Role of viruses in glioblastoma oncogenesis and progression (Review)

1. Introduction

According to the International Agency for Research on Cancer (IARC), > 320,000 individuals were diagnosed with primary central nervous system (CNS) tumors globally in 2022(1). Furthermore, ~80-85% of CNS tumors are gliomas, with glioblastoma (GBM) being the most common and aggressive type (49%) (2). The latest edition of the World Health Organization Classification of CNS tumors ranked GBM as a grade IV glioma characterized by the isocitrate dehydrogenase wild-type gene (3). The GBM incidence increases after the age of 40, reaching its highest rate in individuals aged 75-84, with a clear predominance in men and Caucasian individuals (4). The current standard therapy for newly diagnosed patients involves a combination of neurosurgery, followed by concurrent radiotherapy and temozolomide chemotherapy and subsequently, adjuvant temozolomide (5). Despite this multimodal approach, GBM still exhibits the worst prognosis among primary brain tumors, with a median overall survival of ~15 months and a 5-year survival rate of only 3-5% (6). The etiology of GBM is not entirely known. Yet, an increased risk has been reported in a small number of GBM patients due to some inherited genetic syndromes (such as Li-Fraumeni syndrome) and prior exposure to ionizing radiation (7,8). Given that viruses are implicated in ~10-15% of human cancers, a number of studies have explored their association with GBM (9). Therefore, in the present review, the potential involvement of select viruses in the etiology of GBM was examined, with the aim of providing insight into their possible role in GBM progression.

2. Herpesviruses

Human cytomegalovirus

Human cytomegalovirus (HCMV), also known as human herpesvirus (HHV)-5, is an enveloped, double-stranded DNA virus belonging to the Betaherpesvirinae subfamily. It is a ubiquitous virus that the majority of individuals worldwide encounter during their lifetime. A meta-analysis of seroprevalence data indicated that ~60% of individuals in developed countries and up to 90% in developing countries have been exposed to HCMV (10). Despite its widespread prevalence, HCMV infection is typically asymptomatic in immunocompetent individuals due to an effective immune response. Similar to other herpesviruses (such as HHV-6, Herpes Simplex Virus and Epstein-Barr Virus), HCMV establishes latency following primary infection. Although HCMV exhibits tropism for a wide range of cell types, including endothelial, epithelial and fibroblast cells, its latent reservoir is primarily in hematopoietic cells, particularly CD34+ hematopoietic progenitor cells and myeloid lineage cells. In immunocompromised individuals, reactivation from latency can result in severe, potentially life-threatening disease. During its replicative cycle, HCMV gene expression progresses through three phases: Immediate-early (IE), delayed-early (DE) and late. Proteins expressed during the IE phase regulate cellular and viral gene expression. By contrast, DE proteins are key in viral DNA replication, whereas late proteins are primarily structural and involved in viral capsid assembly. Due to its potential involvement in carcinogenesis, the IARC has classified HCMV in group 2B ‘possibly carcinogenic to humans’ based on limited evidence for human cancers (11).

Association between HCMV infection and GBM

In 2002, Cobbs et al (12) reported for the first time, the presence of HCMV in all examined GBM samples, while no viral presence was detected in the adjacent non-neoplastic brain tissue. This finding was subsequently corroborated by further independent studies, a number of which found a higher prevalence of HCMV in GBM compared with other gliomas, non-neoplastic encephalopathies or epileptic brain tissues (13,14). Despite this, additional research has reported conflicting results: Either a very low detection rate or a complete absence of the virus in GBM samples, even when using validated standard detection techniques that can identify the virus in positive control samples (15-17).

Given the low viral load of HCMV in GBM, the discrepancy of these results may be primarily attributed to method sensitivity and experimental conditions. Factors such as section age, thickness and slide pretreatment and deparaffinization can affect the results of both immunohistochemistry (IHC) and in situ hybridization (ISH) (18). Cobbs et al (12) and additional investigators have emphasized the importance of a protocol that includes optimized antigen retrieval and high antibody concentrations to detect low-level viral proteins (19,20). However, others have argued that this optimized protocol may increase staining artifacts, leading to false positive results (16,21). A number of studies have used other assays, including PCR, ISH, and next-generation sequencing, to further determine IHC results, with findings being frequently consistent with ISH data (12,18,22,23).

Conventional PCR and next-generation sequencing (NGS) often yield negative results, as HCMV copy numbers are below the detection thresholds of these assays. For this reason, more sensitive PCRs, such as nested PCR and droplet digital PCR (ddPCR), have been performed; however, the results either remain negative or show lower prevalence compared with IHC and ISH (18,24-26). The main explanation for this phenomenon is the diversity of HCMV strains and the genetic alterations in the viral genome that may influence primer and probe sensitivity and specificity (14). Despite these divisive findings, a recent meta-analysis including 988 patients from 19 studies reported a pooled prevalence of 68% (95% CI: 0.53-0.80; P<0.0001; I2=91.4%), reinforcing the widespread presence of HCMV in GBM tissue (27). Furthermore, preclinical and clinical studies targeting HCMV have demonstrated improved clinical outcomes in patients with GBM, supporting the hypothesis that HCMV is involved in gliomagenesis. This will be expanded upon in the present review.

Given that HCMV has failed to transform normal human cells in vitro, ‘oncomodulation’ has been proposed as an alternative model to explain how the virus promotes gliomagenesis. This paradigm describes the indirect role of HCMV, where it enhances the oncogenic properties of pre-transformed cells. Despite this, recent findings by El Baba et al (28) and Guyon et al (29) challenged this paradigm. The authors isolated clinical high-risk strains and successfully generated CMV-elicited GBM cells which exhibited GBM-like features, including plasticity, stemness and invasion, capable of forming tumors in xenografted mice (28,29). This evidence suggests that HCMV may also exert a direct oncogenic effect though the precise conditions governing this capacity remain to be defined, warranting further investigation into its role in GBM pathogenesis.

Mechanisms of HCMV in gliomagenesis. All mechanisms as further described are summarized in Fig. 1. i) Effects of HCMV on cell cycle and cellular survival. HCMV contributes to gliomagenesis by promoting cellular proliferation and survival through a number of mechanisms, starting from its attachment to the host cell to its downstream signaling and gene regulation. Platelet-derived growth factor receptor α (PDGFRα), a tyrosine kinase receptor frequently overexpressed in GBM cells, serves a pivotal role in this process. PDGFRα is important in HCMV internalization, gene expression and viral particle production (30). Through its interaction with this receptor, HCMV facilitates the activation of oncogenic pathways that may support tumor development and progression.

