A gain-of-function p53 mutant synergizes with oncogenic NRAS to promote acute myeloid leukemia in mice

We previously demonstrated that a subset of acute myeloid leukemia (AML) patients with concurrent RAS pathway and TP53 mutations have an extremely poor prognosis and that most of these TP53 mutations are missense mutations. Here, we report that, in contrast to the mixed AML and T cell malignancy that developed in NrasG12D/+ p53–/– (NP–/–) mice, NrasG12D/+ p53R172H/+ (NPmut) mice rapidly developed inflammation-associated AML. Under the inflammatory conditions, NPmut hematopoietic stem and progenitor cells (HSPCs) displayed imbalanced myelopoiesis and lymphopoiesis and mostly normal cell proliferation despite MEK/ERK hyperactivation. RNA-Seq analysis revealed that oncogenic NRAS signaling and mutant p53 synergized to establish an NPmut-AML transcriptome distinct from that of NP–/– cells. The NPmut-AML transcriptome showed GATA2 downregulation and elevated the expression of inflammatory genes, including those linked to NF-κB signaling. NF-κB was also upregulated in human NRAS TP53 AML. Exogenous expression of GATA2 in human NPmut KY821 AML cells downregulated inflammatory gene expression. Mouse and human NPmut AML cells were sensitive to MEK and NF-κB inhibition in vitro. The proteasome inhibitor bortezomib stabilized the NF-κB–inhibitory protein IκBα, reduced inflammatory gene expression, and potentiated the survival benefit of a MEK inhibitor in NPmut mice. Our study demonstrates that a p53 structural mutant synergized with oncogenic NRAS to promote AML through mechanisms distinct from p53 loss.

Colony assay and replating assay 5 x 10 4 bone marrow cells were plated in duplicate in semisolid medium MethoCult M3234 (StemCell Technologies) supplemented with mGM-CSF or mIL-3 (Peprotech, Rocky Hill, NJ) according to the manufacture's protocol.The colonies were counted after 7 to 10 days in culture.Then colonies were harvested to repeat the same procedure for serial replating.

RNA-Seq and data analysis
Total RNAs were isolated from 50,000 sorted Lin -cKit + bone marrow cells of age-matched control (n=3), p53 R172H/+ (n=3), Nras G12D/+ (n=3) and moribund NP mut (n=4) mice using RNeasy Micro Kit (Qiagen).RNA-Seq libraries were prepared using SMARTer® Stranded Total RNA-Seq Kit v1 -Pico Input Mammalian (Takara Bio USA/Clontech).Sequencing was performed on an Illumina HiSeq 4000 system at the NUSeq Core facility.Reads from each sample were aligned to the Mus musculus GRCm38.p6genome using STAR v2.6.1d 15.During alignment, the first three nucleotides (the switching oligonucleotides) were clipped from each read as recommended by Takara Bio USA/Clontech.Reads that aligned to the sense strand of genes within the Gencode release M24 genome annotation were then counted.Read counts for each sample were then imported into R v3.6.1 (https://www.Rproject.org/)and normalized using DESeq2 v1.26.0 16.Wald tests within DESeq2 were conducted to assess differential gene expression between groups and the ashr method was used to shrink log2 fold-change values 17.Pre-ranked GSEA was conducted using clusterProfiler v3.14.3 18 using the Wald test statistics as the ranking values and the fgsea method.Gene sets from the Broad Institute's Molecular Signatures Database sets were used.P values from differential gene expression analyses and GSEA were corrected for multiple testing using the Benjamini-Hochberg method.

Quantification of inflammatory cytokines
Serum was collected from 8-weeks old primary NP mut , age matched control, and NP mut recipient mice (8 weeks post transplantation) and diluted 3-fold before assaying with MSD Cytokine Assays Proinflammatory Panel 1 V-Plex (mouse) kits (Meso Scale Discovery, Cat # K15048D).A panel of 10 pro-inflammatory cytokines were measured with the diluted serum sample by first pre-incubating with detection antibodies conjugated with electro chemiluminescent labels, then adding to a plate pre-coated with capture antibodies on independent spots and read on an MSD MESO QUICKPLEX SQ 120 multiplex cytokine plate reader to quantify cytokine levels.

Secondary bone marrow transplantation and drug treatment
Sub-lethally irradiated (4.0 Gy using an X-Rad 320 irradiator, Precision X-Ray) CD45.1 + mice were transplanted with 0.5x10 6 BM cells from moribund NP mut primary recipients with AML phenotype.For treatment with NFB and MEK inhibitors, secondary recipients were randomly separated into four groups at 2 weeks after transplantation and treated with vehicle, bortezomib (0.5mg/kg, every Monday and Thursday), trametinib (0.5mg/kg, once daily), or combined bortezomib and trametinib via oral gavage until the moribund stage.

Figure S3 .
Figure S3.Giemsa staining of blood smears from moribund NP mut mice and age-

Figure S6 .
Figure S6.Reduced lymphopoiesis in NP mut mice.Analyses were performed in moribund NP mut mice and age-matched control (Con), Nras G12D , and p53 mut mice.Quantification of

Figure S7 .
Figure S7.Transcriptional levels of positive and negative regulators of RAS signaling pathway.(A) Volcano plot of positive regulators in NP mut vs control HSPCs (upregulated

Figure S8 .
Figure S8.Btz and combo treatment downregulate inflammation gene expression in NP mut leukemia cells.NP mut cells were transplanted into sublethally irradiated CD45.1 +

Figure
Figure S9.p53 R172H confers increased BM reconstitution through mechanisms distinct from p53 R248W .(A) Total BM cells were isolated from control and p53 mut mice 1-week