SEARCH

SEARCH BY CITATION

Keywords:

  • Notch1;
  • Hematopoietic stem cell;
  • Hematopoiesis;
  • Cancer stem cell;
  • Wnt signaling

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References

In normal hematopoiesis, proliferation is tightly linked to differentiation in ways that involve cell-cell interaction with stromal elements in the bone marrow stem cell niche. Numerous in vitro and in vivo studies strongly support a role for Notch signaling in the regulation of stem cell renewal and hematopoiesis. Not surprisingly, mutations in the Notch gene have been linked to a number of types of malignancies. To better define the function of Notch in both normal and neoplastic hematopoiesis, a tetracycline-inducible system regulating expression of a ligand-independent, constitutively active form of Notch1 was introduced into murine E14Tg2a embryonic stem cells. During coculture, OP9 stromal cells induce the embryonic stem cells to differentiate first to hemangioblasts and subsequently to hematopoietic stem cells. Our studies indicate that activation of Notch signaling in flk+ hemangioblasts dramatically reduces their survival and proliferative capacity and lowers the levels of hematopoietic stem cell markers CD34 and c-Kit and the myeloid marker CD11b. Global gene expression profiling of day 8 hematopoietic progenitors in the absence and presence of activated Notch yield candidate genes required for normal hematopoietic differentiation, as well as putative downstream targets of oncogenic forms of Notch including the noncanonical Wnts Wnt4 and 5A.

Disclosure of potential conflicts of interest is found at the end of this article.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References

Stem cells have an extensive capacity for self-renewal but retain the ability to differentiate into functional progeny. Most cancers are composed of a heterogeneous population of cells with marked differences in their proliferative potential, as well as the ability to reconstitute the tumor upon transplantation. Recent evidence has demonstrated that in leukemias and certain solid tumors, only a minority of cancer cells have the capacity to proliferate extensively and form new tumors [1, 2]. This suggests an interesting possibility that these cells constitute a stem cell population, the so-called cancer stem cells, that either self-renew to generate additional cancer stem cells or differentiate to generate phenotypically diverse cancer cells with limited proliferative potential.

Cancer stem cells have been characterized in the context of human acute myeloid leukemia, breast cancer, and glioblastoma. In each case, surface markers have been identified that distinguish cancer stem cells from cancer cells with more limited proliferative potential [3, [4], [5], [6], [7], [8]9]. In some cases, cancer stem cells might arise from the mutational transformation of normal stem cells, whereas in other cases, mutations might cause restricted progenitors or differentiated cells to acquire properties of cancer stem cells, such as self-renewal potential [10, 11]. Understanding the regulation of normal stem cell self-renewal is fundamental to understanding the regulation of cancer cell proliferation, because cancer can be considered to be a disease of unregulated self-renewal [1, 12, [13]14].

Among the different types of stem cells, hematopoietic stem cells (HSCs) are best characterized [15, [16], [17]18]. The decision of HSCs to self-renew or to differentiate is governed by a complex interplay between both intrinsic signals and stimuli from the surrounding microenvironment. The differentiation of HSCs and their multipotent progenitors to terminally differentiated cell types is tightly regulated by the microenvironment, on the one hand to prevent exhaustion of the stem cell pool and on the other to reduce the likelihood of accumulating transforming mutations. Breakdown in the regulation of self-renewal is likely a key event in the development of cancer, as demonstrated by the fact that several pathways implicated in carcinogenesis also play a key role in normal stem cell self-renewal decisions [1, 12, 19]. Several signaling pathways that regulate normal stem-cell self-renewal cause neoplastic proliferation when dysregulated by mutations. For example, Wnt [20, [21]22], sonic hedgehog (SHH) [23, [24], [25], [26]27], and Notch [28, [29]30] play important roles in the self-renewal of somatic stem cells, as well as neoplastic proliferation in the same tissues when mutated.

Notch receptors and ligands were first discovered in flies and worms, where they were shown to regulate cell proliferation, cell differentiation, and, in particular, binary cell fate decisions in a variety of developmental contexts [31, 32]. The first mammalian Notch homolog was discovered to be a partner in a chromosomal translocation in a subset of human T-cell leukemias. Notch signaling has subsequently been shown to regulate the development and differentiation of multiple hematopoietic cell types, including T cells, B cells, monocytes, macrophages, dendritic cells, osteoclasts, and natural killer cells [33, [34], [35], [36], [37], [38], [39], [40]41]. Finally, Duncan et al. have shown direct evidence for activation of the Notch pathway in HSCs located in the bone marrow niche [42], suggesting a definitive role for Notch signaling in the differentiation and maintenance of HSCs. The same study demonstrated a requirement for intact Notch signaling in Wnt-mediated HSC renewal, thus linking Notch and Wnt signaling in the process of HSC maintenance.