Figure 1 Role of HCMV gene products in GBM. HCMV binds to PDGFRα through gB to enter the cell, leading to PDGFRα phosphorylation and activation of the PI3K/Akt signaling which promotes cell proliferation and survival. IE1-72 enhances cell immortalization and cycle progression by upregulating hTERT expression, inactivating pRb and p53 and stimulating PI3K/Akt. IE2-86 promotes cell survival, migration and angiogenesis by activating AFT5, δ-RON and TSP-1, respectively. US28 induces angiogenesis through the activation of VEGF through the Akt/mTOR/HIF-1 signaling and IL-6/STAT3 axis, it also activates FAK which stimulates cellular migration. pp71 upregulates the expression of the proangiogenic factor SCF and inhibits the surface expression of MHC-I resulting in the cancer cell being unrecognizable by the immune system. HCMV-infected cells secrete cmvIL-10, which causes the conversion of monocytes to an immunosuppressive phenotype (M2) and the expression of PD-L1. Created in BioRender (https://BioRender.com/cqdj1fa). HCMV, human cytomegalovirus; GBM, glioblastoma; gB, glycoprotein B; PDGFRα, platelet-derived growth factor receptor-α; cmvIL-10, cytomegalovirus IL-10; PD-L1, programmed death-ligand 1; JAK, janus kinase; FAK, focal adhesion kinase; SCF, stem cell factor; hTERT, human telomerase reverse transcriptase; ATF5, activating transcription factor 5; pRb, retinoblastoma protein; RON, récepteur d' origine nantais; TSP-1, thrombospondin 1; MHC-I, major histocompatibility complex class I; hnRNP, heterogeneous nuclear ribonucleoprotein; CaMKII, calcium/calmodulin-dependent protein kinase II; IE2-86, 86-kDa immediate-early 2 protein; HIF-1, hypoxia-inducible factor 1.

HCMV binds to PDGFRα through its viral envelope glycoprotein B (gB), leading to PDGFRα phosphorylation and subsequent activation of the PI3K/Akt signaling pathway. This activation enhances cellular proliferative activity and inhibits apoptosis, thereby promoting viral persistence and potentially contributing to oncogenic processes in the host cell (31). At the intracellular level, the viral protein IE1-72 has been shown to activate the PI3K/Akt signaling pathway, contributing to cell survival and proliferation (32). The hyperactivation of this axis is a hallmark of aggressive tumor behavior in GBM, often associated with treatment resistance and poor prognosis (33).

Beyond the dysregulation of growth factor signaling, inactivation of tumor suppressor genes signaling and human telomerase reverse transcriptase (hTERT) gene mutation are frequent alterations that define the genomic landscape of GBM (34). In HCMV infection, IE1-72 was found to promote cell immortalization and genetic instability of GBM cells by dysregulating the oncosuppressor proteins p53 and retinoblastoma protein (pRb) and increasing hTERT expression (32,35,36). In an experimental study using a panel of GBM cell lines, IE1-72 expression exhibited a divergent effect on cell proliferation, triggering either accelerated proliferation with steady-state expression of pRb and the p53 family or cell cycle arrest with high p53 expression. In the context of oncomodulation, this study suggested that the effect of IE1-72 may be conditioned by genetic alterations already present in these cell lines (32). Despite this, a recent study demonstrated that IE1 of a high-risk HCMV strain (HCMV-DB strain) binds to and degrades p53 and pRb through proteasome-mediated pathways, leading to cell cycle entry and inhibition of apoptosis, similar to other viral oncoproteins [such as human papillomavirus (HPV) E6 and E7 oncoproteins] (37). Another marked viral protein is 86-kDa immediate-early 2 protein (IE2-86), which contains domains that bind both pRb and the TATA-binding protein, allowing it to induce cell cycle progression and hijack the cellular transcription machinery (38-40). By contrast, IE2 may also block the cell cycle in the G1 phase independently of p53 and pRb and the underlying mechanism remains to be defined (41). In addition, IE2-86 further enhances cellular survival by inducing acetylation of activating transcription factor 5, thereby increasing its transcriptional activity and inhibiting apoptosis (42,43).

An additional viral protein implicated in HCMV-mediated gliomagenesis is US28, a viral G protein-coupled receptor homologous to the human C-C chemokine receptor type 1. US28 serves a key role in evading host immune responses and has been shown to promote tumorigenic processes by creating a positive loop that activates the IL-6/Janus kinase 1/STAT3 axis initiated by NF-κB activation (44). US28 also activates endothelial nitric oxide synthase, which induces the proliferation of glioma-initiating cells, further contributing to GBM progression (45).

ii) Effects of HCMV on cell invasion and migration. In addition to promoting cellular proliferation and survival, HCMV enhances the invasiveness and migration capacity of GBM cells through numerous molecular mechanisms. The HCMV-encoded protein IE2-86 has been shown to promote GBM migration by upregulating heterogeneous nuclear ribonucleoprotein (hnRNP)-A2B1, a factor involved in precursor-mRNA splicing of the Recepteur d'Origine Nantais (RON) protein (46). hnRNP A2B1 induces the skipping of exon 11 in the mature mRNA, producing a constitutively active δ-RON isoform that promotes cell motility and metastasis (46). In addition, the overexpression of the viral G protein-coupled receptor US28 is associated with increased invasiveness in GBM (45). It was demonstrated that US28 activates certain non-receptor cellular kinases, including focal adhesion kinase (FAK) and proline-rich tyrosine kinase 2, involved in cell migration and invasion (47,48). Furthermore, HCMV infection can accelerate epithelial-to-mesenchymal transition (EMT), a key process in tumor invasiveness, as well as strengthen cell invasion while collaborating with TGF-β1 through JNK pathway activation (49). The evaluation of EMT-related genes expression has exhibited an upregulated expression of MMP-2 and vimentin, as well as a low expression of E-cadherin in HCMV-positive GBM cell lines after TGF-β1 exposure, compared with HCMV-negative cell lines (49). These findings suggest a synergistic role for HCMV and host signaling in promoting the invasive phenotype of GBM.

iii) Effects of HCMV on angiogenesis. Angiogenesis is a marked hallmark of GBM that maintains cell proliferation and ensures invasion and potential metastasis. The main regulator of this process is VEGF, the expression of which can be induced by HCMV through numerous mechanisms, primarily mediated by the viral chemokine receptor US28. The first mechanism involves US28-mediated activation of certain transcription factors, including STAT3, hypoxia inducible factor 1 (HIF-1), specificity protein 1 and activator protein 2 through p38 and p44/42 MAPK signaling pathways, which finally induces VEGF promoter activation and increased VEGF expression (50). Secondly, US28 enhances VEGF production through the IL-6/STAT3 signaling axis, resulting in VEGF secretion and supporting an angiogenic tumor microenvironment (44,45). The third mechanism is also led by US28, which constitutively stimulates Gα/q protein-mediated signaling pathways resulting in the activation of calcium/calmodulin-dependent protein kinase II and the Akt/mTOR cascade to increase the HIF-1α activity. HIF-1α forms a transcription factor with HIF-1β that binds to hypoxia response element regions of different genes, including the VEGF gene (51). In addition, US28 induces the phosphorylation and dimerization of pyruvate kinase M2 (PKM2), a glycolytic enzyme, giving it a transcriptional activity targeting HIF-1 genes. This creates a HIF-1α/PKM2 feedforward loop that promotes tumor cell proliferation, angiogenesis and a glycolytic phenotype characteristic of GBM (51).

