TAF1 plays a critical role in AML1-ETO driven leukemogenesis

AML1-ETO (AE) is a fusion transcription factor, generated by the t(8;21) translocation, that functions as a leukemia promoting oncogene. Here, we demonstrate that TATA-Box Binding Protein Associated Factor 1 (TAF1) associates with K43 acetylated AE and this association plays a pivotal role in the proliferation of AE-expressing acute myeloid leukemia (AML) cells. ChIP-sequencing indicates significant overlap of the TAF1 and AE binding sites. Knockdown of TAF1 alters the association of AE with chromatin, affecting of the expression of genes that are activated or repressed by AE. Furthermore, TAF1 is required for leukemic cell self-renewal and its reduction promotes the differentiation and apoptosis of AE+ AML cells, thereby impairing AE driven leukemogenesis. Together, our findings reveal a role of TAF1 in leukemogenesis and identify TAF1 as a potential therapeutic target for AE-expressing leukemia.

4) TAF1 depletion is associated with partial eviction of AML1-ETO from chromatin to nucleoplasm (Fig. 7a). How can this be reconciled with the fact that AML1-ETO can bind to DNA directly?
5) The authors may wish to discuss possible technical reasons why TAF1 shRNA depletion is nicely observed in total cell extracts and chromatin fractions, but not in cytoplasmic or nucleoplasmic fractions (Fig. 7a).
Reviewer #3 (Remarks to the Author): In Xu et al., the authors propose that TAF1 binds to acetylated K43 on the AML1-ETO fusion transcription factor, and that this recruitment is required for AML1-ETO mediated gene expression, recruitment to chromatin, and ultimately growth of AML1-ETO leukemia. The authors show that other cell types are less dependent on TAF1 suggesting a therapeutic index and that TAF1 KD in AML1-ETO fusion leukemic cells blocks tumor growth and increases overall survival.
Overall this is a nice manuscript with strong phenotypic data. These data provide additional support that targeting of the transcriptional apparatus can be selective especially in tumors that harbor underlying deregulation of transcription.
In its current form, the mechanistic conclusions of the manuscript need to be bolstered. For instance, despite the clear logic linking P300 acetylation of AML1-ETO to TAF1 recruitment, and the well studied role of TAF1 in promoting transcription initiation, the authors suggest a unique function for TAF1 in AML1-ETO where it is able to both activate and repress genes. No evidence is provided to suggest what exactly this function is other than a requirement for the K43 acetyl to TAF1 bromodomain interaction. Given TAF1's various non-PIC bridging functions (its two kinase domains, acetyltransferase activity, and ubiquitin activation/conjugation), the paper's claim that TAF1 is targetable in AE-AML would be significantly strengthened by demonstrating that one of these catalytic functions is actually critical to their proposed mechanism. Mutation studies -as opposed to large deletion studies-targeting the catalytic domains would be helpful. It is also possible that the large numbers of observed up regulated and down regulated genes upon TAF1 or AML1-ETO pertrubation result from a failure to properly interpret gene expression analysis that assume no global changes in gene expression. TAF1, as a general transcription factor is almost certain to have a global role in gene expression and this is masked by the author's analysis.
Finally, the ChIP-seq analysis is mostly limited to overlap analysis and could be further explored to support the mechanisms being proposed. For instance ChIP-seq of AML1-ETO after acute TAF1 depletion would help strengthen the claim that AML1-ETO chromatin recruitment is reduced upon TAF1 KD. The current analysis in Figure 7A needs to show that other chromatin associated proteins are not affected by TAF1 KD. 1

