Figure 1.
Cell populations used for investigation of Notch1 signaling.
(A) Scheme of cell populations and culture conditions for identification of Notch1 target genes. Undifferentiated EB5 NERT or EB5 control ESC were kept under self-renewal conditions (ESCs) and treated with OHT for 4 h. Alternatively, ESC were cultured in ectodermal differentiation medium (ES to ectoderm; ESCe) or mesodermal differentiation medium (ES to mesoderm; ESCm) during OHT treatment. Furthermore, ESC were differentiated for 4 d on Collagen IV to mesodermal cells (Mesoderm), which were then treated by OHT for 4 h. To distinguish direct from indirect targets, all experiments were additionally performed with or without CHX to inhibit protein synthesis. After 4 h OHT treatment, RNA was extracted from the different cell populations and subjected to further analysis. (B) Mesodermal differentiation of ESC and its suppression by activated Notch1. EB5 control or EB5 NERT cells were cultured for 4.5 d in mesodermal differentiation medium on OP9 stromal cells with or without OHT and the percentage of Flk1+ cells was analyzed by flow cytometry. Histograms of the cells stained with anti-Flk-1 antibody are depicted in blue and histograms of the cells stained with isotype matched mouse IgG are depicted in grey. The area left to the dotted line indicates Flk-1+ cells. The percentage of Flk-1+ cells is shown for each histogram. One representative example of eleven (EB5 NERT cells) or four (EB5 controls cells) experiments is shown, respectively. Reduction of Flk+ positive cells is statistically significant (p<0.01). (C) Expression of Rex1, Pax6, and T (brachyury), which characterize undifferentiated ESC, ectodermal cells and mesodermal cells, respectively, in the cell populations used for array analyses. The different cell types described in A were evaluated for RNA expression of key pluripotency and differentiation genes via microarray analysis. The expression was normalized to the highest mean. Error bars indicate standard deviation. Since Pax6 was identified as a Notch1 target, OHT induced samples were not included for calculation of this gene.
Figure 2.
Identification of cell-context specific Notch1 target genes by genome wide gene expression arrays.
(A) 3-dimensional principle component analysis (3D PCA). Notch1 signaling was induced with OHT for 4 h in EB5 NERT or EB5 control cells (Con.) as outlined in Figure 1A. Mesodermal cells (Mesoderm) were additionally treated with or without CHX. The various culture conditions clearly cluster to different areas in the plot, whereas differences induced by Notch1 are rather subtle. An animated presentation of the 3D PCA is available in the supplemental material. (B) Expression heatmap of the Notch1 induced genes. 465 transcripts were identified as differentially expressed due to Notch1 induction in at least one of the conditions (ESCe, ESCm, Mesoderm (Mes.), or Mesoderm+CHX) according to the criteria outlined in Material and Methods. Expression values of the samples were Log2 transformed and normalized within each gene. (C) Venn diagram of Notch1 induced genes in mesodermal cells in the presence or absence of CHX. Of the 262 genes regulated without CHX treatment, 170 are up-regulated and 92 are down-regulated, whereas with CHX treatment 160 of the 179 differentially expressed genes were positively and 19 were negatively regulated. Importantly, there is a significant overlap of 74 up-regulated genes indicating potential direct target genes of Notch1 in mesodermal cells, whereas there is almost no overlap for the down-regulated genes. Thus, positive regulation seems to be the main mechanism of Notch1 gene regulation, whereas the negative regulation observed accounts for indirect effects.
Figure 3.
Cell-context specific Notch1 target genes identified by genome wide gene expression arrays.
Venn diagram of differentially Notch1 induced genes in ESC to ectoderm (ESCe), ESC to mesoderm (ESCm), or mesodermal cells (Mesoderm). Numbers at the labeling indicate the total number of regulated genes. Numbers in the areas indicate the overlap of the different conditions. All numbers in parentheses show the quantity of negatively regulated genes.
Figure 4.
Induction heat maps of Notch1 target genes involved in transcription, apoptosis, cell cycle and cell differentiation.
The identified Notch1 targets were grouped into different Gene Ontology (GO) terms using the DAVID functional annotation analysis tool (http://david.abcc.ncifcrf.gov/). Note that only differentially regulated genes by Notch1 are shown as Log2 transformed values. If no gene symbol was available for the regulated transcript, GenBank accession number was given in parentheses. Significant changes of differentially regulated gene expression according to the criteria outlined in Material and Methods were labeled with a black or white dot for up- and down-regulated genes, respectively. Redundant probe sets for the same gene were removed except when significantly different results were obtained (e.g. Sox9). Pcdha@*) indicates the protocadherin αcluster comprising Pcdha1, Pcdha2, Pcdha3, Pcdha4, Pcdha5, Pcdha6, Pcdha7, Pcdha8, Pcdha9, Pcdha10, Pcdha11, Pcdha12, Pcdhac1, Pcdhac2, and Pcdhgb2. The complete list of regulated genes is available in the Supplement Table S1.
Figure 5.
Induction heat maps of Notch1 target genes involved in regulation of metabolic process, cell adhesion and cell communication.
The identified Notch1 targets were grouped into different Gene Ontology (GO) terms using the DAVID functional annotation analysis tool (http://david.abcc.ncifcrf.gov/). Note that only differentially regulated genes by Notch1 are shown as Log2 transformed values. If no gene symbol was available for the regulated transcript, GenBank accession number was given in parentheses. Significant changes of differentially regulated gene expression according to the criteria outlined in Material and Methods were labeled with a black or white dot for up- and down-regulated genes, respectively. Redundant probe sets for the same gene were removed except when significantly different results were obtained (e.g. Sox9). The complete list of regulated genes is available in the Supplement Table S1.
