Downloaded from genesdev.cshlp.org on October 1, 2021 - Published by Cold Spring Harbor Laboratory Press REVIEW Cancer stem cells in glioblastoma Justin D. Lathia,1,2 Stephen C. Mack,3 Erin E. Mulkearns-Hubert,1 Claudia L.L. Valentim,3 and Jeremy N. Rich2,3 1Department of Cellular and Molecular Medicine, Lerner Research Institute, Cleveland Clinic, Cleveland, Ohio 44195, USA; 2Department of Molecular Medicine, Cleveland Clinic, Lerner College of Medicine of Case Western Reserve University, Cleveland, Ohio 44195, USA; 3Department of Stem Cell Biology and Regenerative Medicine, Lerner Research Institute, Cleveland Clinic, Cleveland, Ohio 44195, USA Tissues with defined cellular hierarchies in development tumors will develop resistance to overcome initially effec- and homeostasis give rise to tumors with cellular hierar- tive therapies that poison driving oncogenes. chies, suggesting that tumors recapitulate specific tissues Glioblastoma (GBM; World Health Organization grade and mimic their origins. Glioblastoma (GBM) is the IV glioma) is the most prevalent and lethal primary in- most prevalent and malignant primary brain tumor and trinsic brain tumor (Stupp et al. 2009). Unlike other solid contains self-renewing, tumorigenic cancer stem cells tumor cell types, GBM widely invades the surrounding (CSCs) that contribute to tumor initiation and thera- brain but rarely metastasizes to other organs. While halt- peutic resistance. As normal stem and progenitor cells ing steps to fight GBM are being made using targeted participate in tissue development and repair, these de- therapies (e.g., bevacizumab) or immunotherapies, GBM velopmental programs re-emerge in CSCs to support the therapy remains focused on achieving maximal surgi- development and progressive growth of tumors. Elucida- cal resection followed by concurrent radiation therapy tion of the molecular mechanisms that govern CSCs has with temozolomide (TMZ; an orally available methyla- informed the development of novel targeted therapeutics tion chemotherapy) and subsequent additional adjuvant for GBM and other brain cancers. CSCs are not self-auton- TMZ therapy. Conventional treatment offers patients omous units; rather, they function within an ecological with GBM additional survival time with generally accept- system, both actively remodeling the microenvironment able quality of life, but a cure is never achieved. GBM and receiving critical maintenance cues from their niches. represents one of the most comprehensively genomi- To fulfill the future goal of developing novel therapies to cally characterized cancer types, leading to recognition collapse CSC dynamics, drawing parallels to other normal of groups of tumors defined by transcription profiles (pro- and pathological states that are highly interactive with neural, neural, classical, and mesenchymal), genetics their microenvironments and that use developmental sig- (mutations of isocitrate dehydrogenase 1 [IDH1]), and epi- naling pathways will be beneficial. genetics (CpG island methylator phenotype [CIMP]) (Weller et al. 2013). Long-term survivors are often, but not exclusively, patients with tumors harboring muta- tions in IDH1, which likely represent a different disease Development is a coordinated summation of the individ- than most GBMs. Beyond IDH1 mutations and a few ual cellular dynamics that build an organ, and the pro- other biomarkers (deletion of chromosomes 1p and 19q grams responsible for this construction are generally in oligodendrogliomas, methyl guanine methyltransfer- preserved in stem cells for organ homeostasis and tissue ase [MGMT] promoter methylation, etc.), the accumulat- repair. Tumors are complex systems that recapitulate the ed genetic characterization of GBMs has failed to impact complexity of organs or tissues with dynamic regulation clinical practice, suggesting that other discovery para- and constituent cellular populations during tumor initia- digms should also be considered. tion, maintenance, and progression (Hanahan and Wein- The brain, like other organs with clearly defined cellu- berg 2011). While many scientists have sought to reduce lar hierarchies in development and homeostasis (e.g., — the complexity of cancer to a one-dimensional process blood, breast, skin, and colon), gives rise to tumors with for example, characterizing cancers solely based on genet- defined cellular hierarchies, suggesting that cancer repli- — ics most advanced cancers unfortunately remain nearly cates ontogeny (Reya et al. 2001). Atop the apex of cellular as lethal since the declaration of the War on Cancer in hierarchies are stem cells, which have been assumed to 1971. Targeted therapeutics offer a transient benefit for be rare, quiescent, self-renewing cells, but several highly some cancer