REVIEW published: 27 June 2022 doi: 10.3389/fnmol.2022.903115 Converging Mechanisms of Epileptogenesis and Their Insight in Glioblastoma Kate E. Hills 1† , Kostas Kostarelos 1,2† and Robert C. Wykes 1,3 *† 1 Nanomedicine Lab, Faculty of Biology, Medicine and Health, University of Manchester, Manchester, United Kingdom, 2 Catalan Institute for Nanoscience and Nanotechnology (ICN2), Edifici ICN2, Campus UAB, Barcelona, Spain, 3 Department of Clinical and Experimental Epilepsy, UCL Queen Square Institute of Neurology, London, United Kingdom Glioblastoma (GBM) is the most common and advanced form of primary malignant tumor occurring in the adult central nervous system, and it is frequently associated with Edited by: epilepsy, a debilitating comorbidity. Seizures are observed both pre- and post-surgical Detlev Boison, resection, indicating that several pathophysiological mechanisms are shared but also Rutgers, The State University of New Jersey, United States prompting questions about how the process of epileptogenesis evolves throughout GBM progression. Molecular mutations commonly seen in primary GBM, i.e., in PTEN and Reviewed by: Chuanzuo Yang, p53, and their associated downstream effects are known to influence seizure likelihood. Zhengzhou University, China Similarly, various intratumoral mechanisms, such as GBM-induced blood-brain barrier Shaunik Sharma, The University of Iowa, United States breakdown and glioma-immune cell interactions within the tumor microenvironment *Correspondence: are also cited as contributing to network hyperexcitability. Substantial alterations to Robert C. Wykes peri-tumoral glutamate and chloride transporter expressions, as well as widespread rob.wykes@manchester.ac.uk dysregulation of GABAergic signaling are known to confer increased epileptogenicity † ORCID: and excitotoxicity. The abnormal characteristics of GBM alter neuronal network function Kate E. Hills to result in metabolically vulnerable and hyperexcitable peri-tumoral tissue, properties the orcid.org/0000-0003-2685-3100 Kostas Kostarelos tumor then exploits to favor its own growth even post-resection. It is evident that there is orcid.org/0000-0002-2224-6672 a complex, dynamic interplay between GBM and epilepsy that promotes the progression Robert C. Wykes orcid.org/0000-0002-6141-6822 of both pathologies. This interaction is only more complicated by the concomitant Specialty section: presence of spreading depolarization (SD). The spontaneous, high-frequency nature of This article was submitted to GBM-associated epileptiform activity and SD-associated direct current (DC) shifts require Brain Disease Mechanisms, a section of the journal technologies capable of recording brain signals over a wide bandwidth, presenting major Frontiers in Molecular Neuroscience challenges for comprehensive electrophysiological investigations. This review will initially provide a detailed examination of the underlying mechanisms that promote network Received: 23 March 2022 Accepted: 25 May 2022 hyperexcitability in GBM. We will then discuss how an investigation of these pathologies Published: 27 June 2022 from a network level, and utilization of novel electrophysiological tools, will yield a more- Citation: effective, clinically-relevant understanding of GBM-related epileptogenesis. Further to Hills KE, Kostarelos K and Wykes RC this, we will evaluate the clinical relevance of current preclinical research and consider (2022) Converging Mechanisms of Epileptogenesis and Their Insight in how future therapeutic advancements may impact the bidirectional relationship between Glioblastoma. GBM, SDs, and seizures. Front. Mol. Neurosci. 15:903115. doi: 10.3389/fnmol.2022.903115 Keywords: glioma, peritumoral border, epilepsy, seizures, spreading depolarizations Frontiers in Molecular Neuroscience | www.frontiersin.org 1 June 2022 | Volume 15 | Article 903115 Hills et al. Epileptogenesis in Glioblastoma INTRODUCTION THE IMPORTANCE OF TIMING Glioblastoma (GBM) is an astrocytic-origin neoplasm There are four main stages associated with clinical GBM categorized as a WHO grade IV glioma (Ohgaki and Kleihues, progression: pre-resection, post-resection, recurrence, and end- 2013). It is generally regarded as the most aggressive form of of-life. The occurrence of seizures relative to these defined primary brain cancer in adults, evidenced by its notoriously periods, i.e., their timing, has potential implications for disease high mortality rate. Median survival after primary diagnosis progression and prognosis. The occurrence of seizures as a is ∼14 months when following standard treatment regimens presenting symptom has historically had a more favorable (Goodwin and Laterra, 2010; Lim et al., 2011; Szopa et al., relationship to prognosis, vs. their appearance post-primary 2017; Khandwala et al., 2021). Long-term survival is so diagnosis. This former connection is primarily based on the exceedingly rare that long-term survivor status is achieved only hypothesis that seizures provoke earlier tumor discovery and 2.5 years post-diagnosis (Thakkar et al., 2014). The presence initiation of anti-tumorigenic treatment, translating to increased and progression of GBM is often associated with seizures, a survival. Yet, recent clinical evidence has indicated a significant debilitating comorbidity otherwise known as tumor-associated association between new-onset seizures post-GBM diagnosis and epilepsy (TAE). Epilepsy associated with GBM is regarded as a mortality in IDH-wildtype GBM patients (Climans et al., 2020; ‘‘secondary focal epilepsy’’ as it progressively develops secondary Wasade et al., 2020; Mastall et al., 2021). This disparity suggests to the nature and presence of the primary lesion (de Curtis et al., that the mechanisms underlying seizure occurrence at these 2015). The high chance of seizure recurrence in GBM, primarily respective time points may slightly differ in their progression. due to the presence of the neoplasm, means that patients are Along the same vein, the recurrence or new-onset occurrence often diagnosed with and treated for epilepsy following initial of seizures post-resection is associated with more negative seizure occurrence. outcomes. In particular, seizure recurrence following first-line Seizures manifest as a presenting symptom in two-thirds of anti-tumor therapy and successful seizure control is closely GBM patients, with the average incidence of epilepsy throughout associated with disease progression (Chaichana et al., 2009; the disease course varying between 30% and 62% (Bruna et al., Vecht et al., 2014). 2013; Kerkhof and Vecht, 2013; Kerkhof et al., 2013; Dührsen It is commonplace that patients with pre-operative seizures et al., 2019). Seizures in GBM are varied in semiology. The will experience seizure re-emergence post-resection. If GBM is most common types seen in the clinic are focal to bilateral tonic considered the singular ‘‘driving force’’ of seizure generation clonic, and either simple or complex focal seizures (Kerkhof et al., then it is reasonable to suggest that removal of the bulk tumor 2013; Liang et al., 2016; Toledo et al., 2017). The initial focal should result in complete seizure cessation. However, this is more onset of the seizures indicates that the transition to seizure is often not the case. Those with lesions previously situated in the influenced by the location of the tumor, and in general GBM frontal lobe carry the highest risk of post-operative seizures; a patients with tumors located in cortical regions are at the highest switch with temporal lobe lesions, which are associated with risk of experiencing seizures (Chaichana et al., 2009; Englot et al., the highest pre-operative seizure likelihood (Liang et al., 2016). 2012; Ertürk Çetin et al., 2017). Lesions situated in the temporal Interestingly, patients with no previous history of TAE have lobe, frontal lobe, and parietal lobe are particularly epileptogenic also been known to develop new-onset seizures post-tumor (Liang et al., 2016). Patients with temporal lobe involvement are resection. Liang et al. (2016) found that 45% of their GBM thought to be most likely to develop pre-operative seizures due cohort without any prior seizure history developed new-onset to the high epileptogenicity of mesial temporal lobe structures, as epilepsy 12 months post-resection. It is well established that observed in epilepsy syndromes such as temporal lobe epilepsy epileptogenesis is an ongoing process that continues after seizure (TLE). GBMs in cerebellar or brainstem regions are thought onset, and in the case of GBM, this process takes place in the to carry a reduced risk of seizures, but these tumors are rare peri-tumoral brain tissue as opposed to the tumor itself (Figure 1; regardless (van Breemen et al., 2007; Lee J. W. et al., 2010; Zhang Köhling et al., 2006). Seizure generation may continue or newly et al., 2020). arise post-tumor removal, as the epileptogenic loci is often still In this review, we will first consider the mechanisms that present in the remaining ‘‘normal’’ parenchyma surrounding underlie epileptogenesis in GBM and the impact of their the resection cavity, otherwise known as the peri-cavity area. resulting bidirectional relationship on disease progression. We Furthermore, the infiltrative nature of GBM confers that small focus on both ‘‘intratumoral’’ and ‘‘peri-tumoral’’ mechanisms populations of tumor cells will be present in the peri-tumoral that involve epigenetic changes, neuroinflammation, BBB border post-resection, and these may continue to promote the dysfunction, and extensive alterations to glutamatergic and cycle of epileptogenesis. The influence of tumor resection on, GABAergic neurotransmission. We then recognize in our and the potential post-resection mechanisms underlying seizure review that our extensive knowledge gap has implications generation will be discussed further in-depth later in our surgical for clinical translation, and address the challenges faced in resection and recurrence section. current preclinical research. We also then discuss the potential The disease’s end-of-life phase often includes the influence of anti-tumorigenic and anti-seizure treatments on development of new-onset seizures in patients with no history GBM-associated epileptogenesis. Finally, we evaluate how of TAE, or recurrence of seizures in patients with previously approaching these pathologies from a network level is a necessary effective anti-epileptic regimens (Oberndorfer et al., 2008; Pace direction for translational future research. et al., 2013; Koekkoek et al., 2014). The non-linear appearance Frontiers in Molecular Neuroscience | www.frontiersin.org 2 June 2022 | Volume 15 | Article 903115 Hills et al. Epileptogenesis in Glioblastoma FIGURE 1 | Seizure occurrence and pathoanatomical changes relative to disease stages in glioblastoma. The different stages of glioblastoma progression are associated with their own relationship to seizure occurrence, of which focal and focal-to-bilateral are the most common semiology. When the bulk tumor is present (glioblastoma), the area where tumor meets “normal brain” is known as the peri-tumoral border (in red). The network organization of this area changes progressively with disease evolution. Upon resection, the peri-tumoral border is now termed the peri-cavity area. Now independent of the glioblastoma, this area may become intrinsically epileptic. When the tumor recurs, areas of the peri-cavity area now merge with the new peri-tumoral border (orange/red). At the endstage of the disease, the peri-cavity area and peri-tumoral border are simultaneously present (orange/red). Therefore, there are a multitude of mechanisms generating seizure activity both in the new peri-tumoral border and the epileptic peri-cavity area. of seizures with GBM progression suggests that the mechanisms (Alzial et al., 2022). Accordingly, the mechanisms contributing by which seizures manifest prior to, or following diagnosis to seizure generation in this patient population (IDH-mutant) (pre-surgical resection) may be different to those resulting in are thought to be markedly different to those occurring in post-resection recurrent or new-onset seizures. And these may IDH-wildtype patients (Liubinas et al., 2014; Berendsen et al., be different still from those occurring in the final phase of the 2016; Toledo et al., 2017; Wasade et al., 2020). Therefore, it is disease. The presence of seizures post-resection infers that whilst crucial that future studies differentiate not only between low and GBM appears to initially influence the development of epilepsy, high-grade gliomas but also subtypes. the mechanisms by which seizures are continuously generated may eventually become independent of the tumor. FACTORS CONTRIBUTING TO The importance of seizure occurrence relative to disease TUMOR-DRIVEN EPILEPTOGENESIS stage has only recently been delineated, and this reflects the historical inconsistencies within the literature when reporting Epileptogenic Molecular Mutations seizure occurrence in GBM. The extent of the relationship The rapid growth of GBM and the presence of areas of between seizures and prognosis has been clouded by studies hemorrhage or necrosis confers that there is a high probability grouping GBM-associated seizures in investigations of low-grade of acute tissue damage, the consequences of which could gliomas. This is exacerbated by the lack of distinction between promote epileptogenesis. Yet, clinical studies have not observed IDH-mutant vs. IDH-wildtype tumors in studies. It is evident a strong correlation between tumor volume, seizures, and mass that the mechanisms underlying network hyperexcitability, and effect. Generally, smaller tumor volume is associated with their effect on disease progression, are extremely varied between seizure presence, and so other tumor-driven mechanisms may tumors despite the umbrella classification of ‘‘glioma’’. The provoke epileptogenesis (Chaichana et al., 2009; Henker et al., inclusion of GBM-associated seizures in studies of lower-grade 2019). GBM’s disorganized, uncontrolled growth inevitably leads gliomas leads to misinterpretation of the role seizures play in to mutations, and consequently, there is substantial genetic GBM progression. Further to this, the IDH subtypes of GBM heterogeneity inter- and even intratumorally (Szopa et al., 2017). show distinct relationships to seizure occurrence and prognosis. Collective mutations in tumor suppressor genes PTEN (Wang Higher rates of seizures are often associated with the IDH-mutant et al., 1997; Han et al., 2016), TP53 (Dittmer et al., 1993; Nagpal GBMs that are known to occur in younger patients, evolve et al., 2006; Zhang et al., 2018), and NF1 (Pearson and Regad, from lower-grade tumors, and have a more favorable prognosis 2017; Soomro et al., 2017; Szopa et al., 2017) are common Frontiers in Molecular Neuroscience | www.frontiersin.org 3 June 2022 | Volume 15 | Article 903115 Hills et al. Epileptogenesis in Glioblastoma in primary GBM. Disruptions to their associated downstream Lin et al. (2017). They utilized their CRISPR-Cas9 in-utero signaling pathways are heavily implicated in tumorigenesis. electroporation (IUE) GBM model to identify glioma-analogs of Interestingly, PTEN, NF1, and TP53 are all also intrinsically astrocyte subpopulations specifically involved in the emergence linked to epilepsy. Recent experimental studies have selected of GBM-associated seizures. One such identified was astrocyte them for deletion/mutation to produce genetic rodent models population C which increased linearly with tumor progression. of GBM that reliably produce epileptogenic tumors (Hatcher This trend was tightly correlated with increased invasion and et al., 2020; Hu et al., 2022). Their commonality in both GBM importantly, seizure onset. Interestingly, astrocyte population and epilepsy indicates that whilst the tumor itself may not C was found to be highly enriched for the expression of be electrographically active, its intrinsic molecular properties synaptic genes, and also for a number of established epilepsy- produce an environment that is both vulnerable to and promotes associated genes. This indicates an additional genetic glial basis mechanisms of hyperexcitability. for epileptogenesis in GBM. These findings also confer that some Physiologically, PTEN negatively regulates PI3K/AKT/mTOR of the most intrinsic properties of the tumor, the genes enriching pathway. However, mutated or deleted forms of PTEN are the glioma cells, directly influence the development of epilepsy. present in GBM resulting in the disinhibition of PI3K/AKT and hyperactivation of mTOR signaling (Endersby and Baker, BBB Dysfunction 2008; Venkatesan et al., 2016). PTEN mutations have been GBM is one of the most vascularized tumors in humans and associated with poorer prognosis in GBM and are also known its associated neoangiogenesis promotes blood-brain barrier to play a role in epilepsy syndromes. Targeted conditional (BBB) disruption. This breakdown allows infiltration of serum deletion of PTEN in mice results in spontaneous seizures proteins into the brain parenchyma, encourages vasogenic with post-mortem brains displaying the hallmark pathologies oedema, and increases intracranial pressure (Noell et al., of temporal lobe epilepsy (Nishio et al., 2000; Backman et al., 2012). Pathologic BBB disruption is readily apparent in GBM 2001; Luikart et al., 2011; Pun et al., 2012; Williams et al., patients using gadolinium contrast MRI, whereby the leaky BBB 2015). NF1, and its protein neurofibromin 1, are known enables gadolinium to diffuse into the tissue and presents as to negatively regulate Ras signaling through GTPase activity ring-enhancing lesions (Wolburg et al., 2012; Dubois et al., 2014). (Pearson and Regad, 2017; Soomro et al., 2017). GBM-induced Morphologically, blood vessels involved in the GBM blood- loss-of-function mutations in NF1 disinhibits the Ras/MAPK tumor barrier are typically non-uniform with altered pericyte pathway, leading to hyperactivation of mTOR and unchecked cell coverage, fenestrations, and reduced tight junctions. For in-depth proliferation, favoring tumor progression. Targeted knockout of reviews focused on tumor progression and its impact on the the NF1 gene is associated with decreased latency to epilepsy BBB, please see Dubois et al. (2014) and Arvanitis et al. (2020). and greater seizure severity in mice (Sabetghadam et al., 2020). It is notable that BBB dysfunction is observed in both human Its mutated form is associated with increased seizure presence and animal studies following acute seizures, and its leakage is in patients with neurofibromatosis-1 (Sorrentino et al., 2021). particularly associated with TLE (van Vliet et al., 2007; Marchi Mutations in TP53, also known as p53, disrupt its ability to et al., 2012). As a corollary, there may be additional effects induce the mechanisms of cell cycle arrest, senescence, and on the BBB in GBM due to the effect of epileptogenesis on apoptosis that prevent proliferation and migration of damaged or pericytes, please see Löscher and Friedman (2020) and Yamanaka neoplastic cells (Kastenhuber and Lowe, 2017). Gain-of-function et al. (2021) for further discussion of this work. Interestingly, (GOF) mut-p53 is a protein highly expressed in GBM that epileptogenesis secondary to BBB injury largely involves the enhances pro-invasive signaling by upregulating the activity of activation of inflammatory responses and reactive astrogliosis, receptor tyrosine kinases e.g., MET and EGFR (Zhang et al., attributes observed in the GBM microenvironment. 