Genetic Events and Signaling Mechanisms Underlying Schwann Cell Fate in Development and Cancer

Abstract

In this review, we describe Schwann cell development from embryonic neural crest cells to terminally differentiated myelinated and nonmyelinated mature Schwann cells. We focus on the genetic drivers and signaling mechanisms mediating decisions to proliferate versus differentiate during Schwann cell development, highlighting pathways that overlap with Schwann cell development and are dysregulated in tumorigenesis. We conclude by considering how our knowledge of the events underlying Schwann cell development and mouse models of schwannoma, neurofibroma, and malignant peripheral nerve sheath tumor can inform novel therapeutic strategies for patients with cancers derived from Schwann cell lineages.

ABBREVIATIONS

GFAP

glial fibrillary acidic protein

HDAC

histone deacetylase complexe

iSC

immature Schwann cell

MAPK

mitogen activation protein kinase

MPNST

malignant peripheral nerve sheath tumor

mpz

myelin protein zero

MRI

magnetic resonance imaging

NF1

neurofibromatosis type 1

NF2

neurofibromatosis type 2

PI3K

phosphoinositide 3-kinase

PNS

peripheral nervous system

PRC2

polycomb repressive complex 2

Ras-GAP

ras GTPase-activating protein

RNA

ribonucleic acid

SCP

Schwann cell precursor

VEGF

vascular endothelial growth factor

The embryonic neural crest is an evolutionarily conserved multipotent cell population that gives rise to a diversity of cell types, including numerous components of the peripheral nervous system (PNS), such as Schwann cells. Anatomically, the neural crest is divided into cranial, vagal, trunk, and sacral components along the anterior-posterior axis, and each segment is regulated via a complex gene regulatory network that dictates neural crest specification, migration, proliferation, and differentiation.¹ The Schwann cell lineage is no different, as these various cellular outcomes must be balanced throughout the developmental process (Figure 1). Following delamination from the dorsal neural tube, trunk neural crest cells migrate along ventral and dorsolateral pathways, and classic lineage tracing experiments demonstrate that the Schwann cell lineage primarily arises from ventrally migrating trunk neural crest cells.² This migratory process is highly coordinated across both cell autonomous transcriptional programs and noncell autonomous signaling events.³ Consistent with the diverse migratory destinations of Schwann cell progenitors, Schwann cell-derived tumors present anatomically throughout the body.

FIGURE 1.

Overview of normal Schwann cell development and transformation to peripheral nerve tumors.

FROM NEURAL CREST TO SCHWANN CELL PRECURSOR: TAKING THE FIRST STEPS TOWARD A GLIAL FATE REQUIRES SOX10 INDUCTION AND NRG1 RESPONSIVENESS

The subsequent transition between migratory trunk neural crest and Schwann cell precursors (SCPs) (Figure 1), the first intermediate population in fate commitment down the Schwann cell lineage, does not appear to constitute a specific spatiotemporal stage but more likely exists within a differentiation continuum. Nonetheless, efforts to identify the key transcriptional programs and noncell autonomous signals mediating the balance between migration, proliferation, and differentiation have revealed a number of factors important for cellular decision-making. Among these, the transcription factor SOX10 is necessary for Schwann cell and melanocytic differentiation^4,5 and serves as a master regulator of neural crest multipotency by repressing neuronal lineages.⁶

In addition, Sox10 cooperates with histone deacetylase complexes (HDACs) to activate Pax3, leading to expression of the Schwann marker myelin protein zero (mpz) and P0 protein, consistent with commitment to a glial fate.^7,8 With regard to noncell autonomous signals, SCPs are dependent on NRG1 (glial growth factor)^9,10 and its partner receptor tyrosine kinases ErbB2 and ErbB3¹¹ for cell proliferation and Schwann commitment, as well as Notch signaling to promote glial differentiation.¹² With regard to crosstalk between cell autonomous and noncell autonomous determinants of cell identity, Sox10 further regulates expression of ErbB3 in Schwann cells in order to maintain responsiveness to NRG1.⁴ Despite the aforementioned mechanisms biasing SCPs toward a glial fate, SCPs remain multipotent¹³ and can give rise to numerous other neural crest derivatives, including melanocytes,¹⁴ neurons,¹⁵ neuroendocrine cells in the adrenal medulla,¹⁶ and dental mesenchymal stem cells.¹⁷ Thus, SCPs comprise an intermediate progenitor in the neural crest hierarchy exhibiting early commitment to the glial lineage, yet retaining broad developmental potential (Figure 1).

