Loss of beta-catenin impairs the renewal of normal and CML stem cells in vivo

Abstract

A key characteristic of stem cells and cancer cells is their ability to self-renew. To test if Wnt signaling can regulate the self-renewal of both stem cells and cancer cells in the hematopoietic system, we developed mice that lack beta-catenin in their hematopoietic cells. Here we show that beta-catenin-deficient mice can form HSCs, but that these cells are deficient in long-term growth and maintenance. Moreover, beta-catenin deletion causes a profound reduction in the ability of mice to develop BCR-ABL-induced chronic myelogenous leukemia (CML), while allowing progression of acute lymphocytic leukemia (ALL). These studies demonstrate that Wnt signaling is required for the self-renewal of normal and neoplastic stem cells in the hematopoietic system.

Figure 1. Conditional Deletion of β-Catenin in the Hematopoietic System

(A) The schematic indicates the strategy for deletion of β-catenin in hematopoietic cells using mice carrying a floxed β-catenin allele crossed to transgenic mice in which expression of the cre recombinase is directed by the vav promoter. (B) Analysis of deletion efficiency by genomic PCR analysis. Genomic DNA from control and β-catenin^−/− whole bone marrow cells was amplified using primer RM68/69, which does not generate a product for the floxed allele but generates a 631 bp product for the floxdel allele (left lanes; 45 relative fluorescence units for −/− and 0.64 RFU for control), and primer RM41/42/43, which generates a product of 324 bp for the floxed allele (1.5 RFU for control and 0.0006 RFU for −/−) and a 500 bp product for the floxdel allele (69 RFU for −/− and 0.32 for control) (Brault et al., 2001). Quantitation of the difference in product formation was carried out by real-time PCR analysis. Results are representative of two experiments. (C) Whole bone marrow lysates from control (+/+) and β-catenin knockout (−/−) mice were analyzed by western blot for presence of β-catenin. Results shown are representative of four experiments. (D) KLS cells were isolated from control (+/+), heterozygous (+/−), and conditional β-catenin knockout (−/−) mice, and RT-PCR analysis was performed to determine expression of β-catenin. Results are representative of three experiments.

Figure 2. β-Catenin Is Required for Long-Term Stem Cell Maintenance and Transplantation In Vivo

(A) Bone marrow cells from control or β-catenin knockout mice were analyzed for frequency of KLSF cells. Dot plots are shown for one representative control (control, left) and β-catenin knockout mouse (β-catenin^−/−, right). (B) Average frequency of KLSF cells in control and β-catenin knockout mice. n = 11 for control, n = 13 for knockout. (C) Five-hundred KLSF cells from control or β-catenin knockout mice were transplanted together with competing bone marrow cells into lethally irradiated congenic recipients, and donor-derived chimerism was monitored for 20 weeks. Plots show representative donor-derived chimerism from individual mice at 20 weeks. (D) Graph of average donor-derived chimerism after long-term reconstitution (5–7 mice in each cohort, p = 0.0079, n = 3). In five independent experiments, the difference in chimerism between control and knockout transplants was 5-fold (500 cells transplanted), 2.8-fold (2000 cells transplanted), and 1.4-fold (6000 cells transplanted). (E) Contribution to differentiated lineages from control and β-catenin knockout cells following long-term bone marrow transplantation in peripheral blood. Results are representative of three independent bone marrow transplantation experiments, with 5–7 mice in each cohort. (F) Limiting-dilution competitive repopulation analysis of control and β-catenin knockout mice. Eight mice were transplanted at each dose (10, 20, 40, and 50 cells/recipient for a total of 64 initial recipients for both genotypes). Peripheral blood cells of all surviving recipients were analyzed 10 weeks after transplantation, and data were analyzed by L-calc software. CRU difference is 4-fold between control and knockout. (G) Secondary limiting-dilution competitive repopulation analysis of control and β-catenin knockout mice was carried out using the 50 cells/recipient group of primary transplanted mice at 21 weeks. Whole bone marrow cells were isolated from primary donors and transplanted into 5–7 recipients at each dose (10,000, 40,000, 80,000, and 160,000 cells/recipient for a total of 48 recipients). Peripheral blood cells of the secondary recipients were analyzed 10 weeks after transplantation, and data were analyzed by L-calc software. CRU difference is 6.3-fold between control and knockout. (H) Relative homing ability of control and β-catenin^−/− bone marrow cells. A total of 2 × 10⁶ bone marrow cells from control or β-catenin^−/− mice were transplanted into irradiated recipients (n = 4 in each group), and the presence of donor-derived cells was analyzed by FACS 6 hr posttransplant. In (B), (D), (E), and (H), error bars represent SEM.

