Kinetics of normal hematopoietic stem and progenitor cells in a Notch1-induced leukemia model

Abstract

The predominant outgrowth of malignant cells over their normal counterparts in a given tissue is a shared feature for all types of cancer. However, the impact of a cancer environment on normal tissue stem and progenitor cells has not been thoroughly investigated. We began to address this important issue by studying the kinetics and functions of hematopoietic stem and progenitor cells in mice with Notch1-induced leukemia. Although hematopoiesis was progressively suppressed during leukemia development, the leukemic environment imposed distinct effects on hematopoietic stem and progenitor cells, thereby resulting in different outcomes. The normal hematopoietic stem cells in leukemic mice were kept in a more quiescent state but remained highly functional on transplantation to nonleukemic recipients. In contrast, the normal hematopoietic progenitor cells in leukemic mice demonstrated accelerated proliferation and exhaustion. Subsequent analyses on multiple cell-cycle parameters and known regulators (such as p21, p27, and p18) further support this paradigm. Therefore, our current study provides definitive evidence and plausible underlying mechanisms for hematopoietic disruption but reversible inhibition of normal hematopoietic stem cells in a leukemic environment. It may also have important implications for cancer prevention and treatment in general.

Figure 1

Experimental design and establishment of a murine leukemia model by Notch1 overexpression. Lin⁻Sca-1⁺ cells (CD45.2⁺) were isolated from BM of C57BL/6J mice. A total of 10⁷ BMNCs (CD45.1⁺) from B6.SJL mice were transplanted into lethally irradiated recipients (10 Gy, C57BL/6J) either with (leukemia group) or without (the control group) 10⁶ cells that were transduced with the MSCV-ICN1-IRES-GFP vector. The transplanted normal hematopoietic cells (CD45.1⁺GFP⁻) from the recipients were analyzed and sorted by flow cytometry at different time points (A). Resultant data are shown in all the figures except Figure 5. An additional control with a higher dose of normal congenic BMNCs was included in some experiments (supplemental Figure 2). A T-ALL feature of the affected mice is shown by an immature T-cell phenotype (supplemental Figure 1). The hosts receiving Notch1-transduced cells displayed leukemia symptoms, including increases of total leukemic cells (CD45.2⁺GFP⁺) in PB (B), WBCs in PB (C), and BM cellularity (D). Values are mean ± SD. *P < .05 (t test). The 100% penetrance of the leukemic development is shown in the survival curve (E). P < .01 (n = 15 each, Kaplan-Meier analysis).

Figure 2

Growth kinetics of normal hematopoietic cells in leukemia-bearing mice. The percentages of normal hematopoietic cells (CD45.1⁺) in the leukemia group were monitored in both PB (A) and BM (B) 2 and 4 weeks after transplantation. In addition, the frequencies of GFP⁻CD45.2⁻CD45.1⁺Lin⁻ and GFP⁻CD45.2⁻CD45.1⁺ LKS cells in BM were measured 1, 2, and 3 weeks after transplantation (C-D). P < .05 (n = 5/each, t test). Frequencies and absolute numbers of different hematopoietic cell populations within GFP⁻CD45.2⁻CD45.1⁺ BMNCs 4 weeks after transplantation are shown in the graphs (E-F). *P < .05 (t test). **P < .01 (t test). The BM cellularity of tibia, femur, iliac, and humerus was calculated as 40% of the total BM cellularity of a young adult C57BL/6J mouse (based on our own data). The data shown here represent 1 of 4 independent experiments (n = 6 or 7/group). Values are mean ± SD.

Figure 3

In vitro clonal growth of normal hematopoietic cells from a leukemic environment. Two weeks after transplantation, the mice were killed and BM was harvested. The normal hematopoietic cells (CD45.1⁺GFP⁻)were doubly sorted with near 100% purity for in vitro clonal assays (A). For the CFC assay to measure committed HPCs (B), the sorted CD45.1⁺GFP⁻ cells were cultured in the defined methylcellulose medium supplemented with a cytokine cocktail. Mix, GM, G, M, and E represent CFC-mix (> 2 lineages), CFC-granulocyte/monocyte, CFC-granulocyte, CFC-monocyte, and BFU-erythrocyte, respectively. Values are mean ± SD. *P < .05 (t test). **P < .01 (t test). In addition, the CAFC assay with limiting dilution at day 35 during the long-term culture was used to measure the more primitive hematopoietic cells, and the result is shown as a representative dataset from 3 experiments with similar results (C).

