The translocation t(8;21) is the most common chromosomal rearrangement in AML and generates the RUNX1/ETO fusion protein. RUNX1/ETO is required for maintaining the leukaemic phenotype influencing both leukaemic clonogenicity and proliferation and is traditionally thought of as a transcriptional repressor through its actions with histone deacetylases and transcriptional co-repressors. Intriguingly our RNAi experiments have also identified many RUNX1/ETO target genes that are transcriptionally upregulated by this fusion protein.
To delve deeper into the involvement of these upregulated genes in RUNX1/ETO driven leukaemia, targeted RNAi screens were performed both in vitro and in vivo, directed at genes found directly bound by RUNX1/ETO and that were found differentially expressed upon RUNX1/ETO knockdown. For the RNAi screens, we used a doxycycline (dox)-inducible lentiviral RNAi library covering each gene with 3 shRNAs. We transduced two t(8;21)-positive AML cell lines, Kasumi-1 and SKNO-1, with this library and performed parallel screens employing colony formation and long-term suspension culture assays in the in vitro arm, and intrafemoral xenotransplantation of immunodeficient NSG mice for the in vivo arm. For comparative purposes, both arms of the RNAi screen were performed in the presence and absence of dox. DNA was isolated throughout both screens and was analysed by Next Generation Sequencing (NGS). Comparison of the changes in level of integrated shRNA coding sequences between dox and no dox groups were made using DESEQ and those genes that are required for RUNX1/ETO driven leukaemia identified.
As expected, RUNX1/ETO shRNA constructs diminished upon induction of shRUNX1/ETO expression by dox, therefore implicating RUNX1/ETO dependency in cell survival. Furthermore non-targeting control shRNA (shNTC) levels were unaffected. Out of the many hits identified (which include previously reported hits such as Pontin (RUBVL1), SKP2 and KIT, Cyclin D2 (CCND2) stood out significantly, since levels of shCCND2 were depleted in all dox samples from both arms of the screen.
CCND2 is a cell cycle regulator whose activity is dependent on its binding to CDK4/6 in G1 phase. Phosphorylation of Rb (Retinoblastoma), by CDK4/6-CCND2, uncouples Rb from E2F allowing transcription of essential S phase genes. Here we show CCND2 knockdown downregulates both cell proliferation and colony formation in t(8;21) positive cells by causing G1 phase cell cycle arrest via a reduction in Rb phosphorylation, which is a phenotype copied by our RUNX1/ETO knockdown. Moreover, inhibition of CDK4/6-CCND2 by palbociclib (PD-0332991) in t(8;21) positive cells similarly reduces cell proliferation and colony formation via a G1 cell cycle arrest. Interestingly, primary CD34+ human stem/progenitor cells (HSPCs) containing an activated form of KIT (N822K) and the RUNX1/ETO 9a isoform are likewise responsive to palbociclib treatment. In addition to in vitro studies, palbociclib was tested in vivo in two separate experimental models: firstly, mice were intravenously injected with murine CD34+ cKit+ HSPCs expressing the RUNX1/ETO9a isoform and secondly, immunodeficient mice were intrahepatically injected with Kasumi-1. Mice were dosed with palbociclib upon confirmation of engraftment. Engraftment was detected using bioluminescence and survival rates recorded. Both studies showed significant increased survival with palbociclib treatment with the first giving an increased median survival of 59 versus 29 days (p<0.005) and the second giving a median survival of 87 and 67 days for palbociclib treated and control mice (p<0.005), respectively.
Finally, Gene Set Enrichment Analysis (GSEA) shows the correlations between palbociclib treatment, CCND2 knockdown and RUNX1/ETO knockdown and shows common gene sets shared between them and other gene sets which include those targeted by MYC and E2F. Interestingly, gene sets that only correlate with knockdown of CCND2 and RUNX1/ETO and evidentially unaffected by CDK4/6 activity will be of interest, in order to target other RUNX1/ETO dependent proteins and pathways alongside of CDK4/6 inhibition. We are currently investigating the use of palbociclib with other drugs in order to develop new drug combinations with reduced toxicity and minimal risk of developing resistance.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.