Expression of activated M-Ras in hemopoietic stem cells initiates leukemogenic transformation, immortalization and preferential generation of mast cells


Cultures of purified hemopoietic stem cells transduced with an activated mutant of M-Ras contained abnormal cells that, despite the presence of only low levels of growth factors, generated large, dense colonies of macrophages and blast cells. Cells from these colonies survived and grew continuously in the absence of growth factors and generated clonal cell-lines that were mainly composed of well-differentiated mast cells, with a low frequency of undifferentiated cells. When transplanted into sublethally irradiated syngeneic mice, four out of four such clones gave rise to a systemic mastocytosis and mast-cell leukemia. However, the donor clones also generated low percentages of cells with the morphological and cell-surface characteristics of erythrocytes, granulocytes, monocytes and T- and B-lymphocytes. These data indicate that signals downstream of activated M-Ras are sufficient to transform hemopoietic stem cells, and while preserving their capacity to generate other cell-lineages in vivo, result in preferential generation of mast cells.


Expression of an activated mutant of M-Ras in normal bone marrow cells resulted in immortalized lines of mast cells that survived and grew continuously in the absence of exogenous growth factors and gave rise to a mastocytosis and mast-cell leukemia when transplanted into mice (Guo et al., 2005). M-Ras is related to the p21 Ras proteins, sharing about 50% overall amino-acid sequence identity, and being identical or similar in the Switch-I region that interacts with effectors, and the Switch-II region that interacts with activators and effectors (Ehrhardt et al., 2002). M-Ras is activated by some of the guanine-nucleotide exchange factors that activate p21 Ras, namely mSos1 and RasGRF (Ohba et al., 2000) but not others such as Ras-GRP-1, the latter accounting for its lack of activation by ligation of the antigen receptors on T- or B-cells (Ehrhardt et al., 2004). In terms of negative regulators, the p21 Ras GTPase-activating factors p120GAP, NF-1 and Gap-1m also inactivated M-Ras (Ohba et al., 2000). M-Ras is expressed in lympho-hematopoietic cells (Ehrhardt et al., 1999; Louahed et al., 1999) and is activated more strongly than H-Ras or N-Ras by growth factors such as interleukin-3 (IL-3) and colony-stimulating factor-1 (Louahed et al., 1999; Ehrhardt et al., 2004). Expression of activated mutants of M-Ras leads to transformation of cell-lines of various types including fibroblasts (Kimmelman et al., 1997; Ehrhardt et al., 1999), mammary epithelial cells (Ward et al., 2004; Zhang et al., 2004) and hematopoietic cells (Ehrhardt et al., 1999; Louahed et al., 1999). M-Ras is overexpressed in carcinomas from breast, uterus, thyroid, stomach, colon, lung, kidney and rectum (Yang et al., 2005). Overexpression of M-Ras had also been reported to correlate with transformation of a murine melanocyte line (Wang et al., 2000).

The leukemogenic effect of expression of activated M-Ras in bone marrow cells indicated that constitutive activity of M-Ras signaling paths was sufficient for transformation of primary cells. However, this raised the question (Passegue et al., 2003) of whether the neoplastic mast cells arose through transformation of committed mast cell progenitors, or of pluripotential hemopoietic stem cells (HSC) that preferentially generated mast-cells through M-Ras dependent signals. Therefore, we have investigated the effects of expressing activated M-Ras in purified HSC.

Purified populations of Sca-1+c-kit+Flk-2 lin bone marrow cells that were highly enriched for primitive hemopoietic stem cells (Christensen and Weissman, 2001) were first cultured in a mixture of steel locus factor (SLF), interleukin-6 (IL-6) and thrombopoietin (Tpo), and transduced with a bicistronic retroviral vector expressing activated M-Ras and GFP in the presence of the same cytokines. At day 4, cells were washed to remove growth factors and transferred to cultures containing Tpo and IL-6, with or without SLF. SLF is absolutely required for normal mast cell development and stimulates the differentiation of mast cells from bone marrow progenitors (Kitamura and Go, 1979). In contrast, while both IL-6 and Tpo can stimulate the entry of stem cells into cell-cycle (Drachman, 2000), neither can induce the in vitro differentiation or proliferation of bone marrow mast cells (Hu et al., 1997; Drachman, 2000). However, even in the absence of SLF, by week 2, GFP-positive cells that resembled mast cells had appeared, and, by week 4, the population consisted mainly of GFP-positive, heavily granulated mast cells that expressed both c-Kit and FcɛRI (Figure 1A). These cultures continued to grow and generate mast cells in the absence of any exogenous growth factors.

