P455: runx1 directly binds and regulates the tcf4 promoter

Background: Aberrations of the RUNX1 transcription factor are frequently observed in hematological malignancies and are incorporated in AML prognostication. Remarkably, AML with a chromosomal t(8;21) translocation, leading to a fusion protein RUNX1-RUNX1T1, has a relatively good prognosis, whereas AML with a point mutation in RUNX1 is considered poor risk. In contrast to t(8;21) AML, the pathophysiology underlying RUNX1 mutated AML is still poorly understood. We have previously shown that TCF4 expression is an independent prognostic factor and that TCF4 was upregulated in RUNX1 mutated AML, while it was downregulated in t(8;21) AML. TCF4 expression was shown to be involved in the effect of RUNX1 mutations on overall survival, as RUNX1 mutational status lost significance when TCF4 expression was included in multivariate analysis. In addition, the impact of t(8;21) on survival diminished when TCF4 was included in the multivariate model. Aims: Here, we aimed to study if RUNX1 directly regulates the TCF4 promoter and how mutant RUNX1 affects this. Methods: To identify RUNX1 binding sites in the TCF4 promoter, we used ChIP-seq data. Subsequently we cloned the relevant promoter sequences in a pGL4 backbone vector in order to measure the transactivation of RUNX1 wildtype and different RUNX1 mutants. Results: First, we analyzed available data of RUNX1 binding locations, identified by ChIP, of different cell lines and primary AMLs with and without a RUNX1 mutation. We observed that in both wildtype (wt) and mutant (mut) RUNX1 cells, there was binding of RUNX1 to the TCF4 promoter. To investigate the ability of RUNX1 to regulate TCF4, we performed transactivation assays using the TCF4 promoter. This revealed that RUNX1 wt binds and represses the TCF4 promoter activity (control set on 1, RUNX1 wt mean=0.14). Next, we tested several clinically relevant RUNX1 mutants in the same conditions. We included 3 missense mutations (R80C, T169A and R201Q), 2 splice site mutations (S199fs and S291fs) and RUNX1-RUNX1T1. For all mutants that we tested, we found that they lost this repressive capacity. In contrast to RUNX1 mut, RUNX1-RUNX1T1 (derived from t(8;21)) still showed repression of the TCF4 promoter (control set on 1, RUNX1-RUNX1T1 mean = 0.19). All of this is in line with the TCF4 expression levels in patients that harbor a RUNX1 mutation or a RUNX1-RUNX1T1 fusion. In order to further delineate the exact location of RUNX1 activation, we divided the promoter in three different parts and tested transactivation of these parts separately. We found that part 2, the part that includes one of the two peaks identified by ChIP-seq, has the most transactivation potential. Therefore we developed a transactivation assay for this specific region. Again we found that RUNX1 wt and RUNX1-RUNX1T1 repress the TCF4 promoter in this assay and clinically relevant RUNX1 mut lost this repressive capacity(RUNX1 wt mean= 0.213, RUNX1 mut mean= 0.79,). In addition, we found that RUNX1 mut had a dominant negative effect over the RUNX1 wt protein, reversing the inhibitory effect of the wt protein on transcriptional activation reflected by its inability to repress the TCF4 reporter in the context of RUNX1 wt overexpression. This is in line with the elevated TCF4 expression levels in patients that harbor a RUNX1 mutation or a RUNX1-RUNX1T1 fusion. Summary/Conclusion: We show that TCF4 is an important downstream target of wild-type and mutated RUNX1. Keywords: Acute myeloid leukemia, Transcriptional regulation, RUNX1