2018). Increased expression of p53, particularly within the The exposure of the brain to blood serum components hippocampus, is found in both experimental models and resected such as fibrinogen, albumin, glutamate, and K+ following tissue samples from patients with drug-resistant TLE (Morrison GBM-induced BBB disruption contributes to changes in et al., 1996; Engel et al., 2007). In seizures, overexpression neuronal excitability. The transformation of astrocytes under of p53 is associated with greater activation of apoptotic these conditions further provokes epileptogenic changes. The processes and neuronal cell death, further contributing to action of albumin at transforming growth factor beta (TGF-β) network imbalances and potentiating mechanisms of excitability receptors induces changes in the expression of critical inwardly (Morrison et al., 1996). Importantly, elevated p53 expression rectifying potassium channels (Kir; Ivens et al., 2007; Marchi has been previously correlated with epileptogenic GBMs (Toledo et al., 2012) and water channels i.e., aquaporin 4 (AQP4; et al., 2017). It is worth noting that TP53 mutations alone are Ikeshima-Kataoka, 2016). In human and animal models of not sufficient to induce the formation of GBM, and require other epilepsy syndromes, it has been shown that serum-exposure genes such as PTEN to also be mutated to drive GBM progression can lead to the direct downregulation or mislocalization of (Zheng et al., 2008). Kir4.1 astrocytic K+ channels consequently resulting in disrupted Whilst there are overt epileptogenic molecular mutations K+ homeostasis (Campbell et al., 2012; Buckingham and Robel, potentiated by GBM growth as discussed above, the epileptogenic 2013; Pallud et al., 2013). Kir4.1 is responsible for buffering potential of a tumor may also be influenced by the particular extracellular K+ and maintaining a negative resting membrane subpopulation of glioma cells that constitute the tumor. potential (RMP; Olsen and Sontheimer, 2004; Hibino et al., An interesting investigation into this relationship is seen in 2010). Therefore, GBM-induced disruption would likely result Frontiers in Molecular Neuroscience | www.frontiersin.org 4 June 2022 | Volume 15 | Article 903115 Hills et al. Epileptogenesis in Glioblastoma in RMP dysregulation of astrocyte processes, accumulation of in TLE and in the chronically epileptic brain following intra- extracellular K+ and consequently a lowered neuronal firing hippocampal kainic acid (IHKA) epilepsy induction (Lee et al., threshold (Table 1; Olsen and Sontheimer, 2004). GBM-induced 2004; Hubbard et al., 2016). This suggests that extremes of AQP4 dysregulation and redistribution is known to be associated AQP4 expression contribute to epileptogenesis (Binder et al., with increased peri-tumoral vasogenic oedema and aberrant BBB 2004a,b). GBM-induced upregulation of AQP4 expression may function (Warth et al., 2007). lead to cell swelling, compensatory efflux of K+ , Cl− , and glutamate, and decreased ECS volume (Table 1). The net effect of Ion Homeostasis which would favor neuronal depolarization and potentiation of Notably, Kir4.1-mediated K+ buffering also plays a crucial role excitatory activity throughout the surrounding neuronal network in glutamate homeostasis as the channel tends to colocalize (Binder et al., 2012). Therefore, GBM-induced alterations to with GLT-1 and AQP4 at tripartite synapses (Nagelhus et al., both AQP4 and Kir4.1 aggressively promote microenvironment 2004; Olsen and Sontheimer, 2008). Glutamate reuptake is changes conducive to both seizures and SDs. suppressed in the case of Kir4.1 inhibition, due to disruption of ionic gradients and astrocytic RMP, with epileptogenic Neuroinflammation consequences. These occur secondary to greater extracellular The extreme infiltrative and epileptogenic nature of GBM can glutamate accumulation, as will be discussed later in glutamate be, at least in part, attributed to its master manipulation of release, reuptake, and post-synaptic action (Kucheryavykh et al., the immune system. GBMs are primarily immunosuppressive 2007; Armbruster et al., 2021). Both downregulation and tumors, and this is primarily due to the high presence of tumor- mislocalization of Kir4.1 has been demonstrated in patient- associated macrophages (TAMs), tumor-associated neutrophils derived GBM samples and cell lines (Zurolo et al., 2012; Madadi (TANs), and regulatory T cells (Tregs). These mediators are et al., 2021). As well as significant electrophysiological changes crucial in maintaining GBM’s extreme immunosuppressive including a depolarized astrocytic RMP and large outward K+ phenotype and are also directly linked to cancer cell proliferation currents (Olsen and Sontheimer, 2004). and the recruitment of surrounding non-neoplastic cells (Fanelli Accumulation of extracellular K+ is known to be a key et al., 2021). However, whilst the net effect of the tumor initiating event in the generation of spreading depolarization microenvironment (TME) is immunosuppressive, this does (SD). SD is characterized by a near-complete regional not preclude the existence of underlying pro-inflammatory depolarization of neurons and glial cells that is accompanied actions. The peri-tumoral environment is inundated with a by the collapse of transmembrane ion gradients and followed variety of pro-inflammatory cytokines (IL-1β, IL-6, IL-8, TNF- by a temporary suppression of neuronal activity (Pietrobon α), chemokines, and extracellular matrix remodeling enzymes and Moskowitz, 2014; Ayata and Lauritzen, 2015). SDs are (MMP-2, MMP-9; Figure 2; DeCordova et al., 2020). Reactive primarily self-terminating in healthy brain tissue as the resources astrocytes, which also produce said mediators (Guan et al., 2018), required to recover from major disruptions to transmembrane are also increasingly present. In particular, IL-1β and IL-6 are ion gradients are readily available. However, in a metabolically well recognized to contribute to oncogenesis and their expression compromised brain, such as one with GBM, there is decreased is abundantly found in GBM cell lines and tissues (Lu et al., availability of these resources. Consequently, repolarization and 2007; Yeung et al., 2012; Liu et al., 2021), IL-6, in particular, recovery is significantly delayed to a point that may exacerbate is associated with poorer prognosis (Hori et al., 2019). IL-1β existing tissue damage (Ayata and Lauritzen, 2015). The complex signaling stimulates activation of NF-kB, ERK, and p38 MAPK relationship between seizures and SDs is well documented, pathway signaling, as well as promoting the release of IL-6 and seizures create an excitotoxic environment conducive to SD IL-8 from GBM cells (Yeung et al., 2013). In turn, the release of generation and vice versa (Berger et al., 2008; Kramer et al., IL-8 is postulated to induce TNF-α secretion from TAMs (Wei 2017). The BBB disruption and subsequent steady increase in et al., 2021). K+ accumulation that occurs within the GBM peri-tumoral Interestingly, the pro-inflammatory molecules seen within border contributes to a progressive reduction in the threshold the TME are historically associated with seizures (Rana and for seizures and SDs (Figure 2). Musto, 2018). It is well recognized that neuroinflammation and Upregulation and redistribution of AQP4 is widely reported associated BBB breakdown are necessary for epileptogenesis in GBM (Figure 2). Interestingly significantly increased diffuse (Marchi et al., 2007). Experimental models of TLE, and and perivascular expression of AQP4 on glioma cells is seen resected hippocampal tissue from TLE patients both display in resected tumor tissue from GBM patients with seizures vs. upregulated expressions of pro-inflammatory molecules, reactive those without (Isoardo et al., 2012). Dysregulation of astrocytic astrocytosis, and activated microglia (Ravizza et al., 2008). AQP4 expression is known to contribute to epileptogenesis Transgenic mice overexpressing IL-6 and TNF-α demonstrate in traditional epilepsy syndromes (Binder and Steinhäuser, increased seizure susceptibility (Campbell et al., 1993; Probert 2006; Binder et al., 2012). AQP4-knockout mice demonstrate et al., 1997; Samland et al., 2003), and treatment with IL-1β increased latency to seizure generation when challenged with prior to kainic acid administration potentiates seizure duration pentylenetetrazole (PTZ). This is largely attributed to expansion (Vezzani et al., 1999). IL-1β exerts its pro-convulsive actions of the extracellular space (ECS), but seizures showed a more through IL-1R (IL-1 receptor)-mediated phosphorylation of severe phenotype once eventually generated (Binder et al., 2004a, the NMDAR GluN2B subunit (Viviani et al., 2003). TNF-α 2006). Conversely, significant AQP4 upregulation is observed upregulates Ca2+ -permeable AMPAR (CP-AMPAR) membrane Frontiers in Molecular Neuroscience | www.frontiersin.org 5 June 2022 | Volume 15 | Article 903115 Frontiers in Molecular Neuroscience | www.frontiersin.org Hills et al. TABLE 1 | Potential biomarkers in GBM-associated epilepsy, their relationship to each pathology, role, and potential targeted therapies. Potential Associated cell types Alterations in GBM (G) and Pathological role Potential targeted therapies References biomarkers epileptogenesis (E) Kir4.1 Astrocytes G and E: Downregulation and Disrupts astrocytic RMP → Gene therapy, VPA (CA) Mukai et al. (2018) and Ohno et al. mislocalization extracellular K+ accumulation; (2021) suppression of glutamate reuptake. Aquaporin 4 Astrocytes, glioma cells G: Upregulated with diffuse Cell swelling → efflux of K+ , Cl− , Gene therapy, acetazolamide (CA) Reiss and Oles (1996), Binder et al. perivascular expression glutamate; decreased ECS volume. (2004a), Alvestad et al. (2013), and E: Extremes of expression, Duan and Di (2017) mislocalization MMP-9 Astrocytes, glioma cells G and E: Increased release Increases NMDAR activity; MMP-inhibitors (e.g., marimastat) Bronisz and Kurkowska-Jastrz˛ebska degrades PNNs surrounding (2016), Mondal et al. (2020), and Pijet FS-PV+ interneurons. et al. (2020) 6 xCT (SLC7A11) Astrocytes, glioma cells G and E: Upregulation Increases glutamate release into the Gene therapy, sulfasalazine (CA) Lewerenz et al. (2014), Leclercq et al. extracellular space. (2019), and Sears et al. (2019) GLT-1 Astrocytes G and E Marked downregulation on Impairs glutamate clearance from Gene therapy, ceftriaxone (CA) de Groot et al. (2005), Peterson et al. astrocytes the extracellular space. (2021), and Ramandi et al. (2021) GABAA R Pyramidal neurons, G: Almost complete absence on Reduces GABAergic Gene therapy, benzodiazepines Houser et al. (2012) and Greenfield glioma cells glioma cells neurotransmission → glioma cell (CA) (2013) E: Dysfunctional membrane proliferation; disinhibits pyramidal trafficking, various subunit neuronal firing alterations NKCC1:KCC2 Pyramidal neurons G and E: NKCC1 upregulated on Reverses the chloride gradient → NKCC1: Bumetanide (CA) Huberfeld et al. (2007), MacKenzie and glioma cells in GBM and pyramidal paradoxical depolarization; cell KCC2: Gene therapy, kenpaullone Maguire (2015), MacKenzie et al. cells, KCC2 downregulated on shrinkage promoting glioma cell (CA) (2016), Duy et al. (2020), and Yeo et al. pyramidal cells proliferation (2021) June 2022 | Volume 15 | Article 903115 CA, clinically approved for use in epilepsy or otherwise; VPA, valproic acid. Epileptogenesis in Glioblastoma Hills et al. Epileptogenesis in Glioblastoma FIGURE 2 | The vicious cycle of hyperexcitability in glioblastoma (GBM) progression. How epileptogenesis is initiated, is influenced by, and contributes to GBM growth. (1) GBM growth and its interaction with neuronal/glial cells initiates the cycle to epileptogenesis. (2) Upregulation of xCT on glioma cells and downregulation of KCC2 on surrounding pyramidal neurons creates an initial imbalance between excitatory and inhibitory neurotransmission. (3) Pro-inflammatory cytokines and MMPs present in the GBM TME act to degrade the PNNs surrounding GABAergic interneurons. (3) Once epileptogenesis has been initiated, further proinflammatory cytokine, K+ , and glutamate release exacerbates BBB disruption and encourages network reorganization creating a microenvironment conducive to SDs. (4) The function of Kir4.1, AQP4 and GLT-1 transporters is disrupted by recurrent seizures and SDs, and via actions of GBM. (5) This dysfunction continues in a cyclical fashion whereby seizures promote more seizures and SDs, and the associated pathological downstream signaling actively potentiates glioma cell proliferation, and vice versa. BBB, blood brain-barrier; TME, tumor microenvironment; MMPs, matrix metalloproteinases; PNNs, perineuronal nets; SDs, spreading depolarizations; APQ4, aquaporin 4. expression and trafficking, enabling greater Ca2+ influx with occurring in conjunction with the epileptogenic mechanisms cytotoxic effects. It also acts to increase the internalization mentioned in the following section, result in seizure generation. of GABAA receptors (GABAA Rs), subsequently reducing These are just a few mechanisms by which GBM promotes the influence of GABAergic inhibitory neurotransmission network hyperexcitability to the point of recurrent seizure (Stellwagen et al., 2005). IL-1β and TNF-α are also known generation. Following which the seizures themselves potentiate to increase glutaminase activity (Bezzi et al., 2001; Ye et al., tumor progression and increased excitability in a malignant 2013), which results in enhanced glutamate release whilst also feed-forward nature. decreasing its re-uptake (Hu et al., 2000; Yeung et al., 2013; Rana and Musto, 2018). All alterations to network function that can SEIZURES, SPREADING be interpreted as epileptogenic. DEPOLARIZATIONS AND PERI-TUMORAL Concomitant to pro-inflammatory cytokine secretion, BORDER GBM cells release matrix metalloproteinase 9 (MMP-9). Chronic MMP-9 production is associated with BBB disruption, Glutamate Release, Reuptake and independent of the insult, and mediates its pathological effects Post-synaptic Action by decreasing the efficiency of BBB tight junctions (Asahi et al., In the GBM peri-tumoral environment, there is a breakdown 2001). MMP-9 also impacts both excitatory and inhibitory in glutamate homeostasis and subsequent pathological network neurotransmission. It acts to increase the activity of NMDARs signaling creates an environment favorable to seizures and SDs. through integrin-β1 signaling (Michaluk et al., 2009), and In GBM, glutamate is primarily released into the synaptic cleft induces the degradation of perineuronal nets (PNNs) that by the cystine/glutamate antiporter system (system xc-), also surround fast-spiking parvalbumin positive (FS-PV+) inhibitory known by its active subunit xCT (Buckingham et al., 2011). interneurons (Figure 2; Tewari et al., 2018). These alterations System xc- primarily mediates the exchange of intracellular increase the excitability of the peri-tumoral network, and when L-glutamate for extracellular L-cystine, a precursor for the Frontiers in Molecular Neuroscience | www.frontiersin.org 7 June 2022 | Volume 15 | Article 903115 Hills et al. Epileptogenesis in Glioblastoma antioxidant glutathione (GSH). Its expression is known to be and SDs may be in part initiated by GBM-induced GLT-1 significantly upregulated in the tumor tissue of epileptic GBM downregulation. patients (Lin et al., 2020). Upregulated expression of system xc- The pathological downstream signaling cascades that occur in GBM cells leads to greater concentrations of glutamate in secondary to extracellular glutamate accumulation in GBM are the extracellular space (Table 1). This promotes hyperexcitability primarily mediated by post-synaptic NMDARs and AMPARs. through activation of post-synaptic glutamatergic receptors Glutamate-mediated prolonged NMDAR and CP-AMPAR on peri-tumoral neurons, subsequently initiating processes activation triggers a wide variety of intracellular events, that of depolarization and excitotoxicity (Figure 2; Buckingham in turn initiate processes of Ca2+ -mediated excitotoxicity and et al., 2011; Neal et al., 2016). As a corollary, increased xCT apoptosis (Zhu et al., 2000; Buckingham and Robel, 2013; Henley expression has been associated with shorter progression-free and Wilkinson, 2016). In both peri-tumoral human and murine survival and poorer overall survival in GBM patients, as well as GBM tissue, there is increased phosphorylation of the NMDAR a more infiltrative phenotype on MRI (Takeuchi et al., 2013). GluN2B subunit, a widely used measure of NMDAR activation Interestingly, the expression of xCT in glioma patients has (Gao et al., 2014). Notably, a relative increase in the NMDAR been found to significantly correlate with epileptic seizures at GluN2B subunit confers slower receptor decay times enabling GBM diagnosis, even denoted as an ‘‘independent biomarker’’ a greater volume of Ca2+ influx through the receptor, thereby (Sørensen et al., 2018). Notably, epileptiform activity can be increasing the risk of Ca2+ -mediated excitotoxicity and neuronal reduced by treatment of both epileptogenic GBM-bearing acute death. NMDARs are also implicated in the propagation of SDs, brain slices and mice with sulfasalazine (SAS), an FDA-approved therefore any alterations to NMDAR activation may potentiate system xc- antagonist (Table 1; Buckingham et al., 2011; SD propagation and alter their waveform (Masvidal-Codina Campbell et al., 2012; Robert et al., 2015; Hatcher et al., et al., 2021). 