types with driving mutations, but even these © 2015 Lathia et al. This article is distributed exclusively by Cold Spring [Keywords: brain tumor; cancer stem cell; glioblastoma; glioma; stem Harbor Laboratory Press for the first six months after the full-issue cell; tumor-initiating cell] publication date (see http://genesdev.cshlp.org/site/misc/terms.xhtml). Corresponding author: [email protected] After six months, it is available under a Creative Commons License Article is online at http://www.genesdev.org/cgi/doi/10.1101/gad.261982. (Attribution-NonCommercial 4.0 International), as described at http:// 115. creativecommons.org/licenses/by-nc/4.0/. GENES & DEVELOPMENT 29:1203–1217 Published by Cold Spring Harbor Laboratory Press; ISSN 0890-9369/15; www.genesdev.org 1203 Downloaded from genesdev.cshlp.org on October 1, 2021 - Published by Cold Spring Harbor Laboratory Press Lathia et al. proliferative organs (e.g., the intestine and skin) contain hierarchies, from their differentiated progeny remains at least two pools of stem cells: one quiescent, and the challenging; however, stem cell biology faces a similar dif- other highly proliferative (Barker et al. 2010). Stem cells ficulty with normal stem cell identification. Of note, the generate transient amplifying cells, which in turn create CSC hypothesis does not claim a stem cell as the cell of or- lineage-restricted progeny that are eventually fated to be- igin for cancers, suggesting that CSCs do not need to ad- come the terminally differentiated effector cells. here to all of the observed features of normal stem cells. Neural stem and progenitor cell (NSPC) pools vary in lo- In the following sections, we provide an update on in- cation and control during development, suggesting that trinsic and extrinsic regulators of the CSC state in GBM different cellular hierarchies may be co-opted by brain tu- and discuss how the integration of genetics, epigenetics, mors (Gibson et al. 2010; Lottaz et al. 2010). Informed and metabolism has shaped our understanding of how by techniques used to enrich and characterize NSPCs, CSCs function to drive GBM growth. We also highlight fu- several groups in parallel demonstrated that gliomas and ture opportunities to further understand the complexity other primary brain tumors contain self-renewing, tumor- of CSC regulation through interaction with other cells (in- igenic cells (Ignatova et al. 2002; Hemmati et al. 2003; cluding immune cells) and how the translation of CSC- Singh et al. 2003, 2004; Galli et al. 2004). The nomencla- based therapies needs to take into account the cellular dy- ture for these cells has been controversial (as discussed be- namics of CSCs, which rely on developmental signaling low), with the dominant choice being cancer stem cells programs. (CSCs) or tumor-initiating cells (importantly, these terms are not identical, as a CSC designation is more restrictive The history of the CSCs but also more informative) (Table 1). Unlike normal brain organization—where the generation of differentiated prog- The heterogeneity of tumor cells has long been appreci- eny is stage-specific (neurons and then glia during devel- ated, but two decades ago, seminal work from Dick’s labo- opment) and derived from rapidly dividing progenitor ratory (Bonnet and Dick 1997) described the isolation of a cells and quiescent, multipotent stem cells that persist leukemia-initiating cell, the first purification of cancer into adulthood and become activated upon differentiation stem-like cells, a population of cells that had originally (Rowitch and Kriegstein 2010)—these populations have been proposed to exist >150 years earlier (Sell 2004). The yet to be delineated in brain tumors. The ability to distin- first prospective isolation of human NSPCs was performed guish between self-renewing cells with stem and progeni- using CD133 (Uchida et al. 2000) and prompted a search tor cell cycle properties and transcriptional signatures is for brain tumor cells that shared the characteristics of likely to provide clarity with respect to nomenclature NSPCs. A burst of studies soon followed describing brain and the functional interplay between cells at the apex of CSCs in anaplastic astrocytoma (Ignatova et al. 2002), me- the hierarchy. The challenges distinguishing CSCs from dulloblastoma, pilocytic astrocytoma, ependymoma, gan- their progeny are derived, in part, from the limited recogni- glioglioma (Hemmati et al. 2003; Singh et al. 2003), and tion of points of relative stability (“attractor states”) in the GBM (Ignatova et al. 2002; Hemmati et al. 2003; Galli landscape of cellular identity that define the stem cell state et al. 2004). Brain CSCs have subsequently been shown and transitions into (dedifferentiation) and out of (differen- to be