The fibrotic phenotype induced by IGFBP-5 is regulated by MAPK activation and egr-1-dependent and -independent mechanisms

Abstract

We have previously shown that insulin-like growth factor (IGF) binding protein- 5 (IGFBP-5) is overexpressed in lung fibrosis and induces the production of extracellular matrix components, such as collagen and fibronectin, both in vitro and in vivo. The exact mechanism by which IGFBP-5 exerts these novel fibrotic effects is unknown. We thus examined the signaling cascades that mediate IGFBP-5-induced fibrosis. We demonstrate for the first time that IGFBP-5 induction of extracellular matrix occurs independently of IGF-I, and results from IGFBP-5 activation of MAPK signaling, which facilitates the translocation of IGFBP-5 to the nucleus. We examined the effects of IGFBP-5 on early growth response (Egr)-1, a transcription factor that is central to growth factor-mediated fibrosis. Egr-1 was up-regulated by IGFBP-5 in a MAPK-dependent manner and bound to nuclear IGFBP-5. In fibroblasts from Egr-1 knockout mice, induction of fibronectin by IGFBP-5 was abolished. Expression of Egr-1 in these cells rescued the extracellular matrix-promoting effects of IGFBP-5. Moreover, IGFBP-5 induced cell migration in an Egr-1-dependent manner. Notably, Egr-1 levels, similar to IGFBP-5, were increased in vivo in lung tissues and in vitro in primary fibroblasts of patients with pulmonary idiopathic fibrosis. Taken together, our findings suggest that IGFBP-5 induces a fibrotic phenotype via the activation of MAPK signaling and the induction of nuclear Egr-1 that interacts with IGFBP-5 and promotes fibrotic gene transcription.

Figure 1

Activation of intracellular signaling cascades induced by IGFBP-5. A: Fibroblasts were stimulated with 500 ng/ml recombinant IGFBP-5. Lysates were analyzed using Western blot. B: Fibroblasts were treated with 500 ng/ml IGFBP-5 or 50 ng/ml IGF-I for 10 minutes and protein phosphorylation was examined by Western blot. C: Fibroblasts were treated with 10 ng/ml IGF-I and increasing amounts of anti-IGF-I antibody. Lysates were harvested 10 minutes post-treatment. D: Fibroblasts were stimulated with 500 ng/ml IGFBP-5 for 10 minutes following a 1-hour pretreatment with anti-IGF-I antibody or rabbit IgG. MAPK activation was assessed using Western blot.

Figure 2

IGFBP-5 induces ECM and α-SMA production in a MAPK-dependent and IGF-I-independent manner. A: Fibroblasts were stimulated with 500 ng/ml IGFBP-5 or vehicle following a 1-hour incubation with MEK inhibitor U0126 or DMSO as a vehicle control. B: Fibroblasts were treated as in A following a 1 hour incubation with neutralizing anti-IGF-I antibody. ECM and α-SMA expression were analyzed by Western blot.

Figure 3

IGFBP-5-induced ECM and α-SMA production is regulated by the MAPK pathway. A: Fibroblasts were infected with Ad-IGFBP-5 (Ad5) or cAd. Activation status of MEK1/2 was analyzed by Western blot. B: Fibroblasts were treated as in A following a 1-hour incubation with 10 μmol/L MEK inhibitor U0126 (Meki) or DMSO as a vehicle control (veh). ECM and α-SMA levels were analyzed by Western blot.

Figure 4

Nuclear translocation of IGFBP-5 is not regulated by MAPK signaling in primary human lung fibroblasts. A: Fibroblasts were infected with Ad-IGFBP-5 (Ad5), Ad-IGFBP-3 (Ad3), or cAd at an MOI of 50. Cytoplasmic and nuclear fractions were extracted. Expression of IGFBP-3 and -5 was examined by Western blot. β-actin is shown as a loading control for cytoplasmic extracts, and histone H3 for nuclear extracts. B: Fibroblasts were infected with Ad-IGFBP-5 or cAd at an MOI of 50 following pretreatment with MEK inhibitor (MEKi) or DMSO as vehicle control (veh). Cytoplasmic and nuclear fractions were extracted. Expression of IGFBP-5 was examined by Western blot.

Figure 5

Egr-1 is up-regulated by IGFBP-5 in a MAPK-dependent manner. Fibroblasts were infected with Ad-IGFBP-5 or cAd. A: Egr-1 mRNA expression was analyzed by RT-PCR. B: Egr-1 protein levels were examined by Western blot in cellular lysates. C: Fibroblasts were incubated with 10 μmol/L MEK inhibitor U0126 (Meki) or DMSO as a vehicle control (veh) before infection. D: Fibroblasts were treated with IGFBP-5 or vehicle following a 1 hour-incubation with neutralizing anti-IGF-I antibody. E: Expression of Egr-1 was examined by immunohistochemistry in mouse lung 3 days after treatment with Ad-IGFBP-5 (Ad5) or control adenovirus (cAd). Magnification, ×400.

