EphrinB2 Regulates Cardiac Fibrosis Through Modulating the Interaction of Stat3 and TGF / 四六文摘

Introduction

Heart failure is a clinical syndrome of a variety of cardiovascular diseases and results in high morbidity and mortality rates through left ventricular remodeling, in which cardiac fibrosis is the major pathological change. Excessive deposition of extracellular matrix in the cardiac interstitium as part of the repair process triggered by the loss of cardiomyocytes is usually referred as replacement fibrosis, whereas reactive fibrosis occurs when there is no significant loss of cardiomyocytes. Replacement fibrosis is mainly observed during the early repair phase after myocardial infarction (MI), whereas reactive fibrosis is typically seen in pressure overload–induced cardiac hypertrophy.¹ Both of these fibrotic responses increase myocardial stiffness and subsequently remodel the ventricular structure, leading to systolic and diastolic dysfunctions, which eventually result in heart failure. The differentiation of cardiac fibroblasts into a unique cell type, myofibroblasts, predominantly contributes to cardiac fibrosis. Myofibroblasts can also potentially be transdifferentiated from a variety of other cell types within an injured heart, resulting in aggravated extracellular matrix production and hence cardiac remodeling.² Because of its key roles in cardiac fibrotic process, cardiac myofibroblast activation, as an important therapeutic target, has attracted many attractions in this field.

Meet the First Author, see p 586

EphrinB2 (erythropoietin-producing hepatoma interactor B2), a cell surface transmembrane ligand for erythropoietin-producing hepatoma B receptors, is a pivotal bidirectional signaling molecule and is ubiquitously expressed in mammals. On binding to erythropoietin-producing hepatoma B receptors residing on adjacent cells, EphrinB2 not only sends a forward signal through the activation of the receptor tyrosine kinase erythropoietin-producing hepatoma B but also triggers intrinsic signaling referred to as reverse signaling.³ Emerging studies have indicated that EphrinB2 reverse signaling is crucial in angiogenesis during development and disease progression. Both global inactivation of EphrinB2 and deletion of the intracellular domain of EphrinB2 impair angiogenesis and result in embryonic lethality in mice.^4,5 Foo et al⁶ observed that the mice with pericyte-derived EphrinB2 deficiency exhibited aberrant fibrosis surrounding abnormal vessels in dermal tissue, which indicates a possible role of EphrinB2 signaling in fibrogenesis. Recently, Kida et al⁷ discovered that EphrinB2 signaling in mouse pericytes protects pericyte-to-myofibroblast transition and prevents myofibroblast activation to limit renal fibrosis after injury, indicating a beneficial role of EphrinB2 in protecting against kidney injury. In contrast, Lagares et al⁸ reported that EphrinB2 was increased in cultured fibroblasts from patients with idiopathic pulmonary fibrosis and was capable of inducing the differentiation of human lung fibroblasts into myofibroblasts in vitro. Taken together, these findings imply that EphrinB2 may have organ-specific effects on fibrosis. However, the role of EphrinB2 in cardiac fibrosis remains unknown. Therefore, in the present study, we aimed to investigate the role and the underlying mechanism of action of EphrinB2 in cardiac fibrosis after injury.

Methods

Human Heart Tissue

The collection and use of human heart tissues were approved by the Human Research Ethics Committee of the Second Affiliated Hospital, Zhejiang University School of Medicine. Normal human hearts were obtained from 3 healthy donors who died from brain death. Failing hearts were obtained from 3 patients with dilated cardiomyopathy who underwent heart transplantation at the Second Affiliated Hospital, Zhejiang University School of Medicine. All recipients and 1 family member of each donor signed informed consent forms. The healthy and failing hearts were collected and flushed with ice-cold cardioplegic solution and immediately transported in an ice box to the laboratory. The left ventricle free walls from these hearts were harvested: half of the heart was frozen in liquid nitrogen for protein extraction, and the other part was fixed in formaldehyde solution for immunohistochemical staining.

