Examining the role of U2AF1 and TET2 mutations in a myeloid malignancy mouse model Free

Splicing factors (SFs) are among the most commonly mutated genes in MDS and are often associated with poor clinical outcome, along with progression to AML. Past studies have shown that SF mutations are among the first to emerge in clonal hematopoiesis and therefore may play an early role in the initiation of oncogenesis [Saez et al. Blood 2017]. However, splicing modulators currently in clinical trials have failed to yield objective therapeutic responses [Pellagatti and Boultwood 2023], highlighting the need to better characterize the pathogenic mechanisms in SF-mutant MDS and AML.

At the same time, modeling SF-mutant myeloid malignancies in vivo has presented challenges. Specifically, U2AF1 S34F mouse models have required a RUNX1 co-mutation and treatment with a mutagen to develop an AML phenotype [Fei D, et al. PNAS 2018], while other U2AF1 S34F models have emulated some characteristics of MDS but without disease progression [Shirai C, et al. Cancer Cell 2015]. TET2, one of the most common co-mutations with U2AF1 in MDS and AML, is heavily involved in DNA methylation, and therefore provides potential to elicit a leukemic phenotype in U2AF1-mutant mice [Badar T, et al. Blood Cancer Journal 2023].

To study the role of U2AF1 mutations in vivo, we first used a U2AF1 S34F mouse model as described in Fei D, et al. PNAS 2018, which relies on Mx1Cre activation in hematopoietic lineages after treatment with poly(I:C). However, we observed low recombination efficiency at 4 weeks post-poly(I:C) injection and loss of mutant cells by 8 weeks post-poly(I:C). To address this, we crossed the same U2AF1 S34F transgene into HSC-SCL Cre-ER, which is activated by tamoxifen specifically in hematopoietic stem cells [Gothert J, et al. Blood 2005]. This new mouse model also contains Cre-dependent TET2 knockout [Moran Crusio K, et al. Cancer Cell 2011] and reporter gene R26-eYFP to track mutant cells [Srivinas S, et al. BMC Developmental Biology 2001]. We analyzed wild-type (WT), U2AF1 heterozygous S34F (U2het), TET2 homozygous knockout (TET2hom), and U2AF1het/TET2hom double mutant (DM) mice over the course of 14 weeks, starting with a 4-week tamoxifen diet. We collected complete blood count (CBC) readings and flow cytometry analysis on peripheral blood and bone marrow to assess general cell populations and importantly, we confirmed retention of mutant U2AF1 cells throughout the study.

We observed an increased MCV, one of the hallmarks of MDS, in both U2het and DM conditions at 6 weeks and even more prominently at 10 weeks and 14 weeks, with a trend towards RBC reduction. In the peripheral blood, we observed decreased myelopoiesis in both U2het and DM conditions at 6 weeks, especially in the monocyte population. At 10 and 14 weeks, this trend faded in U2het mice but persisted in DM mice. However, no significant myeloid trends were observed in the bone marrow, suggesting that U2AF1 mutations may have an early impact on terminal differentiation, while TET2 mutations may be necessary to compound this phenotype. In the bone marrow, DM mice displayed a decrease in lineage-negative cells at 6 weeks relative to all other groups, but this effect seemed to fade by 14 weeks. Moreover, we did not observe differences in LSK and LK populations relative to total lineage-negative cells at 6 weeks or 14 weeks, suggesting that early consequences of U2AF1 mutations are visible in terminal differentiation prior to dysregulation at the stem cell level.

These findings show potential of a new U2AF1/TET2-mutant MDS/AML mouse model. Though analyzed relatively early, our U2AF1-mutant mice show certain phenotypes associated with MDS, as well as impacts on terminal myeloid differentiation that persist specifically in DM mice. Given that splicing factor-mutant mouse models typically develop myeloid malignancies only after aging and/or transplantation, our data prompts investigation of long-term phenotypes in our model. Furthermore, additional work can be done to explain the mechanism through which these mutations produce the observed phenotypes, such as through altered expression of specific genes. Finally, further studies are necessary to better understand why splicing factor mutations are drivers of myeloid malignancies despite providing a competitive disadvantage in mouse models.