PML-RARα: changing myeloid networks

In this issue of Blood, Tan et al identify novel cooperative effects of the promyelocytic leukemia/retinoic acid receptor-α (PML/RARα) chimeric fusion and the growth factor independent-1 (GFI1) transcriptional repressor that underlie aberrant cell fate decisions characteristic of acute promyelocytic leukemia (APL).¹ 

Whether building a home, planning a local social event (physically distanced, of course), or causing leukemia, you’ll be more successful if you work with your neighbors. This is an important message reinforced by the Tan et al article. Investigators at the Shanghai Institute of Hematology have led advances in treating APL that have transformed this subtype of acute myeloid leukemia (AML) into a curable disease. The article by Tan et al increases our understanding of how the PML/RARα chimeric fusion protein remodels transcriptional networks to both impair maturation and to cause the anomalous phenotype of leukemic promyelocytes. Despite several decades of work, we lack a complete understanding of how genetic events that underlie APL result in the cell biological changes expressed as the phenotype particular to APL cells. The article by Tan et al advances this understanding by demonstrating that transactivation of genes by PML/RARα is central to disease pathogenesis. Furthermore, the article shows that regions of chromatin at which myeloid transcription factors cooperatively control cell fate are among the critical areas dysregulated by PML/RARα.

An important element of the study was the generation of an antibody that specifically recognizes the peptide fusion site where PML and RARα are joined in the NB4 APL cell line. This enabled the investigators to perform chromatin immunoprecipitation sequencing (ChIP-seq) analysis that focused specifically on chromatin locations at which the PML/RARα fusion protein was located. By associating the sites found through ChIP-seq with nearby genes showing altered RNA levels after knockdown of PML/RARα, the authors identified 787 genes as candidates for being dysregulated directly by PML/RARα. Surprisingly, given previous evidence that PML/RARα is a strong transcriptional repressor, more than half these genes seemed to be activated by the chimeric fusion. The authors were aware that RNA changes after knockdown could reflect partial differentiation of their NB4 cells (rather than being the result of a direct effect of the fusion), and they therefore further explored whether direct transcriptional upregulation by PML/RARα was in fact occurring, and whether this upregulation might be key to the leukemic phenotype. It is fascinating that 1 of the genes was GFI1 (the expression of which seemed induced by PML/RARα), a gene that is mutated in some patients with severe congenital neutropenia and that is a mediator of neutrophilic cell fate.^2,3  The investigators went on to demonstrate a close relationship between PML/RARα and GFI1. Indeed, the authors used an array of methods to show that activation of GFI1 transcription by PML/RARα is necessary for APL. Their methods included chromatin conformation capture, disruption of PML-RARα binding sites, impairment of PML-RARα oligomerization, and both in vitro and in vivo manipulations of leukemic cells. They also showed that PML/RARα and GFI1 are commonly located together on chromatin, particularly at sites that are predicted to bind additional myeloid transcription factors, including CEBPα and RUNX1. A small but important finding of their work explains why APL expresses exceptionally high levels of myeloperoxidase (which was previously used by pathologists as an identifying feature of the disease): PML/RARα and GFI1 work as neighbors to drive overexpression of the MPO gene.

In light of previous studies indicating that PML/RARα can contribute to leukemia through effects on and interactions with other myeloid transcription factors (eg, CEBPα, PU.1, IRF8),^4-9  the Tan et al study suggests that PML-RARα initiates leukemia by dysregulating the key enhancer regions that govern myeloid cell fate (referred to in the article as super-enhancers). Conversely, the combination of arsenic and all-trans retinoic acid is effective as therapy because it destroys the chimeric protein, which re-enables normal regulatory networks and eliminates the leukemic stem cells for this disease.¹⁰  The authors explore the role of coactivators, corepressors, and histone modification in these changes, but much remains to be done to understand the biochemistry through which PML-RARα may activate abnormal gene expression. Nonetheless, an interesting implication of the work is that targeting GFI1 interactions could potentially be of therapeutic value in other forms of AML.

That both retinoic acid and arsenic trioxide were discovered to be effective treatments for APL has led to numerous discoveries (by scientists and physicians working together across the globe) that have illuminated much of the biological basis for APL leukemogenesis and its reversal. The work of Tan et al is one more step in the work of good neighbors to understand disease and to alleviate human suffering.