In vertebrates, four Notch genes have been identified, Notch1/TAN-1, Notch2, Notch3, and Notch4/int-3 [43]. The TAN1/Notch1 translocation t(7;9)(q34; q34.3) results in an active form of Notch1 mobilized to the T-cell receptor locus and is responsible for the formation of a subset of acute T-lymphoblastic leukemias in humans [44]. More recently, the identification of clustered point mutations in two domains supports a larger role for Notch1 in the initiation or progression of T-cell leukemias [36, 45, 46]. The purpose of the study reported here is to characterize the molecular events triggered by unregulated Notch1 expression in an in vitro model of the hematopoietic stem cell niche. Identification of the downstream targets of Notch1 in HSCs may provide targets for future stem cell and cancer therapies.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References

cDNA Expression Constructs

The truncated form of Notch1 used in our work has been described previously [47, 48]. SRα-ZEDN1 was digested with SpeI and then blunt-ended with T4 DNA ligase. EcoRI digestion was used to release the fragments. The fragments were then purified and cloned into the EcoRI and blunted NotI site of the pUHD10-3 IRES enhanced green fluorescent protein (EGFP) vector. The internal ribosome entry site (IRES)-EGFP vector was a kind gift from Dr. Owen Witte (University of California Los Angeles, Los Angeles, CA).

Cells, Cell Culture, and Transfection

Murine E14Tg2a ES cells and OP9 stromal cells were maintained as previously described [49, [50]51]. Ligand-independent Notch1-inducible ES cell clones (ZEDN1) containing the tetracycline-inducible system were isolated. The expression of ZEDN1 was driven by Tet response promoter (TetO-CMV) [52] and suppressed by the addition of Tet (Tet-Off system). The parental ES cell line (EStTA5-4), expressing tetracycline-transactivator protein and containing the puromycin resistance gene [50, 51], was a kind gift from Dr. Owen Witte (University of California Los Angeles). EStTA5-4 cells were electroporated with pUHD10-3-activated Notch1 IRES-EGFP vector and the neomycin resistance plasmid pCDNA3. Cells were then maintained in the presence or absence of Tet (1 μg/ml) and cloned by limiting dilution. Several Tet-responsive ZEDN1 clones were obtained. Finally we selected three high-EGFP-expressing clones ZEDN1-6 (Z6), ZEDN1-20 (Z20), and ZEDN1-25 (Z25) for our experiments. Clones tightly regulated by Tet were analyzed by EGFP expression by FACScan (Becton, Dickinson and Company, Franklin Lakes, NJ, http://www.bd.com). The EGFP-expressing clones were examined for ZEDN1 expression by Western blotting in Tet-Off and Tet-On conditions.

ES Cell In Vitro Differentiation and Induction of Truncated Notch1 Expression During Hematopoietic Progenitor Development

ES cell in vitro differentiation and induction of truncated Notch1 expression were as previously described [49, [50]51]. Briefly, 6 × 104 undifferentiated ES cells repressed for ZEDN1 expression were cocultured on a confluent layer of OP9 stromal cells in a 10-cm dish in the presence of 100 ng/ml Tet. On day 5 of coculture, both differentiated ES cells and the OP9 stroma were harvested in fresh medium. Cells were replated in new dishes for 20–30 minutes to separate OP9 cells from ES cells. OP9 cells quickly adhered to the dish, and ES cells were harvested. Then, 2 × 106 day 5 ES cells were replated on 10-cm dishes of confluent OP9 cells. These day 5 cocultures were continued in both Notch-Off (Tet-On) and Notch-On (Tet-Off) conditions until day 8 when hematopoietic progenitors were harvested. In some experiments, these cells were transferred to a new layer of OP9 cells and were cultured until day 13 or 14. ZEDN1-expressing hematopoietic cells were identified by expression of EGFP.

Flow Cytometric Analysis of ES Cell-Derived Hematopoietic Progenitors

On day 5, ES clones harvested from coculture with OP9 were stained with anti-Flk-1-PE, CD34-FITC, and CD117-PE (BD Pharmingen, San Diego, http://www.bdbiosciences.com/pharmingen). On day 8, ES cells were stained with the following combination of antibodies, anti-CD117-TC (c-Kit), anti-CD11b-TC, anti-CD34-PE (Caltag, Burlingame, CA, http://www.caltag.com), and anti-Ter119-PE (BD Pharmingen). Day 13/14 ES differentiated cells were stained with anti-Ter119-PE and anti-CD11b-TC. Data were acquired on a FACScan (Becton Dickinson) and analyzed using FCS Express, version 2, (De Novo Software, Thornhill, ON, Canada, http://www.denovosoftware.com). Dead cells were excluded from analysis based on forward- and side-scatter properties.

