Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2019 Aug;13(8):1706-1724.
doi: 10.1002/1878-0261.12504. Epub 2019 Jun 19.

Fibroblast Growth Factor Signals Regulate Transforming Growth Factor-β-Induced Endothelial-To-Myofibroblast Transition of Tumor Endothelial Cells via Elk1

Affiliations
Free PMC article

Fibroblast Growth Factor Signals Regulate Transforming Growth Factor-β-Induced Endothelial-To-Myofibroblast Transition of Tumor Endothelial Cells via Elk1

Yuichi Akatsu et al. Mol Oncol. .
Free PMC article

Abstract

The tumor microenvironment contains various components, including cancer cells, tumor vessels, and cancer-associated fibroblasts, the latter of which are comprised of tumor-promoting myofibroblasts and tumor-suppressing fibroblasts. Multiple lines of evidence indicate that transforming growth factor-β (TGF-β) induces the formation of myofibroblasts and other types of mesenchymal (non-myofibroblastic) cells from endothelial cells. Recent reports show that fibroblast growth factor 2 (FGF2) modulates TGF-β-induced mesenchymal transition of endothelial cells, but the molecular mechanisms behind the signals that control transcriptional networks during the formation of different groups of fibroblasts remain largely unclear. Here, we studied the roles of FGF2 during the regulation of TGF-β-induced mesenchymal transition of tumor endothelial cells (TECs). We demonstrated that auto/paracrine FGF signals in TECs inhibit TGF-β-induced endothelial-to-myofibroblast transition (End-MyoT), leading to suppressed formation of contractile myofibroblast cells, but on the other hand can also collaborate with TGF-β in promoting the formation of active fibroblastic cells which have migratory and proliferative properties. FGF2 modulated TGF-β-induced formation of myofibroblastic and non-myofibroblastic cells from TECs via transcriptional regulation of various mesenchymal markers and growth factors. Furthermore, we observed that TECs treated with TGF-β were more competent in promoting in vivo tumor growth than TECs treated with TGF-β and FGF2. Mechanistically, we showed that Elk1 mediated FGF2-induced inhibition of End-MyoT via inhibition of TGF-β-induced transcriptional activation of α-smooth muscle actin promoter by myocardin-related transcription factor-A. Our data suggest that TGF-β and FGF2 oppose and cooperate with each other during the formation of myofibroblastic and non-myofibroblastic cells from TECs, which in turn determines the characteristics of mesenchymal cells in the tumor microenvironment.

Keywords: Elk1; End-MyoT; End-N-MyoT; EndMT; FGF2; TGF-β2.

Conflict of interest statement

YA is an employee of Nippon Kayaku, Co., Ltd. All other authors declare that they have no competing interests.