Reviewer #1 (Remarks to the Author):
This manuscript by Xu et al describes the function of TAF1 in AML-ETO induced myeloid leukemias. This is an interesting and important question, and the clearly written manuscript presents strong experimental evidence of the requirement of TAF1 for AML1-ETO induced AML.
The experimental models are well developed, data are of high quality, and results are thoughtfully interpreted.
While further questions can be asked to better understand the biochemical basis of the AE-TAF1 binding, its pharmacologic blockade, and the functional effects on gene expression and chromatin structure, I believe that these questions are beyond the scope of the current, already extensive manuscript. However, I would suggest expanding the discussion of the possible mechanisms and functions by which the AE-TAF1 interaction regulates gene expression, with attention to the following questions: Question 1.1 TAF1 is best recognized as a component of TFIID with functions at promoter regions, but as the authors note, a fraction of TAF1 is associated with chromatin outside of promoter regions, and AE itself is mostly localized away from gene promoters. Does this mean that TAF1-AE interaction is restricted to promoter-bound complexes? Answer 1. 1 We thank the reviewer for asking clarification regarding whether the TAF1-AE interaction is restricted to promoter bound complexes. Based on our ChIP-seq data, a majority of TAF1 peaks are present at promoter regions, while AE peaks are distributed at promoter regions, distal intergenic regions and within introns. We estimated the p values using a Monte Carlo simulation of shuffled peaks within either the TSS background or the nonTSS genomic background and found that TAF1 and AE shared binding sites are located at both TSS (p < 1.0e-5) and non-TSS regions (p < 1.0e-3), implying that the interaction of TAF1 and AE is not restricted to promoter regions. In addition, we performed KEGG analysis for those unique AE peaks which spectrometry. In addition to TAF1, we can only detect TAF15, but not other components of TFIID (Supplementary Figure 2d and 2e). It is well known that TAF7 forms a subcomplex with TAF1 in the TFIID complex, and we found TAF7 binding to the AE peptide in our 2011 paper. However, TAF7 was not identified in the recently conducted IP experiment. Studies using human embryonic stem cells have shown that the combination of TAFs in the PIC is both cell context and promoter dependent 1 . Thus, while we can not exclude that limitations in the affinity and specificity of the antibodies used or the co-immunoprecipitation protocol hampered our ability to identify other TAFs in the AE/TAF1 complex, it appears that few TAFs are engaged in the AE/TAF1 complex.
We also analyzed the AE/TAF1 overlapping peaks and the AE unique peaks at non-TSS regions and found that the AE/TAF1 overlapping peaks are adjacent to genes involved in "pathways in  5 function as specialized co-regulators in modulating transcription. We point out that CRISPR screening has found TAF1 is essential for the growth of MV4-11 cells, but not HL-60, MOLM-13 or OCI-AML3 4 . Although the definite roles of TAF1 in MLL-AF4 or CBFβ-MYH11 expressing AML remains unclear, these studies reveal the importance of TAFs in modulating cell behavior in subsets of acute leukemia. Our detailed answer to this query is contained in the sixth paragraph of the discussion. Question 1.4 TAF1 depletion is associated with partial eviction of AML1-ETO from chromatin to nucleoplasm (Fig. 7a) other proteins which could assist in the recruitment of AE to chromatin.

Question 1.5
The authors may wish to discuss possible technical reasons why TAF1 shRNA depletion is nicely observed in total cell extracts and chromatin fractions, but not in cytoplasmic or nucleoplasmic fractions (Fig. 7a).
Answer 1. 5 We would thank the reviewer for asking that we address this issue. In previous Figure   7a, the TAF1 western blot for cytoplasm and nucleoplasm fractions was overexposed. In the new Figure 7a, we added the shorter exposure of TAF1.
Although we could not detect a chromatin fraction of TAF1 in the shorter exposure, we can see the knockdown of TAF1 protein levels in both the cytoplasm and nucleoplasm of a shorter exposure. Since the majority of TAF1 is located in the cytoplasm and nucleoplasm, the influence of TAF1 KD will be easiest to visualize where there is the least amount of TAF1 (on chromatin).

Figure 7a
Reviewer #3 (Remarks to the Author): In Xu et al., the authors propose that TAF1 binds to acetylated K43 on the AML1-ETO fusion transcription factor, and that this recruitment is required for AML1-ETO mediated gene expression, recruitment to chromatin, and ultimately growth of AML1-ETO leukemia. The authors show that other cell types are less dependent on TAF1 suggesting a therapeutic index and that TAF1 KD in AML1-ETO fusion leukemic cells blocks tumor growth and increases overall survival.
Overall, this is a nice manuscript with strong phenotypic data. These data provide additional support that targeting of the transcriptional apparatus can be selective especially in tumors that harbor underlying deregulation of transcription. domain is not required for the chromatin association of AE. In the future, we will continue to investigate whether other TAF1 enzymatic domains or putative DNA binding modules are responsible for the AE recruitment or that TAF1 provides a platform for other proteins which could assist in the recruitment of AE onto chromatin.
3) EMSA assays: We performed EMSAs to identify whether specific domains in TAF1 were critical for modulating AE binding to DNA. However, AE/TAF1 complex bound DNA is too big to be detected by this assay, even using a 4.5% polyacrylamide gel.    (Supplementary Figure 4g) and found that the AE binding signal at AE/TAF1 cobound sites is significantly greater in Kasumi-1 cells with normal levels of TAF1 than in cells with reduced levels of TAF1 (p < 3.3e-17). Thus, both the subcellular fractionation assay (Figure 7a) and the ChIP-seq data show that TAF1 KD significantly reduces the recruitment of AE to chromatin.
To address the second concern, we have added the subcellular fractionation of PU.1 to Figure   7a. The data indicate that the association of PU.1 on chromatin is not affected by TAF1 KD, implying that the impact of AE binding at chromatin by TAF1 KD is specific.