The first discovery of genetic alterations in meningiomas was in the tumor suppressor gene NF2, located on chromosome 22 [17]. Loss of this chromosome is identified in most sporadic meningiomas (40–80%), where almost 60% of these tumors harbor the NF2 gene deficiency, caused by somatic mutations, allelic loss of chromosome 22, aberrant NF2 promoter methylation, or epigenetic NF2 inactivation [18]. The loss of the first allele of chromosome 22 induces the loss of the second, leading to the deletion of the entire chromosome [18]. Although most meningiomas are single and sporadic [19], mutations in the neurofibromin 2 gene can give rise to the familial syndrome neurofibromatosis 2. The hallmark of this syndrome is the growth of benign tumors of the nervous system, namely, schwannomas and meningiomas [20]. NF2 gene produces the merlin protein, also called Schwannomin, a member of the 4.1 family of proteins, which act as a bridge between membrane proteins and the cytoskeleton, cross-linking them. Loss of the merlin protein is characteristic of all NF2 associated meningiomas and nearly half of sporadic cases [21]. Mutations in the NF2 associated meningiomas often yield a truncated, non-functional merlin protein, disturbing the normal cellular growth and intracellular trafficking by disrupting adherent junctions [21, 22]. Recently NF2 was found to be a negative regulator of the FoxM1/Wnt pathway [23]. NF2 mutations potentiated by hypermethylation and PRC inhibition are associated with a functional switch toward the FoxM1 transcriptional program [24]. FoxM1 cooperates with $\beta$-catenin to potentiate Wnt signaling and cancer cell growth, leading to aggressive tumor development and poor prognosis [25]. Whereas about half of sporadic meningiomas harbor NF2 alterations, the remaining NF2-non mutated tumors clearly must be driven by other genetic alterations (Table 1).

In meningiomas devoid of NF2 mutation, a notable gene, the TRAF7, was identified [17]. Twenty-five percent of the meningiomas studied revealed TRAF7 mutations [26]. These mutations are also seen in all meningiomas of the secretory subtype [27]. The TRAF7 houses seven WD40 repeats in its carboxyl terminus, with characteristic repeating units of 40–60 variable residues often terminating in a tryptophan and aspartate (W-D) dipeptides [26, 28, 29]. TRAF7 codes for an enzyme known as an E3 ubiquitin ligase, which is involved in multiple signaling pathways including, K3 through its WD40 repeats [17]. Most TRAF7 mutations appear to be linked to the WD40 domains, which participate in the regulation of JNK and p38 MAPK signaling, besides MEKK3, leading eventually to apoptosis [17, 30, 31] (Table 1).

A fraction of NF2 non-mutated meningiomas harbored alterations in AKT1 (~8%), a component of the PI3K-AKT-mTOR pathway [17]. AKT1, a serine/threonine protein kinase, plays a role in activating the PI3K pathway [32]. The AKT1 p.Glu 17 Lys mutation induces constitutive activation of AKT1 protein. Thus, this mutation strongly supports the overactivation of the PI3K signaling pathway in meningioma, which is one of the essential mediators of growth favoring signals [33, 34] (Table 1).

Meningiomas are devoid of AKT1 and NF2 mutations, which constitute between 1% and 5%, harbor mutations in other genes, such as the SMO gene, coding for smoothened homolog protein. This protein belongs to the Hedgehog signaling pathway [16]. The Hedgehog signaling pathway leads to nuclear translocation of protein GLI and activation of target genes associated with cellular growth, proliferation, and angiogenesis [35] (Table 1).

Benign meningiomas (classified as WHO grade I) can potentially harbor recurrent mutations in KLF4. KLF4 is an important transcription factor in oncogenic activation since it intervenes with cell cycle progression following DNA damage. It mediates this process by activating p53-dependent G1/S cell cycle arrest [36]. Mutations in KLF4 have been detected in as many as 50% of NF2-nonmutated meningiomas, but at an interestingly higher rate with TRAF7 mutations, exclusively in secretory meningioma [27] (Table 1).

Around 8% of non-NF2 meningiomas have mutations in epigenetic modifiers such as KDM5C, KDM6A, and SMARCB1 [37]. KDM5C and KDM6A are histone demethylases, while SMARCB1 is a member of the SWI/SNF chromatin-remodeling complex associated with familial forms of meningioma [38, 39]. Several tumors have aberrant histone methylation of the retinoblastoma zinc-finger (RIZ1) methyltransferase gene which is linked to uncontrolled upregulation of the cell cycle [40] since the normal RIZ1 serves as a checkpoint at the G2/M phase of the cell cycle stalling its progression. In meningiomas, the expression of RIZ1 inversely correlates with the tumor grade, where grade I express the highest level of RIZ1 (87.5%), grade II in 38.9%, and grade III in 23.8% of cases [41]. The replacement of RIZ1 expression with a transactivation fusion protein leads to a decline in abnormal cellular growth and proliferation in primary meningiomas and a reduced burden in xenograft mouse models [42]. The research identified a possible role for STMN1 (Stathmin) in the invasion, metastasis, and resistance to drugs of meningioma cells [43]. STMN1 regulates the microtubule system and is involved in specific cell cycle phase checkpoints [44] (Table 1).