Figure 6.
Induction heat maps of Notch1 target genes involved in mesoderm development, intracellular signalling cascade, cell proliferation and cytoskeleton.
The identified Notch1 targets were grouped into different Gene Ontology (GO) terms using the DAVID functional annotation analysis tool (http://david.abcc.ncifcrf.gov/). Note that only differentially regulated genes by Notch1 are shown as Log2 transformed values. If no gene symbol was available for the regulated transcript, GenBank accession number was given in parentheses. Significant changes of differentially regulated gene expression according to the criteria outlined in Material and Methods were labeled with a black or white dot for up- and down-regulated genes, respectively. Redundant probe sets for the same gene were removed except when significantly different results were obtained (e.g. Sox9). The complete list of regulated genes is available in the Supplement Table S1.
Figure 7.
Notch1 target genes validated by qPCR.
The microarray results were validated by qPCR using ESCs, ESCe, ESCm and Mesoderm. All conditions were tested also in the absence or presence of cycloheximide (+CHX). Oct4 (last gene; also known as Pou5f1) is included as negative non-regulated control. (A) Mean values of relative expression (% of Gapdh expression, Log2 transformed) shown as a heatmap in the absence (−) or presence (+) of OHT. (B) Inductions by activated Notch1 were calculated from relative expression values and were Log2 transformed. Thus, blue areas indicate down-regulation and red regions show up-regulation. *) indicates low level expression of Myf5 near detection limit resulting in artificially high induction values that are not reliable.
Figure 8.
Regulatory transcriptions factors involved in differentiation are induced by activated Notch1 in the absence of protein synthesis.
(A–H) The microarray results were validated by qPCR in ESC cultured in self renewal (ESCs), ectodermal (ESCe), or mesodermal (ESCm) culture medium and in mesodermal cells (Mes.). All conditions were tested in the absence (−) or presence (+) of CHX. Relative expression is shown for untreated (blue bars) and OHT treated samples (red bars) compared to Gapdh. The calculated inductions are indicated by green squares. The error bars indicate the standard deviation from 2 to 6 independent experiments for the cells carrying the inducible Notch1 (NERT). Myf5 (G) is expressed at a very low level in undifferentiated cells (ESCs, ESCe, ESCm). Calculated inductions that do not reach significance above background levels (grey area) are indicated by an asterisk.
Figure 9.
RBP-J binding site sequences of potential direct Notch1 target genes.
Regulatory regions comprising 2000 bp before to 500 bp after the transcription start site of the Notch1 target genes Hes1, Hes5, Sox9, Pax6, Id4, Hey1, Runx1, and Myf5 were screened for potential RBP-J binding sites (red arrow head indicates direction of binding) using the ‘Transcription Element Search System’ (TESS; http://www.cbil.upenn.edu/tess; [67]) and a weight matrix from Ong et al. [68]. As a threshold, a Lg-likelyhood score (La) of 10 had to be exceeded. Binding sites were projected onto respective genomic regions using the UCSC Genome Browser (http://genome.ucsc.edu/) using the Feb. 2006 (NCBI36/mm8) assembly.
Figure 10.
Validating the Notch1 target genes Sox9, Pax6, Id4, Runx1 and Myf5.
Notch1 signaling was induced in ESC and differentiating embryoid bodies by addition of OHT (500 nM). (A) Immunoblot of EB5 NERT ESC (NERT ESCs) and EB5 NERT differentiated embryoid bodies (NERT Mes. 5d) with Sox9, Pax6 and Id4 antibodies. One representative example of three independent experiments is shown. (B) Mesodermal differentiation of Runx1-Venus-NERT ESC or Runx1-Venus control ESC and analysis of the percentage of venus+ and Flk1+ cells by FACS. Error bars indicate standard deviation for three independent experiments. (C) Mesodermal differentiation of Myf5-lacZ-NERT or Myf5-lacZ control ESC and detection of LacZ expression. One representative example of three independent experiments on day 5 is shown. Virtually identical results were obtained when the cells were analyzed 4 or 7 days, respectively, after induction of differentiation. Addition of OHT at day 3 of embryoid body culture similarly yielded in an increased number of lacZ positive cells at days 4, 5 and 7 in the presence of OHT after induction of differentiation.
Figure 11.
Expression of Notch1 target genes is altered in RBP-J deficient ES cells.
mRNA levels were measured by qPCR in ES cells cultured in self renewal medium (ESCs). The error bars indicate the standard deviation from 3 independent experiments. (A) Expression of potentially direct Notch1 target genes is increased in RBP-J deficient ES cells. Fold expression in relation to the expression of RBP-J +/+ D3 cells is shown. (B) The increased expression of Notch1 target genes in RBP-J -/- ES cells is repressed by exogenous RBP-J expression and further increased by a transcriptionally active form of RBP-J (RBP-J-VP16). Fold expression in relation to the expression of cells transfected with the control vector is shown.
Figure 12.
Model summarizing Notch1 cell-context dependent down-stream signaling.
(A) Cell-based model. CoR co-repressor; CoA co-activator; LSTF lineage specific transcription factors; green and red flag activating or repressing chromatin marks. (B) Context-dependent regulatory circuit model. See text for details.