Morphologic Analysis of ES-Derived Hematopoietic Progenitors and Terminally Differentiated Cells

ES cells cocultured with OP9 were harvested on days 5, 8, and 13/14 and transferred to cytospin slides (Thermo Shandon Inc., Pittsburgh, http://www.thermo.com) and visualized with an inverted microscope at a magnification of ×400. Cytospin preparations of ES cells differentiated for 5, 8, and 13–14 days were analyzed by the HEMA 3 stain set solutions according to the manufacturer's protocol (Fisher Scientific, Biochemical Sciences, Swedesboro, NJ, https://www.new.fishersci.com). Cocultures were analyzed by microscopy and photographed.

Western Blot Analysis

Protein lysates were prepared in cold radioimmunoprecipitation assay cell lysis buffer (50 mM Tris, pH 8.0, 150 mM NaCl, 1% Nonidet P40, 0.5% deoxycholate, 0.1% SDS, and protease inhibitor cocktail [complete mini; Roche Diagnostics, Basel, Switzerland, http://www.roche-applied-science.com]) from ZEDN1 clones grown in the absence or presence of 100 ng/ml Tet for 3 days. After electrophoresis of cell lysates on a 7.5% SDS-polyacrylamide gel (Bio-Rad, Hercules, CA, http://www.bio-rad.com), proteins were transferred onto Hybond-ECL nitrocellulose membrane (GE Healthcare/Amersham Biosciences, Piscataway, NJ, http://www.gehealthcare.com). Equivalent loading of cell lysates was confirmed by Ponceau S stain of membrane. The blot was probed with anti-Notch1 antibody 93-4 (kind gift from Dr. Gerry Weinmaster, University of California Los Angeles), generated against the cytoplasmic domain of Notch1 [47]. This was followed by incubation with a biotinylated donkey anti-rabbit IgG secondary antibody (GE Healthcare, Amersham Biosciences) and streptavidin-conjugated with horseradish peroxidase (HRP) (GE Healthcare, Amersham Biosciences) as a tertiary reagent. Chemiluminescence detection (enhanced chemiluminescence [ECL]) was visualized using hyperfilm (GE Healthcare, Amersham Biosciences).

Nuclear Extraction and Western Analysis

Nuclear extraction was performed on day 7 and day 8 differentiated ES ZEDN1 clone 20 cells using the nuclear extract kit (Active motif) according to the manufacturer's recommendations. A 7 × 105 cell equivalent of the specified fraction was loaded in each lane. Electrophoresis was performed as described above and proteins were transferred onto Immobilon-P polyvinylidene difluoride transfer membrane (Millipore, Billerica, MA, http://www.millipore.com). The blot was probed with anti-β-catenin antibody (C2206; Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com), anti-lamin B antibody (sc20682; Santa Cruz Biotechnology Inc., Santa Cruz, CA, http://www.scbt.com), and anti-glyceraldehyde-3-phosphate dehydrogenase antibody (ab9485-100; Abcam, Cambridge, U.K., http://www.abcam.com). This was followed by incubation with goat anti-rabbit IgG HRP conjugate secondary antibody (Upstate, Charlottesville, VA, http://www.upstate.com). Chemiluminescence detection (ECL) was visualized using hyperfilm (GE Healthcare, Amersham Biosciences).

Dual Luciferase Reporter Assay to Detect Truncated Notch1 Activity

Undifferentiated ZEDN1 clones were induced for activated Notch1 expression over a time course of 0, 24, 48, 72, and 96 hours by removing Tet from the culture medium. The ES cells were transfected with firefly luciferase constructs (Promega, Madison, WI, http://www.promega.com) 48 hours prior to harvest. Renilla luciferase was used as the internal control to normalize the results. The manufacturer's protocol was followed, and samples were analyzed using a manual Luminometer (DLReady, TD-20/20; Turner Designs, Sunnyvale, CA, http://www.turnerbiosystem.com).

