Identification of the cellular origin and 'stemness' phenotype of Malignant Rhabdoid Tumours (MRT) may represent a new therapeutic approach in paediatric oncology

Abstract

Introduction: Malignant Rhabdoid Tumours (MRT) are especially lethal cancers that predominantly occur in infants and young children. MRT are caused by biallelic inactivation of a single gene; SMARCB1 which is a component of SWI/SNF chromatin remodelling complex. The tumours can be found almost everywhere in the body and the cell of origin for MRT is still unknown. Expression profiling of primary MRT demonstrates strong overexpression of a large number of stemness/self-renewal genes (PBTG, unpublished). Previous studies have shown that expression of these genes are deregulated by SMARCB1 and/or larger SWI/SNF complex. Therefore, this project was aimed to uncover the potential cell of origin and to examine if expression of self-renewal genes could potentially contribute towards disease aggressiveness. The understanding of other mechanisms involved in MRT tumourigenesis is essential to identify therapeutic targets that can improve the outcomes of MRT patients. Methods: I performed a meta-analysis cross-referencing expression profiles from primary MRT (n= 119), and functional models in which SMARCB1 was re-expressed (n=5 lines) with expression profiles from multiple candidate stem cell types including epithelial, embryonal, neural, mesenchymal and neural crest (n= 446). I developed a lentiviral CRISPR/Cas9 system to efficiently model Smarcb1 mutation in primary candidate cells of origin ex vivo. To understand how SMARCB1 loss is capable of hijacking active stemness/self-renewal machinery to initiate MRT tumourigenesis, an integrated data analysis was performed to identify stemness/self-renewal genes whose expression are SMARCB1-dependent, by cross-referencing a manually curated stemness/self-renewal gene list with existing RNA-seq data from primary MRT and MRT cell lines with and without re-expression of SMARCB1 and with CRISPR/Cas9 genomewide screening data from MRT cells with and without re-expression of SMARCB1. Finally, the expression of stemness/self-renewal genes and their relevance as a therapeutic target in MRT, specifically the protein BMI1, was studied using shRNA and a small molecular inhibitor. ii Results: Bioinformatic analysis of potential cells of origin suggests three putative candidate cell populations within the family of neural crest (NC), neural progenitor and mesenchymal stem cells. The extent of transcriptional overlap clearly reflects differences in tumour location and molecular subgroup. Unlike most cells tested, early neural crest (NC) cells tolerated Smarcb1 mutation and conferred proliferative advantage resembling human MRT molecularly and immunophenotypically. From my integrated bioinformatic analysis of self-renewal/stemness, BMI1 was identified as a functional SMARCB1- dependent gene. Here, I show that targeted disruption of BMI1 by shRNA or pharmacologic inhibition using PTC209 strongly impairs MRT cell growth, suppresses tumour cell self-renewal, induces apoptosis and senescence. Using functional gene expression analysis, I found that BMI1 knockdown in MRT cells transcriptionally activates critical genes in MRT such as p21, p16 and CD44. Conclusion: Our findings suggest NC cells as a potential cell of origin for MRT. I describe a system for ranking putative cells of origin and a practical means for inducing Smarcb1 mutations ex vivo. Furthermore, I show evidence that subtypes within MRT have different cells of origin. Here, I also establish the role of BMI1 in self-renewal and survival of MRT cells. In addition, I show that the effects of BMI1 knockdown recapitulates the effect of SMARCB1 re-expression in MRT cells, demonstrating an attractive therapeutic target for MRT.