Animal Experiments

Male C57BL/6 mice at the age of 8 to 10 weeks were purchased from Shanghai Laboratory Animal Center (SCXK (HU) 2012-0002, Shanghai, China). All animals were fed a chow diet in a 12-hour light/12-hour dark environment at 25°C in the Animal Care Facility of the Second Affiliated Hospital, Zhejiang University School of Medicine. All animal experiments were approved by the Animal Care and Utilization Committee of Zhejiang University (zju201308-1-01-085). The mouse model of MI was established as previously described.⁹ Briefly, mice were anesthetized with sodium pentobarbital (50 mg/kg intraperitoneally). Then MI was induced via permanent ligation of the left anterior descending artery with a 7-0 Prolene suture, whereas sham-operated mice underwent the same procedure but without ligation. At least 5 mice were contained in each group including sham-operated group and MI group. Before the operation, baseline of cardiac function was evaluated via echocardiography; the mean ejection fraction was 85.40±2.56%. We measured ejection fraction 3 days after MI to attest that the MI model was successfully established. The angiotensin II–induced model of cardiac hypertrophy was established as previously described.¹⁰ In brief, primed osmotic minipumps (Alzet model 2004; Alza Corp) were subcutaneously implanted in the dorsal region of mice; pumps contained a 28-day infusion of angiotensin II (1000 ng kg⁻¹ min⁻¹; Sigma-Aldrich) dissolved in sterile saline. Echocardiography was used to assess the myocardial hypertrophy by measuring the wall thickness.

Lentivirus Construction

Lentiviruses expressing full-length mouse efnb2 cDNA and lentiviruses carrying shRNA against efnb2 were constructed by Shanghai Genechem Company (Shanghai, China). The RNAi sequence targeting mouse efnb2 was 5′-GCTAGAAGCTGGTACAAAT-3′. Viruses were amplified and titrated in 293 T cells according to manufacturer’s instructions. Lentiviruses containing empty plasmids (vector) and lentiviruses containing nonspecific shRNA (scramble) were used as controls, respectively.

Plasmid Construction

Expression plasmids for Myc-tagged Smad3 or Flag-tagged Stat3 and Y705E (constitutively active Stat3) referred to the ones previously described.¹¹ Myc-tagged full-length Smad3 or deletion mutants (deletion of 1-144aa, 145-175aa, or 176-425aa from Smad3, respectively) were generated via polymerase chain reactions and cloned into the Xhol and Kpnl sites of pGV219 (CMV-MCS-SV40-Neomycin). Flag-tagged full-length Stat3 or deletion mutants (deletion of 1-120aa, 84-290aa, 141-319aa, 321-573aa, 554-715aa, or 716-770aa from Stat3, respectively) were generated via polymerase chain reaction and cloned into the Xhol and Kpnl sites of pGV141 (CMV-MCS-3FLAG-SV40-Neomycin).

Cell Culture and Transfection

HEK293T cells were cultured in Dulbecco’s modified Eagle’s medium (Corning) with 10% fetal bovine serum (Gibco). HEK293T cells were transfected with Lipofectamine 3000 (Invitrogen) according to the manufacturer’s instructions.

Western Blotting

Proteins were isolated from the snap-frozen heart tissue and cultured cardiac fibroblasts, which were extracted in RIPA solution (Beyotime, Shanghai, China) with a protease inhibitor cocktail (Roche). Next, proteins were quantified using a BCA Protein Assay Kit (Thermo). Then, 20 μg of each protein sample was separated via SDS-PAGE and electrotransferred onto PVDF (polyvinylidene fluoride) membranes. Following blockade with PBST (phosphate-buffered solution with Tween-20) containing 5% BSA, membranes were incubated with primary antibodies against the following molecules overnight at 4°C: EphrinB2, α-SMA (α-smooth muscle actin), Collagen I, TGF-β (Abcam), P-Smad3 (Ser423/425), Smad3, P-Jak2 (Tyr1007/1008), Jak2 (Janus kinase 2), P-Stat3 (Tyr705), and Stat3 (Cell Signaling Technology). Antigen and antibody complexes were detected with an ECL protocol using HRP-conjugated IgG as secondary antibodies. Immunoblots were quantified using Image Laboratory (version 2.0) software.

Isolation of Cardiac Fibroblasts

Cardiac fibroblasts were isolated from neonatal C57BL/6 mice as previously described.¹² Hearts from 1-day-old mice were cut into small chunks. The tissue was digested with 0.05% (wt/vol) tryptase/EDTA (Gibco) and 0.05% (wt/vol) collagenase II (Invitrogen) at 37°C. After centrifugation and resuspension, cells were plated for 1.5 hours, and then the medium was replaced. Most of the adherent cells were cardiac fibroblasts, which were identified on the basis of vimentin-positive expression. Primary cardiac fibroblasts were passaged until cells reached ≈70% to 80% confluence on the plate and were then prepared for further incubation.