Figure 1

Purified hematopoietic stem cells transduced with constitutively active M-Ras generated in vitro lines of factor-independent cells that were predominantly mast cells. Bone marrow cells from Balb/c mice (Guo et al., 2003) were depleted of Mac-1+ cells by magnetic cell-sorting according to the manufacturer's instructions (Miltenyi Biotech) and Sca-1+c-kit+Flk-2 cells negative for the lineage markers CD3, B220, Mac-1, Gr-1 and Ter119 (Christensen and Weissman, 2001) were isolated using the FACSVantage (BD Biosciences). The purified stem cells were cultured in the indicated mixtures of growth factors for 24 h and were transduced with a retroviral vector that expressed activated Q71L M-Ras and GFP from the same mRNA (Guo et al., 2003). Cells were cultured in DMEM with 10% FBS. (A): Sca-1+c-Kit+ Flk-2lin cells were transduced with Q71L M-Ras and GFP as described and 3 days later were washed three times and transferred to medium containing Tpo and IL-6. Shown are nonadherent cells that were harvested at the end of weeks 1, 2, 4 and 8 as indicated, cytospun onto glass slides and stained with Hema 3 stain. Note that typical mast cells dominated the cultures by week 4. (B): Purified sca-1+c-Kit+ Flk-2lin cells were transduced with Q71L M-Ras and GFP by coculture in the presence of 10 ng/ml of each IL-3, IL-6 and SLF. At day 4, 125 μl of medium and nonadherent cells were transferred into 2 ml of medium supplemented with 0.9% methylcellulose but no additional growth factors. After 7 days colonies were plucked and transferred individually into cultures in medium lacking added growth factors. Shown in (a) is a photomicrograph of one such culture 2 weeks later. Note adherent cells and smaller, nonadherent cells that have migrated from the colony seen at bottom right. Nonadherent cells were passaged in medium alone for a further 2 weeks. Shown are (b), smear stained with Hema 3 stain showing granulated cells together with an infrequent ‘lymphocyte-like’ cell (arrow), and (c, d), flow cytometric analyses demonstrating that the bulk of cells exhibited cell-surface expression of FcɛRI and c-Kit.

To determine whether M-Ras had transformed HSC, we next generated cloned lines from HSC that we could test for their ability to generate cells of multiple lineages. We transduced purified HSC with the bicistronic vector expressing activated M-Ras and GFP by co-culture with virus-producing Bosc 23 cells in the presence of IL-3, IL-6 and SLF. At day 4, 125 μl of nonadherent cells and medium were transferred to 2 ml of medium with 0.9% methylcellulose (Stemcell Technology, Vancouver, Canada). We observed the growth of small numbers of dense, spherical colonies, all of which expressed GFP. It was noteworthy that under these conditions – with concentrations of growth factors of <0.6 ng/ml – only transduced cells formed large, compact colonies. There were a small number of GFP-negative colonies but these were contained only small numbers of cells and were diffuse rather than compact. Thus the transduced cells that generated these large colonies were abnormal in their ability to grow and generate progeny in the presence of these low levels of growth factors. We picked both GFP-positive and GFP-negative colonies and transferred them individually to cultures containing medium alone. Cells from the small GFP-negative colonies rapidly died. However, cells from the large GFP-positive colonies survived and proliferated. Microscopic inspection of these cultures revealed that the colonies contained a diversity of cell-types that included adherent cells resembling macrophages and clusters of small nonadherent cells (Figure 1Ba). In 4/5 instances, these nonadherent cells continued to proliferate without apparent limit in the absence of any exogenous growth factors. Examination of these cell populations by histochemistry and flow cytometry revealed that over 99% of these cells exhibited metachromatic granules typical of mast cells and expressed both c-Kit and FcɛRI (Figure 1Bb–d). However, there was also a minor population (<1%) of undifferentiated blast cells (Figure 1Bb). There was no evidence that mast cells expressing activated M-Ras produced growth factors (Guo et al., 2005). Moreover, bone marrow cells that were infected with a retroviral vector expressing activated M-Ras and cultured in the absence of exogenous growth factors without drug-selection, gave rise to only GFP-positive cells, indicating that cells expressing activated M-Ras could not support the survival or growth of normal hemopoietic cells (unpublished observations). Different clones exhibited different levels of GFP (data not shown), indicating that they expressed different levels of the bicistronic mRNA encoding activated M-Ras and GFP and were likely to have been derived from cells in which the retroviral vectors had integrated at different sites. The observation that the colonies initially contained adherent cells resembling macrophages (Figure 1Ba), and later, in liquid cultures, gave rise to mast cells (Figure 1Bb–d), indicated that the expression of activated M-Ras had transformed multipotential cells that could generate multiple myeloid lineages.

