CEBPE activation in PML-RARA cells by arsenic

To the editor:

In a recent perspective, Ablain and de The challenged the classic model of acute promyelocytic leukemia (APL), whereby differentiation-impairment in lineage-committed progenitors causes self-renewal, and instead proposed that APL arises from deregulation of stem cell self-renewal pathways in leukemia initiating cells, with the curative potential of arsenic (As₂O₃) relating to abrogation of these pathways.¹  However, differentiation syndrome occurs at a similar rate with As₂O₃ as with all-trans retinoic acid (ATRA), and furthermore, disseminated intravascular coagulation, the potentially fatal complication frequently triggered by cytotoxic treatment, is not typical.²  Hence, clinical experience suggests that important actions of As₂O₃ include terminating APL proliferation by differentiation. To evaluate this further, the APL cell line NB4 was treated with low concentrations of As₂O₃ which do not induce early apoptosis³  (Figure 1A; supplemental Figure 1, available on the Blood Web site; see the Supplemental Materials link at the top of the online article). These concentrations rapidly activated CEBPE, a key myeloid late-differentiation driver which is repressed by PML-RARA⁴  and which directly represses MYC⁵  (supplemental Figure 2; MYC is an oncogene product that drives myeloid progenitor proliferation⁶ ). As expected therefore, CEBPE activation was accompanied by MYC repression (Figure 1B). MXD1, a down-stream target of CEBPE⁷  and MYC-antagonist, was also activated, but subsequent to activation of CEBPE (Figure 1B). CEBPA, a key lineage-specifying transcription factor that drives CEBPE expression,⁸  is expressed at high levels in NB4 and primary APL cells (supplemental Figure 1B), providing a potential explanation for rapid activation of CEBPE after repressive actions of PML-RARA are inhibited by As₂O₃. Time-course changes in CEBPE and MYC protein levels reiterated the time-course changes in expression of mRNA (Figure 1C). The cyclin-dependent kinase inhibitor p27/CDKN1B mediates cell cycle exit by differentiation and was up-regulated by As₂O₃ at late-time points (Figure 1C); p21/CDKN1A mediates cell cycle exit by differentiation or apoptosis, and was also up-regulated (Figure 1C). Phosphorylation and up-regulation of the master regulator of apoptosis p53 was also observed, but as a late event that occurred subsequent to up-regulation of CEBPE and down-regulation of MYC (Figure 1C). Both ATRA and As₂O₃ were anti-proliferative, with the greater effect from As₂O₃ (Figure 1D). Others have shown that As₂O₃ 0.1-0.5μM induces morphologic differentiation in primary APL and NB4 cells,³  but not to the same extent as ATRA,³  an observation we recapitulated (Figure 1E).

In normal hematopoiesis, MYC-mediated proliferation in progenitors is self-limited by progressive maturation that activates key late-differentiation genes such as CEBPE and MXD1^4–7  (supplemental Figure 2). Thus, aberrant epigenetic repression of CEBPE and other key late-differentiation genes by PML-RARA⁴  (supplemental Figure 2), and as a consequence of cooperating genetic abnormalities (eg, UTX deletion), can potentially contribute to MYC-related self-renewal (proliferation at the same level of differentiation) in progenitors,^4,9,10  is consistent with the MYC up-regulation and amplification that is characteristic of APL,¹¹  and as shown here, is a molecular pathway targeted by As₂O₃ to potentially explain the anti–self-renewal effects of this drug. The curative potential of As₂O₃ compared with ATRA could reflect superior pharmacodynamics, that is, more effective degradation of oncogenic PML-RARA.^3,12  Hence, in important aspects, As₂O₃ could be a differentiation-therapy drug for APL, with a mechanism of action that supports the classic model of this disease, even if it does not recapitulate the full-spectrum morphologic differentiation that is produced by the more physiologic interactions of ATRA with the RARA moiety in PML-RARA.