SCHWANN CELL MATURATION REFLECTS ACTIVATION OF CRITICAL TRANSCRIPTIONAL NETWORKS TO NARROW DEVELOPMENTAL POTENTIAL AND DRIVE TERMINAL DIFFERENTIATION TO IMMATURE AND MATURE SCHWANN CELLS

As the embryo develops, SCPs reach their migratory destination and begin to organize around developing axons, continuing to survive in an axon-dependent manner due to neuronal secretion of NRG1.¹⁸ Intercalated with developing axons, SCPs begin the transition to immature Schwann cells (iSCs), the direct progenitor population for mature Schwann cells (Figure 1).^19,20 A number of shared genes and factors play an important role in both SCP commitment and iSC transition, confounding the study of these 2 processes in isolation. Indeed, Sox10 modulates a positive feedforward loop to promote iSC differentiation by directly activating Oct6 expression through binding to a Schwann cell-specific enhancer, which leads to Oct6-mediated recruitment of HDAC genes.^21,22 In addition, Sox10 and Krox20 cooperatively drive expression of key myelin genes, such as myelin basic protein and mpz.^23,24 From a noncell autonomous perspective, NRG1 is sufficient to drive iSC formation from SCPs,¹¹ and Notch signaling via its transcriptional activator RBP-J is critical for mediating the transition from SCPs to iSCs. In contrast, endothelin signaling²⁵ and the transcription factors Tfap2a²⁶ and Zeb2^27,28 delay progression to iSCs and thus maintain the developmental potential of SCPs. Functionally, iSCs exhibit a significantly narrowed developmental potential compared to SCPs, as iSCs primarily give rise to myelinated and nonmyelinated Schwann cells. Accordingly, iSCs induce key Schwann commitment markers such as S100 and glial fibrillary acidic protein (GFAP),²⁹ gain the ability to survive without mitogenic input from nearby axons, and develop a basal lamina associated with loss N-cadherin.³⁰ The transition to mature Schwann cells is ultimately driven by interactions with axons through radial sorting and paracrine signaling that serves to compartmentalize myelinating Schwann cells with large caliber axons, and nonmyelinating Schwann cells with small caliber axons (Figure 1).²⁰ Radial sorting coincides with exit from the cell cycle and terminal differentiation into mature Schwann cells, completing the developmental trajectory of the Schwann cell lineage, and the decision to form a myelinating or nonmyelinating Schwann cell is primarily driven by axonal secretion of NRG1 in noncell autonomous manner.^31,32 Accordingly, disruption of radial sorting impairs terminal Schwann differentiation, and numerous signaling pathways have been implicated in this process, including NRG1/ErbB signaling, mitogen activation protein kinase (MAPK), and phosphoinositide 3-kinase (PI3K) activation, Notch signaling, and transcription factor activity, including Sox10, Oct6, Egr2, and NF-kB.^33,34 In sum, the Schwann cell lineage arises through an ordered progression from neural crest to differentiated Schwann cell requiring integration of both cell autonomous and noncell autonomous signaling mechanisms.

