Molecular Markers of Therapy-Resistant Glioblastoma and Potential Strategy to Combat Resistance

Glioblastoma (GBM) is the most common primary malignant tumor of the central nervous system. With its overall dismal prognosis (the median survival is 14 months), GBMs demonstrate a resounding resilience against all current treatment modalities. The absence of a major progress in the treatment of GBM maybe a result of our poor understanding of both GBM tumor biology and the mechanisms underlying the acquirement of treatment resistance in recurrent GBMs. A comprehensive understanding of these markers is mandatory for the development of treatments against therapy-resistant GBMs. This review also provides an overview of a novel marker called acid ceramidase and its implication in the development of radioresistant GBMs. Multiple signaling pathways were found altered in radioresistant GBMs. Given these global alterations of multiple signaling pathways found in radioresistant GBMs, an effective treatment for radioresistant GBMs may require a cocktail containing multiple agents targeting multiple cancer-inducing pathways in order to have a chance to make a substantial impact on improving the overall GBM survival.


Introduction
Glioblastoma (GBM) is the most common primary malignant tumor of the central nervous system. With its overall dismal prognosis, GBMs demonstrate a resounding resilience against all current treatment modalities. The estimated overall survival of GBM patients is less than 1.5 years, and the 5-year survival rate is 5% [1][2][3]. The median age of diagnosis of GBM has increased to 64 years over the last decades, and the top incidence is 15.24/100,000 populations diagnosed within the age range of 75-84 years [1][2][3]. While radiation is the only proved cause of GBM, only a minority of patients develop GBMs following exposure to radiation [4]. The etiology of GBM remains to be discovered, and fewer than 5% of patients have a germline mutation which increases the risk for developing GBMs [5,6]. Symptoms at presentation are based on the location of GBMs. Eloquent-area tumors often engender symptoms ranging from numbness, weakness, and visual disturbance to language deficits, while tumors in other areas (including the non-dominant frontal and temporal lobes or the corpus callosum) may induce non-specific symptoms (such as seizures, which can be controlled with anticonvulsant medications in 25% of patients with newly diagnosed GBMs [7]). However, new Int. J. Mol. Sci. 2018, 19, 1765 2 of 23 emerging data have suggested that the administration of anticonvulsants may not be beneficial and can produce significant, undesired effects in GBM patients without seizures [8,9]. The presenting symptoms include headaches (~60%), memory loss (~40%), and cognitive, language, or motor deficits (~40%) [10]. The most common imaging modality to diagnose GBMs is magnetic resonance imaging (MRI) of the brain with and without gadolinium contrast. A heterogeneous ring-enhancement with area of central necrosis is the signature feature of GBMs; infrequently, GBMs can be multi-focal. Headache has been attributed to peritumoral edema, which can cause a major midline shift or mass effect [11]. Steroids such as dexamethasone are commonly employed to provide relief from headache or deficits by reducing the peritumoral edema, generally within 48 h [12,13]. Another therapy aimed at reducing the peritumoral edema is based on the anti-angiogenesis antibody bevacizumab, but it has been shown not to affect the overall survival in patients with newly diagnosed GBMs [14,15]. GBMs having certain prognostic biomarker mutations, such as isocitrate dehydrogenase (IDH), may present, on MRI, with characteristic features, such as a large non-enhancing mass with pial invasion, decreased blood flow, minimal edema and necrosis, and a tendency for the frontal and temporal lobes [16,17]. Following surgery, resected GBM tissues are formalin-fixed and paraffin-embedded prior to undergoing histopathology examinations, which characteristically show palisading necrosis, marked pleomorphism, a high mitotic index, and microvascular proliferation. Additionally, these GBM tissues are also further examined by immunostaining or sequencing for IDH mutations, O6-methylguanine methyltransferase (MGMT) methylation, and other prognostic biomarkers, which will be discussed in detail below [18,19].
The absence of a major progress in the treatment of GBM may be a result of our poor understanding of both GBM tumor biology and the mechanisms underlying the acquirement of treatment resistance in recurrent GBMs. Others have proposed that glioblastoma stem-like cells (GSCs), carrying the cell membrane marker CD133, may play a significant role in the resistance of this cancer to chemotherapy and radiotherapy [20][21][22][23]. The higher expression levels of CD133 have been linked to poorer prognosis [23]. Proteins or signaling pathways that maintain stemness may contribute to the development of therapy-resistant GBMs [24]. Novel druggable targets that have been reported to combat therapy-resistant GBMs include sodium pump α1 subunit, wingless-type MMTV integration site family member (Wnt)/β-catenin, sonic hedgehog/Glioma-associated oncogene (SHH/GLI), oligodendrocyte transcription factor 2(OLIG2), polycomb group RING finger protein 4 (BMI1), NANOG, and inhibitor of differentiation/DNA binding (ID1) [24,25]. More recently, circular RNAs (circRNAs) such as circSMARCA5, whose expression is downregulated in GBM samples as compared to control tissues, has been described to function as a novel tumor-suppressor, regulating the migration of GBM cells by modulating the oncoprotein that modulates cell migration, called RNA binding protein serineand arginine-rich splicing factor 1 (SRSF1) [26].
A comprehensive understanding of established prognostic markers is mandatory for the development of treatments against therapy-resistant GBMs. In addition to discuss the established prognostic markers, this review also provides an overview of a novel marker called acid ceramidase (ASAH1) and its implications in the development of radioresistant GBMs. Multiple signaling pathways were found altered in radioresistant GBMs. Given the global alterations of multiple signaling pathways found in radioresistant GBMs, an effective treatment targeting radioresistant GBMs may require a cocktail containing multiple agents targeting multiple cancer-inducing pathways in order to have a chance to make a substantial impact on improving overall GBM survival.