2020). There are no current clinical trials looking at SAS as a There are marked decreases in the editing of the AMPAR treatment for GBM-related epilepsy. Yet, a phase 2 randomized GluA2 subunit and increases in the expression of GluA1 subunits clinical study evaluating SAS as a primary anti-tumor treatment in human GBM samples and cell lines (Maas et al., 2001; de in recurrent GBM was conducted 2005–2007. This trial was Groot et al., 2008; Venkataramani et al., 2019). This indicates terminated early due to unreasonable occurrence of adverse that there is Ca2+ influx through, and increased activity at, events and lack of tumor response. However, this study was post-synaptic ionotropic glutamate receptors that trends towards conducted in a severely ill, neurologically impaired patient increased neuronal excitability. Epileptiform activity originating population. They also neglected to assess the anti-epileptic from GBM-bearing tissue can be completely abolished by potential of SAS as an adjuvant treatment to standard anti-tumor the administration of NMDAR antagonist APV (Buckingham regimens and therefore this avenue would still be of great interest et al., 2011; Campbell et al., 2012), and AMPAR antagonists (Robe et al., 2009). perampanel and cyanquixaline (CNQX; Venkataramani et al., The disruption in peri-tumoral glutamate homeostasis 2019; Venkatesh et al., 2019; Lange et al., 2020; Mayer et al., is thought to be further exacerbated by GBM-induced 2020). When occurring in GBM, the surrounding cell death downregulation of astrocytic glutamate transporter 1 (GLT-1; occurring secondary to overactivation of NMDARs/AMPARs Table 1). This impairs clearance of glutamate from the synaptic could encourage greater tumor growth by ‘‘making space’’ and cleft and indirectly acts to potentiate its actions at post-synaptic further exacerbate the network imbalance fuelling excitability. NMDARs/AMPARs (Robert and Sontheimer, 2014). GLT-1 The generation of seizures in peri-tumoral neuronal networks (also known as EAAT2) is responsible for 80%–90% of all inevitably disrupts ionic homeostasis. This primarily manifests extracellular glutamate reuptake activity (Lin et al., 2012; as increased extracellular K+ concentration and intracellular Takahashi et al., 2015), but both GBM tumor tissue and cell Na+ , Cl− , and Ca2+ concentrations, as well as a decrease in lines show an almost complete absence of the transporter (Ye intracellular pH (Raimondo et al., 2015). Moreover, recurrent et al., 1999; de Groot et al., 2005). Initial experimental evidence seizures provoke excess glutamate release from pre-synaptic suggests that GLT-1 expression is significantly downregulated neurons. Parallels can be drawn between the signaling pathways in both tumoral and peri-tumoral tissue from GBM patients initiated by seizures/SDs and those potentiated by GBM. with seizures vs. those without (Yuen et al., 2012). Interestingly, The dysfunctional actions of glutamate transporters and parallels are seen in TLE patients with hippocampal sclerosis receptors also create an environment conducive to glioma and drug-refractory seizure phenotypes where GLT-1 expression cell proliferation, migration, and invasion. Blockade of xCT is markedly reduced (Proper et al., 2002; Sarac et al., 2009). with SAS consistently reduces GBM glioma cell migration in Additionally, GLT-1 knockout mice are known to develop multiple cell lines, an anti-tumor effect corroborated in vivo lethal spontaneous seizures (Tanaka et al., 1997), and GLT-1 (Lyons et al., 2007). GLT-1 downregulation also contributes to conditional knock-out mice demonstrate increased susceptibility GBM oncogenesis, in physiological conditions it plays a role in to SD (Aizawa et al., 2020). A reduction in GLT-1 expression growth suppression therefore the absence of this transporter on favors SD propagation by allowing a more rapid increase in glioma cells allows their growth to proceed unchecked. Adeno- extracellular glutamate concentration during the course of associated virus (AAV)-mediated reintroduction of GLT-1 into the phenomenon (Figure 2). Direct links are yet to be fully GBM cells mitigated in vitro growth. Furthermore, flank elucidated in the literature, but current findings support the implantation of U87 cells transfected ex vivo with AAV-GLT-1 theory that the pathological mechanisms underlying seizures into nude mice was associated with significantly decreased tumor Frontiers in Molecular Neuroscience | www.frontiersin.org 8 June 2022 | Volume 15 | Article 903115 Hills et al. Epileptogenesis in Glioblastoma growth (de Groot et al., 2005). As such, evidence suggests that 2014; Schiapparelli et al., 2017; Luo et al., 2020). Epileptic glutamate released by GBM glioma cells may be able to promote discharges from resected human tissue were suppressed by the pro-migration/invasion and proliferative effects through seizure- application of FDA-approved NKCC1 antagonist bumetanide induced alterations to the functionality of GLT-1, NMDARs, and (Table 1; Pallud et al., 2014). Further to this, latency to AMPARs. seizure onset was increased in GBM-bearing mice treated with continuous infusion of bumetanide (MacKenzie et al., 2016). GABAergic Neurotransmission and Evidently, the Cl− dysregulation and consequent defective Chloride Homeostasis GABAergic signaling in GBM facilitates the generation of GBM induces major perturbations to inhibitory GABAergic epileptiform activity. GBM-induced changes in Cl− transporter neurotransmission. These disruptions originate from a loss of expression may be potentially triggered by overstimulation of GABAergic synaptic density on pyramidal neurons, reductions NMDARs and consequent dephosphorylation of KCC2 residue in GABAA R expression, and changes to the Cl− gradient. serine 940 (Ser940); a residue responsible for transporter GBM-induced degradation of the PNNs surrounding FS-PV+ function and stability in the membrane (Lee H. H. C. et al., inhibitory interneurons significantly decreases their frequency of 2010; Lee et al., 2011; MacKenzie et al., 2016). Blockade of action potential firing (Figure 2; Tewari et al., 2018). Significant excess glutamate release from glioma cells with SAS prevents selective loss of FS-PV+ interneurons is also observed within KCC2 downregulation in GBM cells (MacKenzie et al., 2016), the peri-tumoral cortex, secondary to excitoxicity associated implicating increased glutamate signaling in Cl− gradient with increased extracellular glutamate accumulation (Tewari dysregulation and aberrant GABAergic neurotransmission. et al., 2018). Cumulatively, this manifests as reduced GABAergic Additionally, BDNF release from glioma cells or locally neurotransmission and disinhibition of pyramidal neurons activated microglia is known to elicit the downregulation of within the peri-tumoral network. GBM also induces a reduction KCC2 and upregulation of NKCC1 on peri-tumoral neurons in the density of GABA binding sites, such that functional (Pallud et al., 2013). GABAA Rs are postulated to be almost completely absent on Defective GABAergic neurotransmission also confers a glioma cells (Labrakakis et al., 1998; Smits et al., 2012; Jung et al., growth advantage in GBM. Increased GABAergic signaling via 2019). Experimental studies have demonstrated that the majority GABAA Rs has been shown to negatively regulate glioma cell of GBM cells, from both human GBM samples and cell lines, do proliferation (Blanchart et al., 2017). Therefore, the observed not exhibit any electrophysiological response when challenged reduction in functional GABAA Rs expression on glioma cells with the application of GABA. This is suggestive of a lack of would result in decreased GABAergic neurotransmission and functional GABA receptors (Labrakakis et al., 1998). Whilst favor tumor growth (Table 1; Labrakakis et al., 1998). NKCC1 is glioma cells appear to lack functional GABAA Rs, they are still also known to regulate glioma cell volume changes by increasing postulated to be present on surrounding peri-tumoral neurons [Cl− ]i , which results in Cl− efflux upon GABA binding and allowing them to play a role in epileptogenic neurotransmission cell shrinkage to promote glioma cell proliferation, migration, (Campbell et al., 2015; Huberfeld and Vecht, 2016). and invasion of surrounding parenchyma (Table 1; Schiapparelli GBM-induced alterations to the expression ratio of et al., 2017). Both knockdown of NKCC1 and inhibition with Cl− transporters KCC2 and NKCC1 influence GABAergic bumetanide significantly reduces GBM cell migration in vitro neurotransmission by extruding and accumulating intracellular and in vivo, and overexpression of the transporter markedly Cl− ([Cl− ]i ), respectively. They are primarily responsible for the increases GBM cell invasion (Haas and Sontheimer, 2010; pathological reversal of the Cl− gradient that occurs in GBM (Liu Garzon-Muvdi et al., 2012; Schiapparelli et al., 2017). et al., 2020; Zhang et al., 2021). Higher KCC2 expression ensures a low [Cl− ]i concentration such that GABA binding elicits FROM LABORATORY TO CLINICAL Cl− influx and cell hyperpolarization, potentiating inhibitory TRANSLATION GABAergic signaling. Whereas increased NKCC1 expression in mature neurons results in a greater concentration of [Cl− ]i and Clinical Relevance reversal of the chloride gradient, promoting Cl− efflux upon The majority of current GBM literature focuses on either the GABA binding and cell depolarization (Plotkin et al., 1997; complex GBM microenvironment or the development of novel Hübner et al., 2001). Various epilepsy syndromes commonly therapeutics. Ergo, studies on the epileptological aspect of GBM demonstrate alterations in both Cl− cotransporters (Liu are extremely limited, and even less have investigated the impact et al., 2020). Significant downregulation of KCC2 is known of anti-tumorigenic treatment strategies on the relationships to occur within the GBM peri-tumoral cortex (Figure 2), between epilepsy, SDs, and tumor growth. Experimental studies correlating with greater [Cl− ]i concentration and a more of GBM-associated epilepsy primarily employ GBM cell lines positive GABA equilibrium potential (Conti et al., 2011; (human, mouse, or rat) or human primary cultures to generate Pallud et al., 2014; MacKenzie et al., 2016). This has also been intracranial GBM tumors. Wherein a tumor cell suspension shown to significantly correlate with decreased survival in is intracranially injected into brain areas associated with GBM patients (Campbell et al., 2015). Peri-tumoral tissues seizure vulnerability and/or propagation (Table 2). Genetic from patients with epileptogenic GBMs show significantly manipulation of epilepsy-associated known oncogenes via IUE increased expression of NKCC1 vs. non-epileptogenic patients CRISPR/Cas9-mediated deletion or utilization of inducible-Cre- (Conti et al., 2011; Garzon-Muvdi et al., 2012; Pallud et al., LoxP transgenic models has recently been employed to aid Frontiers in Molecular Neuroscience | www.frontiersin.org 9 June 2022 | Volume 15 | Article 903115 Frontiers in Molecular Neuroscience | www.frontiersin.org Hills et al. TABLE 2 | Seminal experimental studies, with a primary objective of GBM-associated epilepsy investigation. Study Model type Time Chronic Seizure frequency Semiology Spreading Additional course EEG? depolarizations? imaging? Köhling et al. (2006) Allograft transplant: nd Y nd EEG: spike wave discharges nd N C6 cells, Wistar rats Buckingham et al. PDX transplant into SCID nd Y In vivo: EEG: spontaneous epileptiform nd N (2011) mice Average activity. Freezing behavior, facial 2.5–6.5 events/hour automatisms, head tremor Campbell et al. (2012) PDX transplant into SCID 21 days N (in vitro In vitro: nd nd N mice only) 3.7 ± 0.6 events/min Campbell et al. (2015) PDX transplant into SCID 14–28 days Y In vivo:1 ± 0.3 events/day EEG: spontaneous epileptiform nd N 10 mice activity. Freezing behavior, facial automatisms, head tremor MacKenzie et al. (2016) Xenograft transplant: 21 days Y In vivo:11.0–18.7 events/week EEG: spontaneous epileptiform nd N C6 cells, nude mice activity. Bouckaert et al. (2020) Allograft transplant: 21 days Y nd EEG: spontaneous epileptiform Y: seen in Y—MRI F98 cells, Fischer rats activity. Ranging from filtered trace, non-convulsive to GTC not acknowledged seizures. Hatcher et al. (2020) Genetic: CRISPR 80 days Y In vivo: EEG: spontaneous epileptiform Y Y—Calcium Cas9-based IUE (deletion Day 80: activity, i.e., cortical spike imaging of Pten, Trp53, Nf1) 3.7–63.6 events/hour discharges. GTC seizures (day 80). PDX, patient-derived xenograft; IUE, in-utero electroporation; EEG, electroencephalography; GTC, generalized tonic clonic; nd, not disclosed. June 2022 | Volume 15 | Article 903115 Epileptogenesis in Glioblastoma Hills et al. Epileptogenesis in Glioblastoma the investigation into genetic mechanisms of GBM-associated effect of RT on seizure activity in high grade (IV) gliomas epileptogenesis (Table 2; Hatcher et al., 2020; Jin et al., 2021). is currently ongoing (Rades et al., 2021). Results from this The majority of existing studies utilize transplantation methods study will aid understanding and inform future decisions (Köhling et al., 2006; Bouckaert et al., 2020), with many seminal pertaining to the application of individual anti-tumorigenic articles employing xenografting (Table 2; Buckingham et al., treatments in GBM patients with seizures. TMZ is an oral 2011; Campbell et al., 2012, 2015). The use of mice lacking alkylating chemotherapeutic that combats tumor growth and an intact immune system, e.g., SCID mice, in investigations progression through DNA methylation and cell cycle arrest. of GBM-associated epileptogenicity is controversial as the The efficacy of TMZ in GBM patients is correlated with neuroinflammatory environment created and maintained by the increased methylation of the MGMT promoter (Perry et al., tumor is a key contributor to the development of network 2017). Interestingly this has also been associated with increased hyperexcitability. As the primary aim of using animal models post-operative seizure control in one clinical study (Feyissa et al., is to accurately recapitulate the disease as seen in humans, 2019). This appears to indicate that increased susceptibility to it stands to reason that GBM should be modeled in an TMZ treatment is associated with greater seizure control and immunocompetent animal to incorporate all aspects of the better survival outcomes. A reduction in seizure burden in immune system that may play any role in seizure generation or TMZ-treated MGMT-methylated epileptogenic GBM patients GBM progression. confers a proportional decrease in the activity of associated The paucity of the literature including seizures as a epileptogenic downstream signaling cascades. These pathways measured variable in oncological studies is a substantial are known to encourage tumor growth, and so their interruption obstacle to collating information about mechanisms underlying would restrict the ability of enduring tumor cells to progress GBM-related epileptogenesis. Where the experimental focus is into fully-fledged recurrent GBMs. However, other recent oncogenesis or novel treatment development, the presence of evidence suggests that the addition of TMZ to RT has a seizures is often overlooked or only mentioned as a general minimal effect on seizure presence in elderly GBM patients indicator of a humane endpoint. Experimental animals typically vs. RT alone (Climans et al., 2020). Climans et al. (2020) also express complex or simple focal seizures characterized by subtle demonstrated that decreased overall survival was significantly behavioral changes such as freezing periods, facial automatisms, associated with seizures. This paradoxically supports the head tremors, or exaggerated startle responses to audiogenic hypothesis that continuous seizure presence and associated stimuli (Köhling et al., 2006; Buckingham et al., 2011; Campbell excitotoxicity may increase the likelihood of tumor recurrence, et al., 2015). These events are completely spontaneous and and subsequently decrease the overall survival of epileptogenic may never occur in the presence of the researcher. Therefore, GBM patients relative to patients without seizures. However, seizures could be easily missed if their investigation is not these conclusions were derived from previously acquired clinical an objective of the study. However, arguably, neglecting trial data and MGMT methylation nor IDH-status were to include seizures as a monitored variable in therapeutic not taken into account in the reanalysis. Moreover, seizure studies has implications for clinical translation. The network outcomes were not reliably recorded in the original study changes occurring secondary to seizures and SDs in GBM potentially confounding results (Perry et al., 2017; Climans et al., undoubtedly impact tumor growth, disease progression, and 2020). potential risk of recurrence. Testing novel treatments in a model of GBM that does not account for seizure presence Anti-Seizure Medication in GBM may confound results by omitting to include a potentially The prevalence of seizures in GBM necessitates the use important influencer of therapeutic efficacy that affects a of anti-seizure drugs (ASDs) as the constant presence of substantial portion of the patient population. Examples of how the tumor increases the likelihood of seizure recurrence. clinically-used anti-tumor treatments affect seizures in GBM GBM-associated seizures present unique challenges as they have been examined in clinical trials (Chalifoux and Elisevich, are often poorly controlled, requiring poly-therapy. However, 1997; Climans et al., 2020), but the literature primarily focuses there are several potential contraindications related to the on low-grade gliomas. use of ASDs in GBM, including increased risk of enhanced The Stupp regimen is the standard of care for GBM, consisting toxicity and undesirable interactions with anti-cancer of resection followed 6 weeks later by post-surgical ionizing treatments. Generally, non-enzyme inducing agents are radiotherapy (RT) with concomitant TMZ chemotherapy and first-line ASD options, i.e., levetiracetam (LEV), as they do maintained by 20 weeks of adjuvant TMZ (Stupp et al., 2005, not interfere with anti-tumor treatment strategies (Rosati 2009; Fernandes et al., 2017). RT induces an acute inflammatory et al., 2010; Kerkhof et al., 2013). Experimental evidence reaction to mediate its anti-tumor effects but, as it targets indicates that certain ASDs may have anti-tumorigenic effects, the vulnerable peri-cavity environment, this may paradoxically supporting a connection between seizures and mortality contribute to network hyperexcitability (Baskar et al., 2014). in GBM. Historically, valproic acid (VPA) treatment has RT has been historically associated