SCHWANN CELL-DERIVED TUMORS SHARE MOLECULAR SIGNATURES AND SIGNALING DEPENDENCIES WITH KEY DEVELOPMENTAL STAGES IN SCHWANN CELL DEVELOPMENT

Given the broad developmental potential of Schwann cells, it is perhaps unsurprising that many tumors are derived from this lineage. Canonically, schwannomas, neurofibromas, and malignant peripheral nerve sheath tumors (MPNSTs) arise from Schwann cells (Figure 1), and recent evidence suggests certain melanomas¹⁴ and sympathoadrenal tumors such as pheochromocytoma, neuroblastoma, and paraganglioma¹⁶ can also arise from the Schwann cell lineage.

SCHWANNOMAS ARE NF2-ASSOCIATED TUMORS MARKED BY RETENTION OF SCHWANN CELL LINEAGE MARKERS AND DYSREGULATED MAP KINASE SIGNALING WITH MINIMAL TUMOR HETEROGENEITY

Schwannomas are the most common cancer of PNS.³⁵ Patients with schwannomas typically present with neurological symptoms secondary to impingement of nearby neurological structures, and workup generally comprises neurological examination and radiological evaluation with magnetic resonance imaging (MRI) (Figure 2), which may reveal the classic finding of bilateral vestibular schwannomas in patients with neurofibromatosis type 2 (NF2) (Figure 2A-2A′). Molecularly, these tumors are primarily associated with loss of the tumor suppressor NF2 and rare gene fusions driving MAPK activation,^36,37 consistent with the role of receptor tyrosine kinase signaling in Schwann cell development (Table 1). Of note, schwannomatosis, defined as the presence of multiple schwannomas in a single patient, appears to involve a multi-hit genetic mechanism, including SMARCB1 and LZTR1 loss, in addition to NF2 loss.³⁸ Microscopically, schwannomas are almost entirely composed of homogeneous neoplastic Schwann cells with spindle-shaped nuclei demonstrating classic biphasic architecture with nuclear palisading (Figure 3). Histologic variants include cellular schwannomas (monophasic with compact spindle cells), melanotic schwannomas (expressing melanin pigment), epithelioid schwannomas (rounded cells arranged in single cells and clusters), or plexiform schwannomas (occurring superficially with a nodular growth pattern).³⁹ By immunohistochemistry, schwannomas typically express S100⁴⁰ and SOX10⁴¹ proteins (Figure 3A-A′) consistent with a homogenously transformed population of Schwann origin, and a subset exhibits GFAP positivity,⁴² mimicking the marker expression patterns observed during normal Schwann cell differentiation. In mice, NF2 misactivation results in a greater range of phenotypes with unclear overlap to human NF2 patients, the latter of which predominantly develop nervous system tumors such as schwannomas, meningiomas, ependymomas, and, rarely, astrocytomas (Table 2).⁴³ NF2 null mice do not complete gastrulation,⁴⁴ and in contrast to human patients, NF2 heterozygotes develop osteosarcomas, fibrosarcomas, and hepatocellular carcinomas that demonstrate high metastatic proclivity and cooperativity with Tp53 loss.⁴⁵ In contrast, conditional genetics has led to the development of mouse models with greater similarity to human NF2 loss. Introduction of pathogenic NF2 variants specifically in the myelinated Schwann cell lineage leads to schwannoma formation,⁴⁶ and conditional NF2 deletion with P0Cre, an SCP marker that directs target gene expression in myelinating Schwann cells, results in the development of schwannomas in mice.⁴⁷ More recently, conditional NF2 mutant mice generated with a Periostin-Cre were shown to develop vestibular schwannomas and hearing impairment similar to human NF2 patients with bilateral schwannomas. With regard to clinical management, schwannomas demonstrate excellent tumor control following surgery or radiation therapy,^48,49 although morbidity from these treatments remains a significant concern.^50-56 In sum, schwannomas are NF2-associated lesions composed of a homogeneous S100- and SOX10-positive Schwann cell population that are managed primarily with local treatment and have limited metastatic potential.

FIGURE 2.

