuan TL, Cantley LC (2008) PI3K pathway alterations in cancer: variations on a theme. Oncogene 27: 5497–5510.
- 8. Dunn GP, Rinne ML, Wykosky J, Genovese G, Quayle SN, et al. (2012) Emerging insights into the molecular and cellular basis of glioblastoma. Genes Dev 26: 756–784.
- 9. Laplante M, Sabatini DM (2012) mTOR signaling in growth control and disease. Cell 149: 274–293.
- 10. Ma XM, Blenis J (2009) Molecular mechanisms of mTOR-mediated translational control. Nat Rev Mol Cell Biol 10: 307–318.
- 11. Tanaka K, Babic I, Nathanson D, Akhavan D, Guo D, et al. (2011) Oncogenic EGFR Signaling Activates an mTORC2-NF-kappaB Pathway That Promotes Chemotherapy Resistance. Cancer Discov 1: 524–538.
- 12. Guertin DA, Sabatini DM (2007) Defining the role of mTOR in cancer. Cancer Cell 12: 9–22.
- 13. Laplante M, Sabatini DM (2012) mTOR Signaling. Cold Spring Harb Perspect Biol 4.
- 14. Sarbassov DD, Ali SM, Kim DH, Guertin DA, Latek RR, et al. (2004) Rictor, a novel binding partner of mTOR, defines a rapamycin-insensitive and raptor-independent pathway that regulates the cytoskeleton. Curr Biol 14: 1296–1302.
- 15. Sarbassov DD, Guertin DA, Ali SM, Sabatini DM (2005) Phosphorylation and regulation of Akt/PKB by the rictor-mTOR complex. Science 307: 1098–1101.
- 16. Sarbassov DD, Ali SM, Sengupta S, Sheen JH, Hsu PP, et al. (2006) Prolonged rapamycin treatment inhibits mTORC2 assembly and Akt/PKB. Mol Cell 22: 159–168.
- 17. Loewith R, Jacinto E, Wullschleger S, Lorberg A, Crespo JL, et al. (2002) Two TOR complexes, only one of which is rapamycin sensitive, have distinct roles in cell growth control. Mol Cell 10: 457–468.
- 18. Jacinto E, Loewith R, Schmidt A, Lin S, Ruegg MA, et al. (2004) Mammalian TOR complex 2 controls the actin cytoskeleton and is rapamycin insensitive. Nat Cell Biol 6: 1122–1128.
- 19. Hagan GN, Lin Y, Magnuson MA, Avruch J, Czech MP (2008) A Rictor-Myo1c complex participates in dynamic cortical actin events in 3T3-L1 adipocytes. Mol Cell Biol 28: 4215–4226.
- 20. McDonald PC, Oloumi A, Mills J, Dobreva I, Maidan M, et al. (2008) Rictor and integrin-linked kinase interact and regulate Akt phosphorylation and cancer cell survival. Cancer Res 68: 1618–1624.
- 21. Martin J, Masri J, Bernath A, Nishimura RN, Gera J (2008) Hsp70 associates with Rictor and is required for mTORC2 formation and activity. Biochem Biophys Res Commun 372: 578–583.
- 22. Zhang F, Zhang X, Li M, Chen P, Zhang B, et al. (2010) mTOR complex component Rictor interacts with PKCzeta and regulates cancer cell metastasis. Cancer Res 70: 9360–9370.
- 23. Gao D, Wan L, Inuzuka H, Berg AH, Tseng A, et al. (2010) Rictor forms a complex with Cullin-1 to promote SGK1 ubiquitination and destruction. Mol Cell 39: 797–808.
- 24. Masri J, Bernath A, Martin J, Jo OD, Vartanian R, et al. (2007) mTORC2 activity is elevated in gliomas and promotes growth and cell motility via overexpression of rictor. Cancer Res 67: 11712–11720.
- 25. Morris LG, Taylor BS, Bivona TG, Gong Y, Eng S, et al. (2011) Genomic dissection of the epidermal growth factor receptor (EGFR)/PI3K pathway reveals frequent deletion of the EGFR phosphatase PTPRS in head and neck cancers. Proc Natl Acad Sci U S A 108: 19024–19029.
- 26. Gibault L, Ferreira C, Perot G, Audebourg A, Chibon F, et al. (2011) From PTEN loss of expression to RICTOR role in smooth muscle differentiation: complex involvement of the mTOR pathway in leiomyosarcomas and pleomorphic sarcomas. Mod Pathol 25: 197–211.
- 27. Srinivas S, Watanabe T, Lin CS, William CM, Tanabe Y, et al. (2001) Cre reporter strains produced by targeted insertion of EYFP and ECFP into the ROSA26 locus. BMC Dev Biol 1: 4.
- 28. Zhuo L, Thesis M, Alvarez-Maya I, Brenner M, Willecke K, et al. (2001) hGFAP-cre transgenic mice for manipulation of glial and neuronal function in vivo. Genesis 31: 85–94.
- 29. Malatesta P, Hack MA, Hartfuss E, Kettenmann H, Klinkert W, et al. (2003) Neuronal or glial progeny: regional differences in radial glia fate. Neuron 37: 751–764.
- 30. Fortin D, Cairncross GJ, Hammond RR (1999) Oligodendroglioma: an appraisal of recent data pertaining to diagnosis and treatment. Neurosurgery 45: 1279–1291; discussion 1191.
- 31. Bromberg JE, van den Bent MJ (2009) Oligodendrogliomas: molecular biology and treatment. Oncologist 14: 155–163.