Figure 3. Loss of β-Catenin Leads to Altered Characteristics of Leukemia Progression In Vivo

(A) Experimental strategy to model BCR-ABL-induced leukemia. c-kit-enriched cells were isolated from control or β-catenin knockout mice and infected with BCR-ABL retroviruses for 48 hr prior to transplantation of 2–5 × 10⁵ cells into lethally irradiated recipient mice. LTR, long terminal repeat. (B) Control and β-catenin knockout c-kit-enriched bone marrow cells were infected with either vector control or BCR-ABL retroviruses, and expression levels of BCR-ABL were analyzed by real-time PCR. Results were normalized to the housekeeping gene β-actin. Data are an average of five independent samples. Error bars represent SEM. (C) Survival curve of mice receiving BCR-ABL-infected control (+/+) or β-catenin knockout (−/−) c-kit-enriched bone marrow cells. Data shown are from one of four independent experiments, with 7–9 mice in each experimental group per experiment. (D–F) Gross examination of mice transplanted with vector-infected control, BCR-ABL-infected control, and β-catenin knockout cells. A majority of mice transplanted with BCR-ABL-infected control cells became weak but did not develop paralysis, unlike several of those transplanted with knockout cells (left panel). Appearance of lung in mice transplanted with vector-infected control, BCR-ABL-infected control, or BCR-ABL-infected knockout cells (middle panel). Relative splenomegaly of mice transplanted with vector-infected control, BCR-ABL-infected control, or β-catenin knockout cells (right).

Figure 4. Loss of β-Catenin Leads to an Altered Pattern of Leukemic Cell Infiltration in the Lung and Liver

Hematoxylin/eosin staining of tissue sections from lung (left panels) and liver (right panels) of mice transplanted with vector-infected control, BCR-ABL-infected control, or BCR-ABL-infected knockout cells. Arrowheads point to infiltration of leukemic cells into tissues. Mice transplanted with BCR-ABL-infected β-catenin knockout cells show either reduced infiltration (bottom row left [lung], black arrowhead) or a different pattern of infiltration (bottom row right [liver], black arrowhead). The scale bar represents 100 μm.

Figure 5. Loss of β-Catenin Impairs In Vivo Progression of CML but Allows Normal Development of ALL

(A) Histologic appearance of leukemic cells in spleens of the majority of mice transplanted with BCR-ABL-infected control and β-catenin knockout cells. Arrowheads point to an increased frequency of immature and mature granulocytic cells in the spleens of mice transplanted with BCR-ABL-infected control cells, and increased lymphocytic cells in the spleens of mice transplanted with BCR-ABL-infected β-catenin knockout cells. Scale bar represents 100 μm. (B) Bone marrow cells from mice receiving BCR-ABL-infected control or β-catenin knockout cells were analyzed by flow cytometry for increased presence of granulocytic (Mac-1⁺Gr-1⁺ cells shown in upper panels) or lymphocytic lineages (B220⁺ cells shown in lower panels). (C) Cumulative data from four experiments indicating type of leukemia induced by BCR-ABL in mice transplanted with control or β-catenin knockout (−/−) cells. Each circle represents an individual mouse. CML, chronic myelogenous leukemia; ALL-B, B cell acute lymphocytic leukemia; ALL-T, T cell acute lymphoblastic leukemia. (D) Distribution of CML and ALL induced by BCR-ABL when using control or β-catenin knockout c-kit-enriched cells. *p = 0.005 (t test from four experimental groups).