MRI appearance of Schwann cell-derived peripheral nerve tumors. A, Enhancing left cerebellopontine angle mass extending into the internal auditory canal along the course of the vestibulocochlear nerve (cranial nerve VIII) with resultant mass effect upon the adjacent pons and cerebellum consistent with a vestibular schwannoma. A′, Bilateral vestibular schwannomas are pathognomonic of NF2. B, T2 hyperintense ovoid lesion extending along the course of the right C7 nerve root consistent with a deep intraneural neurofibroma. B′, Left cervical mass with intense enhancement and multicompartmental extension from the posterior neck and into the prevertebral soft tissues consistent with a plexiform neurofibroma in the setting of NF1. C, Rapid interval growth (upper image at baseline, lower image 12 mo later) of an intensely enhancing nodule within a left cervical plexiform neurofibroma in a patient with NF1. Plexiform neurofibromas are at risk of malignant transformation to MPNSTs, particularly in the setting of rapid growth. C′, Large heterogeneous, enhancing, T2 hyperintense right thigh mass with internal degeneration intimately associated with the sciatic nerve. Size and irregular margins along the posterior thigh muscles are consistent with an MPNST.

TABLE 1.

Key Studies Regarding Genomic Investigation of Human Schwannoma, Neurofibromas, and MPNSTs

Study	Method	Key finding
Schwannoma
Agnihotri et al36  Nature Genetics, 2016	Whole exome (n = 26)	The majority of tumors (n = 20) harbored NF2 mutations with additional low-frequency variants in many other genes
	Targeted sequencing (n = 99)	Exon sequencing of mutations identified from Whole Exome Sequencing (WES) showed recurrent NF2 mutation (n = 76)
	Methylation array (n = 125)	Two epigenomic groups were identified corresponding to vestibular versus spinal schwannomas
	Ribonucleic acid sequencing (n = 41)	Rare SH3PXD2A-HTRA1 fusion (n = 5) observed with tumorigenic and promitogenic functions
Håvik et al37  Journal of Neurosurgery, 2018	Whole exome (n = 46)	The majority of tumors (n = 35) harbored NF2 mutations while the remaining exhibited alterations in genes linked to NF2
Neurofibroma and MPNST
Lee et al79  Nature Genetics, 2014	MPNST whole exome (n = 15)	The majority of tumors harbored recurrent NF1, EED, SUZ12, CDKN2A, and TP53 mutations with only n = 1 sample showing no alterations in any of these genes
	MPNST ribonucleic acid sequencing (n = 15)	Tumors segregated on PCA by PRC2 mutational status with enrichment of a homeobox gene expression signature in PRC2 mutant MPNSTs
	MPNST targeted sequencing (n = 37)	Consistent with WES data, the majority of tumors harbored recurrent NF1, EED, SUZ12, CDKN2A, and TP53 mutations with only n = 2 samples showing no alterations in any of these genes
Zhang et al80  Nature Genetics, 2014	MPNST whole genome/whole exome (n = 8)	The majority of tumors demonstrate inactivating NF1 or SUZ12 mutations (5/8) as well as EED (n = 1) and EPC1 (n = 1) mutations
	MPNST targeted sequencing (n = 42)	Redemontrated recurrent mutations in SUZ12 (n = 11) while further confirming these mutations result in abrogation of SUZ12 expression
	Neurofibroma targeted sequencing (n = 11)	No inactivating SUZ12 mutations were identified in neurofibromas
Nielsen et al81  Am Journal of Pathology, 1999	MPNST IHC and CDKN2A sequencing (n = 11)	Recurrent CDKN2A loss observed in the majority of MPNSTs (n = 10)
	Neurofibroma IHC and CDKN2A sequencing (n = 7)	Intact CDKN2A observed in all non-MPNST lesions

FIGURE 3.

Histopathologic appearance of Schwann cell-derived peripheral nerve tumors. Representative hematoxylin and eosin (H/E) staining A-C as well as S100 A′, B′, C′ and SOX10 A′, B′, C′ immunohistochemistry for A, schwannoma, B, neurofibroma, and C, MPNST.

TABLE 2.

