HOXA9 promotes MYC-mediated leukemogenesis by maintaining gene expression for multiple anti-apoptotic pathways

  1. Ryo Miyamoto
  2. Akinori Kanai
  3. Hiroshi Okuda
  4. Yosuke Komata
  5. Satoshi Takahashi
  6. Hirotaka Matsui
  7. Toshiya Inaba
  8. Akihiko Yokoyama

  • Curated by eLife

    eLife logo

    Evaluation Summary:

    This manuscript is of potential interest to experimental haematologists studying initiation and maintenance factors in leukaemia. Overall, the study is well designed and the data is clearly presented. However, in some places the analysis lacks depth and technological sophistication, and the novel insights are limited without additional experimentation.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

HOXA9 is often highly expressed in leukemias. However, its precise roles in leukemogenesis remain elusive. Here, we show that HOXA9 maintains gene expression for multiple anti-apoptotic pathways to promote leukemogenesis. In MLL fusion-mediated leukemia, MLL fusion directly activates the expression of MYC and HOXA9. Combined expression of MYC and HOXA9 induced leukemia, whereas single gene transduction of either did not, indicating a synergy between MYC and HOXA9. HOXA9 sustained expression of the genes implicated in the hematopoietic precursor identity when expressed in hematopoietic precursors, but did not reactivate it once silenced. Among the HOXA9 target genes, BCL2 and SOX4 synergistically induced leukemia with MYC . Not only BCL2, but also SOX4 suppressed apoptosis, indicating that multiple anti-apoptotic pathways underlie cooperative leukemogenesis by HOXA9 and MYC. These results demonstrate that HOXA9 is a crucial transcriptional maintenance factor that promotes MYC-mediated leukemogenesis, potentially explaining why HOXA9 is highly expressed in many leukemias.

Article activity feed

  1. Version published to 10.7554/elife.64148 on eLife
  2. eLife

    Author Response:

    Reviewer #1:

    Miyamoto and colleagues study the role of various oncogenes including MYC, HOXA9 and SOX4 in transformation of haematopoietic cells in vitro and in vivo. The authors analyze gene expression profiles and characterize leukemogenesis and cell survival resulting from manipulation of MLL-AF10 expression in myeloid leukemias. The experiments largely utilise ectopic over-expression of transgenes; hence results comparing relative "potency" of individual genes must be interpreted with caution due to supraphysiological levels of expression.

    Specific comments:

    1. In Figure 1A, the authors attempt to identify direct target genes of the MLL fusion protein MLL-ENL by performing ChIPseq using an anti-MLL antibody. Whether or not the signal can be attributed to MLL-ENL or wild-type MLL is unclear. Furthermore, genome-wide MLL-occupancy patterns are not shown. The work would be stronger if the authors could reconcile current data with other publicly available datasets for MLL or MLL-fusion protein occupancy in comparable contexts.

    We appreciate the inputs from the Editor and reviewers. Here we provide point-by-point response to the comments.

    We performed ChIP-seq analysis of HB1119 cells in which wildtype MLL, but not MLL-ENL, was specifically knocked down by shRNA (Figure 1A, Figure 1-figure supplement-1B,C), as shown In our previous publication (Okuda et al., 2017). Depletion of wildtype MLL did not affect the ChIP signals. Thus, we concluded that most of the MLL ChIP signals can be attributed to MLL-ENL. These data was presented in our previous report (Okuda et al., 2017) and partially adopted in the revised manuscript. MLL and MLL fusion proteins localize near transcription start sites (TSSs)( Figure 1-figure supplement-1C) because MLL has a CXXC domain that recognizes unmethylated CpGs (Okuda et al., 2014). Such TSS-centric localization of MLL is observed in many other non-MLL-rearranged cell lines such as HEK293T (embryonic kidney) and REH (Leukemia) cells (Miyamoto et al., 2020), in addition to HB1119 cells (MLL-rearranged leukemia cells)(Okuda et al., 2017). We mentioned this in the revised manuscript.

    1. It would appear (based on capitalisation), that the authors are over-expressing human transgenes in mouse cells. This is not necessarily a concern, but should be considered when interpreting the data. Likewise, whether the primers used for qPCR are detecting expression of the transgenes, the endogenous genes or both is important (for some of the figures such as Fig. 1C there seems to be a mix e.g. Myc vs HoxA9/HOXA9).

