circMYBL2, a circRNA from MYBL2, regulates FLT3 translation by recruiting PTBP1 to promote FLT3-ITD AML progression

Abstract

Internal tandem duplication (ITD) mutations within FMS-like tyrosine kinase-3 (FLT3) occur in up to 30% of acute myeloid leukemia (AML) patients and confer a very poor prognosis. The oncogenic form of FLT3 is an important therapeutic target, and inhibitors specifically targeting FLT3 kinase can induce complete remission; however, relapse after remission has been observed due to acquired resistance with secondary mutations in FLT3, highlighting the need for new strategies to target FLT3-ITD mutations. Recent studies have reported that the aberrant formations of circular RNAs (circRNAs) are biological tumorigenesis-relevant mechanisms and potential therapeutic targets. Herein, we discovered a circRNA, circMYBL2, derived from the cell-cycle checkpoint gene MYBL2. circMYBL2 is more highly expressed in AML patients with FLT3-ITD mutations than in those without the FLT3-ITD mutation. We found that circMYBL2 knockdown specifically inhibits proliferation and promotes the differentiation of FLT3-ITD AML cells in vitro and in vivo. Interestingly, we found that circMYBL2 significantly influences the protein level of mutant FLT3 kinase, which contributes to the activation of FLT3-ITD-dependent signaling pathways. Mechanistically, circMYBL2 enhanced the translational efficiency of FLT3 kinase by increasing the binding of polypyrimidine tract-binding protein 1 (PTBP1) to FLT3 messenger RNA. Moreover, circMYBL2 knockdown impaired the cytoactivity of inhibitor-resistant FLT3-ITD+ cells, with a significant decrease in FLT3 kinase expression, followed by the inactivation of its downstream pathways. In summary, we are the first to reveal a circRNA that specifically influences FLT3-ITD AML and regulates FLT3 kinase levels through translational regulation, suggesting that circMYBL2 may be a potential therapeutic target for FLT3-ITD AML.

Graphical abstract

Figure 1.

Identification and characterization of circMYBL2 in FLT3-ITD⁺ AML. (A) Identification of circRNAs having significant differences (P < .01) in FLT3-ITD⁺ AML compared with FLT3-ITD^– AML. (B) Differential expression of circMYBL2 between FLT3-ITD⁺ and FLT3-ITD^– AML patient samples. (C) Structures of the MYBL2 genome and transcript. circMYBL2 is produced by exons 8-9. (D) Identity of the junction point of circMYBL2. (E) RNase R treatment confirmed the circular form of circMYBL2. (F-G) Identification of circMYBL2 cytoplasmic distribution by qRT-PCR analysis and FISH. MTOC1 and MALAT1 were used as the cytoplasmic and nuclear markers, respectively. Cy3 dye and DAPI stain; original magnification ×63. DAPI, 4′,6-diamidino-2-phenylindole; GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

Figure 2.

Requirement for circMYBL2 in FLT3-ITD AML cells. (A) Efficient knockdown of circMYBL2 in MOLM-13 cells. (B) Specific si-circMYBL2 did not influence parental gene expression. (C) Effect of circMYBL2 knockdown on proliferation in FLT3-ITD⁺ cells. (D-E) Flow cytometric analysis of cell-cycle progression in AML cells. The numbers indicate the percentages of cells. (F-G) Flow cytometric analysis of apoptosis. The numbers indicate the percentages of cells. (H) Flow cytometric analysis of apoptosis in primary AML patient cells including FLT3-ITD AML patient cells (#1 and #2), and FLT3-ITD^– patient cells (#4). 7AAD, 7-Aminoactinomycin D; sub-G1, sub cell cycle gap 1 phase.

Figure 3.

circMYBL2 knockdown impairs the cytoactivity of FLT3-ITD AML cells, including inhibitor-resistant cells. (A) Effect of circMYBL2 suppression in methylcellulose cultures of FLT3-ITD⁺ AML MOLM-13 cells. (B) Microscopic analysis of Wright-Giemsa–stained cytospin preparations of MOLM-13 and THP-1 cells. Original magnification ×400. Flow cytometric analysis of myeloid differentiation in MOLM-13 and THP-1 cells. The numbers indicate the percentages of cells. (C) Cell viability was measured with a CCK8 assay 48 hours after the addition of quizartinib at the indicated concentration. (D) Western blot measuring p-FLT3 in MOLM-13 or MOLM-13-RQ exposed to quizartinib at the indicated concentration. (E-F) MOLM-13-RQ cells were as sensitive to circMYBL2 knockdown as parental MOLM-13 cells in terms of apoptosis. (G-H) MOLM-13-RQ cells were as sensitive to circMYBL2 knockdown as parental MOLM-13 cells in terms of cell differentiation. (I) MOLM-13-RQ cells were as sensitive to circMYBL2 knockdown as parental MOLM-13 cells in terms of cell proliferation.