with improved seizure been associated with a potential survival benefit in GBM control in low-grade gliomas (Chalifoux and Elisevich, 1997; patients (Kerkhof et al., 2013; Krauze et al., 2015; Lu et al., Rudà et al., 2013). However, extensive investigations have not 2018). VPA is employed as an ASD as it disrupts GABA yet been reported for GBM and therefore, seizure outcomes degradation and blocks voltage-gated Na+ and Ca2+ channels. may differ. A prospective interventional trial evaluating the Its anti-tumor effects are mediated through its actions as Frontiers in Molecular Neuroscience | www.frontiersin.org 11 June 2022 | Volume 15 | Article 903115 Hills et al. Epileptogenesis in Glioblastoma a histone deacetylase (HDAC) inhibitor and potentiation see Terrone et al. (2020) for a comprehensive discussion of this of the cytotoxicity induced by temozolomide (TMZ; Van work in other types of epilepsy. Nifterik et al., 2012). Application of VPA to a variety of, but not all, GBM cell lines slowed cell cycle progression and Novel GBM Therapeutics and Their Impact inhibits proliferation (Knüpfer et al., 1998; Chavez-Blanco on Seizure Likelihood et al., 2006). However, the beneficial effect of VPA on GBM Despite a wide research field dedicated to anti-GBM therapeutics, progression in the clinic is extremely variable. A recent study no significant positive advances have been made in prognosis for inferred that wild-type p53 status is necessary for VPA to over a decade. As such, novel immunotherapeutic approaches enhance the anti-neoplastic effect of TMZ, demonstrating a such as immune checkpoint inhibitors (ICI), chimeric antigen potential reason for inconsistencies seen within the literature receptor T cell (CAR-T) therapy, myeloid-targeted therapies, (Tsai et al., 2021). and tumor vaccines, have been the focus of GBM literature in In recent years, standards of care have shifted to favor LEV for recent years. GBM mediates an immunosuppressive TME by anti-seizure treatment in GBM, and this has its own association upregulating the expression of immune checkpoint molecules to with anti-cancer effects (Roh et al., 2020). It is currently favored reduce T cell activation and proliferation. This confers an almost as a therapeutic dose that can be achieved quickly in a clinic intrinsic resistance to traditional anti-cancer treatments, making and there is a low likelihood of myelosuppression, hepatotoxicity, immune checkpoint molecules attractive therapeutic targets. In or hyponatremia (Hovinga, 2001). LEV binds to synaptic particular, increased expression of CTLA-4 is seen on Tregs vesicle protein 2A, interfering with neurotransmitter release and cytotoxic T lymphocytes, and PD-L1 on tumor cells and and reducing epileptogenic excitatory neurotransmission (Lynch TAMs (Yu and Quail, 2021). Correspondingly, initial in vivo et al., 2004). Evidence also indicates that LEV exerts a significant evidence demonstrated that the anti-PD-1 ICI nivolumab was inhibitory effect on O6 -methylguanine-DNA methyltransferase remarkably effective at eradicating GL261 tumors. However, (MGMT) in GBM cell lines, conferring reduced resistance whilst essentially dominating the preclinical ICI research field, to TMZ (Bobustuc et al., 2010). LEV also increases MGMT nivolumab’s effect failed to translate clinically (Yu and Quail, transcription in normal astrocytes exhibiting a neuroprotective 2021). Interestingly seizures were the second most common effect (Bobustuc et al., 2010). LEV treatment concurrent with neurological adverse event observed in clinical trials (Lim standard chemo-radiation protocol has been associated with et al., 2017). ICIs provoke an adaptive immune response in increased overall survival in IDH-wildtype GBM patients (Roh the peri-cavity area to increase the activity and infiltration of et al., 2020; Pallud et al., 2022). cytotoxic T cells and pro-inflammatory cytokine secretion. As A third-generation ASD, perampanel (PER), recently gained such, adverse effects have been attributed to ‘‘overshooting’’ attention as it was shown to significantly reduce the seizure T cell activation against antigens expressed on surrounding frequency of previously drug-refractory GBM-related seizures non-neoplastic cells (Roth et al., 2021). This mistargeting could (Vecht et al., 2017; Dunn-Pirio et al., 2018; Izumoto et al., act to promote hyperexcitability in peri-cavity neuronal networks 2018; Coppola et al., 2020; Maschio et al., 2020). Its anti-seizure post-resection, the time period at which systemic treatments are effects are mediated by its actions as a non-competitive typically employed. Interestingly, an adaptive immune response AMPARs antagonist (Hanada et al., 2011; Frampton, 2015; characterized by the infiltration of cytotoxic CD8+ T cells Augustin et al., 2018; Brito da Silva et al., 2021). PER exhibits and associated pro-inflammatory cytokine secretion is known significant anti-proliferative effects on multiple GBM cell lines to give rise to neuronal cell death and seizure induction in at various concentrations, and appears to reduce the overall Rasmussen Encephalitis (Varadkar et al., 2014). Furthermore, concentration of extracellular glutamate (Lange et al., 2019; the epileptogenic potential of ICIs is evidenced in reports Salmaggi et al., 2021). Initial evidence suggests a correlation of paraneoplastic encephalitis, and associated seizures, directly between PER treatment and tumor volume reduction on linked to the administration of anti-PD-1 (Shi et al., 2020) MRI-FLAIR (Izumoto et al., 2018). However, more clinical and anti-CTLA-4 (ipilimumab; Roth et al., 2021). It stands to investigations are required before conclusions can be drawn reason that if a potentially epileptogenic ICI is employed in as to the full extent of PER’s anti-tumor effects and any a seizure-vulnerable patient population, e.g., GBM, this may synergistic interactions it has with chemo-radiation regimens. indirectly encourage tumor progression secondary to increased Notably, PER is associated with various serious adverse effects, epileptiform activity and paradoxically result in decreased primarily psychiatric changes that are often the primary cause of therapeutic efficacy. Unfortunately, there are no reports of the discontinuation due to their prevalence (Rohracher et al., 2018; impact of ICIs on seizure presence and how this affects outcomes Villanueva et al., 2022). Administration of PER concomitant with in GBM specifically. Therefore, the use of particular ICIs in the anti-tumorigenic treatment regimens in a patient population treatment of epileptogenic GBMs should be further investigated already vulnerable to psychiatric changes, secondary to the GBM, preclinically to validate their safe use in the clinic. must be carefully considered as adverse effects may be more CAR-T therapy is another T-cell mediated immunotherapy, frequent or severe. wherein a patient’s own T cells are harvested and engineered It is of note that ASDs are not disease-modifying drugs, they ex vivo to express chimeric antigen receptors (CARs) that do not prevent epileptogenesis but do reduce seizure burden. recognize an identified tumor antigen, e.g., EGFRvIII, before To our knowledge there are no known treatments that act to being re-transplanted (Sterner and Sterner, 2021). CAR-T prevent epileptogenesis in the context of GBM, however, please therapy application in GBM presents significant challenges Frontiers in Molecular Neuroscience | www.frontiersin.org 12 June 2022 | Volume 15 | Article 903115 Hills et al. Epileptogenesis in Glioblastoma namely due to intratumoral heterogeneity in antigen expression MDM repolarization (Carpentier et al., 2006, 2010). Intracranial and the high likelihood of target antigens also being expressed in infusion of litenimod (CpG-28) in a cohort of 34 recurrent normal tissue (Maggs et al., 2021). Pathological overactive T-cell GBM patients was associated with seizures of either generalized responses in CAR-T therapy commonly manifest as seizure- or focal semiology in 14 patients (Carpentier et al., 2010). associated toxicities such as cytokine release syndrome (CRS) The downstream actions of M1-polarized macrophages and or immune effector cell-associated neurotoxicity syndrome aforementioned evidence suggests that TAM reprogramming (ICANS; Shimabukuro-Vornhagen et al., 2018). Headaches, may possess at least an acute epileptogenic capability, and hallucinations, hemiparesis, and seizures are some of the most indicates a clear need for further preclinical investigation. prominent behavioral manifestations observed when CRS and Whilst it is clearly important that seizure presence and ICANS occur secondary to brain-targeted CAR-T therapy prevalence be a monitored variable in preclinical studies of novel (Shimabukuro-Vornhagen et al., 2018). The release of IL-1, TNF- anti-tumorigenic treatment regimens, it ultimately returns to a α, IFN-β, and IL-6 are postulated to play key roles in both balance of risk-benefit when translating these therapeutics into CRS and ICANS pathophysiology as they trigger an almost the clinic. If these therapies effectively eradicate the GBM, despite uncontrolled activation of microglia (Shimabukuro-Vornhagen increasing the risk of epilepsy, then this would be tolerable. et al., 2018; Siegler and Kenderian, 2020). In ICANS, activated However, if these therapies only extend survival by a few months CAR-T cells also provoke the release of pro-inflammatory and this time is associated with debilitating seizures, would this cytokines IL-1 and IL-6, which subsequently activate endothelial still be an acceptable outcome? cells of the BBB to drive its breakdown. The potentiation of significant neuroinflammation in peri-cavity tissue, when Surgical Resection and Recurrence coupled with BBB breakdown, mirrors an environment seen in Another equally important challenge in preclinical research is the GBM-associated seizure generation. Furthermore, CAR-T cells focus on pre-resective models of GBM, all current epileptological can remain in circulation for a prolonged time post-infusion, GBM research has been performed in pre-resection models. and so continued downstream pro-inflammatory signaling These represent a valuable method to investigate underlying may progressively encourage epileptogenesis in the vulnerable mechanisms but may not be as clinically relevant. The majority of peri-cavity area. Whilst CRS/ICANS and the acute neurological GBM patients are eligible for resective surgery to remove the bulk toxicity associated with them are predominantly reversible, tumor, and as a corollary, the most common clinical scenario the potentiation of epileptogenic processes has permanent involves a post-resection brain. The post-resection environment consequences and may result in paradoxical resistance to therapy demonstrates a substantially altered neural network, even vs. secondary to seizure-potentiated tumor growth. pre-resection, and as such the mechanisms of hyperexcitability The therapeutic potential of macrophage and seizure generation may also be altered (Figure 3). reprogramming/repolarization has attracted a great deal of Resection models of GBM have been developed (Kauer et al., interest in recent years, but it also represents an area of 2011; Momiyama et al., 2013; Bianco et al., 2017; Otvos uncertainty in its unknown relationship to epileptogenesis. et al., 2021; Tang et al., 2021), but their use has not been The majority of TAMs are monocyte-derived macrophages reported in the investigation of GBM-related epileptogenesis. (MDMs) that infiltrate the brain through the BBB and they Therefore, there is currently no evidence indicative of how account for up to 50% of the tumor mass (Parmigiani et al., mechanisms of excitability and seizure generation are altered in 2021). M1-polarized macrophages display primarily pro- a post-resection brain. Both pre-resective and recurrent GBMs inflammatory, anti-tumor attributes making them an attractive display considerable intratumoral heterogeneity, but recurrent therapeutic target. Whereas M2-polarized macrophages exhibit GBMs may also have a slightly different molecular phenotype predominantly anti-inflammatory features with pro-tumor than the tumor it originated from (Campos et al., 2016). The effects and are the subtype primarily associated with the GBM recurrent tumor acquires new genetic aberrations as it grows TME (Ma et al., 2021). Repolarization of TAMs within the TME (Kim et al., 2015), potentially resulting in different reactions focuses on therapeutic pro-inflammatory actions either in the to anti-tumor or anti-epileptic therapeutic regimens used to tumor proper, in the absence of resection, or in peri-cavity tissue, treat the pre-resective tumor. This has downstream effects negating the need for systemic application. However, this may on the ability to make research-informed decisions regarding have implications for epileptogenic GBMs as this area is highly anti-seizure treatment regimens post-resection in the clinic, vulnerable to the inflammation-induced alterations in NMDARs, and demonstrates an incomplete understanding of the impact AMPARs, and ion homeostasis that promote hyperexcitability. seizures have on disease progression. The anti-tumor consequences of macrophage repolarization are The recurrence of seizures following GBM resection could be mediated through increased production of inducible nitric oxide attributed to a phenomenon known as secondary epileptogenesis synthase (iNOS), TNF-α, IL-1β, IL-6, and chemokines associated (Figure 3), wherein an extended epileptogenic network enables with cytotoxic T cell recruitment (Duan and Luo, 2021). As epileptogenesis to continue following the removal of a primary previously stated, these are all mediators known to contribute focus (Morrell et al., 1956; Morrell, 1985; Scholly et al., 2017). to epileptogenesis. Whilst there are no reported investigations Morrell (1985) defined secondary epileptogenesis as a three-stage into macrophage reprogramming in epileptogenic GBMs, process in which an actively discharging epileptogenic region studies have acknowledged seizures as a common side effect gradually induces paroxysmal activity in the cellular components in cytosine-guanine oligonucleotide (CpG-ODN) mediated of distally connected normal neuronal network(s). In this Frontiers in Molecular Neuroscience | www.frontiersin.org 13 June 2022 | Volume 15 | Article 903115 Hills et al. Epileptogenesis in Glioblastoma FIGURE 3 | Changes in seizure characteristics during disease progression and treatment. There are a multitude of mechanisms that alter seizure characteristics and phenotype in GBM. The molecular mutations and consequently activated signaling pathways change throughout its progression and as it enters new areas of the brain. Upon resection, these mechanisms are altered again and often accompanied by a period of seizure cessation. Seizures may then be produced in the peri-cavity region due to pre-existing network disruption or tumor regrowth, or in an entirely different region due to secondary epileptogenesis. context, GBM can be considered the primary epileptogenic foci recognized that the peri-tumoral area is a primary source and the generation of seizures is predominantly dependent on of epileptiform activity and hyperexcitability, as previously its presence (stage 1). Following GBM resection, or removal discussed. It has been postulated that, as this area often of the primary focus, there is a period of temporary seizure goes un-resected, post-resection seizures may continue to cessation whilst the network re-adjusts to further reorganization, originate from the pre-existing epileptogenic locus in the aligning with stage 2. Finally, consistent with stage 3, underlying original peri-tumoral border. The process of epileptogenesis epileptiform activity in a secondary previously-connected focus and associated tumorigenic network alterations may potentially persists independently to manifest as post-resection seizures. continue in this peri-cavity area even independent of tumor This is a hypothesis not previously considered in experimental presence. Increased use of intraoperative electrocorticography or clinical literature, and evidently the controversial existence (ECoG) in GBM patients with pre-operative seizure history has of secondary epileptogenesis requires greater attention in the the potential to increase maximal resection margins to include context of GBM. greater portions of peri-tumoral epileptogenic loci. This would New-onset seizures or their recurrence post-tumor resection increase the likelihood of post-operative seizure freedom and is common, and have been independently associated with potentially influence tumor recurrence risk. tumor recurrence, but evidence also suggests a correlation to Additionally, the cycle of epileptogenesis may be potentiated the extent of resection (EOR). Generally, surgical treatment post-resection by the small populations of neoplastic of GBM can range from a minimally invasive surgical biopsy cells that invariably persist in the ‘‘normal’’ parenchyma to a craniotomy involving varying degrees of ‘‘total’’ tumor surrounding the resection cavity. These remaining glioma resection (McGirt et al., 2009; Brown et al., 2016; Tan et al., cells promote network reorganization of the peri-cavity area 2020). Increased EOR is independently associated with increased through further BBB breakdown (Figure 4A), and continued survival rates (McGirt et al., 2009), and this also correlates dysfunctional glutamatergic (Figure 4B), and GABAergic with seizure freedom. Patients who have undergone gross signaling (Figure 4C). Additionally, their release of MMP-9 total tumor resection (GTR; >80%) are much more likely works to degrade FS-PV+PNNs (Figure 4C). Therefore, the to be seizure-free post-operatively than sub-total resection hyperexcitable peri-cavity area in epileptogenic GBM patients is (SubTR; <70%; Liang et al., 2016). Still significantly higher highly likely to actively be undergoing epileptogenesis secondary rates of seizure freedom are seen with supra-total resection to the actions of these enduring glioma cells. These seizures- and (SupraTR; >100%) in the form of anterior temporal lobectomy glioma cell-driven aberrances will result in further progressive (ATL; Borger et al., 2021). Comprehensive investigations reorganization of peri-cavity neuronal networks, working in into the effect of seizures on risk of tumor recurrence and pathological synergy to promote recurrence. The multiple month how this is impacted by EOR have not yet been reported. latency to seizure generation and tumor recurrence following However, interesting