    We used human transgenes in the presented experiments. The qPCR probes for mouse Hoxa9 and Meis1 detected human HOXA9 and MEIS1, respectively. Hence, we described HOXA9/Hoxa9 and MEIS1/Meis1 to clearly indicate that these probes detect both human and mouse genes. The qPCR probe for mouse endogenous Myc did not detect the human MYC transgene. The samples producing qPCR signals for both endogenous murine genes and exogenous human transgenes are highlighted by # and faded color (Figure 1C).

    1. Most of the in vivo transplantation experiments have not been performed using fluorescent reporters or congenic recipients that would enable identification of donor-derived cells. Differences between the groups could be attributed to differential engraftment, or potentially even immune rejection (assuming ectopic expression of human transgenes in an immune-competent context). Disease features in recipient mice (beyond survival) are also not shown and expression of transgenes at end-point not confirmed.

    As for the possibility of immune rejection of the cells expressing human transgenes:

    As shown in Figure 3D, the mouse Myc gene was tested in addition to human MYC and did not induce leukemia in vivo, supporting that the enhanced MYC function alone is insufficient to induce leukemia under these experimental conditions. It has been shown that the mouse Hoxa9 gene is also a weak oncogene in vivo by Kroon et al. whereas it induced leukemia as a combination with Meis1(Kroon et al., 1998). The human HOXA9 transgene phenocopied mouse Hoxa9 in our assays. These results did not support the possibility of immune rejection of the human transgene-expressing cells. We mentioned that in the revised manuscript.

    As for the possibility of different engraftment:

    We did not mean to exclude the possibility of different engraftment as the reason of not inducing leukemia by a certain oncogene. It is likely that HOXA9 promotes engraftment of MYC-transduced cells by conferring survival advantage with BCL2/SOX4-mediated anti-apoptotic properties. It is possible that HOXA9 mediates additional functions to promote engraftment other than providing anti-apoptotic properties. However, we chose to focus on the HOXA9-mediated anti-apoptotic functions in this paper.

    As for the disease features:

    We have added the expression and immune phenotype data in Figure 4-figure supplement-3B and Figure 6-figure supplement-1B.

    In contrast to MLL-AF10 and HOXA9 containing gene sets (HOXA9-MEIS1, HOXA9-MYC), MYC-BCL2 induced lymphoid leukemia in vivo, consistent with the previous report (Luo et al., 2005). We speculate that HOXA9 and SOX4 are more functional in the myeloid lineage, while BCL2 functions more efficiently in the lymphoid lineage than in the myeloid lineage. Consequently, the MYC-BCL2 combination tended to induce lymphoid leukemia.

    As for the expression of the transgene at end-point:

    Regarding the expression of the transgenes in Figure 3D and 4E, we have provided the RT-qPCR data for the transgenes in Figure 4-figure supplement-3B. Regarding the expression of the transgenes in Figure 6B, the protein expression of the transgenes is shown in Figure 6-figure supplement-1A. Regarding the expression of the transgenes in Figure 7A, B, we have provided the RT-qPCR data for the transgenes in Figure 7-figure supplement 2A.

    1. The authors propose that the data in Figure 5B confirms direct regulation of Bcl2, Sox4 and Igf1 by HOXA9. However, the regulation could also be indirect e.g. HOXA9 could regulate a transcription factor that regulates those genes, or HOXA9 depletion could induce differentiation that may result in downregulation of those genes.

    The regulatory mechanisms by which HOXA9 controls the expression of its target genes are of great interest. Indeed, the expression of BCL2 and/or SOX4 could be regulated indirectly by HOXA9. We changed the wording by removing the word “direct” in the revised manuscript.

