Abstract Introduction: Most high stage neuroblastoma initially respond to chemotherapy, but ultimately relapse as therapy-resistant tumor. The mechanisms driving relapse and resistance remain elusive. We investigated whether neuroblastoma tumors include phenotypically and functionally divergent subsets of tumor cells that may underlie its clinical plasticity. Experimental Procedures: Fresh tumor cells were cultured in neural stem cell medium and analyzed by FACS, whole genome sequencing, mRNA profiling, motility assays and chemo-sensitivity assays. Lentivirally transduced inducible gene constructs were used to test state-transitions. Immunohistochemistry was used to define cellular subtypes in tumors. Results: We observed that new neuroblastoma cell lines always include two phenotypically divergent cell types. Both types share the same genetic defects, but have highly divergent phenotypes. One cell type has a neuro-epithelial (NE) phenotype and expresses all classical and diagnostically used neuroblastoma markers. The other type has a mesenchymal (MES) character, lacks all neuroblastoma markers and is highly motile. At low frequency, both cell types can spontaneously transdifferentiate in vitro. Immunohistochemistry (IHC) of primary neuroblastoma detected a small fraction of MES cells in most tumors. To analyze the clinical relevance of MES-type cells, we investigated their sensitivity to chemotherapeutics used in neuroblastoma treatment. In four isogenic pairs, MES cells were more resistant to the drugs than their NE-type counterparts. We investigated whether the chemo-resistance of MES cells may operate in vivo. We analyzed a series of primary neuroblastoma tumors surgically removed immediately after chemotherapy. The viable cells in the post-therapy samples were strongly enriched in MES-type cells as compared to the pre-treatment tumors of the same patients. We also compared primary, pre-treatment tumors with relapses emerging 4-5 years later in the same patients. Most strikingly, also the relapsed neuroblastoma tumors were highly enriched for MES-type cells. As these data suggest a role for MES-type cells in development of therapy-resistant relapses, we analyzed their key regulatory pathways. mRNA profiling of isogenic MES-NE cell line pairs identified consistent mRNA expression differences between both phenotypes. Major signaling routes and transcription factors were highly differentially expressed. MES-type cells had high expression and activation of NOTCH pathway genes and expression of the homeobox gene PRRX1. Induced expression of NOTCH or PRRX1 transgenes in multiple NE-type cell lines converted them into MES-type cell lines, including chemo-resistance. Analysis of the changes in gene expression and activity downstream of NOTCH or PRRX1 allowed reconstruction of the molecular wiring of MES-type cells. This identified several drugable key-players, like MEK and PDGFRβ. Targeting of them with small-molecule inhibitors specifically killed MES cells in vitro. Conclusions: Our data suggest that neuroblastoma is a bi-phasic tumor. MES and NE cells have very different characteristics, but can transdifferentiate into each other. It is tempting to speculate that the MES- and NE-phenotypes recapitulate two developmental stages of neuroblasts: MES cells may correspond to the migrating cell type that has delaminated from the neural crest, while NE cells could correspond to the more differentiated cell in the target organs expressing markers of the adrenalin synthesis route. MES cells strongly accumulate after chemo-therapy and in relapses. They may survive classical therapy and over time seed relapses, that ultimately become heterogeneous again. Targeted elimination of MES cells with small molecule inhibitors shows how cells with a potential key role in relapse development are amenable to therapy. This abstract is also presented as Poster B30. Citation Format: Rogier Versteeg, Tim van Groningen, Bart A. Westerman, Jan J. Molenaar, Ellen M. Westerhout, Mohamed Hamdi, Godelieve A. Tytgat, Jan Koster, Johan van Nes. Neuroblastoma is biphasic and includes classical neuroepithelial cells and chemoresistant mesenchymal cells. [abstract]. In: Proceedings of the AACR Special Conference on Advances in Pediatric Cancer Research: From Mechanisms and Models to Treatment and Survivorship; 2015 Nov 9-12; Fort Lauderdale, FL. Philadelphia (PA): AACR; Cancer Res 2016;76(5 Suppl):Abstract nr PR08.
Read full abstract