initial evidence from a medium-sized initial GBM resection appears to fit with this timeline (Liang cohort study appears to indicate that patients with post-operative et al., 2016). It is pertinent to mention that the cumulative risk of seizures may exhibit less time to tumor recurrence vs. those de novo post-operative epilepsy is generally increased following without, independent of EOR (Liang et al., 2016). It is well all craniotomy procedures, but this is markedly higher in patients Frontiers in Molecular Neuroscience | www.frontiersin.org 14 June 2022 | Volume 15 | Article 903115 Hills et al. Epileptogenesis in Glioblastoma FIGURE 4 | Epileptogenic molecular mechanisms post-GBM resection. (A) The neuroinflammatory environment and BBB disruption originally potentiated by the tumor can also be transiently exacerbated by the resective surgery itself. Serum albumin infiltration evokes Kir4.1 downregulation on reactive astrocytes, enabling K+ accumulation. AQP4 expression is increased on glioma cells, decreasing the ECS. (B) Increased xCT expression on glioma cells and decreased GLT-1 expression on astrocytes leads to extracellular glutamate accumulation in the peri-cavity region. It’s post-synaptic action at NMDARs and CP-AMPARs increases neuronal excitability. (C) The remaining glioma cells continue to release MMP-9 which induces PNN degradation. GABAA R expression is almost absent on glioma cells to aid their proliferation. NKCC1 is upregulated and KCC2 downregulated on surrounding pyramidal neurons, reversing the Cl− gradient. The subsequent decreased inhibitory influence disinhibits pyramidal neuronal firing enabling seizure generation. ECS, extracellular space; BBB, blood-brain barrier; NMDAR, NMDA receptor; CP-AMPARs, Ca2+ -permeable AMPARs; MMPs, matrix metalloproteinases; PNNs, perineuronal nets. undergoing craniotomies for GBM resection (13.9% vs. 23.6% electronics. This represents an obstacle in current preclinical respectively; Giraldi et al., 2021). research, wherein the restricted ability to employ live imaging methods in conjunction with ECoG monitoring hinders the Novel Technologies for investigation of seizures and SDs in GBM. Electrophysiological Investigation in GBM The ideal preclinical scenario would incorporate By nature, the seizures that occur in models of GBM-related MRI-compatible multi-electrode, flexible, penetrating arrays epilepsy are spontaneous and so represent unique challenges. that can be placed at the peri-tumoral border, and are capable of A comprehensive investigation of epileptogenesis in GBM electrographically mapping the changes to network excitability models requires continuous electrophysiological monitoring associated with progressive tumor growth over multiple areas of epileptiform activity. This is typically performed through of the brain. The closest preclinical example of this is seen chronic implantation of electroencephalographic devices, in Hatcher et al. (2020) wherein in vivo Ca2+ fluorescence as it is impossible to predict seizure occurrence without imaging was used in conjunction with EEG monitoring in pro-convulsant intervention by pharmaceutical or stimulatory a CRISPR-Cas9 IUE model of GBM-associated epilepsy, means. Furthermore, the lack of overt behavioral manifestations with great success. However, Ca2+ imaging is a complex of seizure presence in experimental animals renders video technique not readily adopted nor clinically translatable. monitoring insufficient in the absence of accompanying This hurdle presents challenges in both transplantation and EEG recordings. Equipment to record epileptiform activity genetic models of GBM-related epilepsy as tumor volume long-term is widely accessible and readily employed in and growth rates can vary between animals. As a corollary, experimental epilepsy studies, commonly as tethered recordings previous conclusions reached about the causal relationship or subcutaneous battery implantation to power the recording between volume or growth patterns and the development headstage (Wykes et al., 2012). However, these technologies of epileptiform activity are more illustrative than certain are largely unsuitable for use in a progressive GBM model as statements. they primarily employ a single or few electrodes, rather than The presence of SDs in GBM-associated epilepsy further configurations that allow high-density mapping. If a rigid complicates electrophysiological investigations. The massive, electrode is invasively implanted into the brain, this could abrupt change in activity seen in SD results in cortical electrical result in mechanical and physical damage secondary to the potential changes over a wide spectral range. This primarily tumor growth, and a progressive attenuation of the signal as the manifests as slow potentials approaching the DC (or 0 Hz) level. recording site is enveloped by the tumor. Therefore, to effectively Negative potential shift is an important identifier of SD but is not monitor epileptiform activity in GBM, any electrophysiological measured using standard AC-coupled (>0.5 Hz) highpass filtered methods must include some form of multi-electrode array to electrophysiological methods. This can be partially overcome by account for the progressive expansion of the peri-tumoral border. the use of amplifiers capable of recording at lower high-pass filter Furthermore, whilst some electrophysiological techniques, limits (>0.02 Hz), or by using a DC-coupled system. However, i.e., epicortical arrays, can be used in conjunction with imaging substantial attenuation of the amplitude and distortion of the SD techniques, conventional electrodes are poorly suited due to the waveform is evident when using passive metal-based electrodes, presence of overwhelming artifact or magnetic interference with as well as issues with drift and amplifier saturation. The presence Frontiers in Molecular Neuroscience | www.frontiersin.org 15 June 2022 | Volume 15 | Article 903115 Hills et al. Epileptogenesis in Glioblastoma of SD has not yet been investigated intra- or post-operatively may promote such extensive anti-inflammatory signaling to in GBM patients, epileptogenic or otherwise, and only very suppress the immune system, that it indirectly transiently few experimental studies acknowledge its presence (Bouckaert discourages epileptogenesis. This may also offer a further et al., 2020; Hatcher et al., 2020). Current detection of SDs potential explanation for new-onset seizures post-resection. in humans is mediated using platinum electrodes. They allow Wherein the removal of the tumor may provoke such an for the unfiltered recording of DC shifts and have a relatively aggressive inflammatory immune response that it potentiates high signal-to-noise ratio (SNR), but the unfortunate presence seizure generation. of artifacts and transients makes them unsuitable for accurate Recent experimental evidence suggests that tumor-localized monitoring and mapping of SD in GBM. Experimental studies expansion of certain pathological glioma subpopulations employ solution-filled glass micropipettes and silver or silver enriched for epilepsy-associated genes may confer chloride (Ag/AgCl) wires to detect SDs (Hatcher et al., 2020), hyperexcitability secondary to neosynaptogenesis (Lin et al., but these offer little to no spatial resolution. Additionally, the 2017). Therefore, if the proportional expression of certain glioma inherent toxicity of Ag/AgCl makes it inappropriate for use in subpopulations in a tumor differ to not favor this particular clinical studies. population, then the changes incurred by the aforementioned There is a clear need for novel means of electrophysiological peri-tumoral mechanisms may not be cumulatively sufficient to recording that can sustain continuous and simultaneous evoke hyperexcitability and ultimately initiate epileptogenesis. recording of both high frequency synchronous epileptiform This is a link between genetic and acquired mechanisms activity and SD DC-shifts across the brain. Importantly, of GBM-associated epilepsy not previously delineated (Lin these technologies need to be compatible with a variety et al., 2017). The intratumoral heterogeneity of GBM, and of imaging modalities in order to draw direct conclusions its unique microenvironment, support this theory. This is relating to the interaction between tumor presence and indicative that many different pro-epileptogenic mechanisms seizure generation. The application of novel technologies must work in pathological synergy in order for seizures currently in development for preclinical investigations to occur. Further investigations are needed to unravel the of epileptogenesis in GBM are crucial to improving our interactions between genetic predisposition and acquired understanding. One such example of which is graphene- mechanisms of epilepsy in GBM as this is clearly an area that based solution-gated field-effect transistors (gSGFETs). could provide great insight into the determinants of seizure gSGFETs have the potential to fill this technology gap as presence in a population of GBM patients. Investigations in this they offer the means to record SDs, either independently vein would also support the transition to more personalized or concurrently with high-frequency activity, both epi- and therapeutics, as is an important current area of research intra-cortically with high spatiotemporal fidelity (Hébert et al., in GBM. 2018; Masvidal-Codina et al., 2019, 2021; Bonaccini Calia et al., Ultimately, a better understanding of the relationship between 2021). The potential for SDs to exacerbate tissue damage and GBM, epilepsy, and SDs will stem from further experimental promote mechanisms of hyperexcitability in GBM could have investigations. Whilst clinical studies have provided initial consequences for disease progression, prognosis, and resistance evidence connecting tumor location to seizure presence, the to therapies. reluctance of oncology studies to stray from the standard striatal injection site precludes further preclinical investigation. DISCUSSION Incorporating a variety of injection sites, i.e., mesial temporal or frontal lobe structures, into implantation models would It is clear that GBMs create a pro-epileptogenic environment help to further our understanding of how GBM growth in via a multitude of mechanisms, and thus the associated seizures different brain regions predisposes a patient to comorbidities. present unique clinical challenges. However, it is also known that Different areas of the brain are considerably different in some GBM patients do not present with, and never experience how their networks function. Therefore, certain networks tumor-associated seizures or epilepsy. Therefore, we may gain may communicate in ways that have a more favorable more mechanistic insight into what the important essential influence on the relationship between tumor growth and factors are for the development of epilepsy by understanding secondary epileptogenesis. Greater insight into how tumor why some patients do not develop epilepsy despite an aggressive location specifically affects the process of epileptogenesis tumor environment. will undoubtedly have clinical significance. Patients identified The answer may lie somewhere in the complex interaction as predisposed to developing seizures may receive slightly between genetic influences and acquired mechanisms of epilepsy. altered anti-tumorigenic regimens that also simultaneously There is a precedent of diseases that are associated with address the underlying network dysfunction occurring during epilepsy where seizures manifest only following treatment. In tumor-associated epileptogenesis, instead of just using an neurocysticercosis, intracranial cysts evade detection by the adjuvant ASD. immune system through various mechanisms, and it is only It is important to emphasize that novel technologies need to once they are treated that seizures are generated (Carpio be developed to comprehensively investigate the impact tumor et al., 1998; Prodjinotho et al., 2020; Espino et al., 2022). This location has on epileptogenesis. The ability to map seizure and master evasion of the immune system is an aspect mirrored SD propagation across multiple brain areas with high fidelity in GBM. In the GBM TME of some patients, the tumor is an aspect of electrophysiology that has eluded the research Frontiers in Molecular Neuroscience | www.frontiersin.org 16 June 2022 | Volume 15 | Article 903115 Hills et al. Epileptogenesis in Glioblastoma community up until very recently. Initial experimental evidence AUTHOR CONTRIBUTIONS shows that cortical involvement is progressive in preclinical GBM-epilepsy models (Hatcher et al., 2020). Therefore, the KH, KK, and RW developed the concept and designed the initial availability of both epi-dural and intra-cortical arrays is crucial manuscript. KH wrote the manuscript and produced the figures. to spatially map the relationship of epileptiform activity with RW and KK assisted with manuscript revisions. All authors the expansion of the peri-tumoral border. Evidently, the contributed to the article and approved the submitted version. spontaneous nature of epileptiform activity and SDs in this disease model necessitates 24/7 electrophysiological recording. FUNDING However, chronic monitoring is a substantial experimental and technological investment and therefore is not readily performed. This review is supported by the Medical Research Council The inaccessibility, and unavailability, of this technology is (MRC), grant reference MR/W019752/1. KH is sponsored by the a major obstacle to investigating seizures as a measured Graphene NOWNANO PhD programme at the University of variable in oncology studies. Many novel anti-tumorigenic Manchester. RW is supported by a Senior Research Fellowship treatments appear to have pro-epileptogenic potential, but awarded by the Worshipful Company of Pewterers. it is only with increased accessibility of electrophysiological technologies that seizures will be included and monitored in ACKNOWLEDGMENTS these studies. The field of GBM research overwhelmingly focuses on the development of novel treatment options. Therefore, it We would like to thank the institutions and funding agencies is crucially important that the effect seizures and SDs may for their support. We also thank Dr. Thomas Kisby and Ms. have on therapeutic efficacy is evaluated to best inform their Maria Stylianou for their assistance. All figures were created with clinical success. Biorender.com. REFERENCES Berger, M., Speckmann, E.-J., Pape, H. C., and Gorji, A. (2008). Spreading depression enhances human neocortical excitability in vitro. Cephalalgia 28, Aizawa, H., Sun, W., Sugiyama, K., Itou, Y., Aida, T., Cui, W., et al. (2020). 558–562. doi: 10.1111/j.1468-2982.2008.01556.x Glial glutamate transporter GLT-1 determines susceptibility to spreading Bezzi, P., Domercq, M., Brambilla, L., Galli, R., Schols, D., De Clercq, E., depression in the mouse cerebral cortex. Glia 68, 2631–2642. doi: 10.1002/glia. et al. (2001). CXCR4-activated astrocyte glutamate release via TNFα: 23874 amplification by microglia triggers neurotoxicity. Nat. Neurosci. 4, 702–710. Alvestad, S., Hammer, J., Hoddevik, E. H., Skare, Ø., Sonnewald, U., Amiry- doi: 10.1038/89490 Moghaddam, M., et al. (2013). Mislocalization of AQP4 precedes chronic Bianco, J., Bastiancich, C., Joudiou, N., Gallez, B., des Rieux, A., and Danhier, F. seizures in the kainate model of temporal lobe epilepsy. Epilepsy Res. 105, (2017). Novel model of orthotopic U-87 MG glioblastoma resection in athymic 30–41. doi: 10.1016/j.eplepsyres.2013.01.006 nude mice. J. Neurosci. Methods 284, 96–102. doi: 10.1016/j.jneumeth.2017.04. Alzial, G., Renoult, O., Paris, F., Gratas, C., Clavreul, A., and Pecqueur, C. 019 (2022). Wild-type isocitrate dehydrogenase under the spotlight in glioblastoma. Binder, D. K., Nagelhus, E. A., and Ottersen, O. P. (2012). Aquaporin-4 and Oncogene 41, 613–621. doi: 10.1038/s41388-021-02056-1 epilepsy. Glia 60, 1203–1214. doi: 10.1002/glia.22317 Armbruster, M., Naskar, S., Garcia, J., Sommer, M., Kim, E., Adam, Y., et al. Binder, D. K., Oshio, K., Ma, T., Verkman, A. S., and Manley, G. T. (2004a). (2021). Neuronal activity drives pathway-specific depolarization of astrocyte Increased seizure threshold in mice lacking aquaporin-4 water channels. distal processes. bioRxiv [Preprint]. 2021.07.03.450922. doi: 10.1101/2021.07. Neuroreport 15, 259–262. doi: 10.1097/00001756-200402090-00009 03.450922 Binder, D. K., Papadopoulos, M. C., Haggie, P. M., and Verkman, A. S. (2004b). Arvanitis, C. D., Ferraro, G. B., and Jain, R. K. (2020). The blood-brain barrier and in vivo measurement of brain extracellular space diffusion by cortical surface blood-tumour barrier in brain tumours and metastases. Nat. Rev. Cancer 20, photobleaching. J. Neurosci. 24, 8049–8056. doi: 10.1523/JNEUROSCI.2294-04. 26–41. doi: 10.1038/s41568-019-0205-x 2004 Asahi, M., Wang, X., Mori, T., Sumii, T., Jung, J. C., Moskowitz, M. A., et al. Binder, D. K., and Steinhäuser, C. (2006). Functional changes in astroglial cells in (2001). Effects of matrix metalloproteinase-9 gene knock-out on the proteolysis epilepsy. Glia 54, 358–368. doi: 10.1002/glia.20394 of blood-brain barrier and white matter components after cerebral ischemia. Binder, D. K., Yao, X., Zador, Z., Sick, T. J., Verkman, A. S., and Manley, G. T. J. Neurosci. 21, 7724–7732. doi: 10.1523/JNEUROSCI.21-19-07724.2001 (2006). Increased seizure duration and slowed potassium kinetics in mice Augustin, K., Williams, S., Cunningham, M., Devlin, A. M., Friedrich, M., lacking aquaporin-4 water channels. Glia 53, 631–636. doi: 10.1002/glia.20318 Jayasekera, A., et al. (2018). Perampanel and decanoic acid show synergistic Blanchart, A., Fernando, R., Häring, M., Assaife-Lopes, N., Romanov, R. A., action against AMPA receptors and seizures. Epilepsia 59, e172–e178. Andäng, M., et al. (2017). Endogenous GABAA receptor activity suppresses doi: 10.1111/epi.14578 glioma growth. Oncogene 36, 777–786. doi: 10.1038/onc.2016.245 Ayata, C., and Lauritzen, M. (2015). Spreading Depression, Spreading Bobustuc, G. C., Baker, C. H., Limaye, A., Jenkins, W. D., Pearl, G., Depolarizations and the Cerebral Vasculature. Physiol. Rev. 95, 953–993. Avgeropoulos, N. G., et al. (2010). Levetiracetam enhances p53-mediated doi: 10.1152/physrev.00027.2014 MGMT inhibition and sensitizes glioblastoma cells to temozolomide. Neuro Backman, S. A., Stambolic, V., Suzuki, A., Haight, J., Elia, A., Pretorius, J., et al. Oncol. 12, 917–927. doi: 10.1093/neuonc/noq044 (2001). Deletion of Pten in mouse brain causes seizures, ataxia and defects Bonaccini Calia, A., Masvidal-Codina, E., Smith, T. M., Schäfer, N., Rathore, D., in soma size resembling lhermitte-duclos disease. Nat. Genet. 