Figure 1.
TAL1-, LYL1- or LMO2-silencing down regulates ANG-2 expression.
(A, C) HUVECs were transfected with siRNAs targeting TAL1, LYL1, LMO2 or eGFP (used as the control). Whole cell extracts and total RNAs were prepared 48 hours post transfection for expression analysis. (B, C) LLyECs were transduced with pLKO.1 lentiviruses encoding shRNA targeting TAL1, LYL1, LMO2 or control shRNA. Puromycine-resistant populations were analyzed between 12 to 20 days after transduction. (A. B) Top, bottom ANG-2 and LYL1 mRNA levels was analyzed by RT-q-PCR. The signals of ANG-2 and LYL1 mRNA were normalized to those of GAPDH. The data are shown relative to mRNA content from control siRNA-treated cells arbitrarily set at 1. Each bar is the mean +/− SD of three independent experiments. Middle Intracellular ANG-2, TAL1, LMO2 and LYL1 protein expression was analyzed by immunoblotting. β-ACTIN expression was monitored to control protein loading (the bar indicates two independent immunoblots). The images shown are representative of three independent experiments. **, P<0.01; ***, P<0.001. (C) TAL1-, LYL1- or LMO2-depletion reduces constitutive ANG-2 secretion in culture medium. Left: HUVECs were transfected with the indicated siRNA. 24 hours after transfection, medium was changed and ANG2 secretion into culture medium for 24 hours was measured by ELISA. Right: LLyECs were transduced with pLKO.1 lentiviruses encoding the indicated shRNA and puromycine-resistant cell populations were grown to confluence. ELISA was used to measure ANG2 release for 48 hours into the media. ANG-2 production is shown relative to cell number measured at the end of the culture. *, P<0.05; **, P<0.01; ***, P<0.001.
Figure 2.
TAL1, LYL1, LMO2 and GATA2 bind the ANG-2 promoter proximal region.
(A) Alignment of the nucleotide sequence of proximal region of human (top line) and mouse ANG-2 promoter. The conserved E-box, GATA-binding sites and GC-box boxes are indicated. The TATA box is underlined. +1 indicates the transcription start site. Of note, the E box−64 is separated by 11 nucleotides from both the upstream GC-box−85 and the downstream GATA−47. (B) The indicated endothelial cells were transfected with either the wild-type −412/+73 ANG-2 reporter construct (black bars) or the mutated Ebox−64 construct (grey bars). Data are shown relative to the luciferase activity of wt construct (set at 100%) for each cell type. Bars are the means +/− S.D. of three independent experiments, each performed in triplicate. **, P<0.01; *, P<0.05. (C) Chromatin immunoprecipitation assays. Chromatin immunoprecipitations assays (ChIP) were performed on cross-linked chromatin from LLyECs using the indicated antibody. To control chromatin quality, ChIP assays were performed using an antibody against acetylated histone H3. Aliquots of immunoprecipitated DNAs were amplified by PCR with primers (arrows) targeting the proximal ANG-2 promoter (−165; +76) or a non-transcribed genomic region upstream of the c-KIT gene used as a negative control. PCR products were analyzed by gel electrophoresis. Top: Schematic location of conserved elements within the human ANG-2 promoter: G: GATA-binding site; GC: GC-box; E: E-box. Shown images are representative of one of three independent experiments. Bottom: Fold-enrichment of target genomic regions immunoprecipitated by each antibody was normalized to the levels obtained with control IgGs, which was assigned a value of 1, and plotted as the increase over the levels of enrichment at the negative control region. The error bars represent +/− S.D. from three independent ChIPs. **, P<0.01; *, P<0.05.
Figure 3.
Analysis of TAL1, LYL1 and LMO2 interactions in cells.
293T cells were cotransfected with the indicated expression vectors along with eGFP-expressing vector to control transfection efficacy. (A) TAL1 and LYL1 interactions: Equal amounts of whole cell lysates (WCL) were immunoprecipitated with the indicated antibody. Mouse α-Flag- mAb was used to precipitate Flag-LYL1. Input shows levels of tested proteins in untreated WCL (10 µg per lane). Input and immunoprecipitated proteins were analyzed by immunoblot using the appropriate antibodies. (B) E47 induces TAL1 hyperphosphorylation: WCL from cells transfected with combination a or b (indicated in A) were immunoprecipitated with mouse α-TAL1 mAb. Immunoprecipitates coupled to sepharose were treated by Calf Intestinal alcaline Phosphatase (CIP) and analyzed by immunoblot using goat α-TAL1 pAb. (C) LMO2 interactions: Equal amounts of WCL were treated with mouse α-HA-mAb coupled to agarose to precipitate HA-LMO2. Input shows levels of tested proteins in untreated WCL (10 µg per lane). Input and immunoprecipitated proteins were analyzed by immunoblot using the appropriate antibodies.
