In vivo models of subclonal oncogenesis and dependency in hematopoietic malignancy

Abstract

Cancer evolution is a multifaceted process leading to dysregulation of cellular expansion and differentiation through somatic mutations and epigenetic dysfunction. Clonal expansion and evolution is driven by cell-intrinsic and -extrinsic selective pressures, which can be captured with increasing resolution by single-cell and bulk DNA sequencing. Despite the extensive genomic alterations revealed in profiling studies, there remain limited experimental systems to model and perturb evolutionary processes. Here, we integrate multi-recombinase tools for reversible, sequential mutagenesis from premalignancy to leukemia. We demonstrate that inducible Flt3 mutations differentially cooperate with Dnmt3a, Idh2, and Npm1 mutant alleles, and that changing the order of mutations influences cellular and transcriptional landscapes. We next use a generalizable, reversible approach to demonstrate that mutation reversion results in rapid leukemic regression with distinct differentiation patterns depending upon co-occurring mutations. These studies provide a path to experimentally model sequential mutagenesis, investigate mechanisms of transformation and probe oncogenic dependency in disease evolution.

Keywords: genetically engineered mouse models; leukemia; oncogene dependency; sequential mutagenesis.

Figure 1:. Flp-recombinase inducible Flt3^ITD activation.

A, Schematic Flt3^Frt-ITD targeted to the endogenous Flt3 locus replacing wild type (WT) exons. Flp-recombination deletes wildtype (WT) exons and inverts mutant exons resulting in Flt3^ITD expression. B, Peripheral blood chimerism (%Cd45.2) in mice engrafted with WT (Cd45.1) and Flt3^Frt-ITD (Cd45.2) cells following treatment with Tamoxifen (TAM) (red) or control (black). (n=6–7 per group). C, Strip chart of white blood cell (WBC) K/μL in primary WT (black) or Rosa26:FlpoERT2 Flt3^Frt-ITD (red) mice at the indicated timepoint post-TAM. (n=10–19 per group). D, Mixed (1:1) Rosa26:FlpoERT2 Flt3^Frt-ITD Cd45.1 WT bone marrow was transplanted into Cd45.1 recipient mice. TAM was dosed 4 weeks post-engraftment and mice were euthanized 16 weeks post-TAM. Barplot depicts %long term hematopoietic stem cell (LT-HSC) of either the Cd45.1 or Cd45.2 compartment. (n=3–4 per group). E, Row normalized heatmap of RNA-sequencing in WT (grey) or Flt3^Frt-ITD (red) Lin⁻Sca1⁺Kit⁺ cells (LSKs) following TAM. F, Barplots depicting normalized counts WT or Flt3^Frt-ITD LSKs following TAM. Error bars reflect mean ± s.e.m.; and p-values generated by Student’s t-test (B-D). ** p ≤ 0.01 *** p ≤ 0.001 **** p ≤ 0.0001. See also Figure S1 and Tables S1 and S2.

Figure 2:. Flt3^ITD-driven models of leukemogenesis

A, Experimental schematic for simultaneous Npm1^Frt-c Flt3^Frt-ITD induction or staggered Dnmt3a^Lox-R878H (Cre mRNA) and Flt3^Frt-ITD (TAM) activation. B, Flt3^Frt-ITD (red), Dnmt3a^Lox-R878H-Flt3^Frt-ITD (orange), and Npm1^Frt-c-Flt3^Frt-ITD (green) bone marrow from (A) was transplanted into lethally irradiated hosts. Kaplan-Meier survival plot following TAM (n=5–12 per group). C, Bone marrow cells (Cd45.2) from symptomatic Npm1^Frt-c-Flt3^Frt-ITD (green), symptomatic Dnmt3a^Lox-R878H-Flt3^Frt-ITD (orange) or Flt3^Frt-ITD only (red) mutant mice >20w after TAM treatment from (b) were transplanted against Cd45.1 WT bone marrow. Peripheral blood chimerism (%Cd45.2) depicted for individual mice. (n=3–10 per group). D, RNA-sequencing from Npm1^Frt-c-Flt3^Frt-ITD or Dnmt3a^Lox-R878H-Flt3^Frt-ITD LSKs. GSEA gene ranks are depicted as a linechart for each Hallnark gene set. E, Normalized single sample gene set enrichment analysis (ssGSEA) scores (Figure S1M) from sorted LSKs from symptomatic Npm1^Frt-c -Flt3^Frt-ITD (green) or Dnmt3a^Lox-R878H-Flt3^Frt-ITD (orange) mice. F, Barplot depicting % Granulocyte-like (Cluster 1) or % monocyte-dendritic progenitors (MDP)-like (Cluster 11) immunophenotype characterized by CyTOF. (n=3–4 per group). Error bars reflect mean ± s.e.m.; and p-values generated by Fisher’s LSD test (E,F) * p ≤ 0.05 ** p ≤ 0.01 *** p ≤ 0.001 **** p ≤ 0.0001. See also Figure S2 and Tables S3 and S4.

