SETBP1 mutations drive leukemic transformation in ASXL1-mutated MDS

Abstract

Mutations in ASXL1 are frequent in patients with myelodysplastic syndrome (MDS) and are associated with adverse survival, yet the molecular pathogenesis of ASXL1 mutations (ASXL1-MT) is not fully understood. Recently, it has been found that deletion of Asxl1 or expression of C-terminal-truncating ASXL1-MTs inhibit myeloid differentiation and induce MDS-like disease in mice. Here, we find that SET-binding protein 1 (SETBP1) mutations (SETBP1-MT) are enriched among ASXL1-mutated MDS patients and associated with increased incidence of leukemic transformation, as well as shorter survival, suggesting that SETBP1-MT play a critical role in leukemic transformation of MDS. We identify that SETBP1-MT inhibit ubiquitination and subsequent degradation of SETBP1, resulting in increased expression. Expression of SETBP1-MT, in turn, inhibited protein phosphatase 2A activity, leading to Akt activation and enhanced expression of posterior Hoxa genes in ASXL1-mutant cells. Biologically, SETBP1-MT augmented ASXL1-MT-induced differentiation block, inhibited apoptosis and enhanced myeloid colony output. SETBP1-MT collaborated with ASXL1-MT in inducing acute myeloid leukemia in vivo. The combination of ASXL1-MT and SETBP1-MT activated a stem cell signature and repressed the tumor growth factor-β signaling pathway, in contrast to the ASXL1-MT-induced MDS model. These data reveal that SETBP1-MT are critical drivers of ASXL1-mutated MDS and identify several deregulated pathways as potential therapeutic targets in high-risk MDS.

Figure 1

SETBP1 mutations are enriched in ASXL1-mutant de novo MDS patients and inhibit ubiquitination and subsequent degradation of SETBP1. (a) Proportion of SETBP1 mutations among ASXL1-mutated (left, n=64) or ASXL1-wild type (right, n=304) MDS patients are indicated. Red portion shows SETBP1-mutational frequency. (b) Schematic diagram of SETBP1 protein and SETBP1 mutations (asterisk) identified in ASXL1-mutated MDS patients. Three AT hook domains, SKI homologous region, SET binding domain and repeat domain are indicated. (c–d) Cumulative incidence of leukemic transformation (c) and Kaplan-Meier analysis for survival (d) in ASXL1-mutated MDS patients with or without SETBP1 mutations (n=6, solid line, n=58, dashed line, respectively; p-values calculated using a log-rank test).

Figure 2

SETBP1-MT impairs differentiation and apoptosis in vitro. (a–b) IL-3 dependent murine myeloid 32Dcl3 cells transduced with pMYs-IP/pMYs-IG, pMYs-IP-ASXL1-MT/mock, mock/pMYs-IG-SETBP1-WT, mock/pMYs-IG-SETBP1-D868N, pMYs-IP-ASXL1-MT/pMYs-IG-SETBP1-WT, or pMYs-IP-ASXL1-MT/pMYs-IG-SETBP1-D868N were cultured in RPMI-1640 supplemented with 10% FBS. Proportions of viable cells (a) and Annexin V positive cells (b) 24 hours after IL-3 withdrawal are shown. All data with error bars indicate the mean + SEM from duplicated experiments. (c–d) The transduced 32Dcl3 cells were cultured with 50 ng/ml of G-CSF for 6 days. Cytospin preparations of these cells assessed by Wright-Giemsa staining (c) are shown. Original magnification, ×400. Proportions of mature (segmented) cells (%, top) and immature (blastic) cells (%, middle) or mean fluorescent intensity (MFI) of CD11b estimated by FACS analysis (bottom) are shown. All data with error bars indicate the mean + SEM from duplicated experiments. Compared to the control, *p<0.05, ** p<0.01, ***p<0.001, Student’s t test.

Figure 3

Expression of both ASXL1-MT and SETBP1-D868N increased myeloid colony output. Colony-forming potentials of LSKs transduced with indicated plasmids in methylcellulose supplemented with SCF (100 ng/ml) and IL-3 (20ng/ml) are shown. Four columns represent numbers of serially replated colonies from the first (left) to the fourth (right) plating (a). All data with error bars indicate the mean ± SEM from duplicated cultures. Cytospin preparations of cultured cells in the second replating colonies assessed by Wright-Giemsa staining are shown (b). Compared to the control, *p< 0.05, **p<0.01, ***p<0.001, Student’s t test.

