‘All models are wrong, but some are useful’ was a particular phrase that was burned into my mind during my PhD developing methods to model ‘omics time course series. In my field of bioinformatics and applied statistics, this aphorism keeps me on my toes: models can generate answers with great confidence, but does this observation represent the truth? Undeniably, the use of Cre recombinase murine models that allow expression of mutated genes in an inducible or cell type-specific way has revolutionized biomedical research. But everyone working with these models knows that it has its quirks, such as off-target effects, toxicity and lack of fidelity (1). Furthermore, these quirks are not consistent across cell type or induction method and can be hard to spot. And what if these quirks of Cre independently generated the phenotype you were trying to model?

We set out to explore the contribution of FLT3 mutations in acute myeloid leukaemia (AML). AML is an aggressive blood cancer that arises from mutated hematopoietic stem cells. AML is a genetically heterogeneous disease, but one of the most common driver and cooperating oncogenic mutations, occurring in about 30% of patients, involves an internal tandem duplication (ITD) in the juxtamembrane domain of the tyrosine kinase, FLT3. FLT3-ITD is infamous for its adverse prognostic impact in AML (2) and is a frequent driver of relapsed disease (3) and thus, it is important to model this mutation to understand its function in disease progression and therapeutic resistance alone and in combination with cooperating mutations.

Murine models expressing FLT3-ITD constitutively from the endogenous locus have been developed (4, 5). Interestingly, FLT3-ITD mice develop a myeloproliferative disease when homozygous, but only develop AML when crossed with another oncogene-expressing mouse model (6-11).

What do our findings show?

While generating a murine model to study cooperating mutations with FLT3-ITD to drive leukemia we crossed transgenic, FLT3-ITD homozygous mice (5) with mice expressing an inducible loxP targeting Cre recombinase. During this process, Prof. Florian Heidel's group discovered first that induced Cre:FLT3-ITD homozygous mice developed a fully penetrant AML in the absence of a conditional allele and we in Prof. Steven Lane's group confirmed the findings. So we combined forces to embark on a seemingly never ending quest to solve this Cre quirk.

We asked whether this was due to the induction method. AML occurred with both Scl:Cre^ERT induced by tamoxifen and (the notoriously leaky) Mx1:Cre induced by polyinosinic:polycytidylic acid (pIpC), excluding the induction method as the cause of the phenotype. We also asked whether this represented spontaneous acquisition of a cooperating mutation and subsequent monoclonal expansion. By using the R26R-Confetti Cre randomly expressing fluorescent markers, GFP, YFP, CFP or RFP upon the induction of Cre recombinase we determined that the AML produced was polyclonal. Further, whole exome sequencing excluded acquired coding mutations as cooperating leukemic drivers.

We took a closer look at the Flt3 locus. When generating this model, a loxP flanked neomycin resistance (neo-r) cassette was inserted in Flt3 intron 15 and we confirmed its excision mediated by Cre recombinase and concluded that this has possible implications for gene regulation and expression. Supporting this, we found shared gene expression signatures concordant across different Cre genotypes in Kit+ populations, characterised by a loss of myeloid differentiation and enrichment for stem cell regulatory pathways with differential expression of key myeloid transcription factors. Consistent with Cre excision of the neo-r cassette in the other murine FLT3-ITD/ITD model (4) we found upregulation of Flt3 gene and protein expression but only in the Mx1:Cre model. We did however observe suppressor of cytokines (Socs1), a downstream target of Flt3, upregulated and great consistency in different Cres and cell types’ suppression of inflammatory milieu compared to the FLT3-ITD/ITD mice that develop an MPN like phenotype and are characterised by an aberrant inflammatory milieu and disrupted cytokine signaling.

Using ATACSeq we observed an overall increase in chromatin accessibility and loss of genomic regulatory regions associated with myeloid differentiation and PU.1 binding sites consistent with phenotype and gene expression observed. However, we could not find any evidence that this was due to Cre binding to chromatin e.g. no DNA damage, no enrichment for Cre binding motif in differential accessible chromatin.

Additionally, gene expression analysis indicated Myc activation as part of leukemic transformation or maintenance. Myc inhibition using JQ1 was able to significantly reduce leukemic burden and partially resolved differentiation block, confirming Myc activation as having a role in leukemogenesis in this model.