O6-Methylguanine Methyltransferase (MGMT)
Alkylating agents, such as temozolomide (TMZ), attach an alkyl group to the DNA, frequently at the N-7 or O-6 positions of guanine residues ( Figure 1) [27]. This process damages the DNA and triggers cell cycle death, unless the DNA is promptly repaired. O6-methylguanine methyltransferase (MGMT), a DNA repair protein, can hydrolyze the alkyl groups off guanine and impede the effectiveness of such chemotherapeutic agents [28]. Methylation of the MGMT promoter at CpG sites can suppress Presently, molecular testing occurs via either quantitative methylation-specific PCR or pyrosequencing [34]. The former employs methylation-specific primer pairs to probe CpG islands with high methylation density; the latter enumerates the methylation sites at individual CpG sites through the "sequencing-by-synthesis" principle when nucleotides get incorporated by DNA polymerase [35]. To complicate matters, there is no accepted threshold for the number of methylated sites for a tumor to be classified as "methylated"; various detection methods yield methylation rates varying from 33% to 60% for the same group of GBM patients [36]. Moreover, additional research is required to elucidate the impact of the extent of methylation or the patterns of methylation on GBM survival [35].

Epidermal Growth Factor Receptor (EGFR)
Epidermal growth factor receptor (EGFR) is a transmembrane tyrosine kinase that functions as a critical player in pathways linked to cell proliferation, migration, and survival. EGFR activity can be augmented via gene amplification or EGFR variant III deletion mutation (EGFRvIII); the latter results in a truncated receptor that is constitutively active, promoting mitogenic cascades. EGFR amplification occurs in roughly 40-60% of GBM; EGFRvIII, which only occurs in a subset of those GBMs with EGFR amplification, arises in approximately 20-30% of GBM overall [35,[37][38][39][40]. EGFR amplification is assessed via fluorescence in situ hybridization (FISH); EGFRvIII expression can be established by immunohistochemistry (IHC) [38].
Studies regarding the implications of EGFR amplification and EGFRvIII mutation have reported mixed, conflicting results regarding GBM survival [27,37,39]. Given the positive results in other types of cancers, researchers believed that receptor tyrosine kinase inhibitors could play a role in the treatment of GBM [39]. However, clinical trials for GBM designed to target EGFR have been disappointing [39]. The dearth of clinical effectiveness may be due to the inability of the examined drugs to cross the blood-brain barrier and/or the development of resistance through gained mutations [40]. Presently, molecular testing occurs via either quantitative methylation-specific PCR or pyrosequencing [34]. The former employs methylation-specific primer pairs to probe CpG islands with high methylation density; the latter enumerates the methylation sites at individual CpG sites through the "sequencing-by-synthesis" principle when nucleotides get incorporated by DNA polymerase [35]. To complicate matters, there is no accepted threshold for the number of methylated sites for a tumor to be classified as "methylated"; various detection methods yield methylation rates varying from 33% to 60% for the same group of GBM patients [36]. Moreover, additional research is required to elucidate the impact of the extent of methylation or the patterns of methylation on GBM survival [35].

