RUNX1 Is Required For Maintenance Of Established T-ALL Blasts

T-cell acute lymphoblastic leukemia (T-ALL) is a clinically aggressive malignancy of immature T cells. Intensive multiagent chemotherapy achieves cure in 80-90% of pediatric patients, but only 40% of adult patients survive beyond 5 years. Data from recent ChIP-seq studies has shown that RUNX1 binds throughout the T-ALL genome at sites co-occupied by known oncogenic transcription factors including TAL1 and NOTCH1. For this reason, it has been suggested that RUNX1 may be part of a transcriptional activation complex that drives an oncogenic gene expression program in this cell context. In contrast, next generation sequencing studies have recently identified heterozygous point mutations throughout the RUNX1 coding region in T-ALL including some which are predicted to encode truncated polypeptides resembling dominant negative alleles, thus raising the possibility that RUNX1 may function rather as a tumor suppressor in this context. In an effort to explore the functional role of RUNX1 in T-ALL, we examined the effect of RUNX1 knockdown in a broad panel of established human T-ALL cell lines and xenograft-expanded patient biopsy samples.

Cells transduced with lentiviral shRNAs targeting coding and 3’ UTR regions of RUNX1 showed a clear growth disadvantage as compared to either non-transduced cells in the same culture or cells transduced with non-silencing shRNAs in parallel cultures. As well, absolute cell counts of cultures containing only shRUNX1-transduced cells demonstrated dramatically reduced growth rates as compared to either non-transduced or non-silencing shRNA-transduced controls. BrdU incorporation and CFSE dye dilution studies showed that most cell lines exhibited reduced proliferation in response to RUNX1 knock-down, while a subset of lines also showed reduced cell viability. These phenotypes were largely consistent across a panel of over 20 T-ALL cell lines and 4 xenograft-expanded patient samples, including several which harbored either nonsense or missense RUNX1 mutations. These results support the notion that established T-ALL cells are generally dependent on RUNX1 for continued cell growth and, in some cases, also for survival.

We also explored candidate RUNX1 target genes which might be responsible for mediating the observed growth/survival phenotypes. Assembling a short list of the “usual suspects” including genes known to regulate growth of T-ALL cells generally (c-MYC, PTEN), to possess a substantial RUNX1 ChIP-seq peak in T-ALL cells (IGF1R, IL7R), or to be bona fide RUNX1 targets in other cell contexts (p21/WAF1, p27/KIP1), we performed western blot or flow cytometric analyses of multiple shRUNX1-transduced cell lines. We noted consistent regulation of some, but not all of these in a manner consistent with RUNX1 positively supporting cell growth.

From these results, we conclude that RUNX1 plays a pro-oncogenic role in established T-ALL cells. We surmise that the complement of presumed loss-of-function RUNX1 mutations observed in patient T-ALLs may be indicative of its known roles in regulating normal T cell development such that loss-of-function mutations may lead to differentiation arrest and consequently promote tumor initiation. For the majority of T-ALLs that express wild-type RUNX1 proteins, however, our results suggest that RUNX1 acts functionally to support maintenance of the malignant clone by promoting expression of known oncogenic factors and repressing expression known tumor suppressors.