Besides VEGF, HCMV infection can also alter the expression of a number of other angiogenic factors. The HCMV tegument protein pp71 upregulates stem cell factor (SCF) expression, a potent pro-angiogenic factor in GBM, through activation of NF-κB signalling (52). By contrast, IE1-72 downregulates the expression of thrombospondin-1, a well-established anti-angiogenic factor (35). HCMV infection is also associated with the upregulation of endocan, a molecule involved in angiogenesis and vascular remodeling, as well as arginase-2, which enhances vasculogenic mimicry and tumor invasion by modulating MMP-2 and MMP-9 activity (53,54).

iv) Effects of HCMV on stemness. Cancer stem cells are a small subpopulation of quiescent, pluripotent and self-renewing cancer cells identified in numerous malignant tumors, including GBM (55). These cells are known to initiate tumors and exhibit resistance to chemotherapy and radiotherapy, thereby contributing to recurrence and metastasis (56,57). HCMV infection has been shown to promote stemness in GBM cells by upregulating the expression of GBM stem cell markers, namely, CD133, nestin, Notch1, OCT3/4 and Sox2. It also induces sphere-like proliferation, indicative of self-renewal capacity and maintains the cells in an undifferentiated state (58-60). Furthermore, HCMV-infected GBM cells exhibiting a stem cell-like phenotype have been shown to resist the standard-of-care therapeutic agent: Temozolomide. However, that sensitivity can be reversed with anti-viral treatment using ganciclovir (60). Soroceanu et al (61) suggested that the viral IE proteins are implicated in promoting stemness in GBM, as they co-localize with CD133, Sox2 and nestin. Notably, attenuation of their expression was found to be sufficient to suppress neurosphere formation and reduce Sox2 expression. In addition, it was also established that HCMV promotes stemness through the upregulation of enhancer of zeste homolog 2 and Myc expression (22).

v) Effects of HCMV on immune evasion. It has been well established that HCMV has evolved numerous mechanisms to evade host immune detection (62-64). However, it is unclear whether HCMV may also mediate the immune escape of infected GBM cells. A study addressing this question has demonstrated that HCMV tegument protein pp71 blocks the transport of major histocompatibility complex class I molecules to the cell surface, thereby hindering the ability of immune cells to recognize and target GBM cells (65). In addition, HCMV contributes to the establishment of an immunosuppressive microenvironment by modulating the polarization of tumor-associated macrophages from a M1 to M2 phenotype. HCMV-infected GBM cancer stem cells produce a viral homolog of human IL-10 (cmvIL-10), which facilitates the differentiation of recruited monocytes into macrophages or microglia with an immunosuppressive (M2) phenotype and increased expression of programmed death ligand 1 (PD-L1). Consequently, this stimulation activates viral infection in M2 cells with high IE1 and STAT3 expression, which further enhances oncogenesis by supporting VEGF production, TGF-β-mediated immunosuppression and invasion (66). In an additional study, HCMV was also shown to induce M1/M2 polarization through activation of NF-κB and PI3K in infected monocytes/macrophages (67). Furthermore, other immunosuppressive factors, such as TGF-β, IL-6, arginase II and PD-L1, were also found to be upregulated by HCMV, creating a robust immunosuppressive microenvironment that supports the expansion and activation of regulatory T cells and myeloid-derived suppressor cells (54,68,69). In a recent study, Long et al (70) identified FXYD domain-containing ion transport regulator 6+ T cells, an immunosuppressive T cell subpopulation, as well as macrophages co-expressing both cancer cell and macrophage markers in HCMV-infected GBM. HCMV appears to mediate crosstalk between these cell populations, contributing to the development of an immunosuppressive microenvironment.

Targeting HCMV in GBM: Current therapeutic landscape

Elucidating the oncomodulatory effects of HCMV on GBM offers a promising avenue for adopting this virus as a novel therapeutic target. Numerous treatment modalities have been investigated or are currently under clinical assessment for this purpose.

Antiviral drugs. Food and Drug Administration-approved HCMV antiviral agents, typically used to treat immunocompromised patients with HCMV infection, have been tested for efficacy against GBM. Valganciclovir is a prodrug of ganciclovir and a nucleoside analog that blocks viral replication. The first randomized trial (Valcyte Treatment of Glioblastoma Patients in Sweden) didn't show any significant survival benefit of valganciclovir; however, the subgroup analysis demonstrated that patients treated with valganciclovir for >6 months exhibited an improved overall survival (OS) compared with those treated for <6 months (24.1 vs. 13.1 months; P<0.0001) (71). During this trial, new patients requested valganciclovir as an add-on to their standard therapy and a retrospective analysis revealed an unexpectedly higher survival rate, reinforcing the efficacy of valganciclovir in improving patient prognosis (72,73). In addition, a recent meta-analysis of five randomized controlled trials has shown that ganciclovir was associated with a 20% increase in 2 and 4-year survival in the GBM cohort (74). At the molecular level, these antiviral drugs are activated by the viral protein pUL97. Therefore, a mutation of the UL97 gene leads to drug resistance. Surpassing this limitation, cidofovir, an additional HCMV antiviral drug, is activated by host-cell kinases, making it less susceptible to viral resistance. Preclinical evidence has indicated that cidofovir not only downregulates HCMV gene expression and promotes apoptosis but also acts as a radiosensitizer, enhancing the cytotoxicity of radiotherapy by inhibiting DNA repair mechanisms (75).

Viral targeted immunotherapy. Dendritic cells (DC) are antigen-presenting cells known for their important role in controlling the immune response (76,77). Leveraging this, a new treatment modality was developed using these cells to boost immunity within the tumor. The HCMV-targeted DC vaccine is fundamentally patient-derived DCs electroporated with viral pp65 mRNA, enabling them to present pp65 antigen to T cells and induce expansion of antigen-specific immunity after being injected back into the patient. Antitumor Immunotherapy Targeted Against Cytomegalovirus was the first randomized trial to introduce the pp65 DC vaccine in patients with GBM after preconditioning the vaccine site with tetanus-diphtheria toxoid (78). Subsequently, the same group conducted an additional clinical study in a separate cohort to evaluate the combination of the pp65 DC vaccine with a granulocyte-macrophage colony-stimulating factor adjuvant and dose-intensified temozolomide (79). According to the latest follow-up of these trials, both studies showed favorable outcomes: ~33% of cases remained recurrence-free 5 years after diagnosis in the first trial and 36% survived 5 years after diagnosis in the second trial (80).