Epidermal Growth Factor Receptor (EGFR)
Epidermal growth factor receptor (EGFR) is a transmembrane tyrosine kinase that functions as a critical player in pathways linked to cell proliferation, migration, and survival. EGFR activity can be augmented via gene amplification or EGFR variant III deletion mutation (EGFRvIII); the latter results in a truncated receptor that is constitutively active, promoting mitogenic cascades. EGFR amplification occurs in roughly 40-60% of GBM; EGFRvIII, which only occurs in a subset of those GBMs with EGFR amplification, arises in approximately 20-30% of GBM overall [35,[37][38][39][40]. EGFR amplification is assessed via fluorescence in situ hybridization (FISH); EGFRvIII expression can be established by immunohistochemistry (IHC) [38].
Studies regarding the implications of EGFR amplification and EGFRvIII mutation have reported mixed, conflicting results regarding GBM survival [27,37,39]. Given the positive results in other types of cancers, researchers believed that receptor tyrosine kinase inhibitors could play a role in the treatment of GBM [39]. However, clinical trials for GBM designed to target EGFR have been disappointing [39]. The dearth of clinical effectiveness may be due to the inability of the examined drugs to cross the blood-brain barrier and/or the development of resistance through gained mutations [40].

Isocitrate Dehydrogenase (IDH)1/2
Isocitrate dehydrogenase (IDH) is a component of the Krebs cycle that converts isocitrate and cofactor NAD+ to carbon dioxide, NADH, and α-ketoglutarate. Since the initial discovery of IDH mutations in GBM [41], several studies have observed that IDH mutations occurred in approximately 8-13% of all GBMs, including greater than 80% of secondary GBM [42]. The most common mutations are IDH1 R132 and IDH2 R172, comprising roughly 90% of IDH mutations; the former is noted in more than 70% of grade 2/3 gliomas and in GBMs that progressed from these lower-grade tumors [41,43]. The mutations cause a buildup of the onco-metabolite D-2-hydroxy-glutarate, which can disturb DNA methylation, gene transcription, and histone alterations; moreover, mutations may decrease NAPDH formation, promoting oxidative stress and leading to DNA damage [27,35].
Several studies have documented survival benefits (OS and PFS) in gliomas with IDH mutations, ranging from an average of 12 to 30 months [27,[42][43][44]. In addition, IDH mutations convey a higher sensitivity to TMZ and radiotherapy [45][46][47]. IDH mutations have also been correlated with improved MRI-defined enhancing disease, allowing larger resections [48]. At present, IDH mutations can be detected via sequencing or IHC [38]. Preclinical studies have demonstrated that small molecule inhibitors of mutant IDH can lower the intracellular levels of D-2-hydroxy-glutarate, overturn epigenetic dysregulation, and promote cellular differentiation [49].

1p19q Co-Deletion
The unbalanced whole-arm translocation of the centromeric portions between chromosomes 1q and 19q is defined as 1p19q co-deletion. The recent WHO 2016 criteria utilize this co-deletion, along with an IDH mutation, to classify gliomas into the oligodendroglial phenotype. The co-deletion occurs in roughly 60-80% of grade 2 or 3 oligodendrogliomas, 20-50% of grade 2 or 3 oligoastrocytomas, and less than 10% of diffuse gliomas (together with GBM) [50]. For oligodendroglioma, this co-deletion has been associated with favorable survival as well as responsiveness to chemotherapy (PCV and temozolomide) and radiotherapy [51][52][53][54]. The reasoning behind this sensitivity to treatment remains elusive. Studies concerning the co-deletion in GB have reported mixed results [55,56]; however, a meta-analysis by Zhao et al. [57], incorporating 3408 gliomas across 28 studies, noted that 1p/19q co-deletion was associated with improved survival (PFS and OS) irrespective of the histological grade. Frequently, detection of the 1p19q co-deletion is completed through FISH; other methods include microsatellite analysis, PCR, and array comparative genomic hybridization [58].

