Glioblastoma-initiating cells: relationship with neural stem cells and the micro-environment

Abstract

Glioblastoma multiforme (GBM, WHO grade IV) is the most common and lethal subtype of primary brain tumor with a median overall survival of 15 months from the time of diagnosis. The presence in GBM of a cancer population displaying neural stem cell (NSC) properties as well as tumor-initiating abilities and resistance to current therapies suggests that these glioblastoma-initiating cells (GICs) play a central role in tumor development and are closely related to NSCs. However, it is nowadays still unclear whether GICs derive from NSCs, neural progenitor cells or differentiated cells such as astrocytes or oligodendrocytes. On the other hand, NSCs are located in specific regions of the adult brain called neurogenic niches that have been shown to control critical stem cell properties, to nourish NSCs and to support their self-renewal. This "seed-and-soil" relationship has also been adapted to cancer stem cell research as GICs also require a specific micro-environment to maintain their "stem cell" properties. In this review, we will discuss the controversies surrounding the origin and the identification of GBM stem cells and highlight the micro-environment impact on their biology.

Figure 1

Relationship between NSCs’ differentiation and cancer initiation. NSCs (purlple) are able to differentiate in glial (grey) and neural progenitors (yellow). Neural progenitors give rise to neurons (pink) whereas glial progenitors are committed to oligodendrocytes (red), ependymal cells (light blue) and astrocytes (dark blue). Brain cancer formation could arise from the transformation of NSCs into GICs (Brain Cancer Initiating Cells, light purple) which in turn give rise to a more differentiated cancer cell population (green). In the same line, glial progenitors could induce tumor development after malignant transformation of normal progenitor cells (Brain Tumor Progenitor-like Cells, orange). Traditionally, mature cells in the brain such as neurons, oligodendrocytes, ependymal cells and astrocytes are also seen as potential candidates involved in brain cancer tumorigenesis.

Figure 2

The tumor (SVZ) micro-environment. GICs (light purple) are regulated by the activation of oncogenes or by environmental conditions such as hypoxia or acidosis. GICs are able to express pro-angiogenic factors, such as VEGF (vascular endothelial growth factor), able to stimulate the outgrowth of new blood vessels (red) which will in the end help to maintain tumorigenicity. Endothelial cells are also able to produce nitric oxide in the endothelium through the activation of eNOS. Activated Notch signaling by nitric oxide in GICs is notably known to accelerate tumor progression. Defects in growth factor signaling pathways usually result in the activation of downstream growth factor signaling such as the PI3K/Akt signaling pathway which has been shown to be an important regulator of glioma cell survival and GICs’ radioresistant abilities. Neural progenitor cells (yellow) display antitumor effects by secreting tumor suppressors in the niche such as bone morphogenetic protein-7 (BMP7), inducing GICs’ differentiation. Astrocytes (dark blue) are closely associated with the vascular endothelium of the niche and were shown to secrete neurotrophic factors capable of driving the invasive properties of GBM cells (CXCL12, SIP, and GDNF). Astrocytes also express high levels AEG-1. Reports suggest that AEG-1 impacts on brain tumor invasion notably through the activation of MMP-2 and MMP-9.

Figure 3

Relationship between the adult neurogenic niches and glioblastoma-initiating cells. GICs (light purple) have been found to play an important role in GBM aggressiveness and the resistance of the tumor to current treatments, making them an attractive target for therapeutic interventions. However, the design of new GIC inhibitors has proven difficult because their definition is nowadays still confused by the twilight zone of their origin. Indeed, whether GICs are derived from NPCs (yellow) or differentiated/mature cells remains uncertain. In any case, the vascular niche offered by the brain SVZ might be of interest to maintain GICs’ features, even if they are not indigenous. Undoubtedly, particular characteristics of this specific niche and its localization away from the tumor mass make it a significant resource for tumor recurrence. It is expected that future therapies targeting GBM, GICs and their specific environment will help to increase the dramatic survival rate of GBM patients.