Figures

Figure 1
Figure 1
Effects of TGF‐β2, FGF2, and Infigratinib on the expression of mesenchymal markers in TECs. TECs were cultured in the absence (−) or presence (+) of 1 ng·mL−1 of TGF‐β2 in combination with 50 ng·mL−1 of FGF2 or 3 μm of Infigratinib (a pan‐inhibitor of FGF receptors) for 72 h, followed by qRTPCR analysis for the expression of α‐SMA (A), SM22α (B), Col1A1 (C), Fibronectin (D), and Tie2 (E) and immunoblotting analysis for the expression of α‐SMA (F), SM22α (G), and β‐actin (F, G). Concentrations of TGF‐β2 and FGF2 used were determined based on their effects on the expression of various markers (data not shown). Asterisk indicates nonspecific bands generated by anti‐SM22α antibody (G). Error bars represent standard deviation. Student's t‐test with two biological independent replicates was used to determine statistical significance; *< 0.05; N.S., not significant; N.D., not detectable.
Figure 2
Figure 2
Regulation of FGF signals by TGF‐β2 in TECs. (A, B) Effect of TGF‐β2 on the expression of FGF2 in TECs. TECs were cultured in the absence (−) or presence (+) of TGF‐β2 for 72 h, followed by qRTPCR (A) and immunoblot (B) analyses for the expression of FGF2. rFGF2, recombinant FGF2. (C) Effect of TGF‐β2 on the ERK1/2 phosphorylation in TECs. TECs were cultured in the absence (−) or presence (+) of TGF‐β2 for 72 h, followed by immunoblot analysis using phospho‐(P)‐ERK1/2, and total ERK1/2 antibodies. (D, E) Effect of endogenous FGF2 on α‐SMA expression and the ERK1/2 phosphorylation in TECs. TECs were cultured with TGF‐β2 in combination with control‐IgG or anti‐FGF2 antibodies for 72 h, followed by qRTPCR analysis for α‐SMA expression (D) and immunoblot analysis using phospho‐(P)‐ERK1/2, and total ERK1/2 antibodies (E). Error bars represent standard deviation. Student's t‐test with two biological independent replicates was used to determine statistical significance; *< 0.05.
Figure 3
Figure 3
Effects of TGF‐β2 and FGF2 on the myofibroblastic properties of TECs. (A) TECs were preincubated with or without TGF‐β2, FGF2, or combination of both factors for 72 h and embedded into collagen matrices. The mixtures were released from the culture dishes, incubated for additional 2 days, and photographed. Experiments were performed in duplicate. Scale bar: 10 mm. (B) Graphic representation of the relative gel areas quantified by imagej. Error bars represent standard deviation. Student's t‐test with three biological independent replicates was used to determine statistical significance; *< 0.05; N.S., not significant. (C) TECs were preincubated with (+) or without (−) TGF‐β2, FGF2, or combination of both factors for 72 h, followed by fluorescence immunostaining for F‐actin (white) and nuclei (blue). Scale bars: 10 mm (A) and 50 μm (C).
Figure 4
Figure 4
Effects of TGF‐β2 and FGF2 on the migration and proliferation of TECs. (A, B) TECs were preincubated with (+) or without (−) TGF‐β2, FGF2, or combination of both factors for 72 h, followed by chamber migration assay. Cells were allowed to migrate for 6 h, and cells migrated to the bottom side of the chamber were stained (A) and counted under a phase‐contrast microscope (B). Scale bar: 300 μm. (C) TECs were cultured with or without TGF‐β2, FGF2, or combination of both factors for 72 h, followed by direct counting of cell number. Note that treatment of TECs with TGF‐β2, FGF2, or combination of both factors for 6 h did not affect their number (data not shown). Error bars represent standard deviation. Student's t‐test with three biological independent replicates was used to determine statistical significance; *< 0.05; N.S., not significant.
Figure 5
Figure 5
Differential effects of FGF2 on the TGF‐β2‐mediated expression of various markers in TECs. (A) Heatmap and hierarchical clustering of gene expression in TECs treated with TGF‐β2, FGF2, or combination of both factors and analyzed using the Agilent Expression Array data. Results were normalized and log‐transformed. Genes were clustered using the hierarchical method. (B, C) TECs were cultured in the absence (−) or presence (+) of TGF‐β2 in combination with FGF2 or Infigratinib for 72 h, followed by qRTPCR analysis for the expression of Rgs4 (B) and HBEGF (C). Error bars represent standard deviation. Student's t‐test with two biological independent replicates was used to determine statistical significance; *< 0.05; N.S., not significant.
Figure 6
Figure 6
Roles of TECs treated with TGF‐β and FGF2 in the in vivo tumor formation of the A375 human melanoma cell. (A) TECs pretreated either with TGF‐β2 (End‐MyoT TECs) or combination of TGF‐β2 and FGF2 (End‐N‐MyoT TECs) for 72 h were mixed with A375 human melanoma cells in a 3 : 10 ratio and subcutaneously inoculated into immunodeficient mice. Tumor growth was measured using calipers and calculated from minor axis and major radius. One‐way ANOVA followed by the Student–Newman–Keuls test with four (End‐MyoT TEC group) and six (End‐N‐MyoT TEC group) biological independent replicates was used to determine statistical significance (A); *< 0.05. (B–E) Sections of tumors were subjected to immunofluorescence staining with the anti‐Ki67 (purple: B) and anti‐PECAM‐1 antibodies (D: green). Nuclei were counterstained with Hoechst 33342 (blue). Scale bars: 50 μm (B) and 100 μm (D). Levels of proliferation (C) and angiogenesis (E) were quantified. Error bars represent standard error. Student's t‐test with twelve (C) and twenty (E) biological independent replicates was used to determine statistical significance (C, E); *< 0.05.
Figure 7
Figure 7
Effects of Elk1 on the TGF‐β2‐induced expression of myofibroblast markers in TECs. (A) TECs were transfected with negative control siRNA (NC) or siRNA for Elk1 (siElk1 #1 and #2). The expression of Elk1 was determined by qRTPCR analysis. (B–D) TECs transfected with siRNA were cultured in the absence (−) or presence (+) of TGF‐β2 for 72 h, followed by qRTPCR analysis for the expression of α‐SMA (B) and Rgs4 (D), and immunoblot analysis for the expression of α‐SMA and β‐actin (C). (E–H) TECs were infected with lentivirus encoding GFP as control or Elk1, followed by qRTPCR analysis for the expression of Elk1 (E). Elk1‐expressing or control TECs were cultured in the absence (−) or presence (+) of TGF‐β2 for 72 h, followed by qRTPCR analysis for the expression of α‐SMA (F) and Rgs4 (H), and immunoblot analysis for the expression of α‐SMA and β‐actin (G). Error bars represent standard deviation. Student's t‐test with two biological independent replicates was used to determine statistical significance; *< 0.05.
Figure 8
Figure 8
Roles of TGF‐β2, FGF2, Elk1, and MRTF‐A during myofibroblast transition of TECs. (A–C) TECs were cultured with or without TGF‐β2, FGF2, or combination of both factors for the indicated periods, followed by qRTPCR analysis for the expression of MRTF‐A (A), α‐SMA (B), and Elk1 (C). (D) MS‐1 cells were transfected with a luciferase reporter construct containing the α‐SMA promoter fragment along with expression construct encoding MRTF‐A and/or Elk1, in the presence or absence of TGF‐β2 and FGF2, followed by the measurement of luciferase activity. Error bars represent standard deviation. Student's t‐test with two biological independent replicates was used to determine statistical significance; *< 0.05.