α-Thalassemia/Mental Retardation Syndrome X-Linked (ATRX)
The ATRX (α-thalassemia/mental retardation syndrome X-linked) gene encodes a protein involved in genomic stability, chromatin remodeling, and DNA methylation [59]. Inactivation of the gene is highly linked to the ALT (alternative lengthening of telomeres) phenotype. ALT is a mechanism for the regulation of telomere length that is vital to cell survival and proliferation [59]. Its role in glioma biology has only recently been explored. ATRX mutation is frequently associated with IDH mutations, but rarely with 1p19q co-deletions [59]. For anaplastic gliomas, ATRX loss defines a subset of IDH mutants with a significantly longer median time to treatment failure (close to 24 months) [60]. By using a mouse model of ATRX-deficient GBM, Koschmann et al. suggested that ATRX mutations lead to a genetically erratic tumor. With no treatment, the tumor behaved rather aggressively; on the contrary, with treatment directed at double-stranded DNA damage, the overall survival improved [61]. Commonly, detection of ATRX loss is performed via IHC; other methods include PCR, sequencing, and Western blotting [35,38].

Telomerase Reverse Transcriptase (TERT)
Telomeres are nucleoprotein complexes (comprised of hundreds of repetitive nucleotide sequences) that bind the extremes of chromosomes to ensure chromosomal integrity [62]. Each cell division prompts telomere truncation until its depletion, which provokes cell dormancy or death [62]. Telomerase reverse transcriptase (TERT) is a subunit of telomerase, an enzyme that inserts additional nucleotides to telomeres [62]. For normal adult cells, telomerase is typically inactive [62]. Activating mutations in the TERT promoter are frequently reported in grade IV astrocytomas (up to 85% of GBM) and grade 2/3 oligodendrogliomas (close to 80%) [62][63][64]. TERT mutations are strongly correlated with 1p19q co-deletion, but not with either IDH mutations or ATRX loss [64]. Comparisons of groups based on the statuses of IDH, 1p19q, and TERT revealed that TERT mutation bestows better outcomes in gliomas with TERT mutant/IDH mutation/1p19q co-deletion, but poorer survival in GBM with TERT mutant/IDH mutation without 1p19q co-deletion [63]. Currently, detection of TERT mutations is completed via methyl-specific PCR; in addition, rapid intraoperative testing has been reported [65].

Acid Ceramidase (ASAH1) as a Druggable Target to Combat Multiple Therapy-Resistant Cancers
ASAH1, initially discovered in rat brain homogenates and further characterized and purified from human urine in 1995, is a lysosomal cysteine amidase that catalyzes the transformation of ceramide into sphingosine and free fatty acid ( Figure 2) [66][67][68][69][70][71][72]. Following this, sphingosine kinase 1 (SPHK1) or 2 (SPHK2) phosphorylates sphingosine to produce sphingosine-1-phosphate (S1P), which promotes GBM invasiveness via the upregulation of the urokinase plasminogen activator, its receptor, and the pro-invasive molecule CCN1 (cysteine-rich angiogenic protein 61) ( Figure 2) [69,[72][73][74]. On the other hand, high levels of ceramides, carrying fatty acid side chains ranging from 14 to 26 carbons and generated via the action of ceramide synthases (CerS), promote apoptosis in cells that have undergone radio-and chemotherapy via the release of cytochrome c, leading to the activation of caspase-9 and caspase-3 [69][70][71][75][76][77][78][79]. Since its products are involved in the regulation of cell proliferation, multiple studies have linked ASAH1 to multiple cancers such as melanoma, acute myeloid leukemia (AML), and colon and prostate cancers [80][81][82][83][84]. ASAH1 has been proposed as an emerging drug target in AML [85]. Interestingly, over-expression of ASAH1 in prostate cancer promotes resistance to chemotherapy. Prostate cancer upregulates ASAH1 following radiation, which was described as a mechanism enabling the cancer to survive radiation [86]. Consequently, when the activity of ASAH1 is suppressed with an ASAH1 inhibitor named B13, the cells become more sensitive to chemotherapy and radiation as a result of the accumulation of intracellular ceramide up to cytotoxic levels, inducing apoptosis [81,87,88]. Similarly, the acid ceramidase inhibitor ceranib-2 also has activity against the growth of the breast cancer cell lines MCF-7 and MDA MB-231 via the activation of stress-activated protein kinase/c-Jun N-terminal kinase and p38 mitogen-activated protein kinase apoptotic pathways and the inhibition of the Akt pathway [89]. The Wnt/β-catenin signaling pathway appears to play a role in suppressing the proliferation and metastatic potential of cervical cancers when these tumors were treated with a recently identified ASAH1 inhibitor called carmofur [90,91]. The clinical application of carmofur has been attempted and it offered benefits when carmofur was used an adjuvant in patients with early breast cancer in a postoperative setting [92]. division prompts telomere truncation until its depletion, which provokes cell dormancy or death [62]. Telomerase reverse transcriptase (TERT) is a subunit of telomerase, an enzyme that inserts additional nucleotides to telomeres [62]. For normal adult cells, telomerase is typically inactive [62]. Activating mutations in the TERT promoter are frequently reported in grade IV astrocytomas (up to 85% of GBM) and grade 2/3 oligodendrogliomas (close to 80%) [62][63][64]. TERT mutations are strongly correlated with 1p19q co-deletion, but not with either IDH mutations or ATRX loss [64]. Comparisons of groups based on the statuses of IDH, 1p19q, and TERT revealed that TERT mutation bestows better outcomes in gliomas with TERT mutant/IDH mutation/1p19q co-deletion, but poorer survival in GBM with TERT mutant/IDH mutation without 1p19q co-deletion [63]. Currently, detection of TERT mutations is completed via methyl-specific PCR; in addition, rapid intraoperative testing has been reported [65].

