Diverging fates of cells of origin in acute and chronic leukaemia

The large difference in phenotypes among tumour populations may stem from the stochastic origin of tumours from distinct cells – tumour cells are assumed to retain the phenotypes of the cells from which they derive. Yet, functional studies addressing the cellular origin of leukaemia are lacking. Here we show that the cells of origin of both, BCR/ABL-induced chronic myeloid (CML) and B-cell acute lymphoid leukaemia (B-ALL), resemble long-term haematopoietic stem cells (LT-HSCs). During disease-maintenance, CML LT-HSCs persist to function as cancer stem cells (CSCs) that maintain leukaemia and require signalling by the transcription factor STAT5. In contrast, B-ALL LT-HSCs differentiate into CSCs that correspond to pro-B cells. This transition step requires a transient IL-7 signal and is lost in IL-7Rα-deficient cells. Thus, in BCR/ABLp185+ B-ALL and BCR/ABLp210+ CML, the final phenotype of the tumour as well as the abundance of CSCs is dictated by diverging differentiation fates of their common cells of origin.

BCR/ABLp210 + mice (F) and wildtype and moribund BCR/ABLp185 + mice (G). The data show fold-differences above an internal control gene (Ube2d2) and are normalized to the gene expression in wildtype Gr-1 + /Mac-1 + (F) and wildtype CD19 + cells (G). In summary, BCR/ABLp210 + Gr-1 + /Mac-1 + cells display the lowest difference in their gene expression as compared to wildtype Gr-1 + /Mac-1 + cells. Likewise, BCR/ABLp185 + CD19 + cells show the lowest divergence as compared to wildtype CLPs and wildtype CD19 + cells.    Scheme of all experiments performed using BCR/ABLp210-CSCs (A) and BCR/ABLp185-CSCs (B). BM-fractions with presumable CSCs (either LT-HSCs or HSC-depleted leukemic pools) were FACS-purified from each single primary transplanted moribund mouse and retransplanted into secondary recipient mice. The numbers of mice used for secondary transplantation (per each primary diseased mouse) differed depending on the yields of leukemic cells that could be purified from primary transplanted mice and is indicated below.
Total numbers of all secondary transplanted mice employed (n) are indicated in brackets. (C) CSCs from caSTAT5-induced leukemia behave according to model 3 ( Figure 5A). A mixture of caSTAT5 + LT-HSCs with wildtype HSC-depleted BM induces a CML-like disease (upper panel) in 9/9 serially transplanted mice. Mixture of wildtype LT-HSCs with frequent, leukemic caSTAT5 + HSC-depleted BM fails to cause leukemia for 14 months (n=9, lower panel, see also Table S3).
(A) In vitro cultivation of BCR/ABLp185 + precursor-B cells after isolation from moribund primary transplanted mice. One representative experiment is depicted (n=4). (B) BCR/ABLp185 CSCs induce a B-ALL in secondary recipients. One representative set of data is depicted (n=10). (C) Cultivation of BCR/ABLp210 LT-HSCs with SCF/Tpo/IGF-II and FGF-1 and FACS-sorting into forward scatter high (FSC high ) cells allows sustained expansion of these cells. One representative set of data is depicted (n=3). (D) Cumulative cell number of caSTAT5 CSCs and wildtype LT-HSCs during long-term culture in serum-free medium containing SCF/Tpo/IGF-II and FGF-1 and repeated FACS-sorting and replating of forward scatter high (FSC high ) cells upon confluency.  Table S I | LT-HSCs isolated by LSK/Thy1.2 low /Flt3and LSK/CD150 + /Flt3sorting conditions contribute equally well to long-term repopulation in lethally irradiated mice.
1,000 cells from seven different FACS-purified subpopulations (as indicated on top of the figure) were isolated from wildtype mice and transplanted along with 250,000 HSC-depleted BM cells into lethally irradiated wildtype mice. The cases of long-term engraftment are depicted for each transplanted subpopulation. (A) Summary of FACS-analysis of different surface proteins expressed in cultured wildtype LT-HSCs as well as in BCR/ABLp210 + and BCR/ABLp185 + CSCs. BCR/ABLp210 + CSCs (p210 CSCs) express all tested HSC-markers and completely lack mature markers. The same is true for BCR/ABLp185 + pre-leukemic HSC-derived cells (p185 HSC) expanded in medium lacking IL-7. In contrast, BCR/ABLp185 + CSCs isolated from primary B-ALL-diseased mice lack stem cell markers and express B-cell lineage markers (p185 CSCs). (B) Summary of FACS and confocal fluorescence microscopy analysis of cultivated BCR/ABLp210 + CSCs for different surface proteins involved in HSC/niche interactions and homing to the bone marrow.

(A) Summary of FACS-analysis of different surface proteins expressed in cultured wildtype
LT-HSCs (wt HSCs) as well as in caSTAT5-overexpressing CSCs (caSTAT5 CSCs). They express all tested HSC-markers and completely lack mature markers. (B) Summary of FACS and confocal fluorescence microscopy analysis of cultivated caSTAT5 + CSCs for different surface proteins involved in HSC/niche interactions and homing to the bone marrow. The forward scatter high and low fractions differ in CD44, CD105, CD150 and Osteopontin expression (caSTAT5 CSC small = FSC low caSTAT5 CSCs; caSTAT5 CSC big = FSC high caSTAT5 CSCs).

