Schwann cell plasticity regulates neuroblastic tumor cell differentiation via epidermal growth factor-like protein eight

nature COMMUNICATIONS

ARTICLE

Schwann cell plasticity regulates neuroblastic tumor cell differentiation via epidermal growth factor-like protein eight

Adult Schwann cells possess an inherent plastic potential. This plasticity allows Schwann cells to acquire repair-specific functions essential for peripheral nerve regeneration. Here, we investigate whether stromal Schwann cells in benign-behaving peripheral neuroblastic tumors adopt a similar cellular state. We profile ganglioneuromas and neuroblastomas, rich and poor in Schwann cell stroma, respectively, and peripheral nerves after injury, rich in repair Schwann cells. Indeed, stromal Schwann cells in ganglioneuromas and repair Schwann cells share the expression of nerve repair-associated genes. Neuroblastoma cells, derived from aggressive tumors, respond to primary repair-related Schwann cells and their secretome with increased neuronal differentiation and reduced proliferation. Within the pool of secreted stromal and repair Schwann cell factors, we identify EGFL eight, a matricellular protein with so far undescribed function, to act as neuritogen and to rewire cellular signaling by activating kinases involved in neurogenesis. In summary, we report that human Schwann cells undergo a similar adaptive response in two patho-physiologically distinct situations, peripheral nerve injury and tumor development.

Schwann cells are the principal glia of the peripheral nervous system and evolve in close contact with neurons into peripheral nerve fibers. Reciprocal signaling between Schwann cells and neurons regulates the survival, fate decisions, and differentiation of both cell types, but also influences their behavior in regenerative and pathological conditions. Hence, understanding the molecular mechanisms underlying Schwann cell-neuron interaction is of utmost interest to develop effective treatment strategies for injuries and pathologies of the peripheral nervous system.

Despite being necessary for correct nerve development, Schwann cells earned recognition because of their plasticity that allows differentiated Schwann cells, further called adult Schwann cells, to transform into a dedicated repair cell after peripheral nerve injury. The process is referred to as adaptive cellular reprogramming and includes profound transcriptional and morphological changes. This phenotypical switch is mediated by dedifferentiation causing the regain of immature/precursor Schwann cell properties followed by redifferentiation into a repair-specific state. The resulting repair Schwann cell phenotype is characterized by the re-expression of markers known to be upregulated in Schwann cells during development, and by distinct repair functions and repair-associated ligands distinguishing repair Schwann cells from adult Schwann cells or developing Schwann cells. Those repair functions comprise the degradation of myelin debris, attraction of phagocytes, the formation of regeneration tracks for axon guidance, and the expression of cell surface proteins and trophic (neuroprotective and neuritogenic) factors promoting axon survival and re-growth. We have recently provided a comprehensive transcriptomic and proteomic characterization of human repair Schwann cells demonstrating that Schwann cells isolated from excised peripheral nerves adopt the same repair-related phenotype and function in culture as in nerve tissue explants. These included the expression of master transcriptional regulators, such as JUN, as well as myelinophagy, phagocytosis, and antigen processing and presentation via MHC-two. Importantly, transcriptomic signatures of primary repair-related Schwann cell cultures indicated the expression of a variety of neurotrophins and neuritogens and, thus, present an ideal in vitro model to study processes involving nerve repair and neuronal differentiation.

Interestingly, a prevalent stromal Schwann cell population is found in usually benign-behaving subtypes of peripheral neuroblastic tumors. Peripheral neuroblastic tumors originate from trunk neural crest-derived sympathetic neuroblasts and are categorized in neuroblastomas, ganglioneuroblastomas, and ganglioneuromas that represent a spectrum from neuroblastomas, the most aggressive form, to ganglioneuromas, the most benign form, and ganglioneuroblastomas, which exhibit various elements of both. Neuroblastoma and ganglioneuroma subtypes are associated with distinct genomic alterations and strikingly different morphologies. In general, neuroblastomas consist of un- or mostly poorly differentiated tumor cells and cancer-associated fibroblasts, whereas ganglioneuromas are composed of differentiated, ganglionic-like tumor cells scattered within a dominant Schwann cell stroma. The content of Schwann cell stroma was early recognized as a valuable prognostic factor as it correlates with the degree of tumor cell differentiation and a favorable outcome. The ganglionic-like tumor cells also extend numerous neuritic processes that form entangled bundles surrounded by ensheathing stromal Schwann cells. This ganglion-like organoid morphology was assumed to arise from a bi-potent neoplastic neuroblastic precursor cell capable to differentiate along a neuronal and glial lineage. Hence, an active role of stromal Schwann cells in peripheral neuroblastic tumors has been neglected due to their supposed neoplastic origin.

Of note, we and others provided evidence for a non-tumor background of stromal Schwann cells. In a detailed immunohistochemical study, it was shown that the earliest appearance of stromal Schwann cells is confined to the tumor blood vessels and connective tissue septa and not intermingled within the tumor as a clonal origin would imply. Furthermore, we demonstrated the absence of numerical chromosomal aberrations in stromal Schwann cells, while adjacent ganglionic-like tumor cells possessed a typical aneuploid genome. These surprising findings argue against the hitherto presumed model of ganglioneuroblastoma/ganglioneuroma development based on a bi-potent neoplastic cell and support that the tumor cells are able to attract adult Schwann cells from the nervous environment to the tumor.

In detaching the origin of stromal Schwann cells in ganglioneuroblastoma/ganglioneuroma from a neoplastic cell, we realized how little we know about their nature. What is the cellular state of stromal Schwann cells? How do they affect ganglioneuroblastoma/ganglioneuroma development? And why are they not manipulated by the tumor cells to support tumor progression but are associated with a benign tumor behavior/biology? We and others have shown that the aggressiveness of neuroblastoma cell lines, derived from high-risk metastatic neuroblastomas, can be reduced upon exposure to Schwann cells and their secreted factors. Accordingly, a mouse study comparing intra- or extra-fascicularly grown tumor xenografts confirmed that neuroblastomas within the nervous environment were infiltrated by Schwann cells and developed a less aggressive tumor phenotype. However, a comprehensive analysis to assess the origin and functional characteristics of stromal Schwann cells in tumors is still missing.

Based on the inherent plasticity of adult Schwann cells and the yet unresolved nature of Schwann cell stroma, we speculate that ganglioneuroblastoma/ganglioneuroma development could be the result of a reactive/adaptive response of Schwann cells to peripheral neuroblastic tumor cells similar to injured nerve cells. Thus, we here compared the cellular state of stromal Schwann cells in ganglioneuromas to repair Schwann cells in injured nerves by transcriptome profiling of human ganglioneuroma and human injured nerve tissues. Moreover, we analyzed the effect of human primary repair-related Schwann cells and their secreted factors on genetically diverse neuroblastoma cells in co-culture studies and identified a promising candidate factor of therapeutic potential for aggressive neuroblastomas and peripheral nerve injuries.