Acid Ceramidase (ASAH1) as a Druggable Target to Combat Multiple Therapy-Resistant Cancers
ASAH1, initially discovered in rat brain homogenates and further characterized and purified from human urine in 1995, is a lysosomal cysteine amidase that catalyzes the transformation of ceramide into sphingosine and free fatty acid ( Figure 2) [66][67][68][69][70][71][72]. Following this, sphingosine kinase 1 (SPHK1) or 2 (SPHK2) phosphorylates sphingosine to produce sphingosine-1-phosphate (S1P), which promotes GBM invasiveness via the upregulation of the urokinase plasminogen activator, its receptor, and the pro-invasive molecule CCN1 (cysteine-rich angiogenic protein 61) (Figure 2) [69,[72][73][74]. On the other hand, high levels of ceramides, carrying fatty acid side chains ranging from 14 to 26 carbons and generated via the action of ceramide synthases (CerS), promote apoptosis in cells that have undergone radio-and chemotherapy via the release of cytochrome c, leading to the activation of caspase-9 and caspase-3 [69][70][71][75][76][77][78][79]. Since its products are involved in the regulation of cell proliferation, multiple studies have linked ASAH1 to multiple cancers such as melanoma, acute myeloid leukemia (AML), and colon and prostate cancers [80][81][82][83][84]. ASAH1 has been proposed as an emerging drug target in AML [85]. Interestingly, over-expression of ASAH1 in prostate cancer promotes resistance to chemotherapy. Prostate cancer upregulates ASAH1 following radiation, which was described as a mechanism enabling the cancer to survive radiation [86]. Consequently, when the activity of ASAH1 is suppressed with an ASAH1 inhibitor named B13, the cells become more sensitive to chemotherapy and radiation as a result of the accumulation of intracellular ceramide up to cytotoxic levels, inducing apoptosis [81,87,88]. Similarly, the acid ceramidase inhibitor ceranib-2 also has activity against the growth of the breast cancer cell lines MCF-7 and MDA MB-231 via the activation of stress-activated protein kinase/c-Jun N-terminal kinase and p38 mitogen-activated protein kinase apoptotic pathways and the inhibition of the Akt pathway [89]. The Wnt/β-catenin signaling pathway appears to play a role in suppressing the proliferation and metastatic potential of cervical cancers when these tumors were treated with a recently identified ASAH1 inhibitor called carmofur [90,91]. The clinical application of carmofur has been attempted and it offered benefits when carmofur was used an adjuvant in patients with early breast cancer in a postoperative setting [92].

