Hyperactive STAT5 hijacks T cell receptor signaling and drives immature T cell acute lymphoblastic leukemia
Tobias Suske, et al.
Tobias Suske, et al.
View: Text | PDF
Research Article Genetics Hematology

Hyperactive STAT5 hijacks T cell receptor signaling and drives immature T cell acute lymphoblastic leukemia

Abstract

T cell acute lymphoblastic leukemia (T-ALL) is an aggressive immature T cell cancer. Mutations in IL7R have been analyzed genetically, but downstream effector functions such as STAT5A and STAT5B hyperactivation are poorly understood. Here, we studied the most frequent and clinically challenging STAT5BN642H driver in T cell development and immature T cell cancer onset and compared it with STAT5A hyperactive variants in transgenic mice. Enhanced STAT5 activity caused disrupted T cell development and promoted an early T cell progenitor–ALL phenotype, with upregulation of genes involved in T cell receptor (TCR) signaling, even in absence of surface TCR. Importantly, TCR pathway genes were overexpressed in human T-ALL and mature T cell cancers and activation of TCR pathway kinases was STAT5 dependent. We confirmed STAT5 binding to these genes using ChIP-Seq analysis in human T-ALL cells, which were sensitive to pharmacologic inhibition by dual STAT3/5 degraders or ZAP70 tyrosine kinase blockers in vitro and in vivo. We provide genetic and biochemical proof that STAT5A and STAT5B hyperactivation can initiate T-ALL through TCR pathway hijacking and suggest similar mechanisms for other T cell cancers. Thus, STAT5 or TCR component blockade are targeted therapy options, particularly in patients with chemoresistant clones carrying STAT5BN642H.

Authors

Tobias Suske, Helena Sorger, Gabriele Manhart, Frank Ruge, Nicole Prutsch, Mark W. Zimmerman, Thomas Eder, Diaaeldin I. Abdallah, Barbara Maurer, Christina Wagner, Susann Schönefeldt, Katrin Spirk, Alexander Pichler, Tea Pemovska, Carmen Schweicker, Daniel Pölöske, Emina Hubanic, Dennis Jungherz, Tony Andreas Müller, Myint Myat Khine Aung, Anna Orlova, Ha Thi Thanh Pham, Kerstin Zimmel, Thomas Krausgruber, Christoph Bock, Mathias Müller, Maik Dahlhoff, Auke Boersma, Thomas Rülicke, Roman Fleck, Elvin Dominic de Araujo, Patrick Thomas Gunning, Tero Aittokallio, Satu Mustjoki, Takaomi Sanda, Sylvia Hartmann, Florian Grebien, Gregor Hoermann, Torsten Haferlach, Philipp Bernhard Staber, Heidi Anne Neubauer, Alfred Thomas Look, Marco Herling, Richard Moriggl

×

Figure 1

STAT5BN642H occurs in patients with T-ALL and impacts thymic development and architecture.

Options: View larger image (or click on image) Download as PowerPoint
STAT5BN642H occurs in patients with T-ALL and impacts thymic development...
(A) Percentages of patients with somatic STAT5B mutations found in 8 studies (24, 4, 35, 11, 33, 32, 34, 25) in T-ALL and T-LBL. (B) Incidence and localization in the STAT5 protein with amino acid (AA) position of respective somatic mutations in STAT5B found in 1,234 patients with T-ALL or T-LBL from the studies indicated in A. (C) White blood cell (WBC) counts of patients with or without recurrent STAT5B mutations. Data were extracted from ref. 4. (D and E) Flow cytometry analysis and percentages of DN1, DN2, DN3, and DN4 cells as well as DN, DP, SP4, and SP8 cells from STAT5BN642H mice (n = 7) and WT littermates (n = 7) gated on Thy1.2+ thymocytes. Two independent experiments were performed. (F) Representative images and thymus weights in g of 8-week-old WT (n = 7) or STAT5BN642H (n = 7) mice. (G) Representative H&E-stained whole thymi from WT or STAT5BN642H mice, indicating cortical (C) and medullary (M) regions. Original magnification: ×10 (boxed by dashed lines); scale bar: 1 mm. (H) Quantification of medullary-cortical area ratios with n = 6 WT and n = 7 STAT5BN642H using the ImageJ (NIH) software. (I) RPKM values of Runx3, Eomes, Tbx21, Gzmb, Prf1, and Fasl in DN, DP, and SP8 stages of WT or STAT5BN642H thymocytes. In C–F and H, significant differences are indicated as *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 by unpaired 2-tailed Student’s t test. Data are shown as the mean ± SEM. In I, adjusted P values (Padj) of respective comparisons were taken from DESeq2 analysis. Data are shown as the mean ± SEM.