Microarray and Quantitative Real-Time Polymerase Chain Reaction Analysis

Total RNA was extracted from day 8 cell pellets with the RNeasy mini kit (Qiagen, Hilden, Germany, http://www1.qiagen.com) according to the manufacturer's recommendations. Spectrophotometric analysis verified the concentration and purity of RNA. RNA from cells with Notch-Off and Notch-On was used in microarray analysis using mouse Affymetrix chips (MOE-430A). The data were analyzed using GeneSpring software, version 6.0 (Silicongenetics, Redwood, CA, http://www.silicongenetics.com). For real-time polymerase chain reaction (PCR) analysis, The RNA was treated with RNase-free DNase (Promega). Random-hexamer primed cDNA synthesis was performed with Superscript II reverse transcriptase (Invitrogen, Carlsbad, CA, http://www.invitrogen.com) according to the manufacturer's specifications. Quantitative real-time PCR was performed by the SYBR Green system (Molecular Probes Inc., Eugene, OR, http://www.probes.invitrogen.com) with a Bio-Rad iCycler detector. Data were analyzed using the iCycler IQ Optical system software, version 3.0a. Oligonucleotide PCR primers were designed (Primer 3 software; Whitehead Institute for Medical Research, Cambridge, MA, http://www.primer3.sourceforge.net) to have optimal annealing temperatures of 55°C or 60°C and to generate PCR products between 80 and 140 base pairs. Fold changes were calculated using the Pfaffl equation [53].

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References

Regulation of Activated Notch1 Expression in Differentiating ES Cells

To explore the function of Notch signaling in both normal and neoplastic hematopoiesis, we used a tetracycline-inducible system to regulate the expression of a ligand-independent constitutively active form of Notch1 (ZEDN-1) in murine E14Tg2a embryonic stem cells. A total of 14 independent ZEDN1 ES clones were established by transfection and drug selection (details given in Materials and Methods). Western analysis confirmed inducible expression of truncated Notch1, in addition to the expected upregulation of endogenous Notch1 (Fig. 1A). The ZEDN1 construct is transcriptionally active over a time course of 96 hours as confirmed by the dual-luciferase reporter assay, in which eight copies of the CSL binding site are placed upstream of an SV-40 promoter (Fig. 1B).

thumbnail image

Figure Figure 1.. Inducible expression of truncated Notch1 in undifferentiated embryonic stem (ES)-ZEDN1 clones. (A): Western blot analysis of Z-25 and Z-20 grown for 3 days in the presence (+) or absence (−) of 100 ng/ml tetracycline. Arrow indicates inducible expression of truncated Notch1. Arrowhead represents expression of endogenous Notch. Asterisk indicates a background band. Control lane is 293T cell line overexpressing truncated Notch1 (ZEDN1). (B): Kinetics of luciferase reporter expression by ZEDN1 by dual luciferase reporter assay. Undifferentiated ES clones growing in the presence of m-LIF (1,000 U/ml) and Tet (100 ng/ml) were induced for activated Notch1 expression over a time course of 0, 24, 48, 72, and 96 hours by removing Tet from the culture medium. ES-ZEDN1 clones Z6 and Z20 were transfected with firefly and Renilla luciferase constructs 48 hours prior to harvest. The normalized relative luciferase reporter levels correspond to the binding of activated Notch1 to CBF1 protein, thus switching it from a repressor to an activator of transcription, resulting in increased reporter activity. Abbreviations: kd, kilodaltons; Z-20, ES-ZEDN1 clone 20; Z-25, ES-ZEDN1 clone 25.

Download figure to PowerPoint

Development of Hematopoietic Progenitor Cells from ES Lines in the Absence and Presence of Ligand-Independent Activated Notch1 Expression

Previous work has characterized the murine embryonic stem cell /OP9 coculture system as a highly reproducible way to model hematopoiesis in vitro. OP9 is a bone marrow stromal cell line derived from the osteopetrosis mutant mouse that cannot produce macrophage colony-stimulating factor [54]. In the absence of leukemia inhibitory factor (LIF), OP9 cells provide the necessary extrinsic signals for the differentiation of pluripotent ES cells into primitive hemangioblasts (day 4–5). Continued coculture with OP9 stroma produces immature hematopoietic stem and progenitor cells (day 8) and mature myeloid and erythroid cells (days 12–14) (Fig. 2A, 2B) in the absence of exogenous growth factors.