An additional immunological approach proposed to treat GBM is HCMV-directed adoptive T-cell (ATC) therapy. This approach uses autologous T cells that are expanded in vitro after stimulation with HCMV epitopes. Early feasibility studies have demonstrated that infusing these cells to consolidate recurrent GBM treatment was safe and improved patient survival (81,82). In adjuvant settings, a subsequent study extended this strategy to newly diagnosed GBM, determining safety and durable maintenance of HCMV-specific T cells after infusion. This study also reported improved OS among patients treated before recurrence compared with among those treated after progression (23 vs. 14 months; P=0.018) (83). More recently, the Evaluation of Recovery From Drug-Induced Lymphopenia Using Cytomegalovirus-Specific T Cell Adoptive Transfer randomized trial has shown that combining ATC with pp65-loaded DC vaccine increases the polyfunctionality of HCMV-specific CD8+ T cells and this effect was associated with improved OS (84).

Epstein-barr virus

Epstein-Barr virus (EBV), also known as HHV-4, is a ubiquitous γ-herpesvirus first discovered in 1964 by Epstein et al (85) in a biopsy from a patient with Burkitt lymphoma. EBV is an enveloped virus with a linear double-stranded DNA genome of ~175 kb encoding >80 open-reading frames (86). Epidemiological studies have indicated that >90% of the adult population worldwide is chronically infected with EBV (87-92). Transmission occurs mainly through saliva, although it can also be spread through organ transplantation or exposure to other bodily fluids (93). The virus infects a number cell types but typically exhibits tropism for B lymphocytes and epithelial cells (94). Furthermore, EBV was the first identified oncovirus associated with numerous malignancies in humans, primarily B cell lymphomas, nasopharyngeal carcinoma and gastric carcinoma (86). It encodes for certain gene products, including the latent membrane protein 1 oncoprotein and the family of non-coding RNAs microRNA (miR)-BART, responsible for the dysregulation of key signaling pathways involved in oncogenic events (95).

Although not primarily classified as a neurotropic virus, EBV can infect the CNS either directly or indirectly through the infected B cells, leading to numerous CNS disorders, such as Parkinson's disease, multiple sclerosis, meningitis, encephalitis and Alzheimer's disease (96). EBV enters astrocytes primarily through the complement receptor type 2 (or CD21) (97). It has also been implicated in the oncogenesis of primary CNS lymphoma, particularly in immunocompromised individuals, where it is associated with poorer prognosis (96,98,99). Based on these observations, a number of studies have investigated whether EBV may act as an etiological agent in GBM and other gliomas; however, results have been inconsistent. In 2019, Limam et al (100) reported the highest detection rate of EBV in gliomas compared with previous studies (Table I) (24,101-105). Using conventional PCR, this study found EBV DNA in 21.4% of the tested gliomas, all of which were GBM. Similarly, two independent studies investigating glioma subtypes also detected the presence of EBV exclusively in GBM samples, but not in lower-grade gliomas (24,101). Furthermore, two research groups focused solely on GBM specimens and reported comparable EBV detection rates of 21.4 and 20.7%, respectively (26,106). By contrast, other studies employing PCR and reverse transcription (RT)-PCR failed to detect EBV in GBM or other gliomas (17,102), highlighting inconsistencies in the literature. Notably, in research investigating novel miRNA biomarkers for GBM, Herman et al (107) identified differential expression levels of six EBV-derived miRNAs in the blood of patients with GBM compared with healthy controls. These miRNAs are known to serve roles in the switch from latent to lytic infection and in immune evasion.

Table I Studies investigating the presence of EBV in GBM and other types of glioma.

Table I

Studies investigating the presence of EBV in GBM and other types of glioma.

A, Studies including only GBM
First author, year	Glioma type	Sample size	Detection method	Positive results, n (%)	Notes	(Refs.)
Hashida et al, 2015	GBM	39	Reverse transcription PCR (LMP-1)	0 (0.0)		(17)
Ghaffari et al, 2021	GBM	42	Conventional PCR (EBNA3C)	9 (21.4)		(26)
Zavala-Vega et al, 2017	GBM	29	IHC (LMP-1) and ISH (EBER)	6 (20.7)		(106)
Cimino et al, 2014	GBM	21	Next-generation sequencing	5 (24.0)		(108)
			ISH (EBER)	0 (0.0)
B, Studies including multiple types of glioma
First author, year	Glioma type	Sample size	Detection method	Positive results, n (%)	Notes	(Refs.)
Limam et al, 2019	Grade I-IV	112 (82 GBM)	Conventional PCR (BamM region)	24 (21.4)	Detected only in GBM cases	(100)
			IHC (LMP-1)	4 (3.6)
			ISH (EBER)	4 (3.6)
Strojnik et al, 2017	Grade III-IV	45	RT-PCR (ebna)	3 (6.7)	Detected only in grade IV	(101)
Lin et al, 2016	GBM, astrocytoma (grade I-III)	63 (45 GBM)	Multiplex droplet digital PCR (LMP-1)	4 (6.3)	Detected only in FFPE GBM samples	(24)
Cosset et al, 2014	GBM, low-grade astrocytoma and others	28 (20 GBM)	RT-PCR (BALF5)	0 (0)		(102)
Strong et al, 2016	Low-grade and GBM	701 (170 GBM)	WGS RNA-seq RT-PCR	0 (0)	In this virome assessment, EBV was detected in 12 GBM WGS datasets, but this result was not determined due to low viral reads detected and the potential for contamination	(103)
Khoury et al, 2013	Low-grade and GBM	215 (168 GBM)	RNA-seq	0 (0)		(104)
Fonseca et al, 2014	GBM, astrocytoma (grade I-III) and others	75 (18 GBM)	Conventional PCR (BamM region)	11 (14.7)	Only one case of GBM was positive for EBV	(105)

[i] EBV, Epstein-Barr virus; GBM, glioblastoma; EBNA3C, Epstein-Barr nuclear antigen 3C; IHC, immunohistochemistry; ISH, in situ hybridization; LMP-1, latent membrane protein 1; EBER, Epstein-Barr encoded RNA; RT-PCR, reverse transcription PCR; WGS, whole genome sequencing; RNA-seq, RNA sequencing; FFPE, formalin-fixed paraffin-embedded.

Numerous studies have investigated the presence of EBV in glioma specimens using NGS techniques. Cimino et al (108) performed targeted NGS on 21 high-grade gliomas using a panel of 115 selected genes and detected EBV DNA in 5 cases. Similarly, Strong et al (103) analyzed publicly available whole-genome sequencing datasets from patients with low-grade glioma and GBM to perform a virome assessment. This study identified EBV DNA sequences in 13 GBM datasets. However, due to the low number of viral reads and lack of validation with additional methods, these findings remain inconclusive. By contrast, Khoury et al (104), through a bioinformatic analysis of available RNA sequencing (RNA-seq) datasets from low-grade glioma and GBM patients, found no evidence of EBV RNA expression. Notably, to understand GBM pathogenesis, a recent study comparing gene expression between patients with GBM and healthy individuals reported marked enrichment of EBV-associated signaling pathways, such as NF-κB, MAPK, JAK-STAT and Akt, suggesting a potential etiological role for EBV in GBM tumorigenesis (109).

