STAT3 Is Activated By DYRK1A and Is a Potential Therapeutic Target in B-ALL

Signal transducer and activator of transcription 3 (STAT3) is a transcription factor that mediates signal transduction from the extracellular surface to the nucleus. Canonically, STAT3 is phosphorylated at Tyrosine 705 (Y705) by JAK family kinases, which promotes its dimerization and subsequent localization to the nucleus. However, the role of Serine 727 (S727) phosphorylation in regulating STAT3 activity varies across cell types and remains unclear in hematopoietic tissues particularly. Several studies indicate that phosphorylation at S727 is critical for optimal STAT3 function. For example, astrogliogenesis is regulated by enhancing STAT3 activity by phosphorylation of S727 by DYRK1A. Of note, DYRK1A is overexpressed in Down syndrome-acute lymphoblastic leukemia (DS-ALL), and has previously been found to phosphorylate substrates in order to prime them for downstream phosphorylation events. Given these findings, we hypothesized that the DYRK1A phosphorylation of STAT3 at S727 is critical for promoting DS-ALL. Furthermore, certain subtypes of ALL have high rates of JAK2 activation, namely DS-ALL and Philadelphia-like ALL (Ph-like ALL); we propose that STAT3 can effectively be targeted specifically in these subtypes.

In order to elucidate the role of DYRK1A phosphorylation of STAT3, we treated cytokine-deprived murine pre-B cells with EHT1610, a selective DYRK1 inhibitor, or vehicle and then pulsed the cells with JAK-STAT activating cytokines. EHT1610-treated cells had diminished S727 phosphorylation compared to vehicle, regardless of cytokine pulse; however, only vehicle-treated cells regained Y705 phosphorylation after cytokine pulse. This suggests that S727 phosphorylation is cytokine-independent and is critical for maintenance of Y705 phosphorylation. We then generated flag-tagged STAT3 S727 phospho-mimetic (S727D/E) and phospho-deficient (S727A) alleles and transduced them into pre-B cells. We observed that the degree of Y705 phosphorylation is dependent on S727, as cells expressing S727A have reduced Y705 phosphorylation compared to wild-type STAT3. Additionally, overexpression of the phospho-deficient allele conferred a significant proliferative impairment compared to the phospho-mimetic alleles.

As DS-ALL and Ph-like ALL often have JAK2-activating mutations, we next aimed to determine if loss of S727 phosphorylation would decrease ALL cell growth. Indeed, two human Ph-like ALL cell lines, MHH-CALL4 and MUTZ5, displayed decreased proliferation when overexpressing the S727A mutant. These cell lines were also sensitive to treatment with C188-9, a small molecule STAT3 inhibitor that is in clinical trials for various solid tumors. Additionally, we treated primary patient ALL samples with amplification of HSA21 segments ex vivo and found that DS-ALL samples were preferentially sensitive to STAT3 inhibition compared to HD-ALL or iAMP-ALL, suggesting that STAT3 is specifically a target in JAK2-activated ALL.

Our study provides new and significant insights into the regulation of STAT3 by DYRK1A, and presents a new therapeutic target for ALL cells with JAK2 activating mutations.