To further investigate the potential of these clones for self-renewal and differentiation to different lineages, we transplanted 1 × 106 of each of the four clonal cell-lines into sublethally irradiated (8.0 Gy) syngeneic mice. By 5 months, all four groups of mice exhibited wasting and abdominal distension and were sacrificed. All mice exhibited gross splenomegaly and hepatomegaly and these organs, together with the bone marrow, lung and gut were infiltrated with GFP-positive fluorescent cells derived from the injected clone (data not shown). Staining of blood smears and imprints of the spleen with Toluidine Blue revealed mast cells with the typical cytoplasmic granules (data not shown). Flow cytometric analysis of the blood revealed many GFP-positive donor-derived cells, the majority of which expressed the mast cell markers c-Kit and FcɛRI (Figure 2A). However, as exemplified in the analysis of the recipient shown in Figure 2A, among the GFP-positive cells, a minority expressed markers characteristic of other cell-lineages such as the erythroid – (TER119, 7.3%), the myeloid – (Mac-1, 8.8%; Gr-1, 10.1%), the T-lymphocyte – (CD3, 3.4%) and the B-lymphocyte lineage (CD19, 1.2%).

Figure 2

Stem cells transduced with activated M-Ras generate clones giving rise to multiple cell-lineages when transplanted into syngeneic mice. The clones derived from HSC transduced with activated M-Ras described in Figure 1 were cultured in the absence of exogenous growth factors for 12 weeks and 1 × 106 cells of each clone were injected intravenously into each of a group of 4 gamma-irradiated (8.0 Gy) syngeneic mice. These were sacrificed 5 months later. (a) Shows flow-cytometric analysis of blood cells from one typical recipient. Antibodies for the lineage-markers Mac-1, Gr-1, TER119, CD3 and CD19 were directly conjugated with PE, while antibodies for c-Kit were conjugated to Biotin and used with streptoavidin-APC; FcɛRI was detected by incubating cells with mouse Ig E and anti-mouse Ig (Fab)2. Controls were stained with secondary reagents alone. (b): GFP+ cells were sorted from the bone marrow of a typical leukemic recipient. The sorted GFP+ cells were then resorted to ensure that all were GFP+ and the double-sorted GFP+ cells were cytospun onto glass slides and stained with hema 3 stain, revealing mast cells and other cell-types.

To confirm that the donor clones gave rise to cells of multiple lineages, GFP-positive cells were sorted from bone marrow of leukemic mice. This cell population was then resorted to ensure that all cells were GFP-positive. The double-sorted GFP-positive bone marrow cells contained cells with the appearance of mast cells, immature myeloid cells, erythrocytes and megakaryocytes (Figure 2B). Thus, while the clonal cell-lines generated mainly mast cells in vitro, in vivo, they also revealed a capacity to generate cells of other lineages.

When GFP+ cells from the primary recipient mice were transplanted into sublethally irradiated syngeneic mice, the secondary recipients developed a similar disease that was dominated by mast cells. Importantly in the blood, the population of GFP-positive cells again included small populations of cells that were clearly positive for Mac-1, Gr-1, TER119 or B220 cells (data not shown). The original multipotent colony-forming cells from which the clones had been derived were transformed. Thus they were at least partially independent of growth factors for their proliferation and survival in vitro, as were the mast cells that they generated. However, the absence of cells of the erythroid, lymphoid or other myeloid lineages in the clonal cultures suggests that over-activity of the M-Ras pathway was not sufficient to support the survival, proliferation and differentiation of progenitors of these lineages in vitro. The emergence of these lineages upon transplantation of the clones into mice, indicates that the transformed stem cells or their progeny retained responsiveness to normal signals from their environment. Moreover, it is conceivable that, due to cell-type differences in signaling pathways, cells of some of the lineages generated by the leukemic stem cell may behave normally in terms of growth and survival. Interestingly, expression in Sca-1+ Lin cells of BCR-ABL, also resulted in clones that gave rise to only mast cells in vitro, but to multiple lineages in vivo (Jiang et al., 2002). BCR-ABL-induced transformation is blocked by expression of a dominant negative p21 Ras (Cortez et al., 1996) that will also block activation of M-Ras (Ehrhardt et al., 1999, 2002; Louahed et al., 1999), raising the possibility that over-activity of the M-Ras signaling pathway was also involved here.