Figure 4

SETBP1-MT collaborated with ASXL1-MT in inducing AML. (a) Kaplan-Meier analysis for the survival of mice transplants with BM cells transduced with empty vector (Mock, n=6, blue circle), pMYs-ASXL1-MT-IG (ASXL1-MT, n=6, yellow triangle), pMYs-SETBP1-D868N-IG (SETBP1-D868N, n=6, green triangle), and pMYs-SETBP1-D868N-IG and pMYs-ASXL1-MT-INGFR (ASXL1-MT/SETBP1-D868N, n=9, red square). P values were calculated using a log-rank test. (b) Macroscopic findings of sacrificed mice transplanted with BM cells transduced with ASXL1-MT and SETBP1-D868N. Representative photographs are shown. (c) Cytospin preparations of BM cells derived from mice with ASXL1-MT and SETBP1-D868N were stained with Wright-Giemsa. Representative photographs are shown. Original magnification, ×400. (d) Flow cytometric analyses of BM cells derived from mice with ASXL1-MT and SETBP1-D868N. (e–k) Mice transplanted with both ASXL1-MT and SETBP1-D868N (ASXL1-MT/SETBP1-D868N, n=8) expressing cells displayed progressive leukocytosis (e), anemia (f), macrocytosis (g), thrombocytopenia (h), hepatosplenomegaly (i–j) and hyperplastic BM (k) compared with the control mice transplanted with empty vectors (mock, n=6), ASXL1-MT (n=6), or SETBP1-D868N (n=5). BM cells were isolated from the femurs and tibias of the sacrificed mice. *p< 0.05, **p<0.01, ***p<0.001, Student’s t test.

Figure 5

SETBP1 mutation resulted in Pp2a inhibition and increased Hoxa9 and Hoxa10 expression. (a) HEK293T cells were transiently transfected with empty vector, SETBP1-WT-3×TY1 and SETBP1-D868N-3×TY1, followed by IP of TY1 epitope and Western blotting for SET and PP2A. Cell lysates were also subject to immunoblotting with anti-TY1 Ab, anti-SET Ab, anti-PP2A Ab, or anti-tublin Ab. (b) Cell lysates from FACS-purified c-Kit+ cells from the BM of mice transplanted with BM cells transduced with empty vector (control), ASXL1-MT (MDS model), or both ASXL1-MT and SETBP1-D868N (AML model), were subject to immunoblotting with anti-phospho-Akt (T308/S473) Ab, anti-Akt Ab, anti-phopho-PP2A (Y307)Ab, anti-PP2A Ab, or anti-tublin Ab. (c) PP2A phosphatase assay in c-Kit+ cells from a control mouse (Mock) and a leukemic mice (ASXL1-MT/SETBP1-D868N). Results are representative of 3 independent experiments. *p< 0.05. (d) Relative growth rate of BM cells from leukemic mice transplanted with BM cells transduced with ASXL1-MT and SETBP1-D868N in the presence of 1ng/ml IL-3 and indicated concentration of FTY720. (e) Relative expression levels of Hoxa5/Hoxa7/Hoxa9/Hoxa10 were examined by qRT-PCR in CD3- B220- Ter119-negating BM cells derived from normal controls (Control), MDS mice (ASXL1-MT) and AML mice (ASXL1-MT + SETBP1-D868N). The values were normalized by Gapdh mRNA levels. All data with error bars are presented as mean + SEM of 2 independent experiments. (f–g) Quantitative PCR analyses following anti-TY1 ChIP in HEK293T cells transfected with empty vector, SETBP1-WT-3×TY1, SETBP1-D868N-3×TY1 using anti-TY1 antibody. PCR products specific to various regions and diagrams of tested region of HOXA9 (f) or HOXA10 locus (g) were shown in bottom panels. Results are representative of 3 independent experiments. Transcriptional start sites are indicated as arrows. *p< 0.05, **p<0.01, ***p<0.001, Student’s t test.

Figure 6

Combined expression of SETBP1 and ASXL1 mutations induces a stem cell signature and inhibits expression of genes involved in hematopoietic differentiation and TGF-β signaling. (a–e) Gene set enrichment analysis (GSEA) determining specific gene sets or pathways that are positively or negatively regulated by additional expression of SETBP1-D868N. Compared with MDS mice expressing ASXL1-MT alone (right side), AML mice expressing ASXL1-MT and SETBP1-D868N (left side) negatively correlated with gene sets downregulated in hematopoietic stem cells (a), those of mature hematopoietic cells (b), and those of cell lineage (c). The expression profile of the AML mice also inversely correlated with gene sets of TGF-β signaling pathway (d) and associated with those of Myc pathway (e). The normalized enrichment scores (NES) and P-values are given. (f) qRT-PCR validation of genes related to TGF-β signaling pathway in the BM of control, MDS (ASXL1-MT) or AML (ASXL1-MT + SETBP1-D868N) mice. *p< 0.05, **p<0.01, ***p<0.001, Student’s t test. All data with error bars indicate the mean + SEM from duplicated experiments.