Modeling neurofibromatosis type1 neurofibroma composition and formation

Abstract

[eng] Neurofibromatosis type 1 (NF1) is a genetic disease caused by inherited mutations in the NF1 gene. NF1 is a cancer predisposition syndrome, and the disease’s hallmark is the appearance of tumors of the peripheral nervous system called neurofibromas. Neurofibromas are mainly composed of Schwann cells (SCs) and endoneurial fibroblasts (eFbs) and also contain infiltrating immune cells and other cell types. Neurofibromas appear after the complete inactivation of the NF1 gene in SCs or their precursors. Almost all NF1 patients (>99%) develop cutaneous neurofibromas (cNFs), small benign tumors that appear in the skin around puberty and continue appearing throughout the patient’s life. Although cNFs have no risk of becoming malignant, they significantly impact the quality of life of NF1 patients. The only current treatment is surgical removal. Around 50% of patients develop plexiform neurofibromas (PNFs). PNFs form during development and are generally diagnosed either at birth or in early childhood when they grow most rapidly. Some PNFs can progress towards a malignant peripheral nerve sheath tumor, the leading cause of death in NF1.

The three main objectives of this thesis were: (i) to develop an imperishable cell-based model system to study PNFs, (ii) to study the identity of the cells composing PNFs, and (iii) to study the signaling between SCs and Fb concerning cNF growth.

To establish an imperishable cell-based PNF model system, we generated NF1(-/-) and NF1(+/-) induced pluripotent stem cell (iPSC) lines from PNF-derived cells. Then, we set up protocols to differentiate iPSCs into Neural Crest (NC) cells and from them further to SCs, providing a robust NC-SC in vitro differentiation system. We identified differentially expressed genes in stages and timepoints along the in vitro NC-SC lineage that were grouped to establish a NC-SC expression roadmap. Altogether, it provided a framework for analyzing the role of the NF1 gene during SC differentiation. NF1(-/-) differentiating SCs in 2D cultures did not constitute an adequate system due to a high proliferation capacity but poor homogeneous differentiation ability towards SCs. However, using the natural tendency of NF1(-/-) differentiating SCs to form spheres, different 3D models were developed in a multiplexed format. The engraftment of heterotypic spheroids, composed of NF1(-/-) differentiating SC and primary eFbs, represented the most efficient and consistent way of producing human neurofibroma-like tumors in mice.

In addition to being composed of different cell types, we identified the existence of different SC subpopulations within PNFs, by using flow cytometry and single-cell RNA-seq analysis. We characterized at least two distinct SC subpopulations, one expressing markers appearing early in the NC-SC axis, suggesting a precursor-like identity, and another subpopulation expressing both early and late markers along the NC-SC lineage, suggesting a singular committed SC. This heterogeneity may change the current view of PNF development and progression.

We identified differentially upregulated genes produced by the interaction between cNF- derived SCs and eFbs, expressed in both cell types. Analyzing these genes, we identified several enriched signaling pathways potentially participating in SC-Fb crosstalk. Those involved in the infiltration of immune cells were significantly represented and were confirmed by a secretion profile analysis of SC-eFb co-cultures.

3

In conclusion, in the present work, an imperishable iPSC-based in vitro/in vivo PNF model system has been developed, allowing the generation of human neurofibroma-like tumors in mice for the first time. We discovered the presence of different subpopulations of SCs within PNFs that may change the view of how they develop and grow. Finally, we found different promising signaling pathways triggered by the heterotypic interaction between SCs and eFbs that could be relevant for cNF growth.