TSLP-Induced Alterations of Multiple Signaling Pathways in Primary CRLF2 B-ALL Xenografts

B-cell acute lymphoblastic leukemia (B-ALL) with genetic defects leading to overexpression of CRLF2 (CRLF2 B-ALL) is associated with a high relapse rate and poor prognosis. CRLF2 B-ALL comprises approximately half of the high risk B-ALL characterized by a gene expression profile that is similar to that of Philadelphia chromosome-positive ALL (Ph-like B-ALL). In pediatric patients, CRLF B-ALL occurs 5 times more frequently among children of Hispanic and Native American ethnicity and is a major contributor to health disparities in ALL. CRLF2 (cytokine related ligand factor 2) together with the IL-7 receptor alpha chain, forms a receptor complex that is activated by the cytokine, TSLP (Thymic Stromal Lymphopoietin). Activation of CRLF by TSLP leads to downstream JAK-STAT5 and mTOR pathway phosphorylation even in CRLF2 B-ALL harboring activating JAK mutations. We found that primary human marrow (BM) stromal cells express TSLP (RT-PCR and ELISA) and thus provide an in vivo source of TSLP to activate CRLF2 B-ALL cells. Our next step was to develop a xenograft model system to identify the in vivo CRLF2-mediated gene expression profile and disease mechanisms that might contribute to poor prognosis. Unlike most other cytokines, mouse TSLP is species-specific and thus does not activate the human CRLF2 receptor complex. We engineered immune-deficient NOD/SCID IL-2Rγ null (NSG) mice to express normal serum levels (~20 pg/ml) of human TSLP (hTSLP+ mice), as well as control mice that lack human TSLP (hTSLP– mice). Primary human CRLF2 B-ALL were injected into hTSLP+ and hTSLP– mice and expanded for 10 weeks in vivo. Whole genome microarray was performed on CRLF2 B-ALL cells isolated by magnetic separation from the BM of hTSLP+ and hTSLP- xenograft mice. Evaluation of microarray data by Gene Set Enrichment Analysis (GSEA) and Ingenuity Pathway Analysis showed that genes downstream of mTOR pathway activation were upregulated in hTSLP+ as compared to hTSLP- mice, confirming hTSLP activity in the hTSLP+ xenograft mice. Microarray identified 280 genes that are upregulated and 281 genes that are downregulated (> 1.7 fold; p<.05) in vivo in leukemia cells from hTSLP+ as compared to hTSLP– mice. GSEA and Ingenuity Pathway Analysis of these data show increased RAS pathway activation and altered glucose metabolism in CRLF2 B-ALL from hTSLP+ as compared to hTSLP– mice. In addition to increased mTOR pathway activation, these gene expression data implicate altered glucose metabolism and increased RAS pathway activation as potential contributors to the poor prognosis in CRLF2 B-ALL. The hTSLP+ CRLF2 B-ALL xenograft mice described here provide a novel preclinical model for studying disease mechanisms and identifying therapies to target signaling pathways activated by TSLP in CRLF2 B-ALL.