Enhanced targeting of CML stem and progenitor cells by inhibition of porcupine acyltransferase in combination with TKI

Abstract

Tyrosine kinase inhibitor (TKI) treatment of chronic myeloid leukemia (CML) has limited efficacy against leukemia stem cells (LSC) responsible for disease propagation, and most CML patients require continued TKI treatment to maintain remission. LSC maintenance is related, at least in part, to signals from the bone marrow microenvironment (BMM). Our previous studies have shown that Wnt signaling from the BMM contributes to preservation of CML LSC following TKI treatment. Secretion of Wnt ligands requires their modification by the O-acyl transferase Porcupine (PORCN). Here we investigated the activity of a potent and selective PORCN inhibitor, WNT974, against CML stem and progenitor cells. WNT974 efficiently antagonized Wnt signaling in human CML CD34⁺ cells, and in combination with the TKI nilotinib (NIL) significantly enhanced inhibition of proliferation and colony-forming potential of CML stem and progenitor cells and reduced their growth in immunodeficient mice in vivo, in comparison with NIL alone. Treatment of transgenic CML mice in vivo with NIL in combination with WNT974 significantly reduced leukemic stem and progenitor cell numbers, reduced regeneration of leukemic long-term hematopoietic stem cells in secondary transplant recipients, and enhanced survival of mice after discontinuation of treatment, in comparison with NIL alone. CML progenitors demonstrated enhanced sensitivity to Wnt stimulation, associated with increased expression of the FZD4 receptor. FZD4 knockdown inhibited CML progenitor growth. These results support further investigation of PORCN targeting to inhibit Wnt secretion and signaling and enhance targeting of CML stem cells while sparing their normal counterparts.

Figure 1.

WNT974 antagonizes the Wnt signaling pathway in human CML stem/progenitor cells. (A) Wnt secretion was evaluated in WNT1-MSC cultured in the presence of WNT974 for 24 h. Conditioned medium was harvested and added onto 293T-BAR reporter cells. WNT-β-catenin transcriptional activity was then evaluated after a further 24 h (n = 5). (B) HEK 293T cells overexpressing Wnt-1 were treated with WNT974, metabolically labeled with azide-containing palmitic acid and modified palmitoylated proteins detected by labeling with alkyne-containing APC dye using flow cytometry (n = 4). (C) CML CD34⁺ cells were cultured in the presence or absence of human MSC in the presence of WNT974 for 48 h, and immunofluorescence microscopy was performed. CML CD34⁺ cells labeled with antibodies to β-catenin (green) and DAPI (blue) are shown. Two samples were studied. All scale bars represent a size of 10 µM, and at least 200 cells were analyzed for each sample. (D) qPCR analysis for mRNA expression of Wnt target genes in CML CD34⁺ cells (n = 5) cultured as in panel A. Error bars represent mean ± SEM. Ctrl, control; ns, not significant; PPAR, peroxisome proliferator-activated receptors. *P < .05; **P < .01; ***P < .001; ****P < .0001.

Figure 2.

WNT974 in combination with nilotinib inhibits the proliferation and clonogenic potential of human CML primitive and committed progenitor cells. Normal and CML 34⁺ cells (n = 4-5) were labeled with CFSE. CFSE⁺ primitive cells (34⁺38⁻) and committed cells (34⁺38⁺) were sorted by flow cytometry and cultured in the presence or absence of MSC for 6 days with WNT974, Nil (1 µM), or both, or left untreated. A proliferation index was calculated on the basis of reduction in CFSE levels for CML 34⁺38⁻ cells (A) and for normal 34⁺38⁻ cells (B). Cells were also plated in methylcellulose progenitor assays, and colony-forming cell (CFC) frequencies were determined after 14 days for CML 34⁺38⁻ cells (C) and for normal 34⁺38⁻ cells (D). CML 34⁺ cells (n = 3) were labeled with CFSE, CD34, CD38, and CD26, and CD34⁺CD38⁻CD26⁺ and CD34⁺CD38⁻CD26⁻ cells were selected by fluorescence-activated cell sorting (E) and cultured in the presence of MSC for 6 days with WNT974, NIL, or both, or left untreated. (F) Example CFSE plots of one CML sample are shown. Proliferation index was measured on the basis of reduction of CFSE levels and both actual values (G), and fold change in relation to untreated controls (H) are shown. Error bars represent mean ± SEM. FSC, forward scatter; Nil, nilotinib; w, with; w/o, without. *P < .05; **P < .01.

Figure 3.