ASAH1-Induced Radioresistance in GBM
The sphingolipid pathway was initially implicated in GBM in several studies, by showing that S1P augments the migratory response of the GBM cell line U87MG and that S1P level is significantly higher in GBM tissues compared to the normal gray matter [93,94]. We provided further evidence of the important role that the sphingolipid pathway plays in GBM. We showed that ASAH1 level was negatively correlated with GBM survival [95]. To study the role ASAH1 plays in radioresistant GBM, we developed a stable radioresistant GBM model, in which U87 GBM cells were irradiated, and the surviving cells were perpetuated [96]. In this model, we demonstrated that intracellular ASAH1 was upregulated, and its secretion into extracellular space was also increased in the adult GBM cell line U87 and in the pediatric GBM cell line SJGBM2, suggesting that ASAH1 confers radioresistance to GBM ( Figure 3) [96,97]. Our histochemistry data utilizing patient GBM tissues revealed higher levels of ASAH1 in irradiated tissues compared to control tissues [96]. We suggested that ASAH1 may decrease the overall GBM survival and promote recurrence, which is inevitable, by enhancing the survival of irradiated GBMs via the upregulation of ASAH1, leading to decreasing levels of proapoptotic ceramide molecules and increasing levels of prosurvival S1P molecules (Figure 3) [96]. Despite being resistant to radiation, these cells remained sensitive to the ASAH1 inhibitor carmofur, albeit displaying a slightly higher IC 50 value [96]. More importantly, ASAH1 inhibitors have been proposed as radiosensitizers, on the basis of studies that illustrated a greater suppression of the growth of U87 and prostate cancer xenografts when treated with both conventional radiation therapy and ASAH1 inhibitors [87,98]. Carmofur is the only ASAH1 inhibitor that has been used clinically to treat colorectal cancers [99][100][101]. However, carmofur has several issues that need to be addressed before it can be more widely used. It has very low solubility in aqueous solution, an intravenous formula is unavailable, and the extent to which it can penetrate the blood-brain barrier is poorly understood [91]. One strategy to improve the solubility of carmofur is to take advantage of the recently solved crystal structure of acid ceramidase to help predict how carmofur would fit in its active site and perform appropriate modifications to allow the drug to be both more soluble and potent [102]. Another strategy to combat radioresistance induced by secretion of ASAH1 is to induce the immune system to produce autoantibodies against extracellular ASAH1. The benefit of developing auto-ASAH1 antibodies was demonstrated in melanoma patients. The auto anti-ASAH1 antibodies protected the melanoma patients from lymph node metastasis, and the loss of these antibodies could result in melanoma progression [103]. A strategy to promote the development of auto-ASAH1 antibodies is by immunizing patients against ASAH1, and this may mitigate the proliferation and invasion of radioresistant GBM. Further study is needed to examine whether the auto-ASAH1 antibodies can cross the blood-brain barrier, as there is a paucity of data available regarding the benefit of auto-antibodies in treating neurological diseases. S1P augments the migratory response of the GBM cell line U87MG and that S1P level is significantly higher in GBM tissues compared to the normal gray matter [93,94]. We provided further evidence of the important role that the sphingolipid pathway plays in GBM. We showed that ASAH1 level was negatively correlated with GBM survival [95]. To study the role ASAH1 plays in radioresistant GBM, we developed a stable radioresistant GBM model, in which U87 GBM cells were irradiated, and the surviving cells were perpetuated [96]. In this model, we demonstrated that intracellular ASAH1 was upregulated, and its secretion into extracellular space was also increased in the adult GBM cell line U87 and in the pediatric GBM cell line SJGBM2, suggesting that ASAH1 confers radioresistance to GBM (Figure 3) [96,97]. Our histochemistry data utilizing patient GBM tissues revealed higher levels of ASAH1 in irradiated tissues compared to control tissues [96]. We suggested that ASAH1 may decrease the overall GBM survival and promote recurrence, which is inevitable, by enhancing the survival of irradiated GBMs via the upregulation of ASAH1, leading to decreasing levels of proapoptotic ceramide molecules and increasing levels of prosurvival S1P molecules (Figure 3) [96]. Despite being resistant to radiation, these cells remained sensitive to the ASAH1 inhibitor carmofur, albeit displaying a slightly higher IC50 value [96]. More importantly, ASAH1 inhibitors have been proposed as radiosensitizers, on the basis of studies that illustrated a greater suppression of the growth of U87 and prostate cancer xenografts when treated with both conventional radiation therapy and ASAH1 inhibitors [87,98]. Carmofur is the only ASAH1 inhibitor that has been used clinically to treat colorectal cancers [99][100][101]. However, carmofur has several issues that need to be addressed before it can be more widely used. It has very low solubility in aqueous solution, an intravenous formula is unavailable, and the extent to which it can penetrate the blood-brain barrier is poorly understood [91]. One strategy to improve the solubility of carmofur is to take advantage of the recently solved crystal structure of acid ceramidase to help predict how carmofur would fit in its active site and perform appropriate modifications to allow the drug to be both more soluble and potent [102]. Another strategy to combat radioresistance induced by secretion of ASAH1 is to induce the immune system to produce autoantibodies against extracellular ASAH1. The benefit of developing auto-ASAH1 antibodies was demonstrated in melanoma patients. The auto anti-ASAH1 antibodies protected the melanoma patients from lymph node metastasis, and the loss of these antibodies could result in melanoma progression [103]. A strategy to promote the development of auto-ASAH1 antibodies is by immunizing patients against ASAH1, and this may mitigate the proliferation and invasion of radioresistant GBM. Further study is needed to examine whether the auto-ASAH1 antibodies can cross the blood-brain barrier, as there is a paucity of data available regarding the benefit of auto-antibodies in treating neurological diseases.