29, 396–403. Rodríguez-Lucas, E., et al. (2021). Full-bandwidth electrophysiology of seizures doi: 10.1038/ng782 and epileptiform activity enabled by flexible graphene microtransistor depth Baskar, R., Dai, J., Wenlong, N., Yeo, R., and Yeoh, K.-W. (2014). Biological neural probes. Nat. Nanotechnol. 17, 301–309. doi: 10.1038/s41565-021 response of cancer cells to radiation treatment. Front. Mol. Biosci. 1:24. -01041-9 doi: 10.3389/fmolb.2014.00024 Borger, V., Hamed, M., Ilic, I., Potthoff, A.-L., Racz, A., Schäfer, N., et al. (2021). Berendsen, S., Varkila, M., Kroonen, J., Seute, T., Snijders, T. J., Kauw, F., Seizure outcome in temporal glioblastoma surgery: lobectomy as a supratotal et al. (2016). Prognostic relevance of epilepsy at presentation in glioblastoma resection regime outclasses conventional gross-total resection. J. Neurooncol. patients. Neuro Oncol. 18, 700–706. doi: 10.1093/neuonc/nov238 152, 339–346. doi: 10.1007/s11060-021-03705-x Frontiers in Molecular Neuroscience | www.frontiersin.org 17 June 2022 | Volume 15 | Article 903115 Hills et al. Epileptogenesis in Glioblastoma Bouckaert, C., Germonpré, C., Verhoeven, J., Chong, S.-A., Jacquin, L., Mairet- Coppola, A., Zarabla, A., Maialetti, A., Villani, V., Koudriavtseva, T., Russo, E., Coello, G., et al. (2020). Development of a rat model for glioma-related epilepsy. et al. (2020). Perampanel confirms to be effective and well-tolerated as an Int. J. Mol. Sci. 21:6999. doi: 10.3390/ijms21196999 add-on treatment in patients with brain tumor-related epilepsy (PERADET Brito da Silva, A., Pennifold, J., Henley, B., Chatterjee, K., Bateman, D., Study). Front. Neurol. 11:592. doi: 10.3389/fneur.2020.00592 Whittaker, R. W., et al. (2021). The AMPA receptor antagonist perampanel de Curtis, M., Librizzi, L., and Avanzini, G. (2015). ‘‘Chapter 6 - Mechanisms of suppresses epileptic activity in human focal cortical dysplasia. Epilepsia Open Focal Epileptogenesis,’’ in Epilepsy and Brain Tumors, eds H. B. Newton, and [Preprint]. doi: 10.1002/epi4.12549 M. Maschio (Amsterdam: Academic Press), 101–109 Bronisz, E., and Kurkowska-Jastrz˛ebska, I. (2016). Matrix metalloproteinase 9 in de Groot, J. F., Liu, T. J., Fuller, G., and Yung, W. K. A. (2005). The excitatory epilepsy: the role of neuroinflammation in seizure development. Mediators amino acid transporter-2 induces apoptosis and decreases glioma growth Inflamm. 2016:7369020. doi: 10.1155/2016/7369020 in vitro and in vivo. Cancer Res. 65, 1934–1940. doi: 10.1158/0008-5472.CAN- Brown, T. J., Brennan, M. C., Li, M., Church, E. W., Brandmeir, N. J., 04-3626 Rakszawski, K. L., et al. (2016). Association of the extent of resection with de Groot, J. F., Piao, Y., Lu, L., Fuller, G. N., and Yung, W. K. A. (2008). survival in glioblastoma: a systematic review and meta-analysis. JAMA Oncol. Knockdown of GluR1 expression by RNA interference inhibits glioma 2, 1460–1469. doi: 10.1001/jamaoncol.2016.1373 proliferation. J. Neurooncol. 88, 121–133. doi: 10.1007/s11060-008-9552-2 Bruna, J., Miró, J., and Velasco, R. (2013). Epilepsy in glioblastoma patients: basic DeCordova, S., Shastri, A., Tsolaki, A. G., Yasmin, H., Klein, L., Singh, S. K., et al. mechanisms and current problems in treatment. Expert Rev. Clin. Pharmacol. (2020). Molecular heterogeneity and immunosuppressive microenvironment 6, 333–344. doi: 10.1586/ecp.13.12 in glioblastoma. Front. Immunol 11:1402. doi: 10.3389/fimmu.2020.01402 Buckingham, S. C., Campbell, S. L., Haas, B. R., Montana, V., Robel, S., Dittmer, D., Pati, S., Zambetti, G., Chu, S., Teresky, A. K., Moore, M., et al. Ogunrinu, T., et al. (2011). Glutamate release by primary brain tumors induces (1993). Gain of function mutations in p53. Nat. Genet. 4, 42–46. doi: 10.1038/ epileptic activity. Nat. Med. 17, 1269–1274. doi: 10.1038/nm.2453 ng0593-42 Buckingham, S. C., and Robel, S. (2013). Glutamate and tumor-associated epilepsy: Duan, L., and Di, Q. (2017). Acetazolamide suppresses multi-drug resistance- glial cell dysfunction in the peritumoral environment. Neurochem. Int. 63, related protein 1 and p-glycoprotein expression by inhibiting aquaporins 696–701. doi: 10.1016/j.neuint.2013.01.027 expression in a mesial temporal epilepsy rat model. Med. Sci. Monit. 23, Campbell, I. L., Abraham, C. R., Masliah, E., Kemper, P., Inglis, J. D., 5818–5825. doi: 10.12659/msm.903855 Oldstone, M. B., et al. (1993). Neurologic disease induced in transgenic mice Duan, Z., and Luo, Y. (2021). Targeting macrophages in cancer immunotherapy. by cerebral overexpression of interleukin 6. Proc. Natl. Acad. Sci. U S A 90, Signal Transduct. Target Ther. 6:127. doi: 10.1038/s41392-021-00506-6 10061–10065. doi: 10.1073/pnas.90.21.10061 Dubois, L. G., Campanati, L., Righy, C., D’Andrea-Meira, I., de Sampaio e Campbell, S. L., Buckingham, S. C., and Sontheimer, H. (2012). Human glioma Spohr, C. L., Porto-Carreiro, I., et al. (2014). Gliomas and the vascular fragility cells induce hyperexcitability in cortical networks. Epilepsia 53, 1360–1370. of the blood brain barrier. Front. Cell. Neurosci. 8:418. doi: 10.3389/fncel.2014. doi: 10.1111/j.1528-1167.2012.03557.x 00418 Campbell, S. L., Robel, S., Cuddapah, V. A., Robert, S., Buckingham, S. C., Dührsen, L., Sauvigny, T., Ricklefs, F. L., Mende, K.-C., Schaper, M., Matschke, J., Kahle, K. T., et al. (2015). GABAergic disinhibition and impaired et al. (2019). Seizures as presenting symptom in patients with glioblastoma. KCC2 cotransporter activity underlie tumor-associated epilepsy. Glia 63, Epilepsia 60, 149–154. doi: 10.1111/epi.14615 23–36. doi: 10.1002/glia.22730 Dunn-Pirio, A. M., Woodring, S., Lipp, E., Herndon, J. E., Healy, P., Weant, M., Campos, B., Olsen, L. R., Urup, T., and Poulsen, H. S. (2016). A comprehensive et al. (2018). Adjunctive perampanel for glioma-associated epilepsy. Epilepsy profile of recurrent glioblastoma. Oncogene 35, 5819–5825. doi: 10.1038/onc. Behav. Case Rep. 10, 114–117. doi: 10.1016/j.ebcr.2018.09.003 2016.85 Duy, P. Q., He, M., He, Z., and Kahle, K. T. (2020). Preclinical insights into Carpentier, A., Laigle-Donadey, F., Zohar, S., Capelle, L., Behin, A., Tibi, A., et al. therapeutic targeting of KCC2 for disorders of neuronal hyperexcitability. (2006). Phase 1 trial of a CpG oligodeoxynucleotide for patients with recurrent Expert Opin. Ther. Targets 24, 629–637. doi: 10.1080/14728222.2020.1762174 glioblastoma. Neuro Oncol. 8, 60–66. doi: 10.1215/S1522851705000475 Endersby, R., and Baker, S. J. (2008). PTEN signaling in brain: neuropathology and Carpentier, A., Metellus, P., Ursu, R., Zohar, S., Lafitte, F., Barrié, M., et al. tumorigenesis. Oncogene 27, 5416–5430. doi: 10.1038/onc.2008.239 (2010). Intracerebral administration of CpG oligonucleotide for patients Engel, T., Murphy, B. M., Schindler, C. K., and Henshall, D. C. (2007). with recurrent glioblastoma: a phase II study. Neuro Oncol. 12, 401–408. Elevated p53 and lower MDM2 expression in hippocampus from patients with doi: 10.1093/neuonc/nop047 intractable temporal lobe epilepsy. Epilepsy Res. 77, 151–156. doi: 10.1016/j. Carpio, A., Escobar, A., and Hauser, W. A. (1998). Cysticercosis and epilepsy: eplepsyres.2007.09.001 a critical review. Epilepsia 39, 1025–1040. doi: 10.1111/j.1528-1157.1998. Englot, D. J., Berger, M. S., Chang, E. F., and Garcia, P. A. (2012). Characteristics tb01287.x and treatment of seizures in patients with high-grade glioma: a review. Chaichana, K. L., Parker, S. L., Olivi, A., and Quiñones-Hinojosa, A. (2009). Neurosurg. Clin. N. Am. 23, 227–235. doi: 10.1016/j.nec.2012.01.009 Long-term seizure outcomes in adult patients undergoing primary resection Ertürk Çetin, Ö., İşler, C., Uzan, M., and Özkara, Ç. (2017). Epilepsy-related brain of malignant brain astrocytomas. Clinical article. J. Neurosurg. 111, 282–292. tumors. Seizure 44, 93–97. doi: 10.1016/j.seizure.2016.12.012 doi: 10.3171/2009.2.JNS081132 Espino, P. H., Couper, R. G., and Burneo, J. G. (2022). An update on Chalifoux, R., and Elisevich, K. (1997). Effect of ionizing radiation on Neurocysticercosis-related epilepsy. Clin. Neurol. Neurosurg. 213:107139. partial seizures attributable to malignant cerebral tumors. Stereotact. Funct. doi: 10.1016/j.clineuro.2022.107139 Neurosurg. 67, 169–182. doi: 10.1159/000099446 Fanelli, G. N., Grassini, D., Ortenzi, V., Pasqualetti, F., Montemurro, N., Chavez-Blanco, A., Perez-Plasencia, C., Perez-Cardenas, E., Carrasco-Legleu, C., Perrini, P., et al. (2021). Decipher the glioblastoma microenvironment: the first Rangel-Lopez, E., Segura-Pacheco, B., et al. (2006). Antineoplastic effects of the milestone for new groundbreaking therapeutic strategies. Genes (Basel). 12:445. DNA methylation inhibitor hydralazine and the histone deacetylase inhibitor doi: 10.3390/genes12030445 valproic acid in cancer cell lines. Cancer Cell Int. 6:2. doi: 10.1186/1475-2867- Fernandes, C., Costa, A., Osorio, L., Costa Lago, R., Linhares, P., Carvalho, B., 6-2 et al. (2017). ‘‘Chapter 11 Current Standards of Care in Glioblastoma Therapy,’’ Climans, S. A., Brandes, A. A., Cairncross, J. G., Ding, K., Fay, M., Laperriere, N., in Glioblastoma [Internet], ed S. De Vleeschouwer (Brisbane, QLD: Codon et al. (2020). Temozolomide and seizure outcomes in a randomized clinical trial Publications). of elderly glioblastoma patients. J. Neurooncol. 149, 65–71. doi: 10.1007/s11060- Feyissa, A. M., Worrell, G. A., Tatum, W. O., Chaichana, K. L., Jentoft, M. E., 020-03573-x Guerrero Cazares, H., et al. (2019). Potential influence of IDH1 mutation and Conti, L., Palma, E., Roseti, C., Lauro, C., Cipriani, R., de Groot, M., MGMT gene promoter methylation on glioma-related preoperative seizures et al. (2011). Anomalous levels of Cl− transporters cause a and postoperative seizure control. Seizure 69, 283–289. doi: 10.1016/j.seizure. decrease of GABAergic inhibition in human peritumoral epileptic 2019.05.018 cortex. Epilepsia 52, 1635–1644. doi: 10.1111/j.1528-1167.2011. Frampton, J. E. (2015). Perampanel: a review in drug-resistant epilepsy. Drugs 75, 03111.x 1657–1668. doi: 10.1007/s40265-015-0465-z Frontiers in Molecular Neuroscience | www.frontiersin.org 18 June 2022 | Volume 15 | Article 903115 Hills et al. Epileptogenesis in Glioblastoma Gao, X., Wang, H., Cai, S., Saadatzadeh, M. R., Hanenberg, H., Pollok, K. E., Huberfeld, G., and Vecht, C. J. (2016). Seizures and gliomas--towards a single et al. (2014). Phosphorylation of NMDA 2B at S1303 in human glioma therapeutic approach. Nat. Rev. Neurol. 12, 204–216. doi: 10.1038/nrneurol. peritumoral tissue: implications for glioma epileptogenesis. Neurosurg. Focus 2016.26 37:E17. doi: 10.3171/2014.9.FOCUS14485 Huberfeld, G., Wittner, L., Clemenceau, S., Baulac, M., Kaila, K., Miles, R., et al. Garzon-Muvdi, T., Schiapparelli, P., ap Rhys, C., Guerrero-Cazares, H., (2007). Perturbed chloride homeostasis and gabaergic signaling in human Smith, C., Kim, D.-H., et al. (2012). Regulation of brain tumor dispersal temporal lobe epilepsy. J. Neurosci. 27, 9866–9873. doi: 10.1523/JNEUROSCI. by NKCC1 through a novel role in focal adhesion regulation. PLoS Biol. 2761-07.2007 10:e1001320. doi: 10.1371/journal.pbio.1001320 Hübner, C. A., Lorke, D. E., and Hermans-Borgmeyer, I. (2001). Expression of the Giraldi, L., Vinsløv Hansen, J., Wohlfahrt, J., Fugleholm, K., Melbye, M., Na-K-2Cl-cotransporter NKCC1 during mouse development. Mech. Dev. 102, Munch, T. N., et al. (2021). Postoperative de novo epilepsy after craniotomy: 267–269. doi: 10.1016/s0925-4773(01)00309-4 a nationwide register-based cohort study. J. Neurol. Neurosurg. Psychiatry 93, Ikeshima-Kataoka, H. (2016). Neuroimmunological Implications of AQP4 in 436–444. doi: 10.1136/jnnp-2021-326968 Astrocytes. Int. J. Mol. Sci. 17:1306. doi: 10.3390/ijms17081306 Goodwin, C. R., and Laterra, J. (2010). Neuro-oncology: unmasking Isoardo, G., Morra, I., Chiarle, G., Audrito, V., Deaglio, S., Melcarne, A., et al. the multiforme in glioblastoma. Nat. Rev. Neurol. 6, 304–305. (2012). Different aquaporin-4 expression in glioblastoma multiforme patients doi: 10.1038/nrneurol.2010.67 with and without seizures. Mol. Med. 18, 1147–1151. doi: 10.2119/molmed. Greenfield, L. J. Jr. (2013). Molecular mechanisms of antiseizure drug activity at 2012.00015 GABAA receptors. Seizure 22, 589–600. doi: 10.1016/j.seizure.2013.04.015 Ivens, S., Kaufer, D., Flores, L. P., Bechmann, I., Zumsteg, D., Tomkins, O., et al. Guan, X., Hasan, M. N., Maniar, S., Jia, W., and Sun, D. (2018). Reactive (2007). TGF-beta receptor-mediated albumin uptake into astrocytes is involved astrocytes in glioblastoma multiforme. Mol. Neurobiol. 55, 6927–6938. in neocortical epileptogenesis. Brain 130, 535–547. doi: 10.1093/brain/awl317 doi: 10.1007/s12035-018-0880-8 Izumoto, S., Miyauchi, M., Tasaki, T., Okuda, T., Nakagawa, N., Nakano, N., Haas, B. R., and Sontheimer, H. (2010). Inhibition of the sodium-potassium- et al. (2018). Seizures and tumor progression in glioma patients with chloride cotransporter isoform-1 reduces glioma invasion. Cancer Res. 70, uncontrollable epilepsy treated with perampanel. Anticancer Res. 38, 5597–5606. doi: 10.1158/0008-5472.CAN-09-4666 4361–4366. doi: 10.21873/anticanres.12737 Han, F., Hu, R., Yang, H., Liu, J., Sui, J., Xiang, X., et al. (2016). PTEN gene Jin, F., Jin-Lee, H., and Johnson, A. (2021). ‘‘Mouse models of experimental mutations correlate to poor prognosis in glioma patients: a meta-analysis. Onco glioblastoma,’’ in Gliomas, ed W. Debinski (Brisbane, QLD: Exon Publications). Targets Ther. 9, 3485–3492. doi: 10.2147/OTT.S99942 Jung, E., Alfonso, J., Osswald, M., Monyer, H., Wick, W., and Winkler, F. Hanada, T., Hashizume, Y., Tokuhara, N., Takenaka, O., Kohmura, N., (2019). Emerging intersections between neuroscience and glioma biology. Nat. Ogasawara, A., et al. (2011). Perampanel: a novel, orally active, noncompetitive Neurosci. 22, 1951–1960. doi: 10.1038/s41593-019-0540-y AMPA-receptor antagonist that reduces seizure activity in rodent models of Kastenhuber, E. R., and Lowe, S. W. (2017). Putting p53 in context. Cell 170, epilepsy. Epilepsia 52, 1331–1340. doi: 10.1111/j.1528-1167.2011.03109.x 1062–1078. doi: 10.1016/j.cell.2017.08.028 Hatcher, A., Yu, K., Meyer, J., Aiba, I., Deneen, B., and Noebels, J. L. (2020). Kauer, T. M., Figueiredo, J.-L., Hingtgen, S., and Shah, K. (2011). Encapsulated Pathogenesis of peritumoral hyperexcitability in an immunocompetent therapeutic stem cells implanted in the tumor resection cavity induce cell death CRISPR-based glioblastoma model. J. Clin. Invest. 130, 2286–2300. in gliomas. Nat. Neurosci. 15, 197–204. doi: 10.1038/nn.3019 doi: 10.1172/JCI133316 Kerkhof, M., Dielemans, J. C. M., van Breemen, M. S., Zwinkels, H., Hébert, C., Masvidal-Codina, E., Suarez-Perez, A., Calia, A. B., Piret, G., Garcia- Walchenbach, R., Taphoorn, M. J., et al. (2013). Effect of valproic acid on Cortadella, R., et al. (2018). Flexible graphene solution-gated field-effect seizure control and on survival in patients with glioblastoma multiforme. Neuro transistors: efficient transducers for micro-electrocorticography. Adv. Funct. Oncol. 15, 961–967. doi: 10.1093/neuonc/not057 Mater. 28:1703976. doi: 10.1002/adfm.201703976 Kerkhof, M., and Vecht, C. J. (2013). Seizure characteristics and prognostic factors Henker, C., Kriesen, T., Scherer, M., Glass, Ä., von Deimling, A., Bendszus, M., of gliomas. Epilepsia 54, 12–17. doi: 10.1111/epi.12437 et al. (2019). Association between tumor compartment volumes, the incidence Khandwala, K., Mubarak, F., and Minhas, K. (2021). The many faces of of pretreatment seizures and statin-mediated protective effects in glioblastoma. glioblastoma: pictorial review of atypical imaging features. Neuroradiol. J. 34, Neurosurgery 85, E722–E729. doi: 10.1093/neuros/nyz079 33–41. doi: 10.1177/1971400920965970 Henley, J. M., and Wilkinson, K. A. (2016). Synaptic AMPA receptor composition Kim, H., Zheng, S., Amini, S. S., Virk, S. M., Mikkelsen, T., Brat, D. J., et al. in development, plasticity and disease. Nat. Rev. Neurosci. 17, 337–350. (2015). Whole-genome and multisector exome sequencing of primary and doi: 10.1038/nrn.2016.37 post-treatment glioblastoma reveals patterns of tumor evolution. Genome Res. Hibino, H., Inanobe, A., Furutani, K., Murakami, S., Findlay, I., and Kurachi, Y. 25, 316–327. doi: 10.1101/gr.180612.114 (2010). Inwardly rectifying potassium channels: their structure, function and Knüpfer, M. M., Hernáiz-Driever, P., Poppenborg, H., Wolff, J. E., and physiological roles. Physiol. Rev. 90, 291–366. doi: 10.1152/physrev.00021.2009 Cinatl, J. (1998). Valproic acid inhibits proliferation and changes expression Hori, T., Sasayama, T., Tanaka, K., Koma, Y.-i., Nishihara, M., Tanaka, H., et al. of CD44 and CD56 of malignant glioma cells in vitro. Anticancer Res. 18, (2019). Tumor-associated macrophage related interleukin-6 in cerebrospinal 3585–3589. fluid as a prognostic marker for glioblastoma. J. Clin. Neurosci. 68, 281–289. Koekkoek, J. A. F., Dirven, L., Reijneveld, J. C., Postma, T. J., Grant, R., Pace, A., doi: 10.1016/j.jocn.2019.07.020 et al. (2014). Epilepsy in the end of life phase of brain tumor patients: a Houser, C., Zhang, N., and Peng, N. (2012). ‘‘Alterations in the distribution of systematic review. Neurooncol. Pract. 1, 134–140. doi: 10.1093/nop/npu018 gabaa receptors in epilepsy,’’ in Jasper’s Basic Mechanisms of the Epilepsies, Köhling, R., Senner, V., Paulus, W., and Speckmann, E.-J. (2006). Epileptiform 4th Edn., eds J. L. Noebels, M. Avoli, M. A. Rogawski, R. W. Olsen, and activity preferentially arises outside tumor invasion zone in glioma A. V. Delgado-Escueta (Bethesda, MD: Oxford University Press). xenotransplants. Neurobiol. Dis. 22, 64–75. doi: 10.1016/j.nbd.2005. Hovinga, C. A. (2001). Levetiracetam: a novel antiepileptic drug. Pharmacotherapy 10.001 21, 1375–1388. doi: 10.1592/phco.21.17.1375.34432 Kramer, D. R., Fujii, T., Ohiorhenuan, I., and Liu, C. Y. (2017). Interplay between Hu, S., Kao, H.-Y., Yang, T., and Wang, Y. (2022). Early and Bi-hemispheric cortical spreading depolarization and seizures. Stereotact. Funct. Neurosurg. 95, seizure onset in a rat glioblastoma multiforme model. Neurosci. Lett. 1–5. doi: 10.1159/000452841 766:136351. doi: 10.1016/j.neulet.2021.136351 Krauze, A. V., Myrehaug, S. D., Chang, M. G., Holdford, D. J., Smith, S., Hu, S., Sheng, W. S., Ehrlich, L. C., Peterson, P. K., and Chao, C. C. Shih, J., et al. (2015). A phase 2 study of concurrent radiation therapy, (2000). Cytokine effects on glutamate uptake by human astrocytes. temozolomide and the histone deacetylase inhibitor valproic acid for patients Neuroimmunomodulation 7, 153–159. doi: 10.1159/000026433 with glioblastoma. Int. J. Radiat. Oncol. Biol. Phys. 92, 986–992. doi: 10.1016/j. Hubbard, J. A., Szu, J. I., Yonan, J. M., and Binder, D. K. (2016). Regulation of ijrobp.2015.04.038 astrocyte glutamate transporter-1 (GLT1) and aquaporin-4 (AQP4). expression Kucheryavykh, Y. V., Kucheryavykh, L. Y., Nichols, C. G., Maldonado, H. M., in a model of epilepsy. Exp. Neurol. 