Figure 4.
LYL1 is able to form homodimers in vitro and in vivo.
(A) GST pull-down assays: 35S-labeled TAL1 and LYL1 proteins produced in vitro by coupled transcription-translation were incubated with the indicated GST protein coupled to glutathione-Sepharose beads. Captured radiolabeled proteins were analyzed by SDS-PAGE and visualized by autoradiography. (B) LYL1 forms homodimers in cells: 293T cells were transfected with Flag-LYL1 and/or HA-LYL1. Equal amounts of WCL were immunoprecipitated with α-Flag- and α-HA-antibodies to precipitate LYL1-Flag and LYL1-HA respectively. Input shows levels of tested proteins in untreated WCL (10 µg per lane). Input and immunoprecipitated proteins were analyzed by immunoblot using the indicated antibody.
Figure 5.
TAL1- and/or LYL1-containing complexes activate endogenous ANG-2 transcription.
Endothelial HMEC-1 cells or epithelial 293T cells were transfected with the indicated combination of expression plasmids along with pEGFP to control transfection efficiency; GFP-positive cells monitored by FACS analysis varied from 60 to 80% of total living cells. Total RNA and whole cell extracts (WCL) were prepared 48 h post-transfection for gene expression analysis. (A) ANG2 mRNA expression was analyzed by RT-qPCR. The signals of mRNAs were normalized to those of GAPDH. Results are expressed as ANG-2 mRNA levels relative to those of cells transfected with empty vector set at 1 (dotted line). Results are shown as the means ± SD of at least three separate experiments. Asterisks show significant differences with mRNA levels from cells transfected with empty vector. *or #, P<0.05; ** or ##, P<0.01; ***, P<0.001. (B) WCL (10 µg) from the indicated transfected cells were analysed by immunoblot to control ectopic protein expression using the appropriate antibodies. Shown images are representative of a typical experiment; vertical lines indicate the suppression of non-informative lanes from a unique blot.
Figure 6.
ANG-2 gene activation during endothelial tubulogenesis coincides with the upregulation of TAL1 and LMO2 proteins.
Confluent HUVECs were either maintained in basal endothelial medium containing 5% of fetal calf serum (NS: non-stimulated) or primed (P) with VEGF-A and bFGF for tubulogenesis. Primed HUVECs were seeded into collagen-I gels containing SDF-1α, SCF and IL3 to undergo vascular tube morphogenesis over 72 hours. Cells were recovered at the indicated time points to prepare total RNA and/or whole cell extracts. (A) Time-course analysis of ANG-2 gene expression during in vitro tubulogenesis. ANG-2 mRNA expression was analyzed by RT-q-PCR. To analyze ANG-2 primary transcripts (pre-ANG2), ANG-2 sequences from intron 1 were amplified. The signals of mRNAs were normalized to those of GAPDH. Data shown are RNA levels relative to the corresponding RNA levels of non-stimulated cells (NS) arbitrarily set at 1. Each bar is the mean +/− SD. of three independent experiments (*: P values versus NS) *, #: P<0.05; **, ##: P<0.01; ***, ###: P<0.001. (B, C) Time-course analysis of ANG-2, TAL1, LMO2, LYL1, E47 and GATA2 protein expression during tubulogenesis. Whole cell lysates (30 µg) were used to analyze the indicated protein by immunoblot. (B): the images shown are representative of two tubulogenesis experiments. The arrows indicate the different forms of ANG-2: precursor (white), glycosylated monomers (grey); glycosylated dimers (black). The asterisk pinpoints the more slowly migrating isoform of GATA2, indicating post-translational modification. (C) Protein quantification was done by densitometric analysis. The values represent the amounts of the indicated protein normalized to β-Actin and relative to its amounts in non-stimulated cells, arbitrarily set at 1. Each bar is the mean +/− SD of two independent experiments (*: p values versus non-stimulated). *, #: P<0.05; **: P<0.01; (D) Shown photographs illustrate the formation of the tubule network by primed HUVECs at the indicated time points of one experiment.
Figure 7.
A proposed model illustrating how bHLHs may operate in endothelial cells to modulate ANG-2 expression.
The majority of HLH-dimers present in resting endothelial cells are LYL1 homodimers that, associated with LMO2 and GATA2 and with as yet unknown X cofactor(s), participate to constitutive ANG-2 transcription. Upon angiogenic activation, the up-expression of TAL1 and LMO2 leads to the additional formation and binding of the classical TAL1-LMO2-containing complex onto the proximal ANG-2 promoter through its interaction with GATA2 constitutively bound to the promoter, thereby enhancing its transcriptional activity. Higher-order complexes composed of two different bHLH dimers may also be formed due to the ability of LMO2 to dimerize through its interaction with the ubiquitous LDB1 nuclear protein.