Human herpesvirus 6 (HHV-6)

HHV-6 is a lymphotropic β-herpesvirus initially identified by Salahuddin et al (110) in 1986 in patients with lymphoproliferative disorders. Similarly to all other herpesviruses, HHV-6 is highly prevalent, with >90% of adults harboring the virus asymptomatically (111). Initially believed to have two variants: HHV-6A and HHV-6, these two were reclassified as distinct species in 2012 by the International Committee on Taxonomy of Viruses based on their biological, immunological and epidemiological characteristics, as separate (112). HHV-6B is more commonly detected in the peripheral blood leucocytes of healthy individuals, whereas the pathogenic role of HHV-6A remains poorly understood (113,114). Transmission of HHV-6 occurs primarily through saliva and the infection usually occurs through the nasopharynx and olfactory entry routes between 6 months and 2 years of age (115,116). HHV-6 exhibits a broad cellular tropism infecting a wide range of cell types both in vivo and in vitro, including fibroblasts, epithelial cells, neurons, oligodendrocytes and astrocytes (117-121). Despite HHV-6 infection being generally asymptomatic, it can cause roseola infant (Exanthema subitum) and a variety of complications in immunocompromised individuals, such as hepatitis, pneumonitis, encephalitis, multiple sclerosis and epilepsy (122-126). Although HHV-6 is not an established oncovirus, it has been detected in numerous malignancies, including hematological cancers, as well as gastrointestinal, gynecological, neurological and head and neck cancers (127). In 1995, Luppi et al (128) used PCR to assess HHV-6 prevalence in 37 biopsies of neuroglial tumors and 9 necropsies of normal brain tissue. The frequency of HHV-6 DNA was found to be 66% (6/9) in normal tissue but only 16% (6/37) in tumor samples, five of which were GBM. A subsequent study, with a larger cohort, investigated the presence of HHV-6 in 118 brain tumors and 31 normal brain tissues using nested PCR. The results revealed 14 HHV-6 positive cases out of 31 patients with GBM, a frequency of 45% (129). More recently, Crawford et al (130) analyzed the presence of HHV-6 in adult CNS tumors using ISH, nested PCR and IHC. In the large cohort of this study, consisting of >200 adult CNS tumors, 47% of patients were positive for U57 major capsid protein, 24% exhibited the presence of HHV-6A/B early antigen (p41) and 35% had a detectable level of late antigens (glycoprotein 116/64/54), with none of these markers present in any of the 25 control samples. Among grade IV astrocytomas, the respective frequencies were 47, 28, 32%. The detection of both early and late antigens indicated an active HHV-6 infection, as opposed to a latent viral state.

To understand the potential role of HHV-6 in glioma development, a research group evaluated inflammatory cytokines levels in cyst fluid samples from patients with glioma infected with the virus. After determination of the results on astrocytes infected with the isolated HHV-6, this study demonstrated that HHV-6 infection leads to elevated levels of IL-6, IL-8, TNF-α and TGF-β in glioma, suggesting that it may provide a pro-inflammatory environment conducive to gliomagenesis (131). Similarly, the HHV-6 encoded direct repeat 7 protein has been shown to upregulate the expression of pro-inflammatory cytokines such as IL-1β, IL-6 and IL-8 in glioma cells. It can also accelerate cell proliferation and promote cell migration and invasion by facilitating extracellular matrix degradation (132). In summary, molecular and epidemiological studies have suggested a potential role for HHV-6 in gliomagenesis; however, further investigations are needed to determine whether this virus is an oncogenic driver or simply an opportunistic pathogen that reactivates in response to microenvironmental immunodeficiency.

3. Polyomaviruses

John Cunnigham virus (JCV)

JCV is an opportunistic neurotropic polyomavirus originally isolated in 1971 from the brain tissue of a patient with Hodgkin's lymphoma (133). Seroepidemiological studies have indicated that 8-10% of individuals contract the virus before 6 years old, with seroprevalence increasing to ~70% in adulthood (134-138). Fortunately, the infection is asymptomatic; however, JCV can reactivate from latency under severe immunosuppressive conditions, leading to progressive multifocal leukoencephalopathy (139). JCV has a circular double-stranded DNA genome composed of early and late coding regions separated by a regulatory region (140). Although humans are the primary host of this virus, its tumorigenic potential has been well documented in animal models. Numerous studies have demonstrated that JCV induces a wide variety of brain tumors in non-human primates and rodents (141,142).

Large T antigen (Tag) is a viral protein encoded by the early region of the JCV genome. It is considered an oncoprotein due to its key role in tumorigenesis in experimental animal models (143). Tag drives cell cycle progression into the S-phase by interacting with key regulatory proteins, including pRb and the tumor suppressor p53, thereby disrupting their tumor-suppressive functions (144,145). These interactions impair DNA damage repair mechanisms and enhance uncontrolled cellular proliferation.

Similarly to a number of ubiquitous viruses, the association between JCV and human cancers is a divisive topic. Numerous studies have reported the presence of JCV in human CNS malignancies, such as GBM, medulloblastoma and oligoastrocytoma (146-149). However, the prevalence of JCV in GBM tissues varies markedly across studies (Table II). The highest prevalence has been reported by Del Valle et al (150), who examined the presence of JCV DNA sequences and the expression of viral regulatory protein Tag in 85 CNS tumor specimens using PCR and IHC, respectively. JCV DNA was detected in 69% of all samples and 57.1% of GBM cases, whereas the Tag protein expression was observed in only a small subset of cases. In a more recent study, Afshar et al (151) evaluated p53 expression levels in JCV-positive and JCV-negative patients with GBM using reverse transcription PCR (RT-PCR). The results revealed a lower expression of p53 in patients infected with JCV. While these findings support a potential etiological role of JCV in gliomagenesis, two independent investigations suggested no association, reporting very low detection rates of the virus in GBM tissues (3.6 and 2.0%) (152,153). Shifting focus from tumor tissue detection, a serological case-control study assessed the JCV infection history by measuring serum antibodies against the viral capsid protein (VP)-1 in two cohorts. The data revealed no marked association between JCV seropositivity and the risk of all gliomas overall or GBM specifically (154).

Table II Prevalence of JCV in GBM tissue.

Table II

Prevalence of JCV in GBM tissue.