Foundational Genetic Analysis of Mouse Mutants Harboring Mutations in Recurrently Mutated Genes Across Human Schwann Cell-derived Tumors

Genotype	Study	Key finding
NF2
Nf2−/−	McClatchey et al44  Genes & Development, 1997	Homozygous Nf2 mutants fail to complete gastrulation due to an extraembryonic defect
Nf2+/−; Tp53+/−	McClatchey et al45  Genes & Development, 1998	Heterozygous Nf2 mutants develop many distinct tumors (osteosarcoma, lymphoma, lung adenocarcinoma, hepatocellular carcinoma, and fibrosarcoma) with a high metastasis rate in cooperation with P53 loss
P0Cre; Nf2fl/fl	Giovannini et al46  Genes & Development, 2000	Homozygous Nf2 conditional mutations under control of the Schwann cell specific P0Cre leads to numerous human NF2 sequelae, including schwannoma formation
NF1
Nf1−/−	Jacks et al65  Nature Genetics, 1994	Homozygous Nf1 mutants display embryonic lethality at mid gestation while heterozygotes demonstrate increased propensity for pheochromocytoma and myeloid leukemia development
Nf1−/−	Brannan et al66  Genes & Development, 1994	Homozygous Nf1 mutants display embryonic lethality at E14.5 due to cardiac defects with sympathetic ganglia hyperplasia
Nf1+/−; Tp53+/−	Vogel et al82  Science, 1999	Double heterozygous Nf1 and Tp53 mice demonstrate fully penetrant soft tissue sarcoma formation that exhibit loss of heterozygosity and express neural crest markers
Nf1+/−; Tp53+/−	Cichowski et al83  Science, 1999	Compound loss of both Nf1 and Tp53 lead to MPNST development while homozygous Nf1 loss alone primarily results in neurofibroma formation
DhhCre; Nf1fl/fl	Wu et al71  Cancer Cell, 2007	Homozygous Nf1 conditional mutants under control of the Schwann lineage specific DhhCre driving deletion at E12.5 leads to both plexiform and dermal neurofibroma formation, suggesting spatiotemporal pattern of Nf1 loss during development is critical
Nf1+/−; Tp53+/−; Suz12+/−	De Raedt et al92  Nature, 2014	Suz12 loss in addition to Nf1 and Tp53 loss results in increased tumor formation across multiple sites due to potentiation of Ras-driven gene expression by Suz12 loss
Lats1−/−; Lats2−/−	Wu et al93  Cancer Cell, 2018	Hippo signaling pathway activation by homozygous loss of Lats1 and Lats2 lead to MPNST formation, which is abrogated by concurrent inactivation of Hippo pathway activators Taz and Yap