WNT974 in combination with nilotinib inhibits the long-term engraftment of human CML stem cells in immunodeficient mice without affecting normal stem cells. Normal (1 × 10⁵ cells per mouse) and CML 34⁺ cells (2 × 10⁶ cells per mouse) were cultured in the presence of MSC and treated with nilotinib (Nil) (1 µM), WNT974 (1 µM), or both, or were untreated for 4 days; they were then transplanted into NSG mice (6-8 mice per group). Mice were killed after 16 weeks, and bone marrow (BM) content of femurs was obtained. Graph showing normal human CD45⁺ cell engraftment in BM (A); human CML CD45⁺ cell engraftment in BM (B); BCR-ABL mRNA levels in CML CD45⁺ cells engrafted in BM (C); lineage marker expression of normal CD45⁺ hematopoietic cells engrafted in BM (D); and lineage marker expression of CML CD45⁺ hematopoietic cells engrafted in BM (E). In a second experiment, CML CD34⁺ cells were transplanted into sublethally irradiated NSG mice, and 3 weeks posttransplantation, mice were treated with NIL (50 mg/kg), WNT974 (5 mg/kg), or the combination of these for 3 weeks. Representative results of fluorescence-activated cell sorter analysis of BM cells (F), frequency and absolute number of total human CD45⁺ cells (G), and aggregate results of the percentage (H) and absolute number (I) of human myeloid progenitors and lineage marker expression in the BM from treated mice are shown. (J) Human CD45⁺ cells were selected and plated in CFC assays. (K) FISH analysis was performed on interphase nuclei of 200 CD45⁺ cells from each treatment arm. Error bars represent mean ± SEM. *P < .05; **P < .01; ***P < .001; ****P < .0001.

Figure 4.

WNT974 in combination with nilotinib significantly reduces leukemic stem and progenitor cells in the BM and spleen of transgenic BCR-ABL mice. (A) BCR-ABL expression was induced in SCL-tTA-BCR-ABL mice by tetracycline withdrawal. BM cells were obtained 3 weeks after induction and transplanted into wild-type FVB/N recipient mice irradiated at 750 cGy (10⁶ cells/mouse; 6-8 mice/group). Three weeks posttransplantation, treatment with vehicle, NIL (50 mg/kg), WNT974 (5 mg/kg), or a combination was initiated. PB WBC counts (B), PB neutrophil counts (C), BM cellularity (D), spleen size (E), absolute number of LTHSC (F), STHSC (G), MPP1 (H), MPP2 (I), CMP (J), GMP (K), and MEP (L) in the BM were evaluated after 2.5 weeks after drug treatment. Error bars represent mean ± SEM. *P < .05.

Figure 5.

WNT974 in combination with nilotinib significantly inhibits regeneration of CML LTHSC after second transplant and prolongs the survival of BCR-ABL mice. (A) BCR-ABL expression was induced in SCL-tTA-BCR-ABL mice by tetracycline withdrawal. BM cells were obtained 3 weeks after induction and transplanted into wild-type FVB/N recipient mice irradiated at 750 cGy (10⁶ cells/mouse; 6-8 mice/group). Three weeks posttransplantation, treatment with vehicle, nilotinib (50 mg/kg), WNT974 (5 mg/kg), or a combination was initiated and continued for 3 weeks. (B) Mice were followed after discontinuation of treatment. (C) BM cells from mice receiving different treatments as described in Figure 4 were pooled and transplanted (2 × 10⁶ cells per mouse; 8-9 mice/group) into wild-type irradiated FVB/N recipient mice for secondary transplantation. Mice were killed after 12 weeks, and the number of BM LTHSC (D), GMP (E), splenic LTHSC (F), and GMP (G) were analyzed. Error bars represent mean ± SEM. *P < .05.

Figure 6.

Enhanced Wnt sensitivity of CML in comparison with normal stem/progenitor cells. (A) Normal and CML CD34⁺ cells were cultured in recombinant Wnt3A (200 ng/mL), and cells were harvested after 4 h for Western blotting for P-LRP6, LRP6, β-catenin, and GAPDH. (B) Results of densitometry analysis are shown. (C) Normal (CB) and CML CD34⁺ cells (n = 3) cultured in recombinant Wnt3A (200 ng/mL) were plated in methylcellulose assay, following which colonies were counted after 2 weeks of culture. (D) Results of qPCR analysis was performed for FZD genes in normal and CML CD34⁺ cells. CML CD34⁺ cells (n = 3) transfected with control siRNA, FZD1 siRNA, FZD4 siRNA, and FZD5 siRNA were cultured with or without NIL in the absence (E) or presence of MSC (F), for 4 days, following which CFC frequency was analyzed in methylcellulose progenitor assays. Error bars represent mean ± SEM. *P < .05; **P < .01. CB, cord blood.