Figure 1.
Daoy cells respond to SAG and Shh-N by upregulating canonical HH/GLI targets.
(A) Incubation with SAG (100 nM) induced the expression of GLI1 protein and inhibited processing of endogenous GLI3 to its repressor form. Co-incubation with Cyclopamine (cyc, 5 µM) inhibited the SAG-induced GLI1 expression and promoted processing of GLI3 to its repressor (GLI3R) form and inhibited GLI2 expression. Beta-actin shown as Western blot loading control. (B) Transcripts for GLI1 and PTCH were upregulated after stimulation with Shh-N for 24 h, induction of HHIP transcripts was seen after 48 h. Enhanced GLI1, HHIP, and PTCH expression levels were still observed after 72 hours. (C) Schematic presentation of canonical Hedgehog signaling. Left panel illustrates silencing of Hedgehog-mediated signaling via a PTCH-mediated block of SMO so that repressor GLI3 prevails and limits the expression rate of Hedgehog target genes. The right panel illustrates the “Hedgehog-on” state, activating GLI proteins now control the expression of Hedgehog-target genes such as GLI1, PTCH, and HHIP. Upregulation of PTCH and HHIP will eventually result in the downregulation of Hedgehog-signaling. Red color indicates a protein with inactivating properties, white color indicates a protein with activating properties, and grey color indicates RNA transcripts.
Figure 2.
Uptake of EGF by Daoy cells and time-resolved analysis of EGF- downstream signaling.
(A) To measure the EGF receptor binding and depletion of the growth factor from the medium supernatant a fluorescent ELISA was performed. Daoy cells were treated with 5 ng/ml of recombinant EGF, and medium supernatants were collected after the indicated time periods (light grey symbols). To control that the ligand is bound and taken up by the cells, and not just degraded by the incubation conditions within the cell culture incubator, wells without cells served as controls (dark grey symbols). (B) Activation of EGF downstream pathways (PI3K/AKT and MAPK) was analyzed by Western blot with phospho-specific antibodies against AKT and ERK. Prior to the application of 5 ng/ml EGF ligand the cells had been starved in low serum (0.5% FBS) medium o/n to minimize basal pathway activation. After the addition of EGF cell lysates were prepared at the indicated timepoints. One sample remained untreated and served as control. Loading of equal protein amounts was checked by probing the Western blot membranes with ß-Tubulin.
Table 1.
Targets differentially regulated in response to Shh-N/EGF-crosstalk.
Figure 3.
Stimulation of Hedgehog-primed Daoy cells with EGF induced downregulation of canonical HH/GLI targets.
Daoy cells were either treated with control medium (purple), Shh-N conditioned medium (blue), EGF (red) or a combinatorial treatment with Shh-N medium and EGF (green) treated for 24 h. (A) Transcript dynamics for GLI1, PTCH, and HHIP were analyzed by transcript profiling on whole genome arrays. Curves represent mean values of independent biological experiments (n = 3, +/−SEM). (B) Taqman RT-PCR confirmed downregulation of canonical Hedghog targets by EGF.
Figure 4.
Downregulation of GLI1 as canonical Hedgehog-induced transcript was not rescued after inhibition of MEK/ERK and PI3K/AKT signaling.
(A) Impact of PI3K inhibition using LY294002 on the EGF-induced downregulation of GLI1 transcripts after stimulation of Shh-N primed cells for 3 h with EGFR ligands. (B) Impact of MEK1 inhibition using PD98059 on the EGF-induced downregulation of GLI1 transcripts measured 3 h after ligand stimulation of Shh-N primed cells. Figure S3 shows data from a corresponding experiment using U0126 an inhibitor of MEK1/2 and PI103 as a PI3K inhibitor.
Figure 5.
EGF stimulation of Hedgehog-primed Daoy cells strongly upregulated canonical EGF target genes.
Cells were treated with either control medium (purple), Shh-N conditioned medium (blue), EGF (red) or a combinatorial treatment with Shh-N medium and EGF (green). (A) Gene expression profiling data indicated upregulation of MMP7, VEGFA, and IL-8. (B) Taqman Real Time PCR confirmed upregulation of canonical EGF targets in Hedgehog-primed cells. (C) MMP7, VEGFA, and IL-8 protein levels were quantified by ELISA in cell culture supernatants at indicated time points. Average fold-changes or concentrations were obtained as mean of three independent biological experiments (+/−SEM).
Figure 6.
Hedgehog/EGF crosstalk results in maintenance of high GLI1 levels without affecting GLI3 processing.
Samples were obtained after inducing EGFR signaling in Shh-N primed Daoy cells (Shh-N) or unprimed Daoy cells (EGF) for 3 h or 18 h with EGF (A) or AREG (B). Control cells were neither primed with SAG nor exposed to EGF. Western blots were probed with antibodies against GLI1, GLI3 and HSP70. Figure S6 A shows GLI3A/GLI3R ratios for crosstalk induced by EGF and AREG.
Figure 7.
Hedgehog/EGF crosstalk analyzed on the signaling level by RPPA.
Samples were obtained at 14 time points after inducing EGFR signaling in Shh-N primed Daoy cells or control Daoy cells by EGF. Daoy cells were either treated with control medium (purple), Shh-N conditioned medium (blue), EGF (red) or a combinatorial treatment with Shh-N medium and EGF (green) treated for 24 h. Data for selected phosphoproteins are shown for short time signaling (A: 0–6 h) and long term signaling (B: 6–24 h). Certain kinases, e.g. MEK and ERK1/2, clearly depend on fast signals mediated by EGFR, most proteins assessed by RPPA show no significant changes on the phosphoprotein level as shown here for p70S6K and GSK3 signaling. Table 2 summarizes phosphoproteins probed by RPPA.
Figure 8.
. Molecular interactions between Hedgehog and EGF signaling pathways occur on the transcriptional level. Rectangular symbols indicate proteins, oval symbols reflect transcripts. Protein translocation is shown as dashed grey lines, ligand-induced activation of signaling pathways mediated by phosphorylation events is shown as dashed red lines and induction of protein expression as blue lines. Proteins and/or transcripts are produced in response to EGF and/or Shh-N signaling. Release of IL8 is shown as an example of crosstalk induced enhanced secretion of tumor-promoting inflammatory extracellular proteins. GLI1-A and AP-1 interact on the transcriptional level to enhance production of canonical EGF target genes. Numerous phosphoproteins are activated by EGFR-signaling with PI3K/AKT and MEK/ERK signaling as major downstream signaling events. However, both pathways do neither contribute to a stabilization of GLI1 on the protein level nor do they contribute to the observed silencing of canonical Hedgehog target genes on the transcript level. Low activity EGFR signaling events such might play a role in this process or might be mediated by EGFR-induced miRs.
Table 2.
Phosphoproteins quantified by RPPA.