Coimmunoprecipitation

Cultured cardiac fibroblasts or HEK293T cells were resuspended in NP40 lysis buffer (50 mmol/L Tris, 150 mmol/L NaCl, 1% Nonidet P-40, 5 mmol/L EDTA, complete protease inhibitor cocktail, 10 mmol/L NaF, 1 mmol/L sodium orthovanadate, 25 mmol/L β-glycerophosphate). Lysates were incubated overnight at 4°C on a rotating wheel with IP-grade antibodies, including those for Stat3, Smad3 (Cell Signaling Technology), flag (Sigma-Aldrich), and c-myc (Abcam). Then, protein A/G beads (Santa Cruz) were added to the lysates, followed by 2-hour incubation. Beads were centrifuged, and protein complex was separated by SDS-PAGE and detected by Western blotting analysis.¹³

Immunohistochemistry

Paraffin-embedded human heart tissues and mouse heart tissues were sliced into 4-µm sections. Immunohistochemical staining was performed using the antibody for EphrinB2 and α-SMA (Abcam). Quantification of Ephrin B2 immunopositive area was determined by Image-Pro Plus 6.0 indicated by positive staining area/total area.¹⁴

Immunofluorescence Staining

Harvested hearts were sliced while frozen and costained with primary antibodies for EphrinB2 and α-SMA (Abcam) overnight at 4°C. Cy3-conjugated goat anti-rabbit antibody and FITC-conjugated goat anti-mouse antibody (Abcam) were applied as secondary antibodies. Nuclei were stained by DAPI (Sigma-Aldrich). Images were acquired using a fluorescence microscope (Leica, Germany).

Histological Analysis

Mouse hearts were fixed in 4% paraformaldehyde overnight at 4°C and dehydrated in a series of ethanol washes. Samples were subsequently cut into 4-µm-thick sections and stained with hematoxylin and eosin to analyze tissue morphology or with Sirius Red to analyze fibrosis and the collagen content under polarized light.

Wound Healing Assay

Fibroblasts were cultured in gelatin-coated 6-well plates containing the serum-deprived medium for 24 hours before scratching. After wounding, the medium was replaced with fresh serum-deprived medium. The percentage of wound closure was assessed as described in the literature.⁷

Cell Proliferation Assay

The cell proliferation assay was performed using a CCK-8 (cell counting kit-8) assay kit (Beyotime) according to the instructions. Fibroblasts were treated with serum-deprived medium for 24 hours, and 3×10³ cells were then seeded on gelatin-coated 96-well plates and incubated with CCK-8 solution for 24 hours. The absorbance was read using a spectrophotometer.

Statistical Analysis

Data with normal distribution were presented as mean±SEM. For data with small samples, a Mann–Whitney U test was used between 2 groups. Kolmogorov–Smirnov method was used to confirm the normal distribution for the other data. Student t test was applied for comparison between 2 groups, and 1-way ANOVA followed by Tukey comparison test was used for comparison between at least 3 groups. All data were analyzed using SPSS (version 17.0) statistical software. A P<0.05 was considered statistically significant.

Results

Increased EphrinB2 Expression in Failing Hearts of Patients With Severe Cardiac Fibrosis

To detect the expression of EphrinB2 in humans, normal hearts were obtained from healthy donors who experienced traffic accidents, and failing hearts were obtained from patients with advanced heart failure who underwent heart transplantation. Hematoxylin and eosin staining revealed myocardial derangement in failing heart compared with normal heart (Figure 1A). Immunohistochemistry results showed a significant increase in EphrinB2 expression and the expression of myofibroblast marker α-SMA in failing hearts (Figure 1A, 1C, and 1D). Similar results were confirmed by Western blotting, showing that EphrinB2 was increased in parallel with an elevated expression of α-SMA and the production of type I collagen in the failing hearts (Figure 1B and 1E through 1G).

**Figure 1.** **Increased EphrinB2 (erythropoietin-producing hepatoma interactor B2) expression in failing hearts of patients with severe cardiac fibrosis. A**, Representative images of hematoxylin and eosin (HE) staining and immunohistochemical staining of EphrinB2 and α-SMA (α-smooth muscle actin) in the anterior wall of the left ventricle from a normal heart and a heart from a patient with dilated cardiomyopathy (DCM); scale bar represents 100 μm. B, EphrinB2, α-SMA, and Collagen I expressions in the left anterior wall from normal hearts and hearts from patients with DCM were detected via Western blotting, with GAPDH used as a loading control. C and D, EphrinB2 or α-SMA staining area/total area was quantified via immunohistochemical staining, respectively. E–G, EphrinB2, α-SMA, and Collagen I protein levels (the ratio of protein pixel density/GAPDH pixel density) were quantified according to the immunoblots. *P<0.05.