    Reviewer #2:

    The manuscript of Miyamoto et al. describes the synergistic function between HOXA9 and MYC downstream of MLL fusions in myeloid leukemogenesis. They show that MLL-AF10 expression up-regulates both HOXA9 and MYC expression. Gene expression profiles of immortalized cells (IC) indicate that distinct genetic pathways are driven by HOXA9 and MYC. Cooperativity in in vivo leukemogenesis between HOXA9 and MYC is shown. Apoptotic cell death is increased in MYC-IC and it is cancelled by overexpression of BCL2 or SOX4 that are up-regulated in HOXA9-IC but not in MYC-IC, suggesting that these genes are downstream of HOXA9 and responsible for cooperativity between MYC and HOXA9. Moreover, deletion of BCL2 or SOX4 inhibited MLL-AF10- or HOXA9/MEIS1-induced leukemogenesis. This study is well designed and experimental results are clearly presented. These results provide useful information for our understanding the mechanisms of HOX-associated leukemogenesis.

    We appreciate the comments from the reviewer and hope our study is useful for the understanding of leukemogenesis.

  3. eLife

    Evaluation Summary:

    This manuscript is of potential interest to experimental haematologists studying initiation and maintenance factors in leukaemia. Overall, the study is well designed and the data is clearly presented. However, in some places the analysis lacks depth and technological sophistication, and the novel insights are limited without additional experimentation.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)

  4. eLife

    Reviewer #1 (Public Review):
    Miyamoto and colleagues study the role of various oncogenes including MYC, HOXA9 and SOX4 in transformation of haematopoietic cells in vitro and in vivo. The authors analyze gene expression profiles and characterize leukemogenesis and cell survival resulting from manipulation of MLL-AF10 expression in myeloid leukemias. The experiments largely utilise ectopic over-expression of transgenes; hence results comparing relative "potency" of individual genes must be interpreted with caution due to supraphysiological levels of expression.

    Specific comments:

    1. In Figure 1A, the authors attempt to identify direct target genes of the MLL fusion protein MLL-ENL by performing ChIPseq using an anti-MLL antibody. Whether or not the signal can be attributed to MLL-ENL or wild-type MLL is unclear. Furthermore, genome-wide MLL-occupancy patterns are not shown. The work would be stronger if the authors could reconcile current data with other publicly available datasets for MLL or MLL-fusion protein occupancy in comparable contexts.

    2. It would appear (based on capitalisation), that the authors are over-expressing human transgenes in mouse cells. This is not necessarily a concern, but should be considered when interpreting the data. Likewise, whether the primers used for qPCR are detecting expression of the transgenes, the endogenous genes or both is important (for some of the figures such as Fig. 1C there seems to be a mix e.g. Myc vs HoxA9/HOXA9).

    3. Most of the in vivo transplantation experiments have not been performed using fluorescent reporters or congenic recipients that would enable identification of donor-derived cells. Differences between the groups could be attributed to differential engraftment, or potentially even immune rejection (assuming ectopic expression of human transgenes in an immune-competent context). Disease features in recipient mice (beyond survival) are also not shown and expression of transgenes at end-point not confirmed.

    4. The authors propose that the data in Figure 5B confirms direct regulation of Bcl2, Sox4 and Igf1 by HOXA9. However, the regulation could also be indirect e.g. HOXA9 could regulate a transcription factor that regulates those genes, or HOXA9 depletion could induce differentiation that may result in downregulation of those genes.

  5. eLife

    Reviewer #2 (Public Review):

    The manuscript of Miyamoto et al. describes the synergistic function between HOXA9 and MYC downstream of MLL fusions in myeloid leukemogenesis. They show that MLL-AF10 expression up-regulates both HOXA9 and MYC expression. Gene expression profiles of immortalized cells (IC) indicate that distinct genetic pathways are driven by HOXA9 and MYC. Cooperativity in in vivo leukemogenesis between HOXA9 and MYC is shown. Apoptotic cell death is increased in MYC-IC and it is cancelled by overexpression of BCL2 or SOX4 that are up-regulated in HOXA9-IC but not in MYC-IC, suggesting that these genes are downstream of HOXA9 and responsible for cooperativity between MYC and HOXA9. Moreover, deletion of BCL2 or SOX4 inhibited MLL-AF10- or HOXA9/MEIS1-induced leukemogenesis. This study is well designed and experimental results are clearly presented. These results provide useful information for our understanding the mechanisms of HOX-associated leukemogenesis.

  6. Version published to 10.1101/2020.10.20.347765v1 on bioRxiv