First author, year	Sample size	Detection method (target sequence)	Results, n (%)	(Refs.)
Del Valle et al, 2001	21	PCR followed by southern blotting hybridization (early region)	12 (57.1)	(150)
Afshar et al, 2016	199	RT-PCR (Tag)	25 (12.6)	(151)
Limam et al, 2020	82	Conventional PCR (Tag) and ISH (whole genome)	3 (3.6)	(152)
Rollison et al, 2005	102	PCR followed by southern blotting hybridization (Tag) Real-time PCR (Tag)	2(2) 0 (0)	(153)
Boldorini et al, 2003	13	Nested PCR followed by RFLP (Tag)	7 (53.8)	(155)
Delbue et al, 2005	21	RT-PCR (Tag)	11 (52.4)	(156)
Eftimov et al, 2016	51	RT-PCR (Tag) RT-PCR (non-coding control region)	20 (39.2) 2 (3.9)	(157)
Caldarelli-Stefano et al, 2000	5	Nested-PCR (Tag)	0 (0)	(158)

[i] GBM, glioblastoma; RFLP, restriction fragment length polymorphism; RT-PCR, reverse transcription PCR; ISH, in situ hybridization; Tag, T antigen; JCV, John Cunningham virus.

Simian virus 40 (SV40)

SV40 is a polyomavirus that naturally infects rhesus macaques. Human infection with SV40 is uncommon, with seroprevalence of ~2% among healthy adults, and typically requires close contact with infected monkeys (136,159,160). Between the mid-1950s and early 1960s, millions of individuals were accidentally infected by SV40 worldwide due to contaminated polio vaccines prepared using infected monkey kidney cell cultures (161). Shortly after its discovery, SV40 was found to exhibit strong oncogenic potential. It can transform human cells in vitro and induce tumors in animal models, including primary bone and brain cancers, malignant mesotheliomas and lymphomas (162,163). Consequently, a number of investigators have evaluated the presence of SV40 in human tumors to assess its possible role in human carcinogenesis.

Using PCR followed by southern blotting hybridization, Martini et al (164) screened for the presence of the SV40 Tag coding sequences in 83 brain tumors, including 30 GBM. SV40 was detected in 33% of GBM samples, while no viral sequences were found in normal brain tissue (164). Using the same methodology, Huang et al (165) investigated the presence of SV40 Tag DNA sequence across GBM subtypes, including primary, secondary and giant cell GBM. The results of this study revealed a higher prevalence of SV40 sequence in secondary GBM (59%) compared with other subtypes (Table III), with molecular profiles closely resembling low-grade or anaplastic astrocytomas. Since secondary GBM typically arises from the progression of low-grade astrocytomas, the authors suggest that tumors with slower progression and longer clinical history are more likely to be susceptible to SV40 infection (165). In addition, another study conducted on 32 GBM samples detected the regulatory region of SV40 in three elderly patients, with the findings demonstrated by DNA sequencing and ISH following conventional PCR analysis (166). More recently, Limam et al (152) investigated the presence of SV40 DNA in 82 patients with GBM using independent PCR assays targeting the Tag gene, the transcriptional regulatory region and VP1. This approach revealed that 14.5% of GBM samples tested were positive for SV40. Despite these findings suggesting a possible role for SV40 in the initiation or progression of GBM, other studies have reported negative results (153,156,167). In a large series of various brain tumors, Rollison et al (153) investigated the presence of SV40 using two independent methods, followed by southern blotting hybridization compared with RT-PCR conducted in two different laboratories. This analysis revealed that SV40 was absent in all GBM samples, providing evidence against a causal association between SV40 and GBM.

Table III Prevalence of SV40 in GBM tissue samples.

Table III

Prevalence of SV40 in GBM tissue samples.

First author, year	Sample size	Detection method (targets)	Results n (%)	(Refs.)
Limam et al, 2020	82	Conventional PCR (Tag gene, regulatory region and VP1 gene), ISH (whole genome) and IHC (Tag and tAg)	12 (14.6)	(152)
Rollison et al, 2005	102	PCR followed by southern blotting hybridization (Tag) and RT-PCR (Tag)	0 (0)	(153)
Delbue et al, 2005	21	Nested-PCR followed by RFLP (Tag)	0 (0)	(156)
Martini et al, 1996	30	PCR followed by southern blotting hybridization (Tag)	10 (33.0)	(164)
Huang et al, 1999	28 (Primary GBM)	PCR followed by southern blotting hybridization (Tag)	7 (25.0)	(165)
	22 (Secondary GBM)		13 (59.0)
	18 (Giant cell GBM)		2 (11.0)
Kouhata et al, 2001	32	Conventional PCR (regulatory region), DNA sequencing, ISH (mRNA Tag) and IHC (Tag)	3 (9.4)	(166)
Sabatier et al, 2005	20	IHC (Tag)	0 (0.0)	(167)
Zhen et al, 2001	8	Immunoprecipitation followed by silver staining and western blotting (Tag)	4 (50.0)	(169)

[i] GBM, glioblastoma; Tag, large T antigen; tAg, small T antigen; ISH, in situ hybridisation; VP1, viral capsid protein 1; IHC, immunohistochemistry; RT-PCR, reverse transcription PCR; RFLP, restriction fragment length polymorphism; SV40, Simian virus 40.

An additional study evaluated the association between SV40 infection and GBM by comparing the seroprevalence of the virus among patients with GBM, cancer-free individuals and patients with breast cancer. To minimize potential cross-reactivity due to the high degree of protein homology between SV40, JCV and BK viruses (BKV). Mazzoni et al (168) performed indirect ELISA using specific synthetic peptides specifically designed to mimic SV40 VPs. The results revealed a notably higher prevalence of SV40 antibodies in patients with GBM (34%) compared with healthy subjects (15%) and patients with breast cancer (15%).

BK virus (BKV)

Similar to JCV, the human BKV is a ubiquitous polyomavirus that infects >80% of the immunocompetent population (170). It was first isolated in 1971 from the urine of a renal transplant patient (171). BKV infects the renal epithelium and establishes latency. Upon severe immunosuppressive events, this opportunistic virus can reactivate, leading to a variety of clinical manifestations, including hemorrhagic cystitis, ureteral stenosis, pneumonia, encephalitis and colitis. Notably, BKV is also a notable cause of acute interstitial nephritis in ~10% of kidney transplant recipients, which leads to graft loss in 15-100% of affected patients (172).