Figure 3:. Dual-recombinase, reversible models of oncogene activation

A, Schematic depicting knockin Flt3^GL-ITD construct at the endogenous Flt3 locus, replacing exons 12–15. Flanking exons indicated in light grey, Lox2272 sites indicated by red triangles, heterotypic RoxP and Rox12 sites indicated in dark and light blue chevrons respectively. Inverted exons 12–15 are present at baseline with the W51 internal tandem duplication (ITD) encoded in exon 14 (orange). A’, B’ and C’ indicate position of primers used in (B) to detect Dre-inversion and Cre-deletion. B, Lin⁻ bone marrow from WT or Ubc:CreERT2 Flt3^GL-ITD was infected with MSCV:Dre-IRES-GFP retrovirus or mock infection. Cells were treated ± 4-hydroxytamoxifen (4-OHT) to activate Cre, and DNA was isolated 72 hours post treatment. PCR was used to evaluate Dre-inversion and Cre-deletion, visualized by capillary electrophoresis (Qiaxcel). C, Experimental schematic: Ubc:CreERT2-Flt3^GL-ITD mice were treated with TAM to activate Npm1^Lox-C. Eight weeks after TAM treatment, Lin⁻ HSPCs were isolated for Dre mRNA electroporation, and transplanted with support bone marrow. D, WBC (K/μL) from mice in (C). (n=5–6 per group). See also Figure S3.

Figure 4:. Triple-mutant, sequential models of clonal evolution towards acute myeloid leukemia

A, Schematic depicting sequential mutagenesis with Dnmt3a^Los-R878H activated at first transplant with Cre-mRNA, Npm1^Frt-c activated post-transplant with FlpoERT2 activation by TAM, and Flt3^GL-ITD activated with Dre-mRNA at secondary transplant. B, Peripheral blood monocyte count (K/μL) in secondary transplant. (n=5 per group). C, Heatmap depicting relative abundance of indicated cell populations (rows) for the indicated genotypes (columns) in bone marrow isolated from mice either 20 weeks post-transplant (WT, Dnmt3a^Los-R878H (D), Npm1^c-Frt (N), Dnmt3a^Los-R878H Npm1^c-Frt (DN)) or when symptomatic (Npm1^c-Frt 2° (N 2°), Dnmt3a^Los-R878HNpm1^c-Frt 2° (DN 2°), Npm1^c-Frt-Flt3^GL-ITD (NF), or Dnmt3a^Los-R878H-Npm1^c-Frt-Flt3^GL-ITD (DNF)). Secondary transplant indicated by 2°, following either Dre-mRNA or mock electroporation. D, %LSKs of Lin^- Cd45.2 in bone marrow from mice in (C). (n=3–4 per group). E, UMAP of protein abundance on AML patient samples depicting cluster number (top left, color indicate clusters), CD14 (top right), CD45RA (bottom left) or CD90 (bottom right) (n=4). For CD14, CD45RA and CD90, blue indicates low detection and red indicates high detection. F, Stacked barplot indicating fraction of cells present clusters from (E) for DNMT3A-FLT3 (DF), NPM1-FLT3 (NF) or DNMT3A-NPM1-FLT3 (DNF) mutant cells. p ≤ 0.0001 Chi-square test. Error bars reflect the mean ± s.e.m.; and p-values generated by Fisher’s LSD test (D) ** p ≤ 0.01. See also Figure S3 and S4.