Figure 4

Long-term reconstitution of normal hematopoietic cells from leukemic marrow in new nonleukemic hosts. The hematopoietic regeneration of the normal hematopoietic cells from leukemic or control mice were examined in secondary nonleukemic recipients using the cBMT assay, in which equal numbers of test (CD45.1⁺GFP⁻) and competitor cells (CD45.1⁺/.2⁺) were cotransplanted into lethally irradiated congenic recipients (CD45.2⁺). The overall reconstitution levels of normal HSCs from the primary recipients were monitored within 6 months after transplantation (A). Multilineage differentiation of the engrafted cells was analyzed 6 months after transplantation (B). GM, T, and B indicate lineages for myeloid (Mac-1⁺), T (CD3⁺), and B (B220⁺) cells, respectively. Six months after cBMT, the overall representation of CD45.1⁺GFP⁻ cells, multilineage analysis, and different hematopoietic cell subsets in BM were also quantified (C-E). *P < .05; **P < .01 (n = 5-7/group, t test). Data are from 1 of 3 experiments with similar results.

Figure 5

Proliferative response of normal HPCs to leukemic cells in vivo. (A) CFSE assay. A total of 10⁸ BMNCs (C57BL/6J, CD45.2⁺) labeled with CFSE were coinjected with or without 5 × 10⁶ Notch1-induced leukemia cells (CD45.2⁺GFP⁺) into lethally irradiated recipient mice (SJL.B6, CD45.1⁺). BMNCs were harvested 72 hours after the transplantation to assess the number of cell divisions. BMNCs were stained with lineage markers and Sca-1, and CFSE-labeled cells were analyzed in the gate for CD45.1⁺Lin⁻Sca-1⁺. A representative figure of the flow cytometric analysis is shown. Blue peaks on the right represent undivided cells (parent cells); each peak toward the left side, one cell division or generation. The figure shown is from 1 of 4 experiments with similar results. The proliferation index in the CD45.1⁺Lin⁻Sca-1⁺ population is shown in the graph. *P < .05. (B) BrdU assay. BrdU was injected intraperitoneally 72 hours after transplantation. BM cells were analyzed 2 hours later, and the proportion of BrdU⁺CD45.1⁺Lin⁻Sca-1⁺ cells was analyzed with flow cytometry. A representative plot is shown, and the figure shown is from 1 of 6 experiments with similar results. Values are mean ± SD. **P < .01 (n = 3 mice/group, t test). (C) 5-FU assay. A total of 150 mg/g 5-FU was injected 12 hours before BM was harvested, and then CD45.1⁺GFP⁻ cells were isolated for the CFC assay 72 hours after transplantation. The total CFC colonies were counted at day 14 with microscopy. The data shown are from 1 of 2 independent experiments (n = 3-5 wells). **P < .01.

Figure 6

Mitotic quiescence of the primitive hematopoietic cells in the leukemic hosts and expression of several cell-cycle regulators. As illustrated in Figure 1A, 10⁷ BMNCs from B6.SJL mice were transplanted with or without 10⁶ Notch1-induced leukemia cells (CD45.2⁺GFP⁺) into lethally irradiated C57BL/6J recipients. At the 2-week time point, LKS cells from CD45.1⁺ BMNCS were sorted for staining with PI to assess the general cell-cycle profile (G₀/G₁ vs S/G₂ + M) or staining with PY in conjunction with Hoechst 33342 (HO) to specifically determine the portion of cells in G₀ vs G₁ with flow cytometry. An aliquot of the cells was also used to examine the expression of several cell-cycle regulators with real time RT-PCR. (A) PI staining. A representative figure is shown from 1 of 2 experiments with similar results. (B) PY staining. Cells residing in G₀ appear at the bottom of the G₀/G₁ peak as shown in the representative plot. Values are mean ± SD. **P = .017 (t test). (C) Expression of cell-cycle regulators. The CD45.1⁺ LKS cells were sorted directly into lysis buffer for real-time RT-PCR analysis. Data shown are the ratios between leukemic and control groups from 1 of 2 independent experiments with similar results. Values are mean ± SD. *P < .05 (t test). **P < .01 (t test).