NEUROFIBROMAS ARE NF1-ASSOCIATED TUMORS COMPRISED OF NUMEROUS CELL TYPES YET EXHIBIT A DUAL SCHWANN CELL ORIGIN

While the neoplastic population in schwannomas appears homogeneous, neurofibromas exhibit significant cellular heterogeneity.⁵⁷ In addition to neoplastic Schwann cells, neurofibromas are comprised of non-neoplastic Schwann cells, fibroblasts, perineurial cells, axons, and immune cells such as mast cells, macrophages, T cells, and other antigen-presenting cells (Figure 3B) while maintain expression of Schwann markers such as S100 and SOX10 (Figure 3B′-3B′).^58-60 In mice, neurofibroma tumor heterogeneity is critical for oncogenesis, as conditional mutagenesis demonstrates that loss of neurofibromatosis type 1 (NF1) in the Schwann cell lineage is sufficient for neurofibroma formation, and haploinsufficiency of Nf1 in non-Schwann lineages further enhances tumorigenesis.⁶¹ Loss of NF1,⁶² a ras GTPase-activating protein (Ras-GAP) that functions as a negative regulator of key intracellular signaling pathways such as Ras/Raf/MEK/ERK and PI3K,⁶³ is the critical genetic event as often observed in patients with syndromic NF1.⁶⁴ Developmentally, mice lacking Nf1 demonstrate numerous deficits in neural crest-derived tissues and increased propensity for tumor formation.^65,66 Clinically, it is further important to distinguish superficial dermal neurofibromas (small nodular tumors of the skin and subcutaneous tissue arising from small superficial nerves that present with pain and bleeding but not neurological symptoms) from deep intraneural neurofibromas (nerve root-associated lesions causing symptoms from mass effect, including radicular pain or local neurological deficits) (Figure 2B), with plexiform neurofibromas (Figure 2B′) constituting a distinct entity characterized by multinodular continuous masses affecting large nerves and generally considered to be pathognomonic for syndromic NF1. Consistent with this classification, dermal and plexiform neurofibromas exhibit distinct methylation profiles,⁶⁷ and these subgroups further arise from distinct embryologic populations with dermal neurofibromas originating in dermal skin-derived precursors,⁶⁸ while plexiform neurofibromas derive from SCPs and iSCs within a narrow developmental window.⁶⁹ This dual origin of dermal and plexiform neurofibromas may explain, in part, the divergent clinical behavior of these tumors despite their similar appearance on histopathology. Curiously, dermal neurofibromas arising from skin-derived precursors appear to exhibit hormonal dependence on estrogen and progesterone, and indeed, recent work suggests NF1 functions as an estrogen receptor corepressor.⁷⁰ More generally, conditional mutagenesis in mice has shown that Nf1 loss in the embryonic glial progenitor population leads to both dermal and plexiform neurofibroma population, consistent with the notion that adult Schwann derivatives in the skin and nerves share a common developmental origin.⁷¹ The clinical management of neurofibromas is primarily to improve patient morbidity, manage patient-reported symptoms, and obviate local complications due to mass effect, as the risk of metastatic spread is very low. Thus, neurofibromas constitute a second group of benign Schwann cell-derived lesions arising that share key genetic and signaling mechanisms across development and cancer.

MPNSTS DEMONSTRATE LOSS OF SCHWANN LINEAGE MARKER EXPRESSION, SEQUENTIAL GENETIC PERTURBATIONS, AND INCREASED INFLAMMATORY INFILTRATION