Figure 6. Loss of β-Catenin Impairs BCR-ABL-Induced Renewal of CML Stem Cells

(A) A schematic of the strategy to determine relative survival and replating efficiency after BCR-ABL infection in control and β-catenin knockout cells. (B) c-kit-enriched cells from control or β-catenin knockout mice were infected with vector or BCR-ABL and stained with annexin-V and 7-AAD. Annexin-V⁺7-AAD⁻ cells were quantified to determine frequency of apoptotic cells. Results are representative of four independent samples. (C and D) Long-term serial replating. c-kit-enriched control or β-catenin^−/− cells were infected with vector or BCR-ABL, and 5000 GFP-positive cells were sorted into wells containing methylcellulose media to assess primary colony formation. Colony numbers were counted on days 8–10. Cells were then harvested and counted, and 5000 cells were replated for a second, third, and fourth time, and colonies were counted on days 8–10 after each replating. White bars represent B lineage colonies. Results shown are an average of three independent experiments. Error bars show SEM. (E) BCR-ABL-infected control and β-catenin^−/− colonies were collected after the third and fourth plating and stained with the B cell marker B220. The plot shown is a representative FACS from the fourth plating. (F) In vivo self-renewal assay for CML stem cells. Irradiated recipient mice were transplanted with 10,000 KLSGFP⁺ cells (CML stem cells) sorted from control or β-catenin knockout CMLs. Survival curves show the frequency of recipients succumbing to disease after receiving KLS cells from wild-type CML (solid line) or β-catenin null CML (dashed line) stem cells (total number of mice transplanted = 12). In (B), (C), and (D), error bars represent SEM.

Figure 7. Differential Wnt Reporter Activity in CML versus ALL

(A) Leukemic cells from BCR-ABL-induced CML or ALL were sorted for GFP and then infected with a lentiviral Wnt reporter expressing dsREd (BATRED) upon Wnt activation. Cytospins were prepared 48 hr later and observed by immunofluorescence microscopy. (B) Genomic DNA was extracted from Wnt-reporter-infected CML or ALL cells, and viral integration was checked by PCR analysis using specific primers designed to amplify dsREd. The Wnt reporter plasmid served as a positive control. Data shown in both (A) and (B) are representative of three independent experiments. (C) GFP⁺ leukemic cells from β-catenin^−/− ALL were sorted, infected with the BATRED reporter virus, and treated with control media or 1 mM lithium chloride for 48 hr. Cells were then cytospun, and reporter activity was analyzed by immunofluorescence. Data are representative of two independent experiments. In (A) and (C), the scale bar represents 10 μm and 3 μm, respectively.

Figure 8. BCR-ABL-Induced Phosphorylation of Stat5α Is Reduced in the Absence of β-Catenin

(A and C) c-kit enriched control or β-catenin^−/− cells were infected with BCR-ABL for 48 hr. Infected cells (GFP positive) were sorted and cultured for another 3–4 days, and then cytospins were prepared and stained with indicated antibodies. Data are representative of three independent experiments. In (A) and (C), the scale bar represents 10 μm. (B and D) Quantification of fluorescence intensities/cell for phosphorylated (B) and total Stat5α (D) using Metamorph software. In control cells, phosphorylation of Stat5α was significantly induced following BCR-ABL transduction (p = 0.003). In contrast, BCR-ABL transduction in β-catenin^−/− cells did not induce Stat5α phosphorylation over vector and showed reduced phosphorylation compared to BCR-ABL-infected control cells (p = 0.008). (E) Relative levels of protein tyrosine phosphorylation in CML and ALL cells from control or β-catenin^−/− leukemias. GFP⁺ leukemia cells from CML or ALL of each genotype were sorted, and protein lysates were analyzed by western blotting with an anti-phosphotyrosine antibody. (F) Real-time PCR analysis of BCR-ABL mRNA levels in control and β-catenin^−/− leukemia cells. (G) Relative levels of BCR-ABL protein in CML and ALL from control or β-catenin^−/− leukemias. GFP⁺ leukemia cells from CML or ALL of each genotype were sorted, and protein lysate was analyzed by western blotting with an anti-abl antibody. Data are representative of two to three independent experiments. In (B), (D), and (F), error bars represent SEM.