The role of MYC in bone and in the hematopoietic stem cell niche

The nuclear protein c-Myc is one of the most potent proto-oncogenes described. Its expression is found to be deregulated in more than 50 percent of human cancers, where it has been shown to control a number of biological processes including driving proliferation, cell growth and angiogenesis while inhibiting differentiation. The effects of c-Myc are highly cell type specific and include cell autonomous and non-cell autonomous processes. In order to examine the effect of excess amounts of c-Myc in bone, we have generated a mouse line expressing the tTA transactivator under the control of the mouse 2.3 kb col1a1 promoter (named Col1a1::tTA), which directs transgene expression to mature osteoblasts. By combining this transgene with the Tet-O-Myc allele, we generated a Col1a1::tTA; Tet-O-Myc double transgenic mouse model in which it is possible to inducibly express human c-MYC in osteoblasts during development and in the adult. We document that the used 'tet-off' system directs transgene expression to bones and that c-MYC can be completely repressed by treatment with doxycycline. Using this mouse model, we demonstrate that overexpression of c-MYC leads to remarkable changes in mutant bones, which depend greatly in their severity on the developmental time point of c-MYC induction. Unrestricted c-MYC expression, starting with the first emergence of mature osteoblasts in the fetal skeleton (∼E14.0), leads to lethality of the mutants at birth; their bones are shortened and thicker than in controls. Later induction of c-MYC allows survival, but leads to significant changes in bone morphology. Mutant bones develop a very irregular bone structure with signs of excessive bone formation. Surprisingly, this also correlates with definite signs of excessive osteoclast activity, which explains the formation of cavities seen in the bone matrix. Osteoblasts not only produce bone matrix, they are also very important niche cells for hematopoietic cells, especially for hematopoietic stem cells and B cells. We find that c-MYC overexpression severely impairs B lymphopoiesis and also observe signs of extramedullary hematopoiesis in the spleen, which is indicative of a compromised HSC environment in the bone marrow. Most strikingly, these animals develop frank osteosarcomas leading to a much reduced life span. Tumors arise in various bones, with the ribs and hips being the most prominent sites. These c-MYC driven tumors are very aggressive as is evident by their capacity to spontaneously metastasize to lung and liver, a rather rare phenomenon in mouse models, indicating that these tumors share important aspects of human osteosarcomas. Molecular analysis of c-MYC overexpressing osteoblasts has revealed a defect in their terminal differentiation program, rendering them incapable to progress from a relatively immature proliferative to a quiescent, calcified matrix producing state. To our knowledge this is the first report of experimental osteosarcoma formation, originating from a mature and lineage committed osteoblast population indicating that c-MYC may be able to de-differentiate more mature cells. In summary, our novel conditional mouse model showed that c-MYC expression in mature osteoblasts can lead to the formation of highly aggressive and metastatic osteosarcomas, a lethal disease in humans. This model will be the basis for further studies elucidating the molecular effects of c-MYC on de-differentiation and metastasis formation, processes, which remain poorly understood but are critical events in highly malignant cancers.

Prospective identification of hematopoietic and neural stem/progenitor cells using transgenic and lentiviral based approaches