thumbnail image

Figure Figure 2.. Development of hematopoietic cells from embryonic stem lines in the absence and presence of ligand-independent activated Notch1 expression. (A): Progression of tetracycline-regulated embryonic stem (ES)-ZEDN1 clones in coculture. ES cells were grown on a mitotically inhibited layer of OP9 stromal cells for 5 days in the absence of leukemia inhibitory factor (LIF). At day 2, ES cells were observed to be tight clusters with defined borders. On day 5, large colonies of undifferentiated and differentiated hematopoietic colonies containing hemangioblasts were observed. Day 5 ES cells were replated on a fresh layer of OP9 cells and cultured until day 8 to promote differentiation of hematopoietic progenitors. They were observed as two to four cell clusters loosely attached to the OP9 stromal layer. Day 8 cells were continued in a fresh culture until day 13/14 for further differentiation into mature hematopoietic cells. (B): Reversible effect of ZEDN1 expression on myeloid differentiation. ES cells were induced to differentiate by removal of LIF and cultured on OP9 stromal cells between day 0 and day 5 in the presence of Tet (100 ng/ml) to suppress ZEDN1 expression. Day 5 differentiated cells were harvested, washed, and replated on OP9 layers in three different groups. In the first group, cells were continuously cultured in Tet between day 5 and day 14 (Notch-Off 0–14). In the second group, cells harvested on day 5 were recultured in the absence of Tet between day 5 and day 14 (Notch On 5–14). In the third group, cells from day 5 were cultured between day 5 and day 8 without Tet and then added back to suppress the expression of ZEDN1 (Notch On 5–8). Cells harvested from each of the three groups were analyzed by flow cytometry for percentage of erythroid (Ter119+) and myeloid (CD11b+) cell types on day 14. The total cell number per culture is shown.

Download figure to PowerPoint

Induction of Notch1 signaling in hematopoietic stem and progenitor cells has been shown to decrease terminal differentiation [55]. Preliminary experiments showed that induction of ZEDN1 expression in ES cells on day 0 of coculture with OP9 cells inhibited the appearance of day 5 Flk-1-positive hemangioblasts and consequently day 8 hematopoietic cells, as shown previously by Schroeder et al. [56].

Induction of activated Notch expression from day 5 to day 14 dramatically reduced the number of Ter119- and CD11b-positive cells by almost 84% in two independent experiments (Fig. 2B). Tet-regulated expression of activated Notch1 allowed us to investigate whether this decrease in cell number and terminal differentiation could be reversed. When the expression of truncated Notch1 was induced at day 5 and then ended at day 8 (by the addition of Tet), an intermediate profile of cell surface markers was observed, indicating a robust recovery of the cell's proliferative and differentiation potential (Fig. 2B).

In all remaining experiments, ES cells were allowed to differentiate in the presence of tetracycline on OP9 stromal cells for 5 days, at which time the culture was composed of mesodermal cells and hematopoietic precursors, as defined by Flk-1 expression in 50%–65% of cells, c-Kit expression in 55%–60% of cells, and CD34 expression in 6%–10% of cells (Fig. 3). By day 8, approximately 75% of the Flk-1+ progenitor population had differentiated into the hematopoietic lineage, as defined by staining with CD45 and Ter119 (data not shown).

thumbnail image

Figure Figure 3.. Cytospin preparations and flow cytometric analysis from embryonic stem-ZEDN1 clones harvested on day 5 of coculture. (A): Cytospin preparations were stained with HEMA3 stain set solution. The large cells with a wide basophilic cytoplasm are hemangioblasts. They are bipotential precursors of the endothelial and hematopoietic lineage and are positive for markers Flk-1 and c-Kit. (B): Flow cytometric analysis of cells harvested on day 5 indicates high levels of Flk-1 and c-Kit and low levels of the hematopoietic stem cell marker CD34. The results are presented as the mean ± SEM of six independent experiments.

Download figure to PowerPoint

Activated Notch1 Expression at the Onset of Hematopoiesis Affects Cell Survival and Proliferation

To investigate the effect of activated Notch1 on the generation of the earliest hematopoietic precursor cells (seen at maximum levels on day 8), the expression of ligand-independent Notch1 was induced (by removing Tet) on day 5, and cells were harvested for fluorescence-activated cell sorting (FACS) analysis at day 8. A 24-hour exposure to activated Notch1 expression did not significantly alter the total number of cells recovered on day 6 between the induced and uninduced populations. However, apoptosis assays on days 6, 7, and 8 indicated a higher percentage of cells (erythroid and myeloid lineage) in early and late stages of apoptosis in the induced population compared with the uninduced population (data not shown). A 72-hour exposure to activated Notch1 expression (Fig. 4A, 4B; one representative experiment) reduced the total number of day 8 cells in the cocultures by approximately 75% (Fig. 4C; mean of eight independent experiments) as compared with the control population, indicating that exposure to activated Notch1 had resulted in cell death.