Increased EphrinB2 Expression Is Associated With Severe Cardiac Fibrosis in Mice

To investigate the expression of EphrinB2 associated with cardiac fibrosis after injury, mouse models of MI and angiotensin II–induced cardiac hypertrophy were used to represent the 2 different cardiac fibrotic responses mentioned above. In mouse MI model, Sirius Red staining showed a remarkable increase in cardiac fibrosis correlated with the duration of ischemia (Figure 2A). Immunofluorescence staining showed a gradual increase in the expression of EphrinB2, which was colabeled with aggravated α-SMA–marked cardiac fibroblasts in the ischemic border zone of MI (Figure 2B). Furthermore, Western blot analysis indicated a progressive increase in the expression of EphrinB2 protein 28 days after MI in parallel with the activation of myofibroblasts and the deposition of collagen I (Figure 2C and 2D). Similar results were obtained from the mouse model of angiotensin II–induced cardiac hypertrophy. First, we confirmed the successful modeling via echocardiography (Figure 3A through 3C). Immunohistochemistry staining showed an increase in EphrinB2 expression in hypertrophic myocardium, accompanied by aggravated cardiac fibrosis (Figure 3D through 3F). Consistently, Western blot results showed an elevation in EphrinB2 expression in parallel with α-SMA and collagen I expression in the hypertrophic hearts of mice (Figure 3G). Taken together, these results demonstrated that increased EphrinB2 expression was associated with cardiac fibrosis in mice with cardiac injury.

**Figure 2.** **Parallel increases in EphrinB2 (erythropoietin-producing hepatoma interactor B2) with the progression of cardiac fibrosis in a mouse model of myocardial infarction (MI). A**, Representative Sirius Red staining images of transverse sections from a sham heart and from an infarcted heart at the 3, 7, and 28 d after MI; scale bar represents 1 mm. B, Immunofluorescence staining images labeling EphrinB2 (red) and α-SMA (α-smooth muscle actin; green) in a sham mouse and in the border zones of MI mice at 3, 7, and 28 d; scale bars represent 100 and 1 μm in the main images and expanded areas, respectively. C and D, EphrinB2, α-SMA, and Collagen I expressions in infarcted and border zones of mouse hearts were detected via Western blotting and quantified according to the immunoblots. **P<0.01 vs sham group.

**Figure 3.** **Elevated EphrinB2 (erythropoietin-producing hepatoma interactor B2) expression in a mouse model of angiotensin II (Ang II)–induced cardiac hypertrophy. A**, Representative M-mode images of normal and Ang II–treated hearts. B and C, Ejection fraction (EF) and thickness of the interventricular septum during systole (IVS-s) were quantified via echocardiography. D, Gross morphology of a normal heart and an Ang II–treated heart was shown; scale bar represents 1 mm; representative images of HE staining, Sirius Red staining, and immunohistochemical staining of EphrinB2 in normal and Ang II–treated mice; scale bar represents 100 μm. E, Fibrotic area/total area in Sirius Red staining slides from normal mice and Ang II–treated mice were quantified. F, EphrinB2 staining area/total area was quantified by immunohistochemical staining. G, EphrinB2, α-SMA (α-smooth muscle actin), and Collagen I expression levels in the left ventricle of normal and Ang II–treated mice were quantified via Western blotting. Sham group: n=5, Ang II–treated group: n=8; *P<0.05, **P<0.01.

Knockdown of EphrinB2 Ameliorates Cardiac Fibrosis and Improves Cardiac Function In Vivo