Why is this important?

Despite not being able to pinpoint the exact mechanism of how Cre induction drives leukemia in this homozygous FLT3-ITD model, we felt it was important to make the research community aware of this. We only detected this by the painstaking, yet important analysis of Cre-expressing controls in FLT3-ITD/ITD in the absence of a conditional allele, as Cre-negative FLT3-ITD/ITD mice never develop AML.

Although this model shows many characteristics of human cancer like differentiation block and Myc activation, we do not recommend to use this model to study spontaneous leukemic transformation. To date, this murine model has been widely used to study Cre-induced cooperating mutations, including Npm1 and Dnmt3a mostly in the heterozygous setting (8, 9, 12) and we believe that this model remains suitable and appropriate for the testing of cooperative events in AML in the heterozygous setting, given that Cre activity in the heterozygous ITD context did not give rise to AML.

References

1. Schmidt-Supprian M, Rajewsky K. Vagaries of conditional gene targeting. Nat Immunol. 2007;8(7):665-8.
2. Dohner H, Wei AH, Appelbaum FR, Craddock C, DiNardo CD, Dombret H, et al. Diagnosis and management of AML in adults: 2022 recommendations from an international expert panel on behalf of the ELN. Blood. 2022;140(12):1345-77.
3. Stratmann S, Yones SA, Mayrhofer M, Norgren N, Skaftason A, Sun J, et al. Genomic characterization of relapsed acute myeloid leukemia reveals novel putative therapeutic targets. Blood Adv. 2021;5(3):900-12.
4. Li L, Piloto O, Nguyen HB, Greenberg K, Takamiya K, Racke F, et al. Knock-in of an internal tandem duplication mutation into murine FLT3 confers myeloproliferative disease in a mouse model. Blood. 2008;111(7):3849-58.
5. Lee BH, Tothova Z, Levine RL, Anderson K, Buza-Vidas N, Cullen DE, et al. FLT3 mutations confer enhanced proliferation and survival properties to multipotent progenitors in a murine model of chronic myelomonocytic leukemia. Cancer Cell. 2007;12(4):367-80.
6. Meyer SE, Qin T, Muench DE, Masuda K, Venkatasubramanian M, Orr E, et al. DNMT3A Haploinsufficiency Transforms FLT3ITD Myeloproliferative Disease into a Rapid, Spontaneous, and Fully Penetrant Acute Myeloid Leukemia. Cancer Discov. 2016;6(5):501-15.
7. Greenblatt S, Li L, Slape C, Nguyen B, Novak R, Duffield A, et al. Knock-in of a FLT3/ITD mutation cooperates with a NUP98-HOXD13 fusion to generate acute myeloid leukemia in a mouse model. Blood. 2012;119(12):2883-94.
8. Poitras JL, Heiser D, Li L, Nguyen B, Nagai K, Duffield AS, et al. Dnmt3a deletion cooperates with the Flt3/ITD mutation to drive leukemogenesis in a murine model. Oncotarget. 2016;7(43):69124-35.
9. Mupo A, Celani L, Dovey O, Cooper JL, Grove C, Rad R, et al. A powerful molecular synergy between mutant Nucleophosmin and Flt3-ITD drives acute myeloid leukemia in mice. Leukemia. 2013;27(9):1917-20.
10. Dovey OM, Cooper JL, Mupo A, Grove CS, Lynn C, Conte N, et al. Molecular synergy underlies the co-occurrence patterns and phenotype of NPM1-mutant acute myeloid leukemia. Blood. 2017;130(17):1911-22.
11. Vu T, Straube J, Porter AH, Bywater M, Song A, Ling V, et al. Hematopoietic stem and progenitor cell-restricted Cdx2 expression induces transformation to myelodysplasia and acute leukemia. Nat Commun. 2020;11(1):3021.
12. Guryanova OA, Shank K, Spitzer B, Luciani L, Koche RP, Garrett-Bakelman FE, et al. DNMT3A mutations promote anthracycline resistance in acute myeloid leukemia via impaired nucleosome remodeling. Nat Med. 2016;22(12):1488-95.