Final Report Summary - NOREL (In Vitro and in Vivo Screens for Novel Regulators in Leukemia (NOREL))
The aim of the project has been to identify and characterize new factors involved in acute myeloid leukemia. We focused on the families of genes whose products are amenable to inhibition by small molecules and thus attractive as targets for novel anti-cancer therapies. We employed the state-of-the-art shRNA screening approach with high-throughput sequencing as a readout. We performed in vivo and in vitro screens in the pre-clinical model of acute myeloid leukemia driven by MLL-AF9 fusion oncogene. As a counter-screen for the general toxicity of shRNAs, we performed in vitro screen in normal mouse bone marrow cells. This screening approach has led to the identification of several genes whose depletion inhibited the growth of leukemic cells but had no or little effect on normal cells.
We decided to change our original plan for the in vitro screen and used a pooled shRNA library instead of single shRNAs. The pooled shRNA approach turned out to be more reliable and time-efficient. Moreover it allowed us to expand the library from constructs targeting only 48 histone methyltransferases and 24 histone demethylases to a pool of constructs targeting all major families of chromatin writers, erasers and readers (a total of 319 genes).
The in vivo screen was also performed with the expanded library. We were however able to perform the in vivo screening in the MLL-AF9 model only, instead of the originally planned 4 models. This was caused by the fact that in our hands mice transplanted with NRAS(G12D) and AML-ETO9a expressing cells did not develop acute myeloid leukemia, contrary to the data published by other laboratories.
We identified and validated several hits from the screens. The characterization of the first hit, a gene called Jmjd1c, has been already published in the Blood journal. Jmjd1c is required for leukemia maintenance in vitro and in vivo. Human and mouse leukemic cells with depleted levels of Jmjd1c undergo apoptosis, which translates into slower growth and lower colony formation capacity. Down-regulation of Jmjd1c affects such gene expression programs as Myb and Myc target genes and leukemic stem cell signature. Indeed, forced expression of Myc or Myb partially rescue Jmjd1c deficiency phenotype. According to the data available in 2010, Jmjd1c was believed to be an H3K9me2/me1 demethylase. However our data and publications from 2013 showed that Jmjd1c is lacking activity towards H3K9 methyl marks.
We are currently characterizing hits identified in our in vivo screen. This project led to the identification of several potential therapeutic targets in leukemia and their validation in pre-clinical models. Detailed characterization of the remaining hits will be a subject of future projects.
We decided to change our original plan for the in vitro screen and used a pooled shRNA library instead of single shRNAs. The pooled shRNA approach turned out to be more reliable and time-efficient. Moreover it allowed us to expand the library from constructs targeting only 48 histone methyltransferases and 24 histone demethylases to a pool of constructs targeting all major families of chromatin writers, erasers and readers (a total of 319 genes).
The in vivo screen was also performed with the expanded library. We were however able to perform the in vivo screening in the MLL-AF9 model only, instead of the originally planned 4 models. This was caused by the fact that in our hands mice transplanted with NRAS(G12D) and AML-ETO9a expressing cells did not develop acute myeloid leukemia, contrary to the data published by other laboratories.
We identified and validated several hits from the screens. The characterization of the first hit, a gene called Jmjd1c, has been already published in the Blood journal. Jmjd1c is required for leukemia maintenance in vitro and in vivo. Human and mouse leukemic cells with depleted levels of Jmjd1c undergo apoptosis, which translates into slower growth and lower colony formation capacity. Down-regulation of Jmjd1c affects such gene expression programs as Myb and Myc target genes and leukemic stem cell signature. Indeed, forced expression of Myc or Myb partially rescue Jmjd1c deficiency phenotype. According to the data available in 2010, Jmjd1c was believed to be an H3K9me2/me1 demethylase. However our data and publications from 2013 showed that Jmjd1c is lacking activity towards H3K9 methyl marks.
We are currently characterizing hits identified in our in vivo screen. This project led to the identification of several potential therapeutic targets in leukemia and their validation in pre-clinical models. Detailed characterization of the remaining hits will be a subject of future projects.