Didier Surdez

Adult stem cells represent only a minor proportion of a given tissue and are difficult to amplify in vitro without affecting their biological properties. To circumvent this problem, we used a transgenic/lentiviral based approach to isolate and characterize these cells by driving a fluorescent reporter protein under the control of Spi2a, H2Kb or MELK promoter/enhancer elements, which were previously shown to be active in various stem cells. The first chapter focuses on Spi2a, a serine protease inhibitor (Serpin), which is known to be preferentially expressed in hematopoietic stem cells (HSCs). Taking advantage of the previously characterized minimal Spi2a promoter, we were able to determine its activity profile in the hematopoietic system. For this purpose, we infected HSCs with a lentivirus driving d2eGFP expression under the control of the minimal Spi2a promoter. These cells were then injected into lethally irradiated recipient mice that were analyzed at different time points (≥2 months) after transplantation. Despite variable d2eGFP expression in the hematopoietic compartment, we found a population of bright d2eGFP-expressing bone marrow cells (Spi2a0.1%) that were mostly restricted to HSCs and early myeloid progenitors. Spi2a0.1% cells displayed multilineage colony forming potential in vitro and exhibited long-term multilineage repopulating capacity in vivo, when a restricted numbers of these cells (≥ 300 cells) were transplanted into secondary recipient mice. Furthermore, we generated transgenic mice using a similar construct as used for the lentiviral approach. These mice indicated that the Spi2a promoter may be activated in non-hematopoietic adult stem cells, such as in the brain, the testis or the pancreas. In the second chapter, we describe a transgenic mouse line that expresses a green fluorescent protein variant (zFP) under the control of H2Kb promoter/enhancer element. Despite the broad zFP expression, transgenic HSCs expressed exceptionally high levels of zFP, allowing prospective isolation of a population highly enriched in HSCs by sorting the 0.2% of the brightest green cells from the enriched bone marrow of H2Kb-zFP mice. Up to 90% of zFPbright cells were also c-kithigh, Sca-1high, Linneg, Flk-2neg, which is a bona fide phenotype for long-term HSCs. Double-sorted zFPbright HSCs were capable of long-term multilineage reconstitution at a limiting dilution dose of approximately 12 cells, which is comparable to that of highly purified HSCs obtained by conventional multicolor flow cytometry. Thus, the H2Kb-zFP transgenic mice provide a straightforward and easy setup for the simple and highly efficient enrichment for genetically labeled HSCs. The last chapter is dedicated to MELK. This kinase of the snf1/AMPK family was recently identified to regulate the proliferation of multipotent neural progenitor cells and was also found to be expressed in hematopoietic stem cells and in spermatogonia. Using a transgenic mouse approach by expressing eGFP under a 3.5kb MELK promoter/enhancer element, we demonstrated that eGFP-expressing cells were localized in the subventricular zone (SVZ) of the lateral ventricle and in the subgranular zone of the dentate gyrus, the two major regions of adult neurogenesis. Moreover, the eGFP expression level in SVZ-derived cells positively correlated with their ability to form neurospheres. We also observed strong and restricted expression of eGFP in spermatogonia cells, indicating that MELK may play an important role in early spermatogenesis. Additionally, activation of the MELK promoter in hematopoietic stem/progenitor cells suggested a widespread implication of MELK in various adult stem cells. To conclude, we have generated three different reporter strains and one lentiviral based approach that allow for in vivo identification of stem/progenitor cells in different compartment such as the brain, the bone marrow or the testis. With the raising interest of stem-cell-based therapies in the neurodegenerative or the cancer field, these mice should provide novel and alternative tools to study these phenomena in vivo.

EPFL2008

The role of TIF1 gamma during adult murine hematopoiesis

Konstantin Shakhbazov

Transforming Growth Factor beta (TGFβ) signaling plays an important role in a variety of cellular processes during embryonic development, adult tissue homeostasis and cancer. TGFβI ligand is a potent inhibitor of early hematopoietic progenitor [1] cells in vitro, however the exact role and mechanisms of this pathway during adult hematopoietic homeostasis remains poorly understood. The Transcriptional Intermediary Factor 1 gamma (Tif1γ) gene encodes a recently uncovered nuclear protein in the TGFβ transduction pathway. Tif1γ is defective in the zebrafish moonshine mutant, which exhibits primitive and definitive hematopoiesis defects [2]. In addition, Tif1γ has been shown to influence differentiation of human erythroid cells in vitro [3]. Recent work in our laboratory has provided genetic evidence that Tif1γ is required for the generation and survival of hematopoietic stem/progenitor cells during murine development (C.Adolphe, K.Shakhbazov et al., manuscript under submission). Given that Tif1γ function is required for embryonic development and null mutants are embryonic lethal [4] (R.Losson unpublished data), we took advantage of a conditional Tif1γ flox allele (R.Losson unpublished data) and combined it with the inducible MxCre line [5] to generate mice in which Tif1γ function can be specifically ablated during adult hematopoiesis. MxCre:Tif1γKO/flox mutants exhibit a decrease in erythroblasts and dramatic increase in granulocytes within the bone marrow. This phenotype is attributable to a massive increase in granulocyte-monocyte progenitors (GMPs) at the expense of common myeloid progenitors (CMPs) and megakaryocyte- erythrocyte progenitors (MEPs), which are severely decreased in Tif1γ mutant bone marrow. In addition, pre-B cells were dramatically decreased in numbers, attributable to an increase in cell death, indicating a pro-survival role for Tif1γ in B lymphocytes. In addition, we performed deletion of Tif1γ in non-competitive and competitive bone marrow transplantation settings, which indicated that only the CMP/MEP phenotype is a non-cell autonomous effect. Two distinct models have been reported suggesting how Tif1γ functions within the TGFβ pathway: one indicating that Tif1γ binds and acts through Smad2/3 and the other that Tif1γ functions as a Smad4 mono-ubiquitin ligase [3, 6, 7]. To genetically address this apparent discrepancy, we generated double Tif1γ/Smad4 hematopoietic specific mutant mice, whose phenotype was surprisingly similar to that observed in Tif1γ single mutants. These data suggest that Tif1γ function is independent of Smad4 in hematopoietic cells. To address the mechanism by which Tif1γ functions, we performed microarray profiling of normal and mutant progenitors, which revealed prostaglandin D2 synthase as a hub gene linking Tif1γ and TGFβ signaling. Finally, we were able to mimic the hematopoietic-Tif1γ null phenotype by providing 16,16-dimethyl prostaglandin D2 to wild-type hematopoietic cells in vitro. In summary, our data provide genetic evidence for a key role of Tif1γ during early myeloid development as well as pre-B cell survival. In addition, we have identified that the majority of Tif1γ function is Smad4 independent. Finally, we identified a novel link between Tif1γ function and prostaglandin metabolism/signaling as a likely mechanism by which Tif1γ regulates hematopoiesis.