Although schwannomas and neurofibromas are generally well-managed with surveillance and local therapy, these lesions are at risk for transformation to MPNSTs from neurofibromas in NF1 patients or, rarely, from schwannomas in schwannomatosis or NF2 patients.⁷² Accordingly, lesions that demonstrate adverse clinical or radiologic features (Figure 2C-2C′) concerning for progression warrant a more aggressive treatment paradigm.⁷³ In contrast to neurofibromas and schwannomas, patients with MPNSTs have high risk of metastatic progression and poor overall outcomes,⁷⁴ and multimodal therapy consisting of surgery, radiation, and systemic therapy is generally the standard of care, although definitive efficacy trials are lacking. Histologically, MPNSTs present as spindle cell neoplasms with high mitotic rate and gross necrosis (Figure 3C) that can be difficult to distinguish from other aggressive sarcomatous malignancies.⁷⁵ While immunohistochemistry may be positive for markers such as S100⁷⁶ or SOX10,⁷⁷ immunoreactivity for these neural crest lineage markers is often lost in MPNSTs (Figure 3C′-C′), adding to the diagnostic challenge these tumors pose. With regard to the molecular mechanisms underlying MPNST tumorigenesis,⁷⁸ precursor lesions require multiple hits in addition to NF1 loss in order to transform into MPNSTs. In that regard, deoxyribonucleic acid sequencing of MPNSTs in human patients has identified recurrent mutations in SUZ12 and EED, which are components of the polycomb repressive complex 2 (PRC2) family of epigenetic regulators, as well as tumor suppressor loss such as TP53 mutation and CDKN2A deletion.^79-81 Mouse experiments have further demonstrated the importance of combinatorial loss of NF1 and Tp53,^82,83 as well as additional Ink4a/Arf loss, in MPNST pathogenesis.⁸⁴ Moreover, zebrafish lacking Tp53 also develop MPNSTs,⁸⁵ reflecting the conserved important of tumor suppressor loss in MPNST pathogenesis. Consistent with the role of NF1 in this process, receptor tyrosine kinase pathways^86,87 and downstream regulators such as Raf/MEK/ERK⁸⁸ and PI3K signaling⁸⁹ are implicated in MPNST formation. Indeed, Raf and MEK inhibitors have shown efficacy in Vitro and in mouse models of MPNST,^90,91 although clinical trials in humans based on dysregulated pathways in mice are inconclusive, as targeted inhibitors of receptor tyrosine kinase pathways, including Raf, MEK, PI3K (and its mammalian target of rapamycin), epidermal growth factor, and vascular endothelial growth factor (VEGF), show unclear benefit in MPNST.⁷³ Intriguingly, the role of NF1 as a Ras-GAP regulating Ras/Raf/MEK/ERK and observed mutations in PRC2 components such as SUZ12 may converge to drive Ras-mediated transcription and render such tumors sensitive to bromodomain inhibitors.⁹² In addition to receptor tyrosine kinase signaling and PRC2 components, Hippo signaling also appears to be important for MPNST progression via crosstalk with platelet-derived growth factor signaling,⁹³ which may again reflect convergence on Ras-mediated signal transduction mechanisms. Finally, although neoplastic Schwann cells are the cell of origin for MPNSTs, NF1 loss in the Schwann population alone is necessary but not sufficient for transformation to MPNST.⁶¹ Thus, noncell autonomous mechanisms are critical for Schwann cell tumor transformation from a clinically indolent to aggressive neoplasm. Indeed, NF1 deficient Schwann cells mediate a robust inflammatory response through numerous mechanisms, including recruitment of mast cells via the receptor tyrosine kinase Kit,⁹⁴ increased NF1 heterozygous mast cell proliferation through the Rho-GTPase Rac2,⁹⁵ and activation of a CXCR4-CXCL12 axis to drive cell proliferation.⁹⁶ These observations have led to the supposition that Schwann cell injury and neurofibroma transformation are similar biologic processes, and consistent with this hypothesis, nerve crush injury in NF1 mutant mice leads to neurofibroma formation.⁹⁷ Taken together, these studies reveal the cellular heterogeneity and signaling complexity underlying MPNST formation, requiring coordination and corruption of multiple signaling pathways through both cell autonomous and noncell autonomous mechanisms.

The molecular mechanisms underlying Schwann cell development and transformation have informed the selection of targeted agents to treat patients Schwann cell-derived tumors. As alluded to above, the treatment paradigm for schwannomas and neurofibromas primarily comprises local approaches such as surgery and/or radiation therapy. However, for large, rapidly progressive tumors particularly in syndromic cases associated with NF1 or NF2, more aggressive systemic therapy may be warranted to reduce morbidity. In that regard, inhibition of receptor tyrosine kinases or their downstream effectors such as Ras/Raf/MEK/ERK have demonstrated clinical efficacy. For example, the use of bevacizumab, a VEGF inhibitor, for NF2 associated vestibular schwannoma improves hearing loss and inhibits tumor progression.^98,99 More recently, the MEK inhibitor selumetinib demonstrated clinical efficacy for inoperable plexiform neurofibromas in NF1 patients,¹⁰⁰ offering the first potential systemic therapy option to improve morbidity in these patients. Finally, given the need for multimodality treatment and overall poor prognosis for MPNST patients, numerous targeted agents are currently being tested in the clinical trial setting, including inhibitors of MEK (selumetinib), PI3K (sirolimus and everolimus), and BET inhibitors (CPI-0610).¹⁰¹ These efforts illustrate the importance of defining critical developmental pathways underlying tumorigenesis from the Schwann cell lineage in order to rationally select agents to improve outcomes for our patients.