Figure 6

Increased Egr-1 expression in IPF. A: Egr-1 levels in primary fibroblasts from three patients with IPF and three healthy donors were analyzed by Western blot. B: Graphical summary of Egr-1 levels. Intensity of the bands on Western blot was analyzed using ImageJ and the ratio of Egr-1 to GAPDH was calculated. *P < 0.05. C: Egr-1 levels were detected using immunohistochemistry in lung tissues of three patients with IPF and three normal donors. Magnification: ×400; ×800 (Inset).

Figure 7

IGFBP-5-induced Egr-1 translocates to the nucleus in a MAPK-dependent manner, co-localizes and forms a complex with IGFBP-5. A: Primary fibroblasts were infected with Ad-IGFBP-5 (Ad5) or cAd following pretreatment with MEK inhibitor (MEKi) or DMSO as a vehicle (veh). Cytoplasmic and nuclear fractions were extracted after 72 hours. Egr-1 expression was examined by Western blot. β-actin is shown as a loading control for cytoplasmic extracts, and histone H3 for nuclear extracts. B: Expression of Egr-1 (red) and IGFBP-5 (green) was examined by immunofluorescence. DAPI was used to identify nuclei. Arrows show paracrine induction of Egr-1 in non-IGFBP-5-expressing cells. Magnification: ×1200. C: Protein-protein interaction between IGFBP-5 and Egr-1 was examined by immunoprecipitation with anti-Egr-1 antibody or rabbit IgG (IgG).

Figure 8

IGFBP-5 exerts similar effects in primary mouse and human lung fibroblasts. A: IGFBP-5 co-localizes with Egr-1 and IGFBP-5 induces Egr-1 in non-IGFBP-5-expressing mouse fibroblasts in a paracrine manner. Expression of Egr-1 (red) and IGFBP-5 (green) was examined by immunofluorescence. DAPI was used for nuclear counterstaining. Arrows show paracrine induction of Egr-1 in non-IGFBP-5-expressing cells. Images were taken at ×1200 magnification on a confocal microscope. B: IGFBP-5 and Egr-1 co-localize in the peri- and intranuclear cellular regions. Images were reconstructed with X-Z and Y-Z axis using MetaMorph. Yellow signal denotes co-localization. Images were taken at ×2000 on a confocal microscope. C: IGFBP-5 activates MAPK signaling in wild-type and Egr-1 KO mouse lung fibroblasts. Fibroblasts were stimulated with 500 ng/ml recombinant IGFBP-5. Protein phosphorylation was analyzed by Western blot.

Figure 9

IGFBP-5’s effects are Egr-1-dependent and -independent. A: Fibronectin induction by IGFBP-5 is abolished in Egr-1 KO mouse fibroblasts. Fibroblasts were infected with Ad-IGFBP-5 (Ad5) or cAd. Fibronectin expression was analyzed by Western blot. B: Graphical summary from three independent experiments. Intensity of the bands on Western blot was analyzed using ImageJ and the ratio of fibronectin to GAPDH was calculated; *P < 0.05. C: Fibroblasts were infected with Ad-IGFBP-5, and cytoplasmic and nuclear fractions were extracted at 24 to 96 hours. Expression of Egr-1 (nuclear) and fibronectin (cytoplasmic) were examined by Western blot. Histone and GAPDH were used as loading controls in nuclear and cytoplasmic extracts, respectively. D: Fibroblasts were infected with Ad-IGFBP-5. Cytoplasmic and nuclear fractions were extracted after 72 hours. Expression of IGFBP-5 was analyzed by Western blot. Signal intensity was quantified using ImageJ. The number denotes ratio of IGFBP-5 in nuclear extract to that in cytoplasmic extract. E: Egr-1 KO lung fibroblasts were infected with Ad-IGFBP5, Ad-Egr-1, or both at a multiplicity of infection of 25 each. Extracellular matrix was harvested after 72 hours. Fibronectin deposition was assessed in the extracellular matrix using Western blot. Vitronectin served as a loading control.

Figure 10

IGFBP-5 induces migration of mouse lung fibroblasts and splenocytes in an Egr-1-dependent manner. A: Primary mouse lung fibroblasts from wild-type and Egr-1 KO mice were used in migration assays with a vehicle control (veh) or recombinant IGFBP-5 (rBP5). B: Migration of lung fibroblasts from wild-type and Egr-1 KO mice. Data represent mean +SD of 15 independent fields. *P < 9.0 × 10⁻⁴, **P < 5.0 × 10⁻⁵, ***P < 0.05. C: Migration of splenocytes from wild-type and Egr-1 KO mice. Data represent mean + SD of six independent experiments. *P < 0.003, **P < 0.006.