Figure 4.

circMYBL2 regulates FLT3-ITD–dependent signaling pathways by modulating the translation of mutant FLT3 kinase. (A) Western blot showing downregulation of FLT3 protein expression upon circMYBL2 knockdown in MOLM-13 and MV4-11 cells. (B) qPCR measuring the expression of FLT3 mRNA upon circMYBL2 knockdown in MOLM-13 and MV4-11 cells. (C) Downregulation of FLT3 protein expression upon circMYBL2 knockdown in FLT3-ITD AML patient samples. (D) qPCR measuring the expression of FLT3 mRNA upon circMYBL2 knockdown in FLT3-ITD AML patient samples. (E-F) Western blot showing the downregulation of STAT5 signaling activation upon circMYBL2 knockdown in MOLM-13 and MV4-11 cells and in AML patient samples. (G) Flow cytometric analysis of apoptosis in MV4-11 cells upon circMYBL2 knockdown or c-MYC overexpression. The numbers indicate the percentages of cells. (H) Western blot showing the decrease in FLT3 protein, p-STAT5, and c-MYC levels upon circMYBL2 knockdown in MOLM-13 and MOLM-13-RQ cells. (I) circMYBL2 knockdown reduced the levels of FLT3 kinase harboring the D835Y mutation in an AML patient sample. (J) Polysomes in cytoplasmic extracts of sh-NC and sh-circMYBL2 MV4-11 cells were fractionated through sucrose gradients, and the relative levels of FLT3 mRNA were analyzed by qRT-PCR in the gradient fractions. ns, not significant.

Figure 5.

Identification of proteins that interact with circMYBL2. (A) Schematic of circMYBL2 linearization and the tRSA RNA pulldown assay. (B) Identification of translation-associated proteins that interact with circMYBL2 by silver staining and MS. (C) Detection of interactions between circMYBL2 and RBPs by western blot analysis. (D-E) RIP of RBPs using anti-IgG or anti-RBP (PTBP1, DHX9, SYNCRIP, or hnRNPA1) antibodies. The percentage of RIP-enriched circMYBL2 and FLT3 mRNA relative to the input value was calculated by qRT-PCR. (F) Western blot showing the augmented decrease in FLT3 kinase expression upon knockdown of both circMYBL2 and PTBP1 in MOLM-13 and MV4-11 cells. (G) Effect of knockdown of both circMYBL2 and PTBP1 on proliferation. (H-I) Western blot showing downregulation of FLT3 protein expression upon PTBP1 knockdown in MOLM-13 and MV4-11 cells and in FLT3-ITD AML patient samples. SA, streptavidin.

Figure 6.

circMYBL2 interacts with PTBP1 to affect FLT3-ITD AML proliferation through promoting FLT3 kinase translation. (A-B) Western blot showing downregulation of STAT5 signaling activation upon PTBP1 knockdown in MOLM-13 and MV4-11 cells and in FLT3-ITD AML patient samples. (C) Schematic of the circMYBL2 overexpression vector. (D) Western blot showing upregulation of FLT3 protein expression upon overexpression of either circMYBL2 or PTBP1 in MV4-11 and HL60 cells. (E) Polysomes in cytoplasmic extracts of MV4-11 cells with PTBP1 knockdown or control treatment were fractionated through sucrose gradients, and the relative levels of FLT3 mRNA in the gradient fractions were analyzed by qRT-PCR. (F) Western blot showing upregulation of FLT3 protein expression upon overexpression of either circMYBL2 or PTBP1 in 293t cells. (G) Western blot measuring FLT3 protein expression in 293t cells with overexpression of circMYBL2 in the presence or absence of si-PTBP1. (H) RIP of PTBP1 using anti-PTBP1 or anti-IgG antibodies in 293t cells in the presence or absence of circMYBL2 overexpression. (I) Schematic depicting the dominant function of circMYBL2 in facilitating the translational efficiency of FLT3 kinase by enhancing the binding of PTBP1 to FLT3 mRNA, specifically promoting FLT3-ITD AML progression. CMV, cytomegalovirus.

Figure 7.

circMYBL2 knockdown impairs the tumorigenesis and infiltration of FLT3-ITD AML cells in vivo. (A) Wright-Giemsa staining of BM samples isolated from mice engrafted with human MOLM-13 cells cotransfected with sh-NC and sh-circMYBL2. hCD45⁺ cells in the mice are indicated by the black arrows. Original magnification ×400. (B) Representative images of lymph nodes from control and sh-circMYBL2-MOLM-13–treated mice. Reduced lymph node involvement in sh-circMYBL2-MOLM-13–treated mice compared with sh-NC-MOLM-13–treated control mice. (C) H&E staining showing infiltration of leukemic cells in the BM, spleen, and liver of mice engrafted with sh-circMYBL2 cells compared with that in control mice. hCD45⁺ cells in the tissues are indicated by the black arrows. (D) Flow cytometry showing substantially decreased levels of blasts in blood and BM samples from mice treated with circMYBL2-knockdown MOLM-13 cells relative to these levels in control mice. CD11b and CD14 marker expression was dramatically increased in sh-circMYBL2-MOLM-13–treated mice compared with that in sh-NC-MOLM-13–treated mice. (E-F) Scatter plots show the statistical values for panel D. (G) Kaplan-Meier survival curves for mice implanted with sh-NC or sh-circMYBL2 MOLM-13 cells (n = 5 mice per group). P values were calculated using a log-rank (Mantel-Cox) test. (H) Kaplan-Meier survival curves for mice implanted with sh-NC or sh-circMYBL2 MOLM-13-RQ cells (n = 6 mice per group). P values were calculated using a log-rank (Mantel-Cox) test. SSC, side scatter.