Collectively our observations indicate first, that a mast cell leukemia can arise from transformation of a primitive hemopoietic cell that retains the capacity to generate other lineages, including the erythroid, myeloid lymphoid lineages, an observation supported by studies of human mastocytosis (Akin et al., 2000), and second, that over-activity of the M-Ras signaling path appears sufficient to initiate transformation of hemopoietic stem cells.


  1. Akin C, Kirshenbaum AS, Semere T, Worobec AS, Scott LM, Metcalfe DD . (2000). Exp Hematol 28: 140–147.

  2. Christensen JL, Weissman IL . (2001). Proc Natl Acad Sci USA 98: 14541–14546.

  3. Cortez D, Stoica G, Pierce JH, Pendergast AM . (1996). Oncogene 13: 2589–2594.

  4. Drachman JG . (2000). Curr Opin Hematol 7: 183–190.

  5. Ehrhardt A, David MD, Ehrhardt GR, Schrader JW . (2004). Mol Cell Biol 24: 6311–6323.

  6. Ehrhardt A, Ehrhardt GR, Guo X, Schrader JW . (2002). Exp Hematol 30: 1089–1106.

  7. Ehrhardt GR, Leslie KB, Lee F, Wieler JS, Schrader JW . (1999). Blood 94: 2433–2444.

  8. Guo X, Gerl RE, Schrader JW . (2003). J Biol Chem 278: 22237–22242.

  9. Guo X, Schrader KA, Xu Y, Schrader JW . (2005). Oncogene 24: 2330–2342.

  10. Hu ZQ, Kobayashi K, Zenda N, Shimamura T . (1997). Blood 89: 526–533.

  11. Jiang X, Ng E, Yip C, Eisterer W, Chalandon Y, Stuible M et al. (2002). Blood 100: 3731–3740.

  12. Kimmelman A, Tolkacheva T, Lorenzi MV, Osada M, Chan AM . (1997). Oncogene 15: 2675–2685.

  13. Kitamura Y, Go S . (1979). Blood 53: 492–497.

  14. Louahed J, Grasso L, De Smet C, Van Roost E, Wildmann C, Nicolaides NC et al. (1999). Blood 94: 1701–1710.

  15. Ohba Y, Mochizuki N, Yamashita S, Chan AM, Schrader JW, Hattori S et al. (2000). J Biol Chem 275: 20020–20026.

  16. Passegue E, Jamieson CH, Ailles LE, Weissman IL . (2003). Proc Natl Acad Sci USA 100(Suppl 1): 11842–11849.

  17. Wang D, Yang W, Du J, Devalaraja MN, Liang P, Matsumoto K et al. (2000). Oncogene 19: 4647–4659.

  18. Ward KR, Zhang KX, Somasiri AM, Roskelley CD, Schrader JW . (2004). Oncogene 23: 1187–1196.

  19. Yang J, Chatterjee-Kishore M, Staugaitis SM, Nguyen H, Schlessinger K, Levy DE et al. (2005). Cancer Res 65: 939–947.

  20. Zhang KX, Ward KR, Schrader JW . (2004). Mol Cancer Res 2: 242–255.

Download references


We thank Andy Johnson for fluorescence-activated cell sorting; Lea Wong and Samantha Kleczkowski for animal care. This work was supported by grants from The Canadian Institute of Health Research.

Author information

Correspondence to J W Schrader.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Guo, X., Stratton, L. & Schrader, J. Expression of activated M-Ras in hemopoietic stem cells initiates leukemogenic transformation, immortalization and preferential generation of mast cells. Oncogene 25, 4241–4244 (2006).

Download citation


  • M-Ras
  • hematopoietic stem cells
  • leukemia
  • immortalization
  • mast cell

Further reading