Movie I
The movie shows proliferation of large (FSC high ) BCR/ABLp210 CSCs. 3 frames/hour, 72 hours in total, magnification: 200x.
We completely omitted 5-FU pre-treatments of mice in all transplant experiments to avoid a bias to the myeloid lineage or a predominance of the potential target cell that originates cancer. BM from donor mice was flushed from the femurs and tibiae, pooled and filtrated.
BM-subpopulations were infected using stable producer cell lines carrying empty-IRES-GFP, BCR/ABLp210-IRES-GFP, BCR/ABLp185-IRES-GFP and caSTAT5-IRES-GFP as described previously (Moriggl et al, 2005). The stable producer cells (GP+E86 cell lines) were initially derived from mouse embryonic fibroblasts (MEFs) (Markowitz et al, 1988) and therefore represent a supportive microenvironment/niche for hematopoietic cells and embryonic stem (ES) cells. 2-3 x 10 6 infected total BM cells were injected via tail-vein into lethally (10 Gy) irradiated recipient mouse. All mice have successfully engrafted the donor BM during the observed time period. For transduction of LT-HSCs and CLMPs, these cells were FACS-sorted as described in the FACS analysis section and infected with the mentioned producer cell lines on 96-well-plates in analogous way to whole BM (Moriggl et al, 2005).
For secondary transplantations, wildtype or GFP + leukemic LT-HSCs and HSC-depleted wildtype or GFP + leukemic BM were isolated by FACS-sorting from diseased primary transplanted mice, mixed as described in Figure S5A and S5B and injected into lethally irradiated secondary B6/129F1 recipients. We have injected 1,000 GFP + leukemic LT-HSCs supplemented with 250,000 wildtype GFP -HSC-depleted BM cells or 1,000 GFPwildtype LT-HSCs mixed with 250,000 GFP + HSC-depleted leukemic cells. To avoid unwanted differences in leukemogenicity of the sorted leukemic cells, all transplanted mice (primary, secondary and tertiary transplants) were analyzed at the time point when sickness became visible. Similarly, mice that did not become sick in the expected latency time were observed over a period of at least 12 months before sacrificed. For evaluation of the cancer progressing models (cancer stem cell model vs. clonal evolution model), we have injected different limiting dilutions of the GFP + BM fractions isolated from BCR/ABLp210 and BCR/ABLp185 primary transplanted mice (by FACS-sorting). Injection of GFP + (leukemic cell population) cells only ensured that no "healthy cells" were used, so that the actual frequency of CSCs could be determined among all tumor cells.

Statistical analysis
Survival statistics have been analyzed using the log-rank test ( Figure 3B) and the comparison of infection rates was calculated by Tukey's multiple comparison test ( Figure S3A-C) using the GraphPad Prism software. In vivo limiting dilution analysis (Table 1) was performed using the ELDA Software at http://bioinf.wehi.edu.au/software/elda. Paired and unpaired student's t-tests have been used where appropriate. Confocal microscopy was performed on a LSM 510 Meta/Axiovert 200M system from Zeiss.

Preparation of single cell suspensions from mouse organs
The cells were grown on 4-well Lab-Tek Chamber Slides, fixed and stained according to a previously described protocol and mounted in VECTASHIELD (Vector Laboratories) supplemented with DAPI.

Inhibitor treatments
BCR/ABLp210 LT-HSC and BCR/ABLp185 pro-B cells were plated in triplicates at a density of 30,000 cells in 96-well plates (100 µl medium/well) supplemented with different concentrations of imatinib-mesylate, cyclopamine, AG490 or Stattic. On days 7-16, the cells were harvested and counted using a CASY cell counter (Roche).

In vitro cultivation, differentiation and STAT5-deletion of CSCs
Freshly sorted GFP + LT-HSC population from BCR/ABLp210-and caSTAT5-diseased mice was cultivated in serum-free medium (StemPro-34 SFM, GIBCO) containing Nutrient supplement (GIBCO), 2mM L-glutamine, Pen/Strep and supplemented with 10 ng/ml mSCF, 20 ng/ml mTPO, 20 ng/ml IGF-II, 10 ng/ml hFGF-1 (acidic) and 10 ng/ml heparin. Confluent outgrowth of "cobblestone"-like structures ( Figure S6C and Movie S1) was obtained after about 6-7 days. Thereafter, the smaller cells growing on top of the more adherent ones were removed from the 4-well plate by flushing with medium. The adherent cells were washed with 1x PBS, trypsinised and pooled together with the smaller cells. The cells were FACSsorted into a FSC high and a FSC low fraction and seeded at a density of 1,000 cells per 6 cm dish. This procedure of FSC high/low sorting was repeated at the time point when the cells were confluent again. The dish dimensions could be increased at every sorting/replating step.
To specifically delete STAT5 only in LT-HSCs from BCR/ABLp210-infected mice, we isolated the GFP + LT-HSC fraction as described in the sorting conditions and immediately transferred into 4-well plate and cultivated the cells with serum-free medium (StemPro-34 SFM, GIBCO) containing Nutrient supplement (GIBCO), 2mM L-glutamine, Pen/Strep and supplemented with 10 ng/ml mSCF, 20 ng/ml mTPO, 20 ng/ml IGF-II, 10 ng/ml hFGF-1 (acidic) and 10 ng/ml heparin and 10ng/ml IFNα (for activation of Cre-recombinase transcription via the Mx1 promotor and excision of the STAT5 gene) for 18 hours. Then the cells were washed, resuspended in PBS and mixed with HSC-depleted wildtype BM before injecting into lethally irradiated mice.