Linfomi pediatrici di tipo B: studio della riprogrammazione metabolica e della diversità tumorale a livello trascrizionale

Burkitt’s lymphoma (BL) and diffuse large B cell lymphoma (DLBCL) are the most common non-Hodgkin’s lymphomas (NHLs) in the pediatric and young adult population. They originate from the malignant transformation of germinal center (GC) B cells that populate secondary lymphoid organs upon antigen encounter and activation. Despite sharing morphological and phenotypic features with their adult counterparts, pediatric NHLs often present with high-grade and frequent extranodal involvement. However, the clinical course of pediatric patients is more favorable than what is observed in adults. The enhanced enzymatic activity of EZH2 in DLBCL leads to increased deposition of the repressive trimethylation mark on histone 3 lysine 27 (H3K27me3), culminating in the transcriptional silencing of the tumor suppressor PRDM1 and cyclin-dependent kinase inhibitors CDKN1A/B and CDKN2A. The role of TRAP1 in lymphomagenesis has never been investigated before, but its direct inhibition of SDH, followed by succinate accumulation places it at the center of the metabolic and epigenetic changes that occur in DLBCL cells. Pediatric BL is curable in almost all cases following high-dose intense chemotherapy, with an overall survival rate exceeding 90%, when accompanied by appropriate patient-supportive care. However, the overall survival rate for pBLs that are refractory to treatments drops to 20% and there is no consensus on salvage treatments for these patients. Therefore, a more thorough investigation of pBL biological features may help derive which are the factors driving therapy refractoriness. During the first part of this PhD project, TRAP1 expression and functional consequences were investigated in DLCBL. TRAP1 was found to be overexpressed in both published datasets and in a small cohort of primary cases. Stratification of patients based on TRAP1 expression negatively correlated with overall survival, suggesting that TRAP1 may act in concert with other tumorigenic programs to promote disease progression in DLBCL patients. Moreover, functional investigation of TRAP1 role in DLBCL revealed a direct link between TRAP1-induced succinate accumulation and increased levels of H3K27me3, with the consequent downregulation of EZH2 target genes PRDM1 and CDKN2A. Of note, genetic ablation of TRAP1 negatively impacted the ability of cells to form colonies in colony-formation assays. The results of this study indicate that TRAP1 may be involved in the cooperative deregulation of DLBCL cells at the crossroads of metabolism and epigenetics, making it an interesting candidate for patient discrimination at diagnosis and for testing already available TRAP1 inhibitors in DLBCL cell lines. The second part of this PhD project focused on the characterization of transcriptional differences between responder (R) and non-responder (NR) pBL. The results demonstrate that BL is characterized by considerable heterogeneity in its transcriptional programs at the single-cell level. The DZ origin of BL was confirmed with the addition of a level of complexity to the biology of BL cells which were shown to also mimic features of normal GC B cells transitioning from the DZ to the LZ and vice-versa. The comparison between diagnostic specimens from R and NR patients identified molecules involved in the BCR signaling pathway to be consistently upregulated in NR tumors. Among the genes that were differentially expressed between R and NR patients, the top one was TPM2, a member of the tropomyosin actin filament binding protein family. Patient stratification based on TPM2 expression also showed a robust correlation with patient prognosis. The documentation of TPM2 upregulation in this study, along with the upregulation of other members of the cytoskeleton (including PCDH9, PDLIM3, MAP1B) in NR pBL suggests that the cytoskeleton and its features may contribute to therapy resistance in pBL and should be functionally investigated.