c-Myc is required for maintenance of glioma cancer stem cells

Abstract

Background: Malignant gliomas rank among the most lethal cancers. Gliomas display a striking cellular heterogeneity with a hierarchy of differentiation states. Recent studies support the existence of cancer stem cells in gliomas that are functionally defined by their capacity for extensive self-renewal and formation of secondary tumors that phenocopy the original tumors. As the c-Myc oncoprotein has recognized roles in normal stem cell biology, we hypothesized that c-Myc may contribute to cancer stem cell biology as these cells share characteristics with normal stem cells.

Methodology/principal findings: Based on previous methods that we and others have employed, tumor cell populations were enriched or depleted for cancer stem cells using the stem cell marker CD133 (Prominin-1). We characterized c-Myc expression in matched tumor cell populations using real time PCR, immunoblotting, immunofluorescence and flow cytometry. Here we report that c-Myc is highly expressed in glioma cancer stem cells relative to non-stem glioma cells. To interrogate the significance of c-Myc expression in glioma cancer stem cells, we targeted its expression using lentivirally transduced short hairpin RNA (shRNA). Knockdown of c-Myc in glioma cancer stem cells reduced proliferation with concomitant cell cycle arrest in the G(0)/G(1) phase and increased apoptosis. Non-stem glioma cells displayed limited dependence on c-Myc expression for survival and proliferation. Further, glioma cancer stem cells with decreased c-Myc levels failed to form neurospheres in vitro or tumors when xenotransplanted into the brains of immunocompromised mice.

Conclusions/significance: These findings support a central role of c-Myc in regulating proliferation and survival of glioma cancer stem cells. Targeting core stem cell pathways may offer improved therapeutic approaches for advanced cancers.

Figure 1. c-Myc is highly expressed in glioma cancer stem cells.

(A) CD133− and CD133+ cells were isolated from glioma surgical biopsy specimens passaged short-term in immunocompromised mice and briefly cultured. Total RNA was isolated from both CD133− and CD133+ cells. cDNA was prepared by reverse transcription. Expression of c-Myc was then determined by quantitative real-time PCR and normalized to β-actin and HPRT1. Relative mRNA levels of c-Myc in CD133− cells were assigned a value of 1. Data are represented as mean±S.E.M in this and all subsequent graphs (#: p<0.001). (B) Total cellular lysates were resolved by SDS-PAGE. Protein levels of c-Myc and Olig2 were determined by immunoblotting. Actin was blotted as the loading control. (C) Glioma cells were isolated directly from human surgical biopsy specimens, fixed in 4% paraformaldehyde following dissociation, labeled with anti-CD133-APC and anti-c-Myc-FITC, and subjected to FACS analysis. (D) Percentage of cells expressing high levels of c-Myc within either the CD133− fraction or the CD133+ fraction was demonstrated (#: p<0.001). (E) Sections of freshly frozen human glioma surgical biopsy specimens were fixed and co-stained for c-Myc (green) and Nestin (red). Nuclei were counterstained with Hoechst 33342. Representative images (630×) were demonstrated.

Figure 2. c-Myc modulates cell cycle regulators of glioma cancer stem cells.

(A) Early passage (WAF1/CIP1, cyclin D₁ (cycD1) and cyclin D₂ (cycD2) were determined by quantitative real-time PCR 3 days after infection. (B) Protein levels of c-Myc, p53, cyclin D₁, cyclin D₂, cyclin E and p21^WAF1/CIP1 were determined by immunoblotting.

Figure 3. c-Myc induces cell cycle arrest of glioma cancer stem cells.

(A, B) Early passage (0 phase; M2-G₀/G₁ phase; M3-S phase; M4-G2/M phase.

Figure 4. Depletion of c-Myc inhibits growth of glioma cancer stem cells.

(A) T3359, (B) T3832, and (C) T4597 CD133− and CD133+ cells were isolated and infected as described. Cells were then plated in 96-well plates in triplicate at 5000 cells per well for CD133− cells or 1000 cells per well for CD133+ cells. Total viable cell numbers were then determined by the CellTiter-Glo Luminescent Cell Viability Assay (Promega) daily. *, p<0.05; #, p<0.001 with one-way ANOVA comparison of the control groups to the corresponding c-Myc knockdown groups on the same day.

Figure 5. Depletion of c-Myc induces apoptosis in glioma cancer stem cells.

(A, B) T3359 and (C, D) T4597 CD133− and CD133+ cells were isolated and infected as described. (A, C) Six days after infection, cells were trypsinized and quantified for apoptosis using the Annexin V-FITC Apoptosis Detection kit (Calbiochem). The percentage of FITC positive cells was determined by FACS analysis, and dead cells were excluded by propidium iodide staining. (B, D) These cells were also plated in 96-well plates at 5000 cells per well for CD133− cells and 1000 cells per well for CD133+ cells after infection and selection. Six days after infection, combined activities of caspase 3 and caspase 7 were determined by the Caspase 3/7 Luminescence Assay kit (Promega), and were normalized by the viable cell numbers determined by the CellTiter-Glo assays as described in Figure 4. *, p<0.05 with one-way ANOVA comparison of the control groups to the corresponding c-Myc knockdown groups.

Figure 6. Depletion of c-Myc attenuates neurosphere formation by glioma cancer stem cells.

(A) T3359 CD133+ cells were infected with lentivirus and selected as described. Cells were then plated at 100 or 10 cells per well in 24-well plates in the presence of 1 µg/ml puromycin. Eight wells were plated for each group. The numbers of neurospheres in each well were determined in seven days. Spheres that contained more than 20 cells were scored. (B) Percentage of wells without neurospheres formed was quantified. #, p<0.001 with one-way ANOVA comparison of the control groups to the corresponding c-Myc knockdown groups. (C) Representative photographs of neurospheres formed by cells expressing non-targeting shRNA or c-Myc specific shRNA (bars = 50 µm). (D, E) Similar results were demonstrated in T4302 CD133+ cells.

Figure 7. c-Myc knockdown abolishes xenograft tumor formation by glioma cancer stem cells.

(A) T3359 CD133+ cells were infected and selected as described. After selection, cells were injected into brains of athymic BALB/c nu/nu mice (5000 cells per mouse). Four mice were injected for each group. Mice in the control group were sacrificed upon the development of neurologic signs. All the mice bearing c-Myc knockdown glioma cells did not develop neurologic signs and were sacrificed after 100 days without evidence of tumor. Kaplan-Meier survival curves are displayed. (B) Representative photographs of hematoxylin and eosin staining of intracranial xenograft tumors (10×). (C) Xenograft tissue of the control group composed of pleomorphic cells featuring high nuclear to cytoplasmic ratios, prominent nucleoli with minimal cytoplasm, brisk mitotic activity and central geographic necrosis (asterisk, 600×). (D) The control glioma xenograft exhibited focal areas of better differentiated tumor cells with relatively more eosinophilic cytoplasm and cells with eccentric cytoplasmic profiles suggestive of a gemistocytic appearance (arrows, 400×). (E) Xenograft tumor of the control group exhibits infiltration of tumor cells into the surrounding brain tissue along the margin. The mitotically active (arrowhead) infiltrating tumor cells exhibit high nuclear to cytoplasmic ratios and elongated fibrillar cytoplasm (arrow) (600×). (F) Mouse brain injected with T3359 cells expressing c-Myc shRNA showed no evidence of tumor at the needle injection site (arrows, 200×).