SOX11 skews T-cell development and synergizes with LMO2 and MYCN to form T-cell acute lymphoblastic leukemia in mouse models

Abstract

Abstract Objective: T-cell acute lymphoblastic leukemia (T-ALL) is an aggressive hematologic malignancy accounting for 10%–15% of pediatric and 25% of adult ALL cases. While intensified therapy has improved survival, outcomes for relapsed or refractory T-ALL remain poor, emphasizing the need to advance our understanding of T-ALL biology and develop precision oncology therapeutics. SOX11 is a transcription factor involved in embryogenesis and is aberrantly expressed in several tumor types. Although normally absent in thymocytes, SOX11 is expressed in a subset of T-ALL cases, where it may act as an oncogenic driver. Here, we explore the role of SOX11 in T-cell development and T-ALL initiation. Methods: To replicated elevated SOX11 levels, we generated a conditional Rosa26-SOX11 overexpression mouse model and crossed it with CD2-Cre or Lck-Cre drivers to express SOX11 in common lymphoid progenitors (SOX11CD2) or T-cells (SOX11Lck). These were further crossed with Lmo2 or MYCN models to obtain SOX11Lck;Lmo2CD2 and SOX11Lck;MYCNLck mice. CITE-seq was performed on spleens from 12-week-old control (n=3) and SOX11CD2 (n=3) mice. Thymi from 8-week-old SOX11Lck mice were analyzed to assess pre-leukemic changes. RNA-seq was conducted sorted CD4-CD8- double negative (DN) or CD4+CD8+ double negative (DP) cells from control, SOX11Lck, Lmo2CD2, and SOX11Lck;Lmo2CD2 mice. Thymic transplantation in sublethally irradiated mice was used to evaluate self-renewal of SOX11-overexpressing thymocytes. We analyzed publicly available transcriptomic data from 1,309 T-ALL patients to evaluate SOX11 expression and identify common genomic features in SOX11high T-ALL. Additionally, we transduced Lck-Cre+ bone marrow progenitors from control or SOX11Lck mice with either an empty vector or Cre-dependent MYCN-expression contstructs, and transplanted them into immunocompromised recipients to evaluate leukemogenic potential. Results: We analyzed the expression levels of SOX11 in 1309 T-ALL cases that were categorized according to the classifying driver or subtype. SOX11 was highly expressed in LMO1/2, TAL1/2, and KMT2A driver groups and in STAG&LMO2, KMT2A, LMO2 γδ-like, and TAL1 αβ-like subtypes. In contrast, SOX11 was low or absent in most thymocyte populations. scRNA-seq of SOX11CD2 mice revealed a block in αβ T-cell development and skewing toward B-cell and γδ T-cell fates, suggesting a potential role for SOX11 in LMO2 γδ-like T-ALL development, but also that SOX11 may exert its oncogenic role post T-cell commitment. To further dissect the role of SOX11 during T-cell development, we used a T-cell-restricted Lck-Cre driver in subsequent studies. Similar to a previously published Lmo2 model, SOX11Lck mice displayed an increase in immature DN3 thymocytes and a decrease of DP thymocytes. However, thymic transplantation assays demonstrated that, unlike Lmo2, SOX11 does not confer self-renewal capacity to thymocytes. Notably, co-expression of SOX11 and Lmo2 synergized to expand DN3 thymocytes and accelerate T-ALL onset, with median survival dropping from 332 days in Lmo2CD2 to 163 days in SOX11Lck;Lmo2CD2mice. To dissect downstream targets, we performed RNA-seq on FACS-sorted pre-leukemic and leukemic DN3 thymocytes. Enrichement analysis identified MYCN as a top SOX11 target. Strikingly, SOX11high T-ALL were enriched for MYCN P44L mutations (13/20 cases), which increase MYCN stability via disruption of a degradation motif. We validated MYCN as a functional SOX11 effector in genetic and transplantation models and found combined SOX11 and MYCN expression accelerated T-ALL onset. Conclusions: SOX11 is aberrantly expressed in a subset of T-ALLs and associated with MYCN P44L mutations. In mice, SOX11 induces a developmental arrest at the DN3 stage and synergizes with Lmo2 or MYCN to drive T-ALL formation.