Stat5 Feedback Control Switches MYC- and BCL6-Dependent Cell Fates in Normal and Transformed B Cells

Background: During early B cell development, IL7-dependent large cycling pre-B cells undergo dramatic phenotypic changes and differentiate into small resting pre-B cells in the absence of IL7. Inhibition of oncogenic tyrosine kinases in pre-B acute lymphoblastic leukemia (ALL) results in very similar changes. Both IL7 cytokine signaling in pre-B cells and oncogenic tyrosine kinase activity in pre-B ALL induces strong activation of Stat5. STAT5 induces its own negative feedback through transcriptional activation of CISH and SOCS proteins. Interestingly, SOCS and CISH are highly expressed in pre-B ALL cells and represent the focus of this study.

Results: mRNA levels of SOCS2 and CISH were 10-fold and 3-fold upregulated in ALL compared to normal pre-B cells respectively. Importantly, we found that high expression levels of SOCS/CISH are predictive of poor outcome in one clinical trial for ALL patients. These results indicate that inhibitory signaling from SOCS2 and CISH may contribute to the disease progression of ALL. To study the function of Socs2, Socs3 and Cish in genetic experiments, we transformed pre-B cells from mutant mice. Deletion of Socs2, Socs3 or Cish strongly increased phosphorylation of Stat5, increased ROS and accumulation of Arf, p53 and p21, ultimately resulting in leukemia cell death. Studying BCR-ABL1 transformed Cish^fl/fl ALL cells in vivo, we found that genetic deletion of Cish resulted in rapid leukemia regression and prolonged survival of recipient NOD-SCID mice (13 days vs 30 days; n=7; p<0.001). To elucidate the underlying mechanisms, we transduced a constitutively active form of STAT5 (Stat5^CA) or a non-phosphorylatable form of STAT5 (Stat5^Y694F) into B cell lineage ALL cells. While Stat5^CA increased cell size by nearly 2 fold, Stat5^Y694F caused rapid cell shrinkage. RNA-seq and Western blot results showed that Stat5^CA increased MYC and glycolytic gene expression but decreased BCL6 and fatty acid oxidation (FAO)-related genes. Conversely, Stat5^Y694F upregulated BCL6 expression and a FAO-related gene expression program, but decreased MYC and glycolytic gene expression. Metabolomics analyses confirmed that Stat5^CA increased glucose consumption, ATP production and protein synthesis. While Stat5^Y694F decreased glucose consumption, ATP production, and increased fatty acid oxidation through BCL6 and AMPK-activation. ChIP-seq revealed both forms of STAT5 can directly bind to MYC and BCL6 promoters as well as glycolytic and FAO targets, albeit with antagonizing effects. Inhibition of mTOR signaling by Rapamycin or protein synthesis inhibitors or knocking down MYC rescued Stat5^CA-induced toxicity in the context of dramatically enlarged cell size. On the other hand, overexpression of MYC rescued Stat5^Y694F-transduced cells by increasing glucose consumption and ATP production in cells that would otherwise dramatically shrink.

Conclusion: Our results suggest that STAT5 feedback control is required to keep a balance between p-STAT5 (CA) and non-p-STAT5 (Y694F), which promote divergent, MYC (CA) and BCL6 (Y694F)-dependent transcriptional programs. A single mutation affecting the phosphorylation status of STAT5-Y694 has far reaching impact on cell fate, differentiation status, cell size and metabolic behavior (glycolysis vs FAO), which need to be balanced by STAT5-feedback control.