thumbnail image

Figure Figure 4.. Activated Notch1 expression at the onset of hematopoiesis affects cell survival and proliferation. (A): Morphology of cytospin preparations of cells harvested on day 8 of coculture. Cytospin preparations were stained with HEMA3 stain set solution. Slides were photographed at a magnification of ×1,000 using an oil-immersion lens. (B): Induction of truncated Notch1 on day 5 led to its expression in 46% of cells by day 8 as measured by enhanced green fluorescent protein fluorescence. The flow cytometric data are shown for a representative experiment. (C): The effect of ZEDN1 expression on proliferation of hematopoietic progenitors was analyzed by counting the total number of embryonic stem (ES) day 8 cells following Notch1 activation on day 5 (Notch-On). ES cells cultured continuously in the presence of Tet served as control (Notch-Off). The results represent the mean ± SEM of eight independent experiments. (D): Effect of Notch1 induction on the percentage of markers for multipotential, myeloid progenitors (CD11b) and erythroblasts (Ter119) on day 8 of ES in vitro differentiation. The results represent mean ± SEM of eight independent experiments. (E): Total cell counts of day 8 ES-derived hematopoietic progenitors combined with percentages of multipotent progenitors (c-Kit+, CD-34+), myeloid progenitors (CD11b+), and erythroblasts (Ter119+) were used to obtain cell numbers of specific populations. Results represent the mean ± SEM of eight independent experiments. ∗, p < .05; ∗∗, p < .01; ∗∗∗, p < .001. Abbreviations: GFP, green fluorescent protein; X, fold.

Download figure to PowerPoint

The relative percentages of c-kit+, CD34+ and CD11b+ cells decreased in cells with activated Notch1 expression, in contrast to a 1.7-fold increase in the percentage of erythroblasts generated (Fig. 4D; mean of eight independent experiments). When the cell counts and FACS analysis are combined, the expression of activated Notch1 significantly reduced the c-Kit and CD34 progenitor population by 84% each (Fig. 4E). The myeloid progenitor population was substantially reduced in the Notch-On population by 87%. However, despite this increase in the percentage of erythroblasts, there was still a 38% decrease in the absolute number of erythroblasts (Fig. 4E).

Global Gene Expression Profiling in Differentiating ES Cells Expressing Activated Notch1

Analysis of the Flk1+ cells cultured in the presence of activated Notch1 for 72 hours demonstrated a reduction in cell numbers, indicating a role of Notch1 in both cell survival and proliferation. The reversible effect of Notch expression on cell proliferation and hematopoietic commitment is consistent with preservation of a multipotent state. Conversely, genes that are upregulated in the control population likely play a role in hematopoietic specification. Global gene expression profiling makes it possible to pose such mechanistic questions without prior bias. Three independent coculture experiments were performed using ZEDN1 clone 20 (two experiments) and ZEDN1 clone 6. Cells were cultured in the absence (Notch-On) or presence (Notch-Off) of Tet from day 5 to day 8 and harvested on day 8 for preparation of mRNA. Gene expression profiling was conducted as described in Materials and Methods.

Three experiments performed using two different ZEDN1 clones showed significant changes in the expression of 158 genes, of which 136 were upregulated and 22 were downregulated in the presence of activated Notch1. Genes of interest were further analyzed using real-time quantitative PCR. As expected, Notch1 expression levels were upregulated almost 30-fold upon activation of Notch signaling (Fig. 5A). Known downstream targets of Notch signaling Hey1 and Nrarp also showed a similar profile upon Notch1 activation. Hey2 and Hes1 were also upregulated, although to a lesser degree.

thumbnail image

Figure Figure 5.. Activated Notch1 expression at the onset of hematopoiesis upregulates expression of known downstream targets of Notch signaling in hematopoietic progenitors (day 8). Notch1 signaling was activated in hemangioblasts on day 5, and its effect was analyzed on day 8 hematopoietic progenitors (Notch-On 5–8). Day 8 precursors uninduced for Notch1 signaling served as controls (Notch-Off). (A): Quantitative real-time polymerase chain reaction (PCR) analysis for Notch1, Nrarp, Hey1, Hey2, and Hes1 in Notch-On and Notch-Off day 8 precursors showed an upregulation of these genes following activation of Notch signaling. (B): Quantitative real-time PCR analysis for Wnt4 and Wnt5A showed an upregulation of these genes following activation of Notch signaling. GAPDH was used as the normalizing control. Fold changes were calculated based on the Pfaffl equation [53]. (C): Western blot analysis for β-catenin in Nuc and Cyt fractions on day 7 and day 8 differentiated clone Z20 embryonic stem cells showed a downregulation of β-catenin following Notch activation. Lamin B served as the Nuc marker, and GAPDH was used as a Cyt marker. A 7 × 105 cell equivalent of the fraction was loaded in each lane. Abbreviations: Cyt, cytoplasmic; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; Nuc, nuclear.

