FLT3-ITD gets by with a little help from PRMT1

Kira Gritsman

In this issue of Blood, He et al discuss their discovery that methylation of fms-like tyrosine kinase 3-internal tandem duplication (FLT3-ITD) at arginines 972 and 973 by protein arginine N-methyltransferase 1 (PRMT1) potentiates oncogenic signaling and that inhibition of PRMT1 can improve the efficacy of kinase inhibitors in FLT3-ITD acute myeloid leukemia (AML).1

The loss of autoinhibition of FLT3-ITD leads to phosphorylation (P) of several tyrosines (Y) in the cytoplasmic domain. This leads to constitutive kinase activation, which can be inhibited with FLT3 kinase inhibitors. Methylation (me) of arginines 972 and 973 (R972/R973) by PRMT1 cooperates with phosphorylation at tyrosine 969 (Y969) by enhancing the association with SH2 domains of adaptor proteins such as GRB2. This leads to increased activation of downstream signaling effectors, such as PI3K/AKT and STAT5. PRMT inhibitors prevent methylation of R972/973 and can thereby cooperate with FLT3 inhibitors in blocking activation of downstream signaling pathways.

AML is still a devastating disease with poor overall survival. In AML patients, genetic alterations that affect epigenetic regulation are often detected together with mutations in signaling molecules, and these combinations occur in a stereotypic and nonrandom fashion.2 A fascinating conundrum of AML biology is how these 2 disparate classes of genetic alterations could cooperate to lead to hematopoietic transformation.

Mutations in the receptor tyrosine kinase FLT3 are seen in approximately 30% of patients with newly diagnosed AML.3 The most common FLT3 mutation observed in AML patients is the ITD of the juxtamembrane domain (FLT3-ITD), which results in kinase activation as a result of loss of autoinhibition of FLT3 kinase activity.4 The presence of FLT3-ITD portends a particularly poor prognosis, and most of these patients relapse after initial therapy. Several kinase inhibitors that target FLT3 have been developed for AML patients with FLT3 mutations.5 Two of these FLT3 inhibitors, midostaurin and gilteritinib, were recently approved by the US Food and Drug Administration (FDA), and several others are currently in development.5

Despite the clinical benefit demonstrated from the pharmacologic targeting of FLT3 in AML, many of the responses are short-lived, and relapses are still frequent.5 One potential contributing factor to high relapse rates in AML patients is the persistence of leukemia-initiating cells (LICs; also called leukemic stem cells) in the bone marrow after initial treatment. It has been reported that in FLT3-ITD AML, LICs express FLT3-ITD.6 Therefore, devising strategies to eradicate LICs is an important goal in AML therapy, and more efficient targeting of FLT3-mutated cells could help to achieve this goal. Persistent activation of downstream signaling pathways, such as PI3K/AKT, MAPK, and STAT5, has also been shown to contribute to intrinsic resistance to FLT3 inhibitors in FLT3-mutated AML.5

The elegant study by He et al uncovers a novel mechanism that highlights how epigenetic modifying enzymes and activated kinases can cooperate in AML and may explain the incomplete or short-lived response of FLT3-ITD AML cells to kinase inhibitors. The authors uncover a novel role for PRMT1 in regulating the signaling activity of FLT3-ITD. PRMT1 was previously described as playing a role in transcriptional regulation via its histone arginine methyltransferase function in other subtypes of AML.7,8 The authors report particularly high levels of PRMT1 RNA and protein expression in FLT3-ITD AML cells and show that PRMT1 catalyzes methylation of arginines 972 and 973 (R972/973) on FLT3-ITD, a process that is unaffected by kinase inhibitors (see figure). Interestingly, they demonstrate that dimethylation of R972/973 potentiates the kinase activity of FLT3-ITD by enhancing its physical interaction with SH2 domains of adaptor signaling proteins such as GRB2, leading to enhanced activation of AKT and STAT5 even more than activating kinase mutations in FLT3 (see figure). Remarkably, the authors demonstrate that point mutations of the arginines 972 and 973 of FLT3-ITD to lysines is sufficient to prolong survival in a mouse model of AML driven by MLL-AF9 and FLT3-ITD, and to a greater extent than a mutation at tyrosine 969, which inhibits activation of downstream signaling by FLT3-ITD. This suggests that methylation of FLT3 at R972/973 may have additional roles in tumorigenesis in addition to enhanced activation of downstream signaling.

Furthermore, He et al show that combined inactivation of FLT3 kinase activity with the FLT3 inhibitor AC220 (quizartinib) and PRMT1 activity with a PRMT1 inhibitor leads to increased apoptosis of FLT3-ITD AML cells and improved survival in several mouse models of FLT3-ITD AML more effectively than either drug alone. Most remarkably, this drug combination resulted in prolonged survival in a patient-derived xenograft model of FLT3-ITD AML. This effect was maintained even after drug discontinuation and also significantly compromised the engraftment of human leukemic cells in a secondary transplantation assay. This suggests that this drug combination not only decreases the disease burden but may also be able to target LICs in FLT3-mutated AML. Thus, the work of He et al proposes an innovative and promising therapeutic strategy to improve the efficacy of FLT3 inhibitors through coinhibition of PRMT1.

The He et al study raises several interesting questions. First, does dimethylation at R972/973 preexist in some patient cells before treatment with an FLT3 inhibitor? Or does it increase as a resistance mechanism after treatment? If it is usually preexisting, as is suggested in the He et al article, then could R972/973 dimethylation be used as a biomarker to predict for poor response to FLT3 inhibitors and to select the patients that would most benefit from treatment with a combination of FLT3 inhibitors and PRMT1 inhibitors?

Second, some of the data presented by the authors in FLT3-ITD AML mouse models suggest that PRMT1 may have additional effects on FLT3-ITD AML cells in addition to altering the effects of FLT3-ITD on downstream signaling. The effects on the self-renewal of LICs suggest that PRMT1 inactivation may also exert its effects by altering transcriptional activation of self-renewal genes through its histone arginine methyltransferase function, as has been described in other subtypes of AML.8 If this is true, it could explain the potential synergistic effects of FLT3 inhibitors and PRMT1 inhibitors on AML LICs.

Finally, the He et al study raises the question of how generalizable this mechanism might be to other activating kinase mutations in hematologic malignancies, such as those in C-KIT, PDGFR, JAK2, and others. The authors have shown that PRMT1 is the main PRMT that catalyzes R972/973 dimethylation on FLT3-ITD, but could PRMT1 or other PRMTs also alter the signaling activity of other kinases? Could combined kinase inhibition and PRMT inhibition be a common paradigm to combat resistance to kinase inhibitors in cancer? These questions will open new areas for future investigation and hopefully lead to the development of more effective therapeutic strategies for patients with AML driven by activating kinase mutations.


  • Conflict-of-interest disclosure: The author declares no competing financial interests.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.