283, 85–96. doi: 10.1016/j.expneurol.2016. Baksi, K., Reichenbach, A., et al. (2007). Downregulation of Kir4.1 inward 05.003 rectifying potassium channel subunits by RNAi impairs potassium transfer Frontiers in Molecular Neuroscience | www.frontiersin.org 19 June 2022 | Volume 15 | Article 903115 Hills et al. Epileptogenesis in Glioblastoma and glutamate uptake by cultured cortical astrocytes. Glia 55, 274–281. Lu, V. M., Texakalidis, P., McDonald, K. L., Mekary, R. A., and Smith, T. R. (2018). doi: 10.1002/glia.20455 The survival effect of valproic acid in glioblastoma and its current trend: a Labrakakis, C., Patt, S., Hartmann, J., and Kettenmann, H. (1998). Functional systematic review and meta-analysis. Clin. Neurol. Neurosurg. 174, 149–155. GABAA receptors on human glioma cells. Eur. J. Neurosci. 10, 231–238. doi: 10.1016/j.clineuro.2018.09.019 doi: 10.1046/j.1460-9568.1998.00036.x Lu, T., Tian, L., Han, Y., Vogelbaum, M., and Stark, G. R. (2007). Lange, F., Hartung, J., Liebelt, C., Boisserée, J., Resch, T., Porath, K., et al. Dose-dependent cross-talk between the transforming growth factor-β and (2020). Perampanel add-on to standard radiochemotherapy in vivo promotes interleukin-1 signaling pathways. Proc. Natl. Acad. Sci. U S A 104, 4365–4370. neuroprotection in a rodent f98 glioma model. Front. Neurosci. 14:598266. doi: 10.1073/pnas.0700118104 doi: 10.3389/fnins.2020.598266 Luikart, B. W., Schnell, E., Washburn, E. K., Bensen, A. L., Tovar, K. R., Lange, F., Weßlau, K., Porath, K., Hörnschemeyer, J., Bergner, C., Krause, B. J., and Westbrook, G. L. (2011). Pten knockdown in vivo increases et al. (2019). AMPA receptor antagonist perampanel affects glioblastoma excitatory drive onto dentate granule cells. J. Neurosci. 31, 4345–4354. cell growth and glutamate release in vitro. PLoS One 14:e0211644. doi: 10.1523/JNEUROSCI.0061-11.2011 doi: 10.1371/journal.pone.0211644 Luo, L., Guan, X., Begum, G., Ding, D., Gayden, J., Hasan, M. N., et al. (2020). Leclercq, K., Liefferinge, J. V., Albertini, G., Neveux, M., Dardenne, S., Mairet- Blockade of cell volume regulatory protein NKCC1 increases TMZ-induced Coello, G., et al. (2019). Anticonvulsant and antiepileptogenic effects of glioma apoptosis and reduces astrogliosis. Mol. Cancer Ther. 19, 1550–1561. system xc- inactivation in chronic epilepsy models. Epilepsia 60, 1412–1423. doi: 10.1158/1535-7163.MCT-19-0910 doi: 10.1111/epi.16055 Lynch, B. A., Lambeng, N., Nocka, K., Kensel-Hammes, P., Bajjalieh, S. M., Lee, H. H. C., Deeb, T. Z., Walker, J. A., Davies, P. A., and Moss, S. J. Matagne, A., et al. (2004). The synaptic vesicle protein SV2A is the binding (2011). NMDA receptor activity downregulates KCC2 resulting in site for the antiepileptic drug levetiracetam. Proc. Natl. Acad. Sci. U S A 101, depolarizing GABAA receptor-mediated currents. Nat. Neurosci. 14, 736–743. 9861–9866. doi: 10.1073/pnas.0308208101 doi: 10.1038/nn.2806 Lyons, S. A., Chung, W. J., Weaver, A. K., Ogunrinu, T., and Sontheimer, H. Lee, T. S., Eid, T., Mane, S., Kim, J. H., Spencer, D. D., Ottersen, O. P., et al. (2004). (2007). Autocrine glutamate signaling promotes glioma cell invasion. Cancer Aquaporin-4 is increased in the sclerotic hippocampus in human temporal lobe Res. 67, 9463–9471. doi: 10.1158/0008-5472.CAN-07-2034 epilepsy. Acta Neuropathol. 108, 493–502. doi: 10.1007/s00401-004-0910-7 Ma, J., Chen, C. C., and Li, M. (2021). Macrophages/microglia in the glioblastoma Lee, H. H. C., Jurd, R., and Moss, S. J. (2010). Tyrosine phosphorylation regulates tumor microenvironment. Int. J. Mol. Sci. 22:5775. doi: 10.3390/ijms22115775 the membrane trafficking of the potassium chloride co-transporter KCC2. Mol. Maas, S., Patt, S., Schrey, M., and Rich, A. (2001). Underediting of glutamate Cell. Neurosci. 45, 173–179. doi: 10.1016/j.mcn.2010.06.008 receptor GluR-B mRNA in malignant gliomas. Proc. Natl. Acad. Sci. U S A 98, Lee, J. W., Wen, P. Y., Hurwitz, S., Black, P., Kesari, S., Drappatz, J., et al. (2010). 14687–14692. doi: 10.1073/pnas.251531398 Morphological characteristics of brain tumors causing seizures. Arch. Neurol. MacKenzie, G., and Maguire, J. (2015). Chronic stress shifts the GABA reversal 67, 336–342. doi: 10.1001/archneurol.2010.2 potential in the hippocampus and increases seizure susceptibility. Epilepsy Res. Lewerenz, J., Baxter, P., Kassubek, R., Albrecht, P., Van Liefferinge, J., 109, 13–27. doi: 10.1016/j.eplepsyres.2014.10.003 Westhoff, M.-A., et al. (2014). Phosphoinositide 3-kinases upregulate system MacKenzie, G., O’Toole, K. K., Moss, S. J., and Maguire, J. (2016). Compromised x−c via eukaryotic initiation factor 2α and activating transcription factor 4 - GABAergic inhibition contributes to tumor-associated epilepsy. Epilepsy Res. a pathway active in glioblastomas and epilepsy. Antioxid. Redox. Signal. 20, 126, 185–196. doi: 10.1016/j.eplepsyres.2016.07.010 2907–2922. doi: 10.1089/ars.2013.5455 Madadi, A., Wolfart, J., Lange, F., Brehme, H., Linnebacher, M., Bräuer, A. U., et al. Liang, S., Zhang, J., Zhang, S., and Fu, X. (2016). Epilepsy in adults with (2021). Correlation between Kir4.1 expression and barium-sensitive currents supratentorial glioblastoma: incidence and influence factors and prophylaxis in rat and human glioma cell lines. Neurosci. Lett. 741:135481. doi: 10.1016/j. in 184 patients. PLoS One 11:e0158206. doi: 10.1371/journal.pone.0158206 neulet.2020.135481 Lim, S. K., Llaguno, S. R. A., McKay, R. M., and Parada, L. F. (2011). Glioblastoma Maggs, L., Cattaneo, G., Dal, A. E., Moghaddam, A. S., and Ferrone, S. (2021). CAR multiforme: a perspective on recent findings in human cancer and mouse T cell-based immunotherapy for the treatment of glioblastoma. Front. Neurosci. models. BMB Rep. 44, 158–164. doi: 10.5483/BMBRep.2011.44.3.158 15:662064. doi: 10.3389/fnins.2021.662064 Lim, M., Omuro, A., Vlahovic, G., Reardon, D. A., Sahebjam, S., Cloughesy, T., Marchi, N., Angelov, L., Masaryk, T., Fazio, V., Granata, T., Hernandez, N., et al. et al. (2017). Nivolumab (nivo) in combination with radiotherapy (2007). Seizure-promoting effect of blood-brain barrier disruption. Epilepsia (RT) ± temozolomide (TMZ). Updated safety results from CheckMate 48, 732–742. doi: 10.1111/j.1528-1167.2007.00988.x 143 in pts with methylated or unmethylated newly diagnosed glioblastoma Marchi, N., Granata, T., Ghosh, C., and Janigro, D. (2012). Blood-brain barrier (GBM). Ann. Oncol. 28: v109. doi: 10.1093/annonc/mdx366 dysfunction and epilepsy: pathophysiologic role and therapeutic approaches. Lin, C.-L. G., Kong, Q., Cuny, G. D., and Glicksman, M. A. (2012). Glutamate Epilepsia 53, 1877–1886. doi: 10.1111/j.1528-1167.2012.03637.x transporter EAAT2: a new target for the treatment of neurodegenerative Maschio, M., Zarabla, A., Maialetti, A., Giannarelli, D., Koudriavtseva, T., diseases. Future Med. Chem. 4, 1689–1700. doi: 10.4155/fmc.12.122 Villani, V., et al. (2020). Perampanel in brain tumor-related epilepsy: Lin, W., Wang, C., Liu, G., Bi, C., Wang, X., Zhou, Q., et al. (2020). SLC7A11/xCT observational pilot study. Brain Behav. 10:e01612. doi: 10.1002/brb3.1612 in cancer: biological functions and therapeutic implications. Am. J. Cancer Res. Mastall, M., Wolpert, F., Gramatzki, D., Imbach, L., Becker, D., Schmick, A., et al. 10, 3106–3126. (2021). Survival of brain tumour patients with epilepsy. Brain 144, 3322–3327. Lin, C.-C. J., Yu, K., Hatcher, A., Huang, T.-W., Lee, H. K., Carlson, J., et al. (2017). doi: 10.1093/brain/awab188 Identification of diverse astrocyte populations and their malignant analogs. Masvidal-Codina, E., Illa, X., Dasilva, M., Calia, A. B., Dragojević, T., Vidal- Nat. Neurosci. 20, 396–405. doi: 10.1038/nn.4493 Rosas, E. E., et al. (2019). High-resolution mapping of infraslow cortical Liu, H., Sun, Y., Zhang, Q., Jin, W., Gordon, R. E., Zhang, Y., et al. (2021). Pro- brain activity enabled by graphene microtransistors. Nat. Mater. 18, 280–288. inflammatory and proliferative microglia drive progression of glioblastoma. doi: 10.1038/s41563-018-0249-4 Cell Rep. 36:109718. doi: 10.1016/j.celrep.2021.109718 Masvidal-Codina, E., Smith, T. M., Rathore, D., Gao, Y., Illa, X., Prats-Alfonso, E., Liu, R., Wang, J., Liang, S., Zhang, G., and Yang, X. (2020). Role of NKCC1 and et al. (2021). Characterization of optogenetically-induced cortical spreading KCC2 in epilepsy: from expression to function. Front. Neurol. 10:1407. depression in awake mice using graphene micro-transistor arrays. J. Neural doi: 10.3389/fneur.2019.01407 Eng. 18:55002. doi: 10.1088/1741-2552/abecf3 Liubinas, S. V., D’Abaco, G. M., Moffat, B. M., Gonzales, M., Feleppa, F., Mayer, J., Kirschstein, T., Resch, T., Porath, K., Krause, B. J., Köhling, R., et al. Nowell, C. J., et al. (2014). IDH1 mutation is associated with seizures (2020). Perampanel attenuates epileptiform phenotype in C6 glioma. Neurosci. and protoplasmic subtype in patients with low-grade gliomas. Epilepsia 55, Lett. 715:134629. doi: 10.1016/j.neulet.2019.134629 1438–1443. doi: 10.1111/epi.12662 McGirt, M. J., Chaichana, K. L., Gathinji, M., Attenello, F. J., Than, K., Olivi, A., Löscher, W., and Friedman, A. (2020). Structural, molecular and functional et al. (2009). Independent association of extent of resection with survival alterations of the blood-brain barrier during epileptogenesis and epilepsy: a in patients with malignant brain astrocytoma. J. Neurosurg. 110, 156–162. cause, consequence, or both? Int. J. Mol. Sci. 21:591. doi: 10.3390/ijms21020591 doi: 10.3171/2008.4.17536 Frontiers in Molecular Neuroscience | www.frontiersin.org 20 June 2022 | Volume 15 | Article 903115 Hills et al. Epileptogenesis in Glioblastoma Michaluk, P., Mikasova, L., Groc, L., Frischknecht, R., Choquet, D., and isocitrate dehydrogenase wildtype glioblastoma in adults: an observational Kaczmarek, L. (2009). Matrix metalloproteinase-9 controls NMDA receptor study. Neurology 98, e125–e140. doi: 10.1212/WNL.0000000000013005 surface diffusion through integrin beta1 signaling. J. Neurosci. 29, 6007–6012. Pallud, J., Le Van Quyen, M., Bielle, F., Pellegrino, C., Varlet, P., Cresto, N., et al. doi: 10.1523/JNEUROSCI.5346-08.2009 (2014). Cortical GABAergic excitation contributes to epileptic activities around Momiyama, M., Hiroshima, Y., Suetsugu, A., Tome, Y., Mii, S., Yano, S., et al. human glioma. Sci. Transl. Med. 6:244ra89. doi: 10.1126/scitranslmed.3008065 (2013). Enhanced resection of orthotopic red-fluorescent-protein-expressing Parmigiani, E., Scalera, M., Mori, E., Tantillo, E., and Vannini, E. (2021). Old stars human glioma by fluorescence-guided surgery in nude mice. Anticancer Res. and new players in the brain tumor microenvironment. Front. Cell. Neurosci. 33, 107–111. 15:709917. doi: 10.3389/fncel.2021.709917 Mondal, S., Adhikari, N., Banerjee, S., Amin, S. A., and Jha, T. (2020). Matrix Pearson, J. R. D., and Regad, T. (2017). Targeting cellular pathways in glioblastoma metalloproteinase-9 (MMP-9). and its inhibitors in cancer: a minireview. Eur. multiforme. Signal Transduct. Target. Ther. 2:17040. doi: 10.1038/sigtrans. J. Med. Chem. 194:112260. doi: 10.1016/j.ejmech.2020.112260 2017.40 Morrell, F. (1985). Secondary epileptogenesis in man. Arch. Neurol. 42, 318–335. Perry, J. R., Laperriere, N., O’Callaghan, C. J., Brandes, A. A., Menten, J., doi: 10.1001/archneur.1985.04060040028009 Phillips, C., et al. (2017). Short-Course radiation plus temozolomide in Morrell, F., Roberts, L., and Jasper, H. H. (1956). Effect of focal epileptogenic elderly patients with glioblastoma. N. Engl. J. Med. 376, 1027–1037. lesions and their ablation upon conditioned electrical responses of the doi: 10.1056/NEJMoa1611977 brain in the monkey. Electroencephalogr. Clin. Neurophysiol. 8, 217–236. Peterson, A. R., Garcia, T. A., Cullion, K., Tiwari-Woodruff, S. K., Pedapati, E. V., doi: 10.1016/0013-4694(56)90115-8 and Binder, D. K. (2021). Targeted overexpression of glutamate transporter-1 Morrison, R. S., Wenzel, H. J., Kinoshita, Y., Robbins, C. A., Donehower, L. A., reduces seizures and attenuates pathological changes in a mouse model of and Schwartzkroin, P. A. (1996). Loss of the p53 tumor suppressor gene epilepsy. Neurobiol. Dis. 157:105443. doi: 10.1016/j.nbd.2021.105443 protects neurons from kainate-induced cell death. J. Neurosci. 16, 1337–1345. Pietrobon, D., and Moskowitz, M. A. (2014). Chaos and commotion in the doi: 10.1523/JNEUROSCI.16-04-01337.1996 wake of cortical spreading depression and spreading depolarizations. Nat. Rev. Mukai, T., Kinboshi, M., Nagao, Y., Shimizu, S., Ono, A., Sakagami, Y., et al. Neurosci. 15, 379–393. doi: 10.1038/nrn3770 (2018). Antiepileptic drugs elevate astrocytic kir4.1 expression in the rat limbic Pijet, B., Konopka, A., Rejmak, E., Stefaniuk, M., Khomiak, D., Bulska, E., et al. region. Front. Pharmacol. 9:845. doi: 10.3389/fphar.2018.00845 (2020). The matrix metalloproteinase inhibitor marimastat inhibits seizures Nagelhus, E. A., Mathiisen, T. M., and Ottersen, O. P. (2004). Aquaporin-4 in the in a model of kainic acid-induced status epilepticus. Sci. Rep. 10:21314. central nervous system: cellular and subcellular distribution and coexpression doi: 10.1038/s41598-020-78341-y with KIR4.1. Neuroscience 129, 905–913. doi: 10.1016/j.neuroscience.2004.08. Plotkin, M. D., Snyder, E. Y., Hebert, S. C., and Delpire, E. (1997). 053 Expression of the Na-K-2Cl cotransporter is developmentally regulated in Nagpal, J., Jamoona, A., Gulati, N. D., Mohan, A., Braun, A., Murali, R., et al. postnatal rat brains: a possible mechanism underlying GABA’s excitatory (2006). Revisiting the role of p53 in primary and secondary glioblastomas. role in immature brain. J. Neurobiol. 33, 781–795. doi: 10.1002/(sici)1097- Anticancer Res. 26, 4633–4639. 4695(19971120)33:6<781::aid-neu6>3.0.co;2-5 Neal, A., Yuen, T., Bjorksten, A. R., Kwan, P., O’Brien, T. J., and Morokoff, A. Probert, L., Akassoglou, K., Kassiotis, G., Pasparakis, M., Alexopoulou, L., (2016). Peritumoural glutamate correlates with post-operative seizures in and Kollias, G. (1997). TNF-α transgenic and knockout models of supratentorial gliomas. J. Neurooncol. 129, 259–267. doi: 10.1007/s11060-016- CNS inflammation and degeneration. J. Neuroimmunol. 72, 137–141. 2169-y doi: 10.1016/s0165-5728(96)00184-1 Nishio, S., Morioka, T., Hisada, K., and Fukui, M. (2000). Temporal lobe epilepsy: Prodjinotho, U. F., Lema, J., Lacorcia, M., Schmidt, V., Vejzagic, N., a clinicopathological study with special reference to temporal neocortical Sikasunge, C., et al. (2020). Host immune responses during Taenia solium changes. Neurosurg. Rev. 23, 84–89. doi: 10.1007/pl00021698 neurocysticercosis infection and treatment. PLoS Negl. Trop. Dis. 14:e0008005. Noell, S., Ritz, R., Wolburg-Buchholz, K., Wolburg, H., and Fallier- doi: 10.1371/journal.pntd.0008005 Becker, P. (2012). An Allograft glioma model reveals the dependence of Proper, E. A., Hoogland, G., Kappen, S. M., Jansen, G. H., Rensen, M. G. A., aquaporin-4 expression on the brain microenvironment. PLoS One 7:e36555. Schrama, L. H., et al. (2002). Distribution of glutamate transporters in the doi: 10.1371/journal.pone.0036555 hippocampus of patients with pharmaco-resistant temporal lobe epilepsy. Oberndorfer, S., Lindeck-Pozza, E., Lahrmann, H., Struhal, W., Hitzenberger, P., Brain 125, 32–43. doi: 10.1093/brain/awf001 and Grisold, W. (2008). The end-of-life hospital setting in patients with Pun, R. Y. K., Rolle, I. J., Lasarge, C. L., Hosford, B. E., Rosen, J. M., Uhl, J. D., et al. glioblastoma. J. Palliat. Med. 11, 26–30. doi: 10.1089/jpm.2007.0137 (2012). Excessive activation of mTOR in postnatally generated granule cells is Ohgaki, H., and Kleihues, P. (2013). The definition of primary and secondary sufficient to cause epilepsy. Neuron 75, 1022–1034. doi: 10.1016/j.neuron.2012. glioblastoma. Clin Cancer Res. 19, 764–772. doi: 10.1158/1078-0432.CCR-12- 08.002 3002 Rades, D., Witteler, J., Olbrich, D., Trillenberg, P., Schild, S. E., Tvilsted, S., et al. Ohno, Y., Kunisawa, N., and Shimizu, S. (2021). Emerging roles of astrocyte (2021). A prospective interventional study evaluating seizure activity during kir4.1 channels in the pathogenesis and treatment of brain diseases. Int. J. Mol. a radiotherapy course for high-grade gliomas (SURF-ROGG). BMC Cancer Sci. 22:10236. doi: 10.3390/ijms221910236 21:386. doi: 10.1186/s12885-021-08121-y Olsen, M. L., and Sontheimer, H. (2004). Mislocalization of Kir channels in Raimondo, J. V., Burman, R. J., Katz, A. A., and Akerman, C. J. (2015). Ion malignant glia. Glia 46, 63–73. doi: 10.1002/glia.10346 dynamics during seizures. Front. Cell. Neurosci. 9:419. doi: 10.3389/fncel.2015. Olsen, M. L., and Sontheimer, H. (2008). Functional implications for 00419 Kir4.1 channels in glial biology: from K+ buffering to cell differentiation. Ramandi, D., Salmani, M. E., Moghimi, A., Lashkarbolouki, T., and Fereidoni, M. J. Neurochem. 107, 589–601. doi: 10.1111/j.1471-4159.2008.05615.x (2021). Pharmacological upregulation of GLT-1 alleviates the cognitive Otvos, B., Alban, T. J., Grabowski, M. M., Bayik, D., Mulkearns-Hubert, E. E., impairments in the animal model of temporal lobe epilepsy. PLoS One Radivoyevitch, T., et al. (2021). Preclinical modeling of surgery and steroid 16:e0246068. doi: 10.1371/journal.pone.0246068 therapy for glioblastoma reveals changes in immunophenotype that are Rana, A., and Musto, A. E. (2018). The role of inflammation in the development associated with tumor growth and outcome. Clin. Cancer Res. 27, 2038–2049. of epilepsy. J. Neuroinflammation 15:144. doi: 10.1186/s12974-018-1192-7 doi: 10.1158/1078-0432.CCR-20-3262 Ravizza, T., Gagliardi, B., Noé, F., Boer, K., Aronica, E., and Vezzani, A. Pace, A., Villani, V., Di Lorenzo, C., Guariglia, L., Maschio, M., Pompili, A., et al. (2008). Innate and adaptive immunity during epileptogenesis and spontaneous (2013). Epilepsy in the end-of-life phase in patients with high-grade gliomas. seizures: evidence from experimental models and human temporal lobe J. Neurooncol. 111, 83–86. doi: 10.1007/s11060-012-0993-2 epilepsy. Neurobiol. Dis. 29, 142–160. doi: 10.1016/j.nbd.2007.08.012 Pallud, J., Capelle, L., and Huberfeld, G. (2013). Tumoral epileptogenicity: how Reiss, W. G., and Oles, K. S. (1996). Acetazolamide in the treatment of seizures. does it happen? Epilepsia 54, 30–34. doi: 10.1111/epi.12440 Ann. Pharmacother. 