EPFL2009

The Cell Differentiation Status Influences the Outcome of Notch-Induced Malignancy

Caroline Poisson

The Notch signaling cascade is a well conserved pathway that regulates many aspects of embryonic development and plays a crucial role in the differentiation process and tissue homeostasis in multiple organs of adult species. Several human pathological disorders, including cancer, arise as a consequence of the deregulation of the Notch signaling pathway. Notch deregulation, up to date, has been best characterized in acute T-cell lymphoblastic leukemia (T-ALL). Indeed, in T-ALL pediatric patients activating Notch1 mutations have been identified in more than 50% of the cases. Increased knowledge of the mechanisms underlying T-ALL development and maintenance is clearly needed as one out of 5 patients with T-ALL has a very poor prognosis. Moreover, the high toxicity of chemotherapeutic regimens has been associated with elevated risks of developing severe secondary health problems. In the present study, we characterized the role of Notch in T-ALL induction and maintenance. Using transgenic mouse models, we could demonstrate that the oncogenic potential of Notch is dependent on RBP-Jκ transcriptional activity. Indeed, the canonical Notch signaling pathway, mediated by the RBP-Jκ transcription factor, is essential for both Notch-mediated T-ALL induction and maintenance, as well as for the induction of a Notch-mediated lymphoproliferative disorder. Moreover, constitutive overexpression of Notch1 at discrete developmental stages within the hematopoietic system, led us to identify a correlation between the differentiation status of a cell, within the hierarchy of the hematopoietic tree, and its sensitivity to Notch1-induced transformation. We demonstrated that forced Notch1 expression in hematopoietic stem cells was not sufficient to induce T-ALL. Nevertheless, aberrant Notch1 expression in the bone marrow resulted in the development of a lethal lymphoproliferative disorder. Overexpression of Notch1 in thmus at the DN2 to DN3 stage of T-cell delvelopment generated an aggressive transplantable T-ALL with important metastatic potential. This suggested that progenitors already committed to the T-cell lineage are highly susceptible to Notch-mediated transformation, which is relevant to the human disease. Nevertheless, forced Notch1 expression after the β-selection checkpoint in the thymys, was no longer sufficient to induce transformation of late DN3 to DN4 thymocytes. We therefore established a murine genetic system that allowed us to dissect the mechanisms of aberrant Notch1 signaling in a lymphoproliferative disorder and a true T-ALL model. We were able to assess the genetic signature that is involved in the transplantability and metastatic potential of leukemic cells and generated a list of candidate genes implicated in these processes. In addition, we correlated genes identified in our murine models with genes associated as well in human T-ALL profiles to identify genes implicated in T-ALL pathogenesis in both mouse and human.

EPFL2011