Figure 5:. Chemically inducible orthogonal recombinase activation

A, Schematic depicting DreSTAPL-ODC: treatment with NS3 protease inhibitor results in complex formation of the split N and C-terminus of Dre. Without NS3-inhibition, proteolytic cleavage reveals the ODC degron causing degradation of the N-terminus. B, Lin⁻ HSPCs from fluorescent reporter mice (Lox, Ai14), or (Rox-RLTG) were infected with GFP+ retroviruses encoding CreER or DreSTAPL-ODC. Cultures were treated with DMSO, 4-OHT (400nM), or grazoprevir (GZV) (10μM). Barplot depicts recombination in infected population (%TdTomato+ of GFP+ cells). (n=3 per group). C, DreSTAPL-ODC knockin mice were crossed to the Rosa26:RLTG reporter. Recombination was assessed by %TdTomato+ in the blood 2 weeks after GZV treatment. (n=3–5 per group). D, Experimental schematic: Lin⁻ HSPCs from Dnmt3a^Los-R878H-Npm1^c-Frt-Flt3^GL-ITD Rosa26:FlpoERT2; Vav1:DreSTAPL-ODC mice were electroporated with Cre mRNA and transplanted into CD45.1 hosts. Eight weeks later, mice were treated with TAM or GZV to activate Npm1^c-Frt or Flt3^GL-ITD respectively; mice were treated with the alternative ligand 8 weeks later. E, Kaplan-Meier survival curve on experimental schematic in (D). DNF (red) refers to the mutation order of Dnmt3a, Npm1, then Flt3, DFN refers to Dnmt3a, Flt3, then Npm1. (n=4 per group) F, %LSKs of Cd45.2 bone marrow from symptomatic mice depicted in (E) (n=4 per group). G, UMAP of scRNA-seq from either DNF or DFN whole bone marrow in (D) (n=2). Color indicates cell type. H, Barplot indicating pathway analysis from scRNA-seq data in (G) on all HSPC populations pooled together. Y-axis indicates gene set, X-axis indicates −log10(pvalue) *sign(fold change). Positive scores indicate enrichment in DNF and negative scores indicates enrichment in DFN. Error bars reflect the mean ± s.e.m.; and p-values generated by Student’s T-test (F) * p ≤ 0.05. See also Figure S5.

Figure 6:. Reversible Flt3 mutagenesis in acute myeloid leukemia

A, Experimental schema for primary mice with Ubc:CreERT2, Vav1:DreSTAPL-ODC, Flt3^GL-ITD and either Idh2^Lox-R140Q or Npm1^Lox-C. Schema depicts dosing schedule of either TAM or GZV to activate Cre and Dre. B, Peripheral blood monocyte (K/μL) counts in Flt3^GL-ITD (red), Idh2^Lox-R140Q Flt3^GL-ITD (pink) or Npm1^Lox-C Flt3^GL-ITD (green) mice following GZV. (n=4–13 per group). For (C-I) Npm1^Lox-C Flt3^GL-ITD (red) or Idh2^Lox-R140Q Flt3^GL-ITD (blue) from (B) were engrafted into lethally irradiated recipients with Cd45.1 WT cells. Mice were monitored for disease development (Npm1^Lox-C Flt3^GL-ITD 4 weeks, or Idh2^Lox-R140Q Flt3^GL-ITD 10 weeks), and then euthanized following 7 days of TAM treatment (C-H). C, WBC (K/μL) before and after TAM treatment. (n=3–7 per group) D, Spleen mass in control and TAM treated mice. (n=3–7 per group). E, Bone marrow H&E from either control mice or after 7 days of TAM (400x, scale bar 50μm). F, Dot chart depicting %LSKs of Cd45.2 cells in bone marrow from control and TAM treated mice. (n=3–4 per group). G, RNA-sequencing from Lin^- HSPCs purified from 7-day TAM treated Npm1^Lox-C-Flt3^GL-ITD or control mice. Gene ranks from GSEA are depicted as a line chart for the indicated Hallmark gene set. H, Immunohistochemical staining on bone marrow for Ki67 (left) and phopsho-ERK1/2 (right) for untreated (top) and 7-day TAM treated (bottom) Npm1^Lox-C-Flt3^GL-ITD mice. I, Kaplan-Meier survival curve of mice engrafted with Cd45.1 WT cells and bone marrow cells from symptomatic Npm1^Lox-C-Flt3^GL-ITD mice from Figure (B). Mice were treated with control chow (grey). placed on a TAM-chow diet (red) for 4 weeks, or treated with gilteritinib for 4 weeks (green); treatment was withdrawn at the indicated time point (n=7–10). Error bars reflect the mean ± s.e.m.; and p-values are calculated by Dunn’s test (C,D), Fisher’s LSD test (F) and Log-Rank test (I). * p ≤ 0.05 ** p ≤ 0.01 *** p ≤ 0.001 **** p ≤ 0.0001. See also Figure S6 and Table S5 and S6.