Recent studies have suggested that co-activators are good drug targets for cancer therapy, as exemplified by BRD4, the small molecule inhibitors of which have already entered clinical trials. The high potency of targeting BRD4 in leukemia cells mainly rises from suppressing the functions of multiple lineage specific transcription factors. However, such pleiotropic effects of Brd4 on different transcription factors might be problematic, considering many of the perturbed transcription factors are equally required by normal cells. This prompted us to explore more co-activators as potential drug targets, aiming to identify the candidates that discretely disturb the function of oncogenic transcription factors but not the general ones. In this study, we explored the roles of one of the largest co-activator complexes TFIID in the maintenance of acute myeloid cells.

To address this, we designed an shRNA library targeting each TAF subunit of TFIID and tested the growth effect of these shRNAs in both mouse leukemia cell line and primary bone marrow culture. We found that multiple shRNAs targeting Taf12 shows strong toxicity in leukemia cells but not normal cells in vitro. Knocking down Taf12 also inhibits leukemia progression in vivo. To gain insight into the toxicity of inhibiting Taf12 in different organs, we generated doxycycline inducible transgenic Taf12 shRNA mice, where Taf12 can be inhibited in different tissues specifically at adult stage. In consistent to our in vitro observation, normal myeloid cells are not affected by Taf12 knockdown. Instead, the lymphoid lineage cells are affected while removing doxycycline can reverse the phenotype. To our surprise, all the other tissues that have been examined and show significant Taf12 knockdown do not have any defects.

At molecular level, we found that inhibition of Taf12 suppresses the transcription of core targets of Myb, but not other transcription factors, like Pu.1, Erg, Fli-1, which is in contrast to the pleiotropic effects of Brd4. Interestingly, it has been shown before that leukemia cells are hypersensitive to Myb inhibition compared to normal cells. Furthermore, similar to Taf12 mice, mice with hypomorphic level of Myb show defects in lymphoid but not myeloid cells. Taf12 is a small protein with only one histone fold domain (HFD), which can dimerize with the HFD of Taf4a within TFIID complex. We found that the histone fold dimer of Taf12 and Taf4a can physically interact with the transactivation domain (TAD) of Myb. Finally, we showed that the HFD of Taf4a can be utilized to target Taf12 to inhibit growth of leukemia cells. Similar to Taf12 shRNA, function of Myb is disturbed by overexpressing Taf4a HFD. More strikingly, overexpressing Taf4a HFD is sufficient to cause regression of leukemia in vivo.

Citation Format: Yali Xu, Joseph Milazzo, Yusuke Tarumoto, Chris Vakoc. Selection inhibition of acute myeloid leukemia by targeting Taf12 [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 1510. doi:10.1158/1538-7445.AM2017-1510