Somatic mutations in U2AF1, a spliceosome gene involved in pre-mRNA splicing, occur in up to 11% of MDS patients. While we reported that mice expressing mutant U2AF1(S34F) have altered hematopoiesis and RNA splicing, similar to mutant MDS patients, the role of wild-type U2AF1 in normal hematopoiesis has not been studied. U2AF1mutations are always heterozygous and the wild-type allele is expressed, suggesting that mutant cells require the residual wild-type (WT) allele for survival. A complete understanding of the role of wild-type U2AF1 on hematopoiesis and RNA splicing will enhance our understanding of how mutant U2AF1 contributes to abnormal hematopoiesis and splicing in MDS.

In order to understand the role of wild-type U2af1 in normal hematopoiesis, we created a conditional U2af1 knock-out (KO) mouse (U2af1flox/flox). Homozygous embryonic deletion of U2af1using Vav1-Cre was embryonic lethal and led to reduction in fetal liver hematopoietic stem and progenitor cells (KLS and KLS-SLAM, p ≤ 0.05) at embryonic day 15, suggesting that U2af1 is essential for hematopoiesis during embryonic development. To study the hematopoietic cell-intrinsic effects of U2af1 deletion in adult mice, we performed a non-competitive bone marrow transplant of bone marrow cells from Mx1-Cre/U2af1flox/flox, Mx1-Cre/U2af1flox/wtor Mx1-Cre/U2af1wt/wtmice into lethally irradiated congenic recipient mice. Following poly I:C-induced U2af1deletion, homozygous U2af1 KOmice, but not other genotypes (including heterozygous KO mice), became moribund. Analysis of peripheral blood up to 11 days post poly I:C treatment revealed anemia (hemoglobin decrease >1.7 fold) and multilineage cytopenias in homozygous U2af1 KOmice compared to all other genotypes(p ≤ 0.001, n=5 each).Deletion of U2af1 alsoled to rapid bone marrow failure and a reduction in the absolute number of bone marrow neutrophils (p ≤ 0.001), monocytes (p ≤ 0.001), and B-cells (p ≤ 0.05), as well as a depletion of hematopoietic progenitor cells (KL, and KLS cells, p ≤ 0.001, n=5 each). Next, we created mixed bone marrow chimeras (i.e., we mixed equal numbers of homozygous KO and wild-type congenic competitor bone marrow cells and transplanted them into lethally irradiated congenic recipient mice) to study the effects of U2af1 deletion on hematopoietic stem cell (HSC) function. As early as 10 days following Mx1-Cre-induction, we observed a complete loss of peripheral blood neutrophil and monocyte chimerism of the U2af1 KOcells, but not U2af1 heterozygous KO cells, and at 10 months there was a complete loss of homozygous U2af1 KObone marrow hematopoietic stem cells (SLAM, ST-HSCs, and LT-HSCs), neutrophils, and monocytes, as well as a severe reduction in B-cells and T-cells (p ≤ 0.001, n=3-4 for HSCs. p ≤ 0.001, n=9-10 for all other comparisons). The data indicate that normal hematopoiesis is dependent on wild-type U2af1expression, and that U2af1 heterozygous KO cells that retain one U2af1 allele are normal.

Next, we tested whether mutant U2AF1(S34F) hematopoietic cells require expression of wild-type U2AF1 for survival. To test this, we used doxycycline-inducible U2AF1(S34F) or U2AF1(WT) transgenic mice. We generated ERT2-Cre/U2af1flox/flox/TgU2AF1-S34F/rtTA(S34F/KO), and ERT2-Cre/U2af1flox/flox/TgU2AF1-WT/rtTA,(WT/KO) mice, as well as all other single genotype control mice. We then created 1:1 mixed bone marrow chimeras with S34F/KO or WT/KO test bone marrow cells and wild-type competitor congenic bone marrow cells and transplanted them into lethally irradiated congenic recipient mice. Following stable engraftment, we induced U2AF1(S34F) (or WT) transgene expression with doxycycline followed by deletion of endogenous mouse U2af1 using tamoxifen. As early as 2 weeks post-deletion of U2af1, S34F/KO neutrophil chimerism dropped to 5.4% indicating loss of mutant cells, while WT/KO neutrophil chimerism remained elevated at 31.6% (p = 0.01, n=6-8). The data suggest that mutant U2AF1(S34F) hematopoietic cells are dependent on expression of wild-type U2af1 for survival. Since U2AF1mutant cells are vulnerable to loss of the residual wild-type U2AF1allele, and heterozygous U2af1KO cells are viable, selectively targeting the wild-type U2AF1allele in heterozygous mutant cells could be a novel therapeutic strategy.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.