CONCLUSION

The Schwann cell lineage is a remarkable population that undergoes a complex developmental trajectory requiring coordination of cell autonomous and noncell autonomous inputs to undergo differentiation, migration, and proliferation. In many ways, this process happens in reverse in human patients who develop Schwann cell-derived tumors, as evidenced by progressive loss of Sox10, S100, and GFAP positivity during the transformation from benign to malignant cancers (Figure 1). Furthermore, as observed in human patients and interrogated in mouse genetic models, NF1 or NF2 loss in Schwann cells is necessary but not sufficient for transformation, as additional genetic hits and noncell autonomous contributions play a critical role in tumorigenesis.

With respect to the broader parallels between Schwann cell development and tumorigenesis, there are a number of questions that remain to be investigated. First, it will be important to define the effect of observed genetic aberrations in how noncell autonomous signals are interpreted by Schwann cells, and whether the genotype or transcriptome of Schwann cells fully dictates their cellular outcome. Second, single cell analyses will be critical to better understand the degree of cell type heterogeneity within these tumors, and how patterns of mutation co-occurrence are distributed across these populations to influence cellular phenotypes. Third, it will be important to test how the sequential order of mutations observed in human patients affects Schwann cell transformation. Finally, in addition to the canonical Schwann cell fates outlined above, cellular descendants of the early Schwann cell lineage can give rise to nonglial cell types, highlighting the remarkable plasticity of this population. Understanding the mechanisms underlying this process will provide insight into how related tumors, such as melanoma, take on aggressive features. Such analyses will pave the way for a deeper understanding of the mechanisms underlying transformation and metastasis in all neural crest lineages, which will ultimately improve outcomes for patients who develop these tumors.

Funding

H.N.V. is supported by a Children's Tumor Foundation Young Investigator Award.

Disclosures

The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.

COMMENTS

Similar to Mirsky and Jessen's recent review of the neurobiology of regenerating Schwann cells,¹ I feel that this manuscript will become a highly-cited reference moving forward. The first figure alone is gorgeous. To summate, the authors have provided an in-depth yet accessible review that details the molecular events of both Schwann cell development and fate determination, in addition to what is currently known regarding the genetic events that surround tumors of Schwann cell origin.

Of particular value, the authors analyze both the human and experimental animal literature regarding the latter topic, including potential therapeutics. We get an optimistic impression that the advent of conditional genetics in mouse models has led to a greater understanding of the biology of these tumors, allowing testing of new therapeutic strategies based in molecular medicine. The authors nicely summarise these findings in several tables that are a concise introduction to the key literature of this field.

Joey Kevin Grochmal

Calgary, Canada

This papers is an excellent and current review of the Schwann cell, with an emphasis on developmental aspects, and maldevelopment which heralds oncogenesis potential.

Many tumors recapitulate their developmental programs during progression, and peripheral nerve tumors are of no exception. In the era of precision medicine, understanding the molecular pathways leading to Schwann cell differentiation, as summarized nicely in this review article, and equally important, de-differentiation will certainly provide the starting point for novel therapies.

As the authors point out, the fate and diversity of peripheral nerve tumors from relatively benign schwannoma and neurofibroma to malignant peripheral nerve sheath tumor is likely the combinatorial outcomes from cell and non-cell autonomous events. These include gene mutations (NF1 vs NF2), different stages along the Schwann cell lineage and interactions with neighboring cells in the microenvironment. In fact, the review highlights the notion that the Schwann cell is not a homogenous population but instead a pool of heterogeneous cells with specialized functions. One example is the newly discovered capacity of Schwann cells in skin that enable nociceptive function, with the ability to initiate and transmit pain signal to adjacent axons and nerve bundles.¹ The heterogeneity is, however at the moment, masked by the lack of rigorous markers for individual subtypes.

With increasing ability to examine cell genotype and phenotype at high molecular and genomic resolution, we may one day achieve personalized medicine for different types of nerve tumors.

Tak-Ho Chu

Rajiv Midha

Calgary, Canada