- 32. Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, Burger PC, et al. (2007) The 2007 WHO classification of tumours of the central nervous system. Acta Neuropathol 114: 97–109.
- 33. Turner CE (2000) Paxillin and focal adhesion signalling. Nat Cell Biol 2: E231–236.
- 34. Ding H, Shannon P, Lau N, Wu X, Roncari L, et al. (2003) Oligodendrogliomas result from the expression of an activated mutant epidermal growth factor receptor in a RAS transgenic mouse astrocytoma model. Cancer Res 63: 1106–1113.
- 35. CGARN (2008) Comprehensive genomic characterization defines human glioblastoma genes and core pathways. Nature 455: 1061–1068.
- 36. Fraser MM, Zhu X, Kwon CH, Uhlmann EJ, Gutmann DH, et al. (2004) Pten loss causes hypertrophy and increased proliferation of astrocytes in vivo. Cancer Res 64: 7773–7779.
- 37. Dasgupta B, Yi Y, Chen DY, Weber JD, Gutmann DH (2005) Proteomic analysis reveals hyperactivation of the mammalian target of rapamycin pathway in neurofibromatosis 1-associated human and mouse brain tumors. Cancer Res 65: 2755–2760.
- 38. Johannessen CM, Johnson BW, Williams SM, Chan AW, Reczek EE, et al. (2008) TORC1 is essential for NF1-associated malignancies. Curr Biol 18: 56–62.
- 39. Lee da Y, Yeh TH, Emnett RJ, White CR, Gutmann DH (2010) Neurofibromatosis-1 regulates neuroglial progenitor proliferation and glial differentiation in a brain region-specific manner. Genes Dev 24: 2317–2329.
- 40. Quinones-Hinojosa A, Sanai N, Soriano-Navarro M, Gonzalez-Perez O, Mirzadeh Z, et al. (2006) Cellular composition and cytoarchitecture of the adult human subventricular zone: a niche of neural stem cells. J Comp Neurol 494: 415–434.
- 41. Doetsch F, Caille I, Lim DA, Garcia-Verdugo JM, Alvarez-Buylla A (1999) Subventricular zone astrocytes are neural stem cells in the adult mammalian brain. Cell 97: 703–716.
- 42. Doetsch F, Garcia-Verdugo JM, Alvarez-Buylla A (1997) Cellular composition and three-dimensional organization of the subventricular germinal zone in the adult mammalian brain. J Neurosci 17: 5046–5061.
- 43. Gage FH (2000) Mammalian neural stem cells. Science 287: 1433–1438.
- 44. Guertin DA, Stevens DM, Saitoh M, Kinkel S, Crosby K, et al. (2009) mTOR complex 2 is required for the development of prostate cancer induced by Pten loss in mice. Cancer Cell 15: 148–159.
- 45. Bar-Sagi D, Hall A (2000) Ras and Rho GTPases: a family reunion. Cell 103: 227–238.
- 46. De Virgilio C, Loewith R (2006) Cell growth control: little eukaryotes make big contributions. Oncogene 25: 6392–6415.
- 47. Abel TW, Clark C, Bierie B, Chytil A, Aakre M, et al. (2009) GFAP-Cre-mediated activation of oncogenic K-ras results in expansion of the subventricular zone and infiltrating glioma. Mol Cancer Res 7: 645–653.
- 48. Doetsch F, Petreanu L, Caille I, Garcia-Verdugo JM, Alvarez-Buylla A (2002) EGF converts transit-amplifying neurogenic precursors in the adult brain into multipotent stem cells. Neuron 36: 1021–1034.
- 49. Jackson EL, Garcia-Verdugo JM, Gil-Perotin S, Roy M, Quinones-Hinojosa A, et al. (2006) PDGFR alpha-positive B cells are neural stem cells in the adult SVZ that form glioma-like growths in response to increased PDGF signaling. Neuron 51: 187–199.
- 50. Otaegi G, Yusta-Boyo MJ, Vergano-Vera E, Mendez-Gomez HR, Carrera AC, et al. (2006) Modulation of the PI 3-kinase-Akt signalling pathway by IGF-I and PTEN regulates the differentiation of neural stem/precursor cells. J Cell Sci 119: 2739–2748.
- 51. Setoguchi T, Kondo T (2004) Nuclear export of OLIG2 in neural stem cells is essential for ciliary neurotrophic factor-induced astrocyte differentiation. J Cell Biol 166: 963–968.
- 52. Nagy A, Gertsenstein M, Vintersten K, Behringer R (2003) Manipulating the mouse embryo: a laboratory manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press.
- 53. Soriano P (1999) Generalized lacZ expression with the ROSA26 Cre reporter strain. Nat Genet 21: 70–71.
- 54. Shannon P, Sabha N, Lau N, Kamnasaran D, Gutmann DH, et al. (2005) Pathological and molecular progression of astrocytomas in a GFAP:12 V-Ha-Ras mouse astrocytoma model. Am J Pathol 167: 859–867.
- 55. Suzumura A, Bhat S, Eccleston PA, Lisak RP, Silberberg DH (1984) The isolation and long-term culture of oligodendrocytes from newborn mouse brain. Brain Res 324: 379–383.
- 56. Chin L, Tam A, Pomerantz J, Wong M, Holash J, et al. (1999) Essential role for oncogenic Ras in tumour maintenance. Nature 400: 468–472.
- 57. Ricci MS, Jin Z, Dews M, Yu D, Thomas-Tikhonenko A, et al. (2004) Direct repression of FLIP expression by c-myc is a major determinant of TRAIL sensitivity. Mol Cell Biol 24: 8541–8555.