Download figure to PowerPoint

Two members of the Wnt signaling pathway, Wnt4 and Wnt5A, were also upregulated in the hematopoietic progenitor population following activation of Notch1 signaling in hemangioblasts (Fig. 5B). A similar trend was observed in the levels of E-cadherin and CdkIn1c, whereas the level of cyclin D1 was downregulated, following activation of Notch signaling in hemangioblasts. Control experiments showed that these values could not be due to contamination with OP9 stromal cells (data not shown). To determine whether the autocrine/paracrine Wnt signaling was mediated by the canonical or noncanonical pathway, β-catenin levels were measured in both cytoplasmic and nuclear extracts in the presence and absence of activated Notch1. Protein levels of β-catenin were unchanged in the cytoplasmic extracts but were found to be significantly reduced in the nuclear extracts from cells expressing activated Notch1 (Fig. 5C). This is consistent with the notion that noncanonical Wnt signaling can downregulate the canonical pathway [57].

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References

Although it is generally acknowledged that the Notch and Wnt pathways play important roles in maintaining hematopoietic stem cells, the precise mechanisms by which this occurs have not been elucidated [58, [59]60]. Stem cell signaling is bidirectional and heterotypic, involving complex interactions between the stem cells and membrane-bound and secreted proteins on stromal elements. We and others hypothesize that the interactions of Notch and its stromal cell ligands serve to prevent stem cell exhaustion throughout the life of the organism. To date, most studies on the role of Notch in hematopoiesis have focused on isolated cell populations or knockout models [61]. Current technology does not allow stimulation of stem cell self-renewal without differentiation in vitro. Well-characterized in vitro systems do exist in which murine ES cells will differentiate to mature hematopoietic cells on cultured stroma [49, [50]51, 54].

The role of Notch in proliferation and differentiation depends on the context in which Notch signaling occurs. To understand the molecular mechanisms involved in neoplastic transformation, it is becoming increasingly important to identify the cellular factors that determine whether Notch signaling will be oncogenic or tumor suppressive. To identify the molecular targets of Notch and examine the effects of Notch1 activation, we used the ES/OP9 coculture system, which recapitulates hematopoietic development without exogenous growth factors [50, 51, 54]. The tetracycline-inducible system enables us to precisely control the expression levels of activated Notch1 to mimic the effects of expression of oncogenic forms of Notch in leukemia and other malignancies. The relative homogeneity of the differentiating ES cells over time abrogates problems associated with heterogeneous bone marrow samples. Notch ligands have been identified in both human [62] and murine [45, 63] bone marrow stroma, whereas Notch receptors have been found in many blood cell types, enabling us to study this heterotypic interaction between membrane-bound receptors and ligands in a system that models some aspects of the stem cell niche.

The activation of Notch signaling in hemangioblasts (day 5) has two dramatic effects. First, the total number of cells is reduced on day 8, and second, the balance of myeloid to erythroid progenitors is reversed. The reduction in cell numbers on day 8 is the result of Notch-induced cell death and growth arrest across all lineages (data not shown), over and above that caused by the end of the first wave of erythropoiesis [64]. The resultant arrest in cell proliferation probably causes the shift in balance between erythroid and myeloid progenitors. All of these results are consistent with activated Notch1 affecting cell survival and proliferation and inhibiting differentiation to hematopoietic progenitors by keeping them in a less differentiated state. This is consistent with results published previously by our group using normal human bone marrow cells plated on stromal cell lines in the presence and absence of the Notch ligand Jagged [55, 65].

In both vertebrate and invertebrate nervous system development, activation of Notch results in activation of expression of the enhancer of split group of genes, hey/hes, that inhibit neural fate determination [43]. Consistent with this, we observed high levels of hey1 and hes1 expression in day 8 cells following Notch1 activation in hemangioblasts on day 5, which might be responsible for blocking subsequent differentiation. Sustained high levels of Notch1 are required to maintain the block, which is readily reversible upon addition of Tet (Fig. 2B). This suggests that blocking the effects of activated or oncogenic forms of Notch could help to reverse the malignant phenotype.

To understand the mechanism of Notch1-mediated block in cell proliferation and differentiation during HSC fate determination, we conducted microarray experiments on day 8 cells in the absence or presence of activated Notch1 from the hemangioblast stage, on day 5 of coculture. We did three independent microarray experiments with two different ZEDN1 clones and found several genes that were upregulated/downregulated in Notch-On versus Notch-Off conditions. These observations were further confirmed by quantitative real-time PCR. The experiments were designed to detect both direct and indirect targets of the Notch pathway. As expected, levels of Notch1 were high following activation of Notch signaling. We also observed robust activation of hey1, hey2, hes1, and nrarp genes, which are known downstream targets of Notch signaling. This confirms the function of activated Notch construct. By contrast, genes expressed at higher levels in the control cells (Tet-Off) on day 8, when the levels of CD34 and c-kit are the highest, could be inferred to play a role in generation or maintenance of “normal” HSC.

