Background
Neurofibromatosis type 2 (NF2) is an autosomal dominant tumor syndrome characterized by the growth of multiple neoplasms within the central nervous system. Although bilateral vestibular schwannomas are the hallmark of NF2, meningiomas are the second most frequent intracranial tumor, and occur in about 52% of NF2 patients [
1,
2]. Benign meningiomas (WHO grade I) feature a 5-year tumor recurrence rate of 5% as compared to 50–80% for anaplastic meningiomas (grade III), highlighting the importance of elucidating the molecular mechanisms which contribute to tumor progression [
3].
The most common genetic mutation in meningiomas is
NF2 inactivation, which is observed not only in NF2-associated tumors, but also in 47 to 72% of sporadic meningiomas, and is thus considered an integral step for meningioma tumor initiation [
4‐
6]. Recent studies utilizing high throughput whole-exome and whole-genome sequencing have identified two distinct subtypes of sporadic meningiomas: tumors with or without an inactivated
NF2 gene [
7,
8]. Sporadic meningiomas with disrupted
NF2 tend to display greater genomic instability (including several cases of chromothripsis) and higher grades than non-
NF2 meningiomas. Non-
NF2 tumors have been shown to contain recurrent oncogenic mutations in
AKT1,
KLF4,
TRAF7 and
SMO, indicating the alternate involvement of the PI3K-AKT and Hedgehog signaling pathways.
NF2-associated meningiomas are rarer than their sporadic counterparts and far fewer studies have investigated the genetics underlying their initiation and progression. Two case series evaluated meningiomas from NF2 patients only for the allelic imbalances most commonly observed in sporadic meningiomas, and confirmed frequent somatic inactivation of the
NF2 gene, as well as losses of chromosome arms 1p, 6q, 9p, 10q, 14q and 18q [
9,
10]. A more recent study used single nucleotide polymorphism array analysis to report increased chromosomal instability with increasing grade in NF2-associated meningiomas [
11].
Here, we present an in-depth genomic study of grade I and grade II meningiomas that resided in close proximity in the brain of an NF2 patient. The tumors contained the same NF2 germline mutation and similar somatic hits affecting the normal remaining copy of the gene, yet differed drastically in genomic architecture and growth rate. The tumors were investigated using whole-exome sequencing complemented with SKY and SNP-array copy-number analysis.
Discussion
To our knowledge, this is the first whole-exome sequencing study of NF2-associated grade I and grade II meningiomas. Besides chromosome 22 loss, the genome of the grade I meningioma closely resembled that of a normal diploid cell, while the genome of the grade II tumor contained several chromosomal rearrangements previously observed in meningiomas, including losses in 1p, 2p, 2q, 3p, 3q, 6q, 12p, 14q, 18q, Xp, gain in 1q [
21‐
25], and multiple translocations. Our observations confirm previous findings that inactivation of
NF2 is likely to be the primary step in NF2-associated meningioma formation [
26]. In addition, we show that both benign and atypical tumors had a low somatic mutation burden. Although limited to a single patient, this data permits speculation that tumor progression to a higher grade likely occurs through multiple chromosomal gains, losses and translocations and to a lesser extent from the accumulation of point mutations and small indels.
Chromosomal translocations leading to the disruption of tumor suppressors or activation of proto-oncogenes are common in many neoplasms [
27,
28]. Limited evidence suggests that chromosomal translocations may also be present in meningiomas [
29] and systematic studies addressing this mechanism of tumorigenesis in meningiomas are emerging. We observed numerous chromosomal translocations (both balanced and unbalanced) as well as one case of highly irregular, shattered chromosomes. Interestingly, similar to the observation made by Brastianos et al. [
7], close examination of SNP-array plots of chromosome 1 in the tumor revealed deletion of the 5′-half of the
NEGR1 gene (not shown). These findings suggest that structural aberrations might be more frequent than previously believed in
NF2-driven familial and sporadic meningiomas, and could represent one of the mechanisms of genetic instability and routes of tumor progression to higher grades.
By analyzing the genomic architecture and somatic mutations in multiple fragments of the grade II tumor, we gained insight into the clonal evolution of this fast growing neoplasm. We observed not only a remarkably uniform pattern of chromosomal gains and losses, but also the consistent presence of the only two potentially pathogenic mutations, in ADAMTSL3 and CAPN5, in all four fragments. These findings indicate that the aberrations were likely present in the initial cell undergoing fast clonal expansion, and that any of these aberrations/mutations could impact tumor progression and accelerate growth rate.
Germline mutations in
CAPN5 (Calpain 5), which encodes a calcium-dependent endopeptidase, have been associated with neovascular inflammatory vitreoretinopathy [
30]. Though the role of the protein in neoplastic transformation is unclear, a recent study reported association of CAPN5 with promyelocytic leukemia nuclear bodies, which are involved in transcriptional regulation, cell differentiation, apoptosis, and cell senescence [
31]. The protein encoded by
ADAMTSL3 (A Disintegrin And Metalloproteinase with TromboSpondin Like 3) is involved with extracellular matrix function and to cell–matrix interactions, and is frequently mutated and under-expressed in colorectal cancer [
32]. The gene belongs to a large family of proteins associated with microfibrils in the extracellular matrix, thus mediating sequestration of the TGFB superfamily of proteins and affecting wide array of cellular functions such as adhesion, migration, proliferation and angiogenesis [
33,
34].
The majority of meningiomas are benign and asymptomatic tumors that require little or no treatment [
35,
36]. However, a subset of tumors becomes more clinically aggressive as they evolve toward atypical and anaplastic stages, causing increased morbidity and mortality. Remarkably, the tumors we investigated had the same
NF2 germline mutation, the same genetic background, similar chromosome 22 LOH and were residing within a few millimeters from one another in the patient’s brain, yet one remained as a slowly growing asymptomatic grade I meningioma and the other evolved into a fast growing grade II tumor. This observation underscores the importance of stochastic factors in meningioma progression, which are still poorly understood.
Conclusions
We performed an in-depth genomic study of NF2-associated benign and atypical meningiomas. Both tumors had inactivated second copies of NF2 and a low burden of somatic mutations. However, unlike the benign tumor, the atypical meningioma presented with widespread genomic aberrations, implying that chromosomal instability may be a key driving force in tumor progression. In addition, we identified two candidate driver genes, CAPN5 and ADAMTSL3, which could contribute to the elevated growth rate of the grade II meningioma. Future efforts should be focused on understanding the mechanistic links between NF2 deficiency and genomic instability.
Acknowledgements
This study was supported by the Intramural Research Programs of the National Institute of Neurologic Disease and Stroke (NINDS), the Division of Cancer Epidemiology and Genetics of the National Cancer Institute (NCI), and the National Human Genome Research Institute (NHGRI).