Identification of Novel Drug Targets to Combat Radioresistant GBM
The current standard treatment regimen for GBM includes maximal safe surgical resection, followed by radiation therapy combined with concomitant and adjuvant temozolomide [30,104].
However, recurrence of GBM-characterized by radioresistance-remains inevitable [105,106]. The absence of a major progress in the treatment of GBM maybe a result of our poor understanding of both GBM tumor biology and the mechanisms underlying the acquirement of treatment resistance in recurrent GBMs. In support of this view, very little data about the radiation effects on global gene expression at the messenger ribonucleic acid (mRNA) level in a stable radioresistant GBM model are available. Ma et al., in their transcriptome analysis of glioma within hours following irradiation, suggested that the development of radioresistance of glioma may be due to the inactivation of early proapoptotic molecules and to the late activation of antiapoptotic genes [107]. To identify radiation-responsive genes that may enable GBM cells to acquire resistance to radiation, we performed complete RNA sequencing (RNA-seq) of control tissues and our recently established stable, radioresistant U87-based GBM model [96,108]. Our study revealed that the aberrant gene expression observed in irradiated U87-10gy cells regarded, in particular, genes involved in enhancing tumor malignancy and invasion. In irradiated U87-10gy cells, we observed the upregulation of antiapoptotic genes (BNIP3 and SOD2), of genes promoting epithelial to mesenchymal transition, of genes with metalloendopeptidase activity, and of genes involved in the response to hypoxia (Tables 1 and 2) [108]. Metalloproteases are known to promote tumor invasion and metastasis of many cancers by degrading the extracellular matrix [109,110]. MME, MMP2, MMP3, MMP7, MMP12, ADAM9, and ADAM12 were shown to be upregulated in radioresistant GBMs (Tables 1 and 2) [108]. Epithelial to mesenchymal transition, a process characterized by increased cell motility and resistance to chemo-and radiotherapy, is typically induced by TGFB3, which was also upregulated in irradiated U87-10gy cells [108,111,112]. Hypoxia, which is frequent in GBM, induces hypoxia-inducible factor 1-alpha (HIF-1α) and carbonic anhydrase 9 expressions, which in turn promote angiogenesis, migration, cell survival, proliferation, epithelial to mesenchymal transition, and radio-and chemoresistance [111,113,114]. HIF-1α and carbonic anhydrase 9 were upregulated in irradiated GBM cells [108].
On the other hand, we found that the downregulated genes were enriched in tumor suppressors, in genes positively regulating the immune response, in genes involved in p53-dependent apoptosis, and in cell adhesion genes. Suppressing the apoptotic potential through gene expression regulation in the irradiated cells was a proposed mechanism that explained the radioresistant nature of the irradiated GBM cells [107,108]. Many apoptotic genes discovered in our study were known to play major roles in attenuating GBM apoptosis, especially, BBC3, DCC, BEX2, CASP1, IL1B, and SFRP2 [115][116][117][118][119][120]. GBM cells produce an immunosuppressive microenvironment to escape immune surveillance and enhance their own survival, and this can be accomplished through the secretion of transforming growth factor β (TGF-β) to block T cell activation and proliferation [121]. We identified many other downregulated genes involved in the activation of the immune system, especially genes mediating T cell antigen processing and presentation that may enable immune evasion in the radioresistant GBM cells [108].
Considering these global alterations of multiple biological pathways observed in irradiated GBM cells, an effective treatment targeting radioresistant GBM may require a cocktail containing multiple agents targeting multiple implicated pathways in order to have a chance to make a substantial impact on improving the overall GBM survival. Table 1. Upregulated genes of selected enriched gene ontology categories following irradiation are shown on the basis of sets of statistically significant changes (p < 0.05) [108].