Similarly to polyomaviruses, BKV has demonstrated oncogenic potential by inducing tumors in experimental animal models and transforming normal cells in vitro (173). Its genome encodes two key oncoproteins: Small Tag and large Tag, which trigger cell proliferation and immortalization by interacting with tumor suppressor genes, namely p53 and Rb (173-175). Despite this proven oncogenic potential of BKV, its association with human cancer types remains inconclusive. A number of studies have detected BKV in different tumors, including brain, bladder, kidney, prostate and cervix cancers, while others did not (176-181). In GBM, studies have showed contradictory results (Table IV). The highest prevalence has been reported by De Mattei et al (182), who found that 88% of GBM specimens were positive for BKV Tag DNA sequences. In addition, a study conducted by Corallini et al (183) demonstrated the presence of BKV DNA sequence in 50% of GBM cases using southern blotting hybridization. However, contrasting findings have emerged from other studies, which detected the virus in GBM at low frequencies. For example, Delbue et al (156) identified BKV Tag DNA sequence in 9.5% of GBM samples using nested PCR. In a different study including three GBM subtypes, only one primary GBM specimen (4%) was positive for BKV, while all secondary and giant cell GBM samples were negative (165). In line with these low detection rates, Rollison et al (153) investigated the presence of BKV in a large series of patients by performing two independent protocols in separate laboratories. Only three GBM samples (2.9%) exhibited positivity for BKV in one laboratory, while no positive cases were detected in the other laboratory. By contrast, Grossi et al (180) indicated the absence of BKV DNA sequences in all human tumors tested, including GBM, by applying DNA-DNA reassociation kinetics and blotting hybridization (180). Consistent with this finding, Limam et al (152) recently reported the absence of BKV DNA in all GBM tested by PCR assay.

Table IV Prevalence of BKV in GBM tissue samples.

Table IV

Prevalence of BKV in GBM tissue samples.

First author, year	Sample size	Detection method	Results n (%)	(Refs.)
Limam et al, 2020	82	Conventional PCR (Tag)	0 (0)	(152)
Rollison et al, 2005	102	PCR followed by southern blotting hybridization (Tag)	3 (2.9)	(153)
		RT-PCR (Tag)	0 (0.0)
Delbue et al, 2005	21	Nested PCR (Tag)	2 (9.5)	(156)
Huang et al, 1999	28 (Primary GBM)	PCR followed by southern blotting hybridization (Tag)	1 (4.0)	(165)
	22 (Secondary GBM)		0 (0.0)
	18 (Giant cell GBM)		0 (0.0)
De Mattei et al, 1995	17	PCR followed by southern blotting hybridization (Tag)	15 (88.0)	(182)
Corallini et al, 1987	18	Southern blotting hybridization (whole viral genome)	9 (50.0)	(183)
Dörries et al, 1987	5	RFLP followed by southern blotting and hybridization (whole viral genome)	1 (20.0)	(185)

[i] GBM, glioblastoma; RT-PCR, reverse transcription PCR; Tag, large T antigen; BKV, BK virus; RFLP, restriction fragment length polymorphism.

To address the discrepancy in previous findings, two serological studies compared the prevalence of BKV antibodies in patients with glioma and healthy controls to assess the potential association between BKV infection and glioma risk. In 2003, Rollison et al (184) conducted a nested case-control study to evaluate the potential association between BKV infection and the risk of primary brain tumors. The study analyzed sera collected years before diagnosis from brain tumor patients and matched controls, assessing the presence of BKV-specific antibodies. The findings revealed a non-marked inverse association between BKV seropositivity and the risk of developing brain tumors, including GBM. In a larger recent case-control study conducted within two prospective cohorts, the research group measured BKV antibody levels in adult patients with glioma and cancer-free controls. The seroprevalence analysis demonstrated no notable association between BKV infection and glioma risk in both cohorts. These results remained consistent even when the analysis was restricted to GBM cases (154).

4. HPV

Papillomaviruses are a large group of DNA viruses that infect the epithelial cells of both animals and humans. To date, >300 genotypes of papillomaviruses have been identified and fully sequenced, including >200 exclusively infecting humans (186). Among them, HPV is responsible for ~4.5% of human malignancies and is the primary cause of 95% of cervical cancer cases (187,188). Notably, not all HPV genotypes are oncogenic. According to the IARC classification, only 12 types are considered carcinogenic: HPV-16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58 and 59(189). The early region of the HPV genome encodes two major oncoproteins, E6 and E7, serving a notable role in cellular transformation by binding to tumor suppressor proteins p53 and pRb, respectively, which results in their degradation and promotes DNA damage and uncontrolled cell cycle progression (190).

Over the last decade, emerging evidence has reported the presence of HPV DNA sequence in GBM tissue, suggesting a positive association between GBM and HPV infection (Table V). In 2013, Vidone et al (191) analyzed tumor samples from 52 patients with GBM using nested-PCR targeting the major capsid protein L1 gene. Positive genes were subsequently genotyped through PCR product sequencing. HPV DNA was detected in 12 cases (23%), including nine cases positive for the low-risk HPV-6 genome and three cases harboring the high-risk HPV-16 genome. This research group also assessed the association between HPV and the prognosis of GBM, finding that the prognosis was markedly worse in HPV-positive cases aged 50 or older compared with their HPV-negative counterparts (191). Similarly, a study conducted with Japanese patients detected the presence of HPV in 21% of GBM, with HPV-16 being the most prevalent genotype, followed by HPV-18(17). More recently, Adnan Ali et al (192) also detected HPV positivity in 28% of GBM from Pakistani patients, among which 20% had HPV-18,16% HPV-16 and 11 cases showed coinfection with both genotypes. In addition, this study observed that HPV-positive patients exhibited longer survival compared with non-infected cases, but the difference was not statistically significant. The highest prevalence of HPV in GBM was reported by Limam et al (193), who detected HPV DNA in 53.7% (44 out of 82) of Tunisian patients. Genotyping showed that 50% of positive samples carried high-risk genotypes and five samples showed simultaneous infection with HPV18 and HPV16. In a recent study, Arsene et al (194) found that HPV was markedly more prevalent in GBM compared with normal brain tissues, determining the results of previous studies (17,191-193), suggesting a potential association between HPV infection and gliomagenesis.

Table V Studies detecting HPV in GBM tissue specimens.

Table V

Studies detecting HPV in GBM tissue specimens.

First author, year	Detection method (targeted sequence)	Prevalence, % (positive cases/ all patients with GBM)	Genotyping method	Genotyping results	(Refs.)
Hashida et al, 2015	RT-qPCR, nested-PCR (E6 of HPV-16 and L1 region of HPV-18) and IHC (L1 region)	20.5 (8/39)	RT-qPCR Nested-PCR (E6 of HPV-16 and L1 region of HPV-18)	HPV-16 (n=6) HPV-18 (n=2)	(17)
Vidone et al, 2014	Nested-PCR (L1 region)	23.1 (12/52)	PCR and Sanger Sequencing)	HPV-6 (n=9) HPV-16 (n=3)	(191)
Adnan Ali et al, 2019	Conventional PCR (L1 region)	28.0 (31/112)	PCR and sanger sequencing	HPV-18 (n=22) HPV-16 (n=18) Other types (n=2)	(192)
Limam et al, 2020	Conventional PCR (L1 region) and DNA ISH	53.7 (44/82)	Conventional PCR (E7 of HPV-16 and HPV-18)	HPV-16 (n=14) HPV-18 (n=13)	(193)
Arsene et al, 2022	Conventional PCR (L1 region)	29.5 (13/44)	N/A	N/A	(194)

[i] HPV, human papillomavirus; ISH, in situ hybridisation; RT-qPCR, reverse transcription-quantitative PCR; GBM, glioblastoma.