Similar articles

See all similar articles

Cited by 2 articles

References

    1. Akita S, Akino K, Imaizumi T and Hirano A (2008) Basic fibroblast growth factor accelerates and improves second‐degree burn wound healing. Wound Repair Regen 16, 635–641. - PubMed
    1. Anderberg C, Cunha SI, Zhai Z, Cortez E, Pardali E, Johnson JR, Franco M, Paez‐Ribes M, Cordiner R, Fuxe J et al (2013) Deficiency for endoglin in tumor vasculature weakens the endothelial barrier to metastatic dissemination. J Exp Med 210, 563–579. - PMC - PubMed
    1. Augsten M (2014) Cancer‐associated fibroblasts as another polarized cell type of the tumor microenvironment. Front Oncol 4, 62. - PMC - PubMed
    1. Besnard A, Galan‐Rodriguez B, Vanhoutte P and Caboche J (2011) Elk‐1 a transcription factor with multiple facets in the brain. Front Neurosci 5, 35. - PMC - PubMed
    1. Castets M, Coissieux MM, Delloye‐Bourgeois C, Bernard L, Delcros JG, Bernet A, Laudet V and Mehlen P (2009) Inhibition of endothelial cell apoptosis by netrin‐1 during angiogenesis. Dev Cell 16, 614–620. - PubMed

Publication types

MeSH terms

LinkOut - more resources

Feedback