Treatment of a TP53 isogenic AML Cell Line Model with Decitabine and All Trans-Retinoic Acid (ATRA) Targets MYC for Antileukemic Activity in a p53-Independent Fashion

Background

The DECIDER randomized phase II trial showed that the addition of ATRA to decitabine (DAC) significantly improves survival and overall response rate (ORR) in newly diagnosed elderly AML patients (Lübbert et. al., J Clin Oncol. 2020), with a higher ORR in TP53-mutated (mostly heterozygous) compared to TP53 wildtype. Mechanisms of this in vivo synergism are under study, involving changes in global transcriptome and chromatin accessibility in both TP53wt and TP53mut AML cell lines (Meier et. al., Blood Cancer J. 2022). To further unravel the mechanism of action of DAC+ATRA (D+A) in AML, these studies are now extended to an isogenic TP53 model representing wt vs knock-out.

Methods

TP53 isogenic MOLM13 cells ( TP53+/+, TP53-/-, Boettcher et al., Science 2019) were treated with 100nM DAC by 3 daily pulses, 250nM ATRA on day 4, and studied for proliferation (EdU), by RNA-sequencing (RNA-seq) and immunofluorescence (IF) at 24, 48 and 72h post ATRA addition. RNA-seq data were analyzed after pre-processing (Cutadapt) and annotation (RNA star) by DEseq2 (adj p-value <0.05, fold change >2), functional enrichment analysis by pyGSEA on ‘50 Hallmark pathways’ using T-statistics on differentially expressed genes (DEGs).

Results

D+A reduced cell proliferation and increased apoptosis by 88% and 44% in TP53 +/+ cells, by 50% and 15% in TP53 -/- cells relative to control at 48h. By RNA-seq, the global transcriptomic profile of D+A was distinct from single-agent treatment at all 3 timepoints, as the majority of the DEGs were unique to the combination. D+A had 2250 unique DEGs (uDEGs) in TP53 +/+ cells (compared to 192 uDEGs for DAC, 249 uDEGs for ATRA); D+A had 1209 uDEGs in TP53 -/- cells (compared to 328 uDEGs with DAC and 197 uDEGs with ATRA) at 72h. By functional enrichment analysis, we found that D+A requires TP53 for antileukemic activity, as most proliferative pathways (e.g. hallmark G2M checkpoint, hallmark E2F targets) were depleted exclusively in a p53-dependent manner. Interestingly, D+A was able to induce an interferon response and deplete MYC target genes and mTOR signalling independent of p53.

As MYC is a known target depleted by D+A, we investigated MYC (by IF) as an additional target for antileukemic activity. We found that global MYC levels were significantly depleted upon ATRA addition to DAC. ATRA depleted MYC levels independent of the TP53 status, while DAC depleted MYC in a p53-dependent manner (Fig. 1A). DAC was able to induce MYC ^T58 phosphorylation in both TP53 +/+ and TP53 -/- cells (Fig. 1B) but was only able to reduce MYC global levels in TP53+/+ cells, pointing towards a possible p53-dependent perturbation of proteasomal activity.

FBXW7 (SCF complex), responsible for ubiquitinating pMYC ^T58, and USP28 (deubiquitinase) expression levels analysis revealed that USP28 is depleted upon DAC treatment, suggesting that FBXW7 should be able to mark pMYC ^T58 for ubiquitination and degradation, and further investigation is ongoing to determine proteasomal perturbation and how p53 loss is associated with it.

Conclusions

Our results represent a mechanistic exploration of how the addition of ATRA was able to enhance DAC's antileukemic activity in two polar opposite TP53 states (intact vs. null). D+A treatment induces a unique molecular signature compared to either drug and targets p53 for an antileukemic effect, but not exclusively, as MYC depletion as well as interferon induction is independent of p53. We also identified p53-dependent perturbation of proteasomal activity, as high amounts of pMYC ^T58 accumulate in TP53-/- cells upon D+A treatment, and hypothesize that pMYC ^T58 accumulation results in development of resistance to D+A. These studies will be extended to the most clinically relevant heterozygous TP53 hotspot mutations.