The majority of meningioma tumor cells have deletions on Chromosome 22, located at 22q12. This locus is also responsible for making NF2, mentioned above [45, 46]. The second most identified chromosomal abnormality is the deletion of chromosome 1p. FISH studies of 1p may help predict the prognosis of meningioma patients. Genes on 1p are classified as tumor suppressors. Alkaline phosphatase gene (ALPL) is in the deletions’ sites 1p36, 1p35–p32. Its loss was associated with tumor progression [47, 48, 49, 50]. In parallel, loss of heterozygosity of chromosome 14 was identified in 12% to 55% of meningioma cases [51]. Large series of patients identified deletions at the 14q32.2 regions, in which ELK-2, AKT, and TCI-1 are coded [52]. The specific role of these genes is to be studied. Also, deletions on chromosome 10 almost equally occur in meningioma as deletions on 14q [51]. They are attributed to tumor progression rather than formation and are abundantly detected in MII and MIII (40%–50%) than in MI (5%–12%). Loss of heterozygosity at microsatellites predicted higher tumor grades and faster recurrence of the tumor [53].

Spinal meningiomas are the most widely recognized spinal tumors in adults, representing up to 38% of intradural spinal tumors. However, they account for merely 6.5% of total craniospinal tumors in adulthood [54]. Comparable histological subtypes are seen in intracranial and spinal meningiomas, including meningothelial, psammomatous, transitional, atypical, metaplastic, and clear cell types. The former three subtypes are the most widely recognized spinal meningiomas but intriguingly, for unknown reasons, demonstrate a lower hazard for reappearance than their intracranial equivalents [55, 56]. In addition, though cancerous alterations of spinal meningiomas, similar to their intracranial counterparts, are observed, these alterations account for 3% only of reported cases [56]. Additionally, associations were found between several genes and spinal meningiomas: mainly chromosome 22q and its associated gene NF2[57, 58, 59]. Sayagues et al. demonstrated in a microarray-based study of spinal versus intracranial meningiomas, that homozygous deletions on chromosome 22 were more related to spinal rather than intracranial meningiomas [58, 59]. Additionally, spinal meningiomas had a higher likelihood to originate from a single cell clone than from multiple cells [59, 60]. This work also showed a variable expression of 1555 genes in spinal and intracranial meningiomas, among which 35 genes were highly expressed in the spinal counterpart. These genes include Hox genes, the NR4 family of genes, KLF4, FOSL2, and TCF8. Genes involved in intra/extracellular signaling such as RGS16, DUSP5, DUSP1, SOCS3, CMKOR, L6, TGFB1I4, IL1B, CYR61, and CDH2[58, 60] (Fig. 1).

Fig. 1.

Meningioma tumor locations and their subsequent genetic alterations.

Arslantas et al. [57] histological analyses of 16 patients presented cells from tumors with either complete or partial loss of chromosome 22, with damage of 1p, 9p, and 10q and addition of 5p and 17q, in comparison to chromosomes derived from the patients’ white blood cells. Sayagues et al. [58] also demonstrated that the above 35 genes, while differentially expressed in spinal versus intracranial meningiomas, show no significant association with any of the histological subtypes of meningiomas. Thus, these genes do not explain the predominance of psammomatous, meningothelial, and transitional histological subtypes among spinal meningiomas.

In terms of how spinal and intracranial meningiomas relate, two possibilities are known: (1) The two tumors may correspond to the same tumor type, but at different points in the severity progression, or (2) the two tumors are completely genetically different with shared genetic mutations [58]. The first possibility is supported by the prevalence of chromosome 22 abnormalities among spinal and intracranial meningiomas with intracranial tumors having even more complex and heterogenous abnormalities, which could explain why they have a worse clinical outcome [59]. However, this does not explain the predominance of spinal meningioma in females or its preferred histological subtypes. In addition, many other tumor characteristics remained of higher prevalence among intracranial tumors (proliferative rate, chromosomal abnormality other than with chromosome 22, and tumor cell clones’ number). This supports the possibility that spinal and intracranial meningiomas are different genetically and not just in severity [58]. Table 2 summarizes the genes and pathways that are involved in spinal meningiomas.