In our system, we observed an interaction between Notch and Wnt signaling pathways. Following activation of Notch signaling in hemangioblasts, we observed an upregulation of Wnt4 and Wnt5a, which are associated with noncanonical Wnt signaling [57, 66]. Signaling through the canonical Wnt pathway disrupts the association of β-catenin with the axin/APC/GSK-3β complex. This results in the accumulation of β-catenin in the cytosol and its translocation into the nucleus. Once in the nucleus, β-catenin binds to the LEF-1/TCF family of transcription factors and activates expression of target genes such as cyclin D1, c-myc etc. β-Catenin levels in the cytosol are also kept low by its interaction with the intracellular domain of E-cadherin at the plasma membrane. Therefore, cadherin-mediated adhesion may act as a negative regulator of Wnt signaling because it can sequester β-catenin at the cell surface and thereby deplete it from the cytosolic pool [67]. Our microarray data and the follow-up real-time PCR analysis show an upregulation in the levels of E-cadherin and cdkIn1c and a slight downregulation in the levels of cyclin D1. In addition, we observed a significant decrease in the levels of β-catenin protein in the nucleus following activation of Notch signaling. The expression pattern of the above genes and β-catenin protein is in contrast to what is expected as a result of canonical Wnt signaling. Previous studies have shown that noncanonical Wnt signaling can antagonize the canonical Wnt pathway, although the precise mechanism of inhibition is still in question [57, 68, 69]. β-Catenin, a mediator of the Wnt signaling pathway, as well as purified Wnt proteins, can promote self-renewal of murine and human HSCs in vitro and increase the ability of murine HSCs to reconstitute the hematopoietic system of lethally irradiated mice [22, 58]. Conversely, overexpression of Axin, which enhances β-catenin degradation and thereby inhibits Wnt signaling, causes inhibition of HSC growth in vitro and decreased reconstitution of irradiated mice [58].

In other developmental systems, there is considerable cross-talk between the Wnt and Notch signaling pathways [59, 60]. In particular, in the developing Drosophila wing disc, the expression of Wnt is transcriptionally regulated by the Notch pathway, which then diffuses across the wing disc to regulate its global patterning. Interestingly, our preliminary analysis of 1.5 kilobases upstream of both these genes shows that it contains sites that can potentially bind to CBF-1 (unpublished results). It is possible that CBF-1/Notchintra complex binds to these sites and promotes the upregulation of Wnt5a. Taken together, these observations suggest that activation of Notch results in the activation of Wnt4 and 5A, which subsequently activate the noncanonical Wnt pathway, resulting in inhibition of the canonical Wnt pathway. These observations and the above-mentioned role of Notch and Wnt during the differentiation of HSC suggest that Notch and Wnt signals could be functioning in concert to maintain the balance between self-renewal and differentiation of hematopoietic stem cells.

Much of cancer research is focused on the identification of therapies that target essential genes or pathways critical to the development of cancer. If only a rare subset of tumor stem cells drives tumor formation then there are major implications in the study, diagnosis, and treatment of cancer. Traditional treatments for cancer nondiscriminately kill proliferating cells. Therefore, agents selectively killing the cancer stem cells are likely overlooked in screening methods that rely on rapid reduction of tumor size. Another factor to be taken into consideration is the potential difference in drug sensitivity between the tumorigenic and nontumorigenic populations [70]. Normal stem cells from various tissues tend to be more resistant to chemotherapeutics than mature cell types from the same tissues [71]. The reasons for this are not clear, but they may relate to high levels of expression of antiapoptotic proteins [72, [73], [74]75] or ABC transporters, such as the multidrug resistance gene [76]. If cancer stem cells have similar properties, then one would predict that these cells would be more resistant to chemotherapeutics than tumor cells with limited proliferative potential. It is becoming apparent that treatments that directly target those pathways involved in malignant stem-cell self-renewal would have a significantly greater chance of success. Our future aim is directed toward understanding the relationship between Notch and Wnt and the effect their interaction has on the downstream targets. A better understanding of the signaling pathways involved in normal stem-cell self-renewal will be important in furthering our knowledge of the events that lead to the development of cancer.

Disclosure of Potential Conflicts of Interest

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References

The authors indicate no potential conflicts of interest.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References

We thank Habib Hamidi for assistance with statistical analysis of data and Encarnacion Montecino Rodriguez for assistance with flow cytometry. Microarray experiments were performed with the assistance of the staff in the UCLA DNA microarray core facility at the Jonsson Comprehensive Cancer Center. Funding for these studies was provided by the Bergman Foundation.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References