Conversely, a bioinformatics analysis conducted by Strong et al (103) utilizing publicly available RNA-seq and whole genome sequencing datasets, detected the presence of HPV in four cases of low-grade glioma, but none in GBM (103). Similarly, an additional study analyzed RNA-seq datasets to screen for viral sequences across numerous tumor types and failed to detect any HPV transcripts in GBM (104). These findings suggest that, unlike certain other malignancies, HPV is unlikely to play a role in the pathogenesis of GBM.

5. Human endogenous retroviruses (HERVs)

Differing from the aforementioned viruses, HERVs are proviruses that represent ~8% of the human genome (195). They are the remnants of ancient germinal retroviral infections that occurred millions of years ago and have been transmitted vertically through successive generations (196,197). HERVs are considered defective viruses that have lost the ability to replicate and produce infectious particles. A typical proviral sequence comprises viral gag (group-specific antigen), pol (polymerase) and env (envelope) genes framed by long terminal repeats (LTRs). Despite this, some HERVs lack the env gene, while others have lost the entire coding sequence during homologous recombination events (196). The genetic expression of HERVs is suppressed by a number of epigenetic mechanisms in normal adult cells, such as chromatin remodeling and DNA methylation (198,199). Yet, in pathological conditions including cancer, the epigenetic regulators are altered, leading to upregulation of HERVs-encoded transcripts and retroviral proteins. Furthermore, growing evidence has suggested that exogenous viral infections, such as HCMV and EBV, are also implicated in HERVs transactivation (200). Initially dismissed as ‘junk DNA’, HERVs were initially thought to serve no functional role in the human genome. Yet, subsequent discoveries have revealed that certain HERVs have been co-opted by the host for beneficial purposes. A notable example is the HERV-W env gene, which encodes the protein Syn-1, a key component in placental development (201).

HERV-K is a β-retrovirus-like group that includes the youngest and most active HERV subgroup, human MMTV-like-2 (HML-2) (202). Since HML-2 is the most recent integrated HERV, its genomic sequence remained nearly intact, capable of expressing functional viral proteins and producing retrovirus-like particles (203). These gene products were shown to participate in oncogenesis in different ways. In in vitro experiments, env proteins induced cell-cell fusion and interacted with numerous signaling pathways implicated in carcinogenesis, such as c-Myc and Notch pathways. Furthermore, due to non-allelic recombination, HERV-K LTRs can serve as enhancers or alternative promoters for oncogenes (204). Whereas Kessler et al (205) suggested that HERV-K does not serve a marked role in gliomagenesis, a recent study discovered that HML-2 transcripts are markedly upregulated in GBM specimens compared with matched normal brain tissue and that high expression of HML-2 full-length retroviral sequence is associated with poor clinical outcome (205,206). Using single-cell RNA-seq, the same research group analyzed cellular subpopulations of GBM and found that HML-2 transcripts were primarily upregulated in neural progenitor cells that trigger cellular plasticity. To further investigate the association between HML-2 expression and stemness in glioma, the depletion of HML-2 utilizing CRISPR interference was conducted in order to assess its functional impact on tumor biology. These results exhibited decreased cellular viability and adhesion, along with reduced neurosphere formation, indicating diminished GBM stemness properties. On the other hand, the transfection of normal glial stem cells with the HML-2 consensus sequence revealed an upregulation of the stem cell marker OCT4. Since HML-2 LTR5Hs contains a binding motif for the nuclear transcription factor OCT4, it may be suggested that OCT4 sustains HML-2 expression, which creates a positive feedback loop that locks GBM cells in a stem-like state. As stemness in GBM is responsible for tumor invasiveness, dissemination and resistance to therapy, targeting this hallmark can notably improve patient survival. In this manner, an antiretroviral nucleoside retroviral inhibitor (abacavir) that abrogates HML-2 expression was tested to evaluate its impact on GBM outcome. The data indicated a marked decrease in OCT4 and env transcript, cellular viability and improved survival in GBM mouse models. Based on these findings, the researchers suggested that HML-2 can serve as a therapeutic target in GBM (206).

An additional HERV family member has also been implicated in GBM tumorigenesis. Reiche et al (207) demonstrated that HERV-W, a member of the γ-retrovirus-like supergroup, expresses its env protein in microglia cells within the microenvironment of the glioma, including GBM tissues. To elucidate the effect of the HERV-W env protein on GBM cell line behavior, this study co-cultured microglial cells with GBM cells and subsequently examined changes in gene expression, morphology, migration and proliferation in GBM cells. In the presence of microglia, the stimulation of GBM cells by HERV-W env protein upregulated the expression of a number of genes, including monocyte chemoattractant protein (MCP)-1, MMP9, colony stimulating factor 1, IL6, IL-1β, TNFα, MCP3, inductible nitric oxide synthase and VEGF, involved in tumor growth, cell survival, angiogenesis, extracellular matrix degradation and invasion. Furthermore, enhanced tumor cell proliferation, migration velocity and tumor cell cluster formation by HERV-W env were indicated by this study. However, no marked effect was observed in apoptosis or necrosis (207).

6. Conclusion and future directions

GBM is one of the most aggressive malignancies and its etiology is not fully understood, making it hard to establish predictive and preventive strategies. Over the past decades, numerous studies have suggested that the aforementioned viruses may serve a role in GBM carcinogenesis. Although viral prevalence in GBM varies across studies, HCMV and HERVs were clearly shown to be active participants in the oncogenic process. Current evidence indicates that these viruses act as oncomodulators, driving key GBM hallmarks. Since HCMV may be responsible for HERV reactivation, investigating the crosstalk between these two in GBM could reveal a multi-modal model of gliomagenesis and identify novel therapeutic targets. Furthermore, the viral prevalence discrepancy in GBM needs to be addressed by furthering our understanding of the technical factors involved, in order to establish standardized detection protocols. Finally, current evidence is not sufficient to classify HCMV as an oncovirus; further studies are warranted to conclusively determine its role in GBM initiation.

Acknowledgements

Not applicable.

Funding

Funding: No funding was received.

Availability of data and materials

Not applicable.

Authors' contributions

FG conceptualized the present study and provided overall supervision. IS and NI developed the methodology. IS and PM wrote the original draft. FG, IC and RA reviewed and edited the manuscript. IS created the figure. All authors read and approved the final version of the manuscript. Data authentication is not applicable.

Ethics approval and consent to participate

Not applicable.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References