30, 514–519. doi: 10.1177/106002809603000515 Pallud, J., Huberfeld, G., Dezamis, E., Peeters, S., Moiraghi, A., Gavaret, M., Robe, P. A., Martin, D. H., Nguyen-Khac, M. T., Artesi, M., Deprez, M., Albert, A., et al. (2022). Effect of levetiracetam use duration on overall survival of et al. (2009). Early termination of ISRCTN45828668, a phase 1/2 prospective, Frontiers in Molecular Neuroscience | www.frontiersin.org 21 June 2022 | Volume 15 | Article 903115 Hills et al. Epileptogenesis in Glioblastoma randomized study of Sulfasalazine for the treatment of progressing malignant Sørensen, M. F., Heimisdóttir, S. B., Sørensen, M. D., Mellegaard, C. S., gliomas in adults. BMC Cancer 9:372. doi: 10.1186/1471-2407-9-372 Wohlleben, H., Kristensen, B. W., et al. (2018). High expression of cystine- Robert, S. M., Buckingham, S. C., Campbell, S. L., Robel, S., Holt, K. T., Ogunrinu- glutamate antiporter xCT (SLC7A11) is an independent biomarker for epileptic Babarinde, T., et al. (2015). SLC7A11 expression is associated with seizures seizures at diagnosis in glioma. J. Neurooncol. 138, 49–53. doi: 10.1007/s11060- and predicts poor survival in patients with malignant glioma. Sci. Transl. Med. 018-2785-9 7:289ra86. doi: 10.1126/scitranslmed.aaa8103 Sorrentino, U., Bellonzi, S., Mozzato, C., Brasson, V., Toldo, I., Parrozzani, R., Robert, S. M., and Sontheimer, H. (2014). Glutamate transporters in the biology of et al. (2021). Epilepsy in NF1: Epidemiologic, genetic and clinical features. a malignant gliomas. Cell. Mol. Life Sci. 71, 1839–1854. doi: 10.1007/s00018-013- monocentric retrospective study in a cohort of 784 patients. Cancers (Basel) 1521-z 13:6336. doi: 10.3390/cancers13246336 Roh, T. H., Moon, J. H., Park, H. H., Kim, E. H., Hong, C.-K., Kim, S. H., et al. Stellwagen, D., Beattie, E. C., Seo, J. Y., and Malenka, R. C. (2005). Differential (2020). Association between survival and levetiracetam use in glioblastoma regulation of AMPA receptor and GABA receptor trafficking by tumor necrosis patients treated with temozolomide chemoradiotherapy. Sci. Rep. 10:10783. factor-alpha. J. Neurosci. 25, 3219–3228. doi: 10.1523/JNEUROSCI.4486-04. doi: 10.1038/s41598-020-67697-w 2005 Rohracher, A., Zimmermann, G., Villanueva, V., Garamendi, I., Sander, J. W., Sterner, R. C., and Sterner, R. M. (2021). CAR-T cell therapy: current limitations Wehner, T., et al. (2018). Perampanel in routine clinical use across and potential strategies. Blood Cancer J. 11:69. doi: 10.1038/s41408-021 Europe: pooled, multicenter, observational data. Epilepsia 59, 1727–1739. -00459-7 doi: 10.1111/epi.14520 Stupp, R., Hegi, M. E., Mason, W. P., van den Bent, M. J., Taphoorn, M. J. B., Rosati, A., Buttolo, L., Stefini, R., Todeschini, A., Cenzato, M., and Padovani, A. Janzer, R. C., et al. (2009). Effects of radiotherapy with concomitant and (2010). Efficacy and safety of levetiracetam in patients with glioma: a clinical adjuvant temozolomide versus radiotherapy alone on survival in glioblastoma prospective study. Arch. Neurol. 67, 343–346. doi: 10.1001/archneurol.2009.335 in a randomised phase III study: 5-year analysis of the EORTC-NCIC trial. Roth, P., Winklhofer, S., Müller, A. M. S., Dummer, R., Mair, M. J., Gramatzki, D., Lancet Oncol. 10, 459–466. doi: 10.1016/S1470-2045(09)70025-7 et al. (2021). Neurological complications of cancer immunotherapy. Cancer Stupp, R., Mason, W. P., van den Bent, M. J., Weller, M., Fisher, B., Treat. Rev. 97:102189. doi: 10.1016/j.ctrv.2021.102189 Taphoorn, M. J. B., et al. (2005). Radiotherapy plus concomitant and Rudà, R., Magliola, U., Bertero, L., Trevisan, E., Bosa, C., Mantovani, C., et al. adjuvant temozolomide for glioblastoma. N. Engl. J. Med. 352, 987–996. (2013). Seizure control following radiotherapy in patients with diffuse gliomas: doi: 10.1056/NEJMoa043330 a retrospective study. Neuro Oncol. 15, 1739–1749. doi: 10.1093/neuonc/not109 Szopa, W., Burley, T. A., Kramer-Marek, G., and Kaspera, W. (2017). Diagnostic Sabetghadam, A., Wu, C., Liu, J., Zhang, L., and Reid, A. Y. (2020). Increased and therapeutic biomarkers in glioblastoma: current status and future epileptogenicity in a mouse model of neurofibromatosis type 1. Exp. Neurol. perspectives. Biomed. Res. Int. 2017:8013575. doi: 10.1155/2017/8013575 331:113373. doi: 10.1016/j.expneurol.2020.113373 Takahashi, K., Foster, J. B., and Lin, C.-L. G. (2015). Glutamate transporter Salmaggi, A., Corno, C., Maschio, M., Donzelli, S., D’Urso, A., Perego, P., et al. EAAT2: regulation, function and potential as a therapeutic target for (2021). Synergistic effect of perampanel and temozolomide in human glioma neurological and psychiatric disease. Cell. Mol. Life Sci. 72, 3489–3506. cell lines. J. Pers. Med. 11:390. doi: 10.3390/jpm11050390 doi: 10.1007/s00018-015-1937-8 Samland, H., Huitron-Resendiz, S., Masliah, E., Criado, J., Henriksen, S. J., and Takeuchi, S., Wada, K., Toyooka, T., Shinomiya, N., Shimazaki, H., Nakanishi, K., Campbell, I. L. (2003). Profound increase in sensitivity to glutamatergic- but et al. (2013). Increased xCT expression correlates with tumor invasion not cholinergic agonist-induced seizures in transgenic mice with astrocyte and outcome in patients with glioblastomas. Neurosurgery 72, 33–41. production of IL-6. J. Neurosci. Res. 73, 176–187. doi: 10.1002/jnr.10635 doi: 10.1227/NEU.0b013e318276b2de Sarac, S., Afzal, S., Broholm, H., Madsen, F. F., Ploug, T., and Laursen, H. (2009). Tan, A. C., Ashley, D. M., López, G. Y., Malinzak, M., Friedman, H. S., and Excitatory amino acid transporters EAAT-1 and EAAT-2 in temporal lobe Khasraw, M. (2020). Management of glioblastoma: state of the art and future and hippocampus in intractable temporal lobe epilepsy. APMIS 117, 291–301. directions. CA Cancer J. Clin. 70, 299–312. doi: 10.3322/caac.21613 doi: 10.1111/j.1600-0463.2009.02443.x Tanaka, K., Watase, K., Manabe, T., Yamada, K., Watanabe, M., Takahashi, K., Schiapparelli, P., Guerrero-Cazares, H., Magaña-Maldonado, R., Hamilla, S. M., et al. (1997). Epilepsy and exacerbation of brain injury in mice lacking the Ganaha, S., Goulin Lippi Fernandes, E., et al. (2017). NKCC1 regulates glutamate transporter GLT-1. Science 276, 1699–1702. doi: 10.1126/science. migration ability of glioblastoma cells by modulation of actin dynamics and 276.5319.1699 interacting with cofilin. EBioMedicine 21, 94–103. doi: 10.1016/j.ebiom.2017. Tang, B., Foss, K., Lichtor, T., Phillips, H., and Roy, E. (2021). Resection of 06.020 orthotopic murine brain glioma. Neuroimmunol. Neuroinflammation 8, 64–69. Scholly, J., Staack, A. M., Kahane, P., Scavarda, D., Régis, J., Hirsch, E., et al. doi: 10.20517/2347-8659.2020.28 (2017). Hypothalamic hamartoma: epileptogenesis beyond the lesion? Epilepsia Terrone, G., Balosso, S., Pauletti, A., Ravizza, T., and Vezzani, A. (2020). 58, 32–40. doi: 10.1111/epi.13755 Inflammation and reactive oxygen species as disease modifiers in epilepsy. Sears, S. M. S., Hewett, J. A., and Hewett, S. J. (2019). Decreased epileptogenesis in Neuropharmacology 167:107742. doi: 10.1016/j.neuropharm.2019.107742 mice lacking the System x-c transporter occurs in association with a reduction in Tewari, B. P., Chaunsali, L., Campbell, S. L., Patel, D. C., Goode, A. E., and AMPA receptor subunit GluA1. Epilepsia Open 4, 133–143. doi: 10.1002/epi4. Sontheimer, H. (2018). Perineuronal nets decrease membrane capacitance of 12307 peritumoral fast spiking interneurons in a model of epilepsy. Nat. Commun. Shi, S., Jaoube, J. A., Kanwar, R., Jin, M. C., Amorin, A., Varanasi, V., et al. (2020). 9:4724. doi: 10.1038/s41467-018-07113-0 Neurological adverse effects due to programmed death 1 (PD-1) inhibitors. J. Thakkar, J. P., Dolecek, T. A., Horbinski, C., Ostrom, Q. T., Lightner, D. D., Neurooncol. 148, 291–297. doi: 10.1007/s11060-020-03514-8 Barnholtz-Sloan, J. S., et al. (2014). Epidemiologic and molecular prognostic Shimabukuro-Vornhagen, A., Gödel, P., Subklewe, M., Stemmler, H. J., review of glioblastoma. Cancer Epidemiol. Biomark. Prev. 23, 1985–1996. Schlößer, H. A., Schlaak, M., et al. (2018). Cytokine release syndrome. doi: 10.1158/1055-9965.EPI-14-0275 J. Immunother. Cancer 6:56. doi: 10.1186/s40425-018-0343-9 Toledo, M., Sarria-Estrada, S., Quintana, M., Maldonado, X., Martinez-Ricarte, F., Siegler, E. L., and Kenderian, S. S. (2020). Neurotoxicity and cytokine Rodon, J., et al. (2017). Epileptic features and survival in glioblastomas release syndrome after chimeric antigen receptor T Cell therapy: presenting with seizures. Epilepsy Res. 130, 1–6. doi: 10.1016/j.eplepsyres.2016. insights into mechanisms and novel therapies. Front. Immunol. 11:1973. 12.013 doi: 10.3389/fimmu.2020.01973 Tsai, H.-C., Wei, K.-C., Chen, P.-Y., Huang, C.-Y., Chen, K.-T., Lin, Y.-J., et al. Smits, A., Jin, Z., Elsir, T., Pedder, H., Nistér, M., Alafuzoff, I., et al. (2012). GABA- (2021). Valproic acid enhanced temozolomide-induced anticancer activity A channel subunit expression in human glioma correlates with tumor histology in human glioma through the p53-PUMA apoptosis pathway. Front. Oncol. and clinical outcome. PLoS One 7:e37041. doi: 10.1371/journal.pone.0037041 11:722754. doi: 10.3389/fonc.2021.722754 Soomro, S. H., Ting, L. R., Qing, Y. Y., and Ren, M. (2017). Molecular biology van Breemen, M. S. M., Wilms, E. B., and Vecht, C. J. (2007). Epilepsy in of glioblastoma: classification and mutational locations. J. Pak. Med. Assoc. 67, patients with brain tumours: epidemiology, mechanisms and management. 1410–1414. Lancet Neurol. 6, 421–430. doi: 10.1016/S1474-4422(07)70103-5 Frontiers in Molecular Neuroscience | www.frontiersin.org 22 June 2022 | Volume 15 | Article 903115 Hills et al. Epileptogenesis in Glioblastoma Van Nifterik, K. A., Van den Berg, J., Slotman, B. J., Lafleur, M. V. M., Sminia, P., Ye, L., Huang, Y., Zhao, L., Li, Y., Sun, L., Zhou, Y., et al. (2013). IL-1β and Stalpers, L. J. A. (2012). Valproic acid sensitizes human glioma cells for and TNF-α induce neurotoxicity through glutamate production: a potential temozolomide and γ-radiation. J. Neurooncol. 107, 61–67. doi: 10.1007/s11060- role for neuronal glutaminase. J. Neurochem. 125, 897–908. doi: 10.1111/ 011-0725-z jnc.12263 van Vliet, E. A., da Costa Araújo, S., Redeker, S., van Schaik, R., Aronica, E., and Ye, Z.-C., Rothstein, J. D., and Sontheimer, H. (1999). Compromised glutamate Gorter, J. A. (2007). Blood-brain barrier leakage may lead to progression of transport in human glioma cells: reduction-mislocalization of sodium- temporal lobe epilepsy. Brain 130, 521–534. doi: 10.1093/brain/awl318 dependent glutamate transporters and enhanced activity of cystine-glutamate Varadkar, S., Bien, C. G., Kruse, C. A., Jensen, F. E., Bauer, J., Pardo, C. A., et al. exchange. J. Neurosci. 19, 10767–10777. doi: 10.1523/JNEUROSCI.19-24- (2014). Rasmussen’s encephalitis: clinical features, pathobiology and treatment 10767.1999 advances. Lancet Neurol. 13, 195–205. doi: 10.1016/S1474-4422(13)70260-6 Yeo, M., Chen, Y., Jiang, C., Chen, G., Wang, K., Chandra, S., et al. (2021). Vecht, C., Duran-Peña, A., Houillier, C., Durand, T., Capelle, L., and Huberfeld, G. Repurposing cancer drugs identifies kenpaullone which ameliorates (2017). Seizure response to perampanel in drug-resistant epilepsy with gliomas: pathologic pain in preclinical models via normalization of inhibitory early observations. J. Neurooncol. 133, 603–607. doi: 10.1007/s11060-017 neurotransmission. Nat. Commun. 12:6208. doi: 10.1038/s41467-021 -2473-1 -26270-3 Vecht, C. J., Kerkhof, M., and Duran-Pena, A. (2014). Seizure prognosis in brain Yeung, Y. T., Bryce, N. S., Adams, S., Braidy, N., Konayagi, M., tumors: new insights and evidence-based management. Oncologist 19, 751–759. McDonald, K. L., et al. (2012). p38 MAPK inhibitors attenuate doi: 10.1634/theoncologist.2014-0060 pro-inflammatory cytokine production and the invasiveness of human Venkataramani, V., Tanev, D. I., Strahle, C., Studier-Fischer, A., Fankhauser, L., U251 glioblastoma cells. J. Neurooncol. 109, 35–44. doi: 10.1007/s11060-012 Kessler, T., et al. (2019). Glutamatergic synaptic input to glioma cells drives -0875-7 brain tumour progression. Nature 573, 532–538. doi: 10.1038/s41586-019- Yeung, Y. T., McDonald, K. L., Grewal, T., and Munoz, L. (2013). Interleukins in 1564-x glioblastoma pathophysiology: implications for therapy. Br. J. Pharmacol. 168, Venkatesan, S., Lamfers, M. L. M., Dirven, C. M. F., and Leenstra, S. (2016). 591–606. doi: 10.1111/bph.12008 Genetic biomarkers of drug response for small-molecule therapeutics targeting Yu, M. W., and Quail, D. F. (2021). Immunotherapy for glioblastoma: current the RTK/Ras/PI3K, p53 or Rb pathway in glioblastoma. CNS Oncol. 5, 77–90. progress and challenges. Front. Immunol. 12:676301. doi: 10.3389/fimmu.2021. doi: 10.2217/cns-2015-0005 676301 Venkatesh, H. S., Morishita, W., Geraghty, A. C., Silverbush, D., Gillespie, S. M., Yuen, T. I., Morokoff, A. P., Bjorksten, A., D’Abaco, G., Paradiso, L., Finch, S., Arzt, M., et al. (2019). Electrical and synaptic integration of glioma into neural et al. (2012). Glutamate is associated with a higher risk of seizures in patients circuits. Nature 573, 539–545. doi: 10.1038/s41586-019-1563-y with gliomas. Neurology 79, 883–889. doi: 10.1212/WNL.0b013e318266fa89 Vezzani, A., Conti, M., De Luigi, A., Ravizza, T., Moneta, D., Marchesi, F., Zhang, Y., Dube, C., Gibert Jr, M., Cruickshanks, N., Wang, B., Coughlan, M., et al. (1999). Interleukin-1beta immunoreactivity and microglia are enhanced et al. (2018). The p53 pathway in glioblastoma. Cancers (Basel) 10:297. in the rat hippocampus by focal kainate application: functional evidence doi: 10.3390/cancers10090297 for enhancement of electrographic seizures. J. Neurosci. 19, 5054–5065. Zhang, X., Zheng, L., Duan, J., Li, Z., and Tang, Y. (2020). Clinical characteristics doi: 10.1523/JNEUROSCI.19-12-05054.1999 of brain tumor-related epilepsy and factors influencing the identification Villanueva, V., D’Souza, W., Goji, H., Kim, D. W., Liguori, C., McMurray, R., of epilepsy-associated tumors. Acta Epileptol. 2:25. doi: 10.1186/s42494-020- et al. (2022). PERMIT study: a global pooled analysis study of the effectiveness 00034-w and tolerability of perampanel in routine clinical practice. J. Neurol. 269, Zhang, S., Zhou, J., Zhang, Y., Liu, T., Friedel, P., Zhuo, W., et al. (2021). 1957–1977. doi: 10.1007/s00415-021-10751-y The structural basis of function and regulation of neuronal cotransporters Viviani, B., Bartesaghi, S., Gardoni, F., Vezzani, A., Behrens, M. M., NKCC1 and KCC2. Commun. Biol. 4:226. doi: 10.1038/s42003-021-01750-w Bartfai, T., et al. (2003). Interleukin-1beta enhances NMDA receptor-mediated Zheng, H., Ying, H., Yan, H., Kimmelman, A. C., Hiller, D. J., Chen, A.-J., et al. intracellular calcium increase through activation of the Src family of kinases. (2008). p53 and Pten control neural and glioma stem/progenitor cell renewal J. Neurosci. 23, 8692–8700. doi: 10.1523/JNEUROSCI.23-25-08692.2003 and differentiation. Nature 455, 1129–1133. doi: 10.1038/nature07443 Wang, S. I., Puc, J., Li, J., Bruce, J. N., Cairns, P., Sidransky, D., et al. (1997). Zhu, J. J., Esteban, J. A., Hayashi, Y., and Malinow, R. (2000). Postnatal synaptic Somatic mutations of PTEN in glioblastoma multiforme. Cancer Res. 57, potentiation: delivery of GluR4-containing AMPA receptors by spontaneous 4183–4186. activity. Nat. Neurosci. 3, 1098–1106. doi: 10.1038/80614 Warth, A., Simon, P., Capper, D., Goeppert, B., Tabatabai, G., Herzog, H., et al. Zurolo, E., de Groot, M., Iyer, A., Anink, J., van Vliet, E. A., Heimans, J. J., et al. (2007). Expression pattern of the water channel aquaporin-4 in human gliomas (2012). Regulation of Kir4.1 expression in astrocytes and astrocytic tumors: a is associated with blood-brain barrier disturbance but not with patient survival. role for interleukin-1 β. J. Neuroinflammation 9:280. doi: 10.1186/1742-2094- J. Neurosci. Res. 85, 1336–1346. doi: 10.1002/jnr.21224 9-280 Wasade, V. S., Viarasilpa, T., Balki, I., Osman, G., Gaddam, A., Dharaiya, D., et al. (2020). Effect of seizure timing on long-term survival in patients with brain Conflict of Interest: KK would like to declare he is co-founder and holds equity in tumor. Epilepsy Behav. 111:107307. doi: 10.1016/j.yebeh.2020.107566 INBRAIN Neuroelectronics (Barcelona, Spain). Wei, Q., Singh, O., Ekinci, C., Gill, J., Li, M., Mamatjan, Y., et al. (2021). TNFα secreted by glioma associated macrophages promotes endothelial activation The remaining authors declare that the research was conducted in the absence of and resistance against anti-angiogenic therapy. Acta Neuropathol. Commun. any commercial or financial relationships that could be construed as a potential 9:67. doi: 10.1186/s40478-021-01163-0 conflict of interest. Williams, M. R., DeSpenza Jr, T., Li, M., Gulledge, A. T., and Luikart, B. W. (2015). Hyperactivity of newborn Pten knock-out neurons results from increased Publisher’s Note: All claims expressed in this article are solely those of the authors excitatory synaptic drive. J. Neurosci. 35, 943–959. doi: 10.1523/JNEUROSCI. and do not necessarily represent those of their affiliated organizations, or those of 3144-14.2015 the publisher, the editors and the reviewers. Any product that may be evaluated in Wolburg, H., Noell, S., Fallier-Becker, P., Mack, A. F., and Wolburg-Buchholz, K. this article, or claim that may be made by its manufacturer, is not guaranteed or (2012). The disturbed blood-brain barrier in human glioblastoma. Mol. Aspects endorsed by the publisher. Med. 33, 579–589. doi: 10.1016/j.mam.2012.02.003 Wykes, R. C., Heeroma, J. H., Mantoan, L., Zheng, K., MacDonald, D. C., Copyright © 2022 Hills, Kostarelos and Wykes. This is an open-access article Deisseroth, K., et al. (2012). Optogenetic and potassium channel gene therapy distributed under the terms of the Creative Commons Attribution License (CC BY). in a rodent model of focal neocortical epilepsy. Sci. Transl. Med. 4:161ra152. The use, distribution or reproduction in other forums is permitted, provided the doi: 10.1126/scitranslmed.3004190 original author(s) and the copyright owner(s) are credited and that the original Yamanaka, G., Takata, F., Kataoka, Y., Kanou, K., Morichi, S., Dohgu, S., et al. publication in this journal is cited, in accordance with accepted academic practice. (2021). The neuroinflammatory role of pericytes in epilepsy. Biomedicines No use, distribution or reproduction is permitted which does not comply with 9:759. doi: 10.3390/biomedicines9070759 these terms. Frontiers in Molecular Neuroscience | www.frontiersin.org 23 June 2022 | Volume 15 | Article 903115