To elucidate the important role of EphrinB2 in cardiac fibrosis after injury, we generated lentiviruses carrying EphrinB2-shRNA and the control scramble plasmids. After mice were subjected to left anterior descending artery ligation, lentiviruses were injected into myocardium at multiple locations surrounding the infarct border zone. Sham mice underwent identical procedures but without ligation and received the same viral dose in the heart (Figure 4A). To evaluate whether cardiac function was affected by EphrinB2 knockdown, echocardiography was performed. Significant improvement in ejection fraction was observed in MI mice injected with the EphrinB2-shRNA compared with those injected with control scramble viruses (Figure 4B and 4C). To assess the extent of fibrosis after EphrinB2 knockdown, a polarizing microscope was used to distinguish the collagen types in the fibrotic scar of the infarcted zone. Although the area of fibrotic scarring did not differ between the Scramble MI group and the EphrinB2-shRNA MI group, the EphrinB2 knockdown showed an attenuated deposition of collagen I, which was visible as compact fibers with strong refraction under polarized light (Figure 4D and 4E). In accordance with the results above, the concomitant expressions of EphrinB2 and α-SMA were greatly increased in the MI mice injected with the scramble viruses (Scramble MI group) compared with the expression levels in the sham groups (Figure 4F). However, MI mice with EphrinB2 knockdown (EphrinB2-shRNA MI group) exhibited a dramatic reduction in α-SMA expression compared with that in the Scramble MI group, suggesting a pivotal role of EphrinB2 in cardiac myofibroblast activation in vivo (Figure 4F). Taken together, these results demonstrated that knockdown of EphrinB2 ameliorated cardiac fibrosis and improved cardiac function after MI.

**Figure 4.** **Knockdown of EphrinB2 (erythropoietin-producing hepatoma interactor B2) ameliorates cardiac fibrosis and improves cardiac function in vivo. A**, Schema of the procedure for injection of lentiviruses carrying EphrinB2-shRNA. B and C, Representative M-mode images of Scramble Sham, EphrinB2-shRNA Sham, Scramble myocardial infarction (MI), and EphrinB2-shRNA MI mice; ejection fraction (EF) was quantified via echocardiography. D, Representative Sirius Red staining images photographed under white light (scale bar represents 1 mm) and under polarized light (scale bar represents 100 μm) were shown, representing the extent of cardiac fibrosis and the variety of collagen components in the infarcted and border zones in Scramble MI and EphrinB2-shRNA MI mice. E, Infarcted area size/total area size was quantified and compared between Scramble MI and EphrinB2-shRNA MI mice. F, EphrinB2 and α-SMA (α-smooth muscle actin) expression levels were quantified by Western blotting. n=5 in each sham group, n=6 in each MI group; *P<0.05. LAD indicates left anterior descending.

EphrinB2 Regulates Cardiac Myofibroblast Activation In Vitro

Cardiac fibroblasts were cultured from neonatal mice and were infected with EphrinB2-overexpressing lentiviruses or EphrinB2-shRNA lentiviruses to further explore the effects of EphrinB2 in cardiac fibrosis and the underlying mechanisms. EphrinB2 overexpression promoted the conversion of fibroblasts toward myofibroblasts, indicated by increased α-SMA and collagen I expression (Figure 5A), and enhanced migration and proliferation capabilities (Figure 5B and 5C). Conversely, EphrinB2 knockdown inhibited myofibroblast activation (Figure 5A, 5B, and 5D).

**Figure 5.** **EphrinB2 (erythropoietin-producing hepatoma interactor B2) alone induces myofibroblast activation in vitro. A**, EphrinB2, α-SMA (α-smooth muscle actin), and Collagen I levels in cultured cardiac fibroblasts infected with EphrinB2-overexpressing lentiviruses or EphrinB2-shRNA lentiviruses were detected via Western blotting. B, Representative images of wound healing in the vector group, EphrinB2-overexpressing group, scramble group, and EphrinB2-shRNA group; scale bar represents 20 μm. C and D, Proliferation abilities of cells in the EphrinB2-overexpressing group and EphrinB2-shRNA group were analyzed via CCK-8 assay. Each experiment was repeated 3×, **P<0.01.

Next, we investigated the expression of EphrinB2 in cardiac fibroblasts at different time points after hypoxia exposure. EphrinB2 expression increased significantly after 24 hours of hypoxia and remained at a high level for 48 hours, which was consistent with the expression pattern of α-SMA expression (Figure 6A and 6B). To further define whether EphrinB2 mediated hypoxia-induced activation of myofibroblasts, cardiac fibroblasts were maintained under hypoxic conditions for 24 hours after infection with EphrinB2-shRNA–carrying lentiviruses. Similar to the in vivo observation, hypoxia induced a remarkable elevation in both EphrinB2 and α-SMA expression compared with that in the control groups, whereas EphrinB2 knockdown markedly inhibited α-SMA expression, suggesting that EphrinB2 knockdown protected against hypoxia-induced myofibroblast activation (Figure 6C and 6D). These results suggested that EphrinB2 regulated cardiac myofibroblast activation in normoxic and hypoxic conditions in vitro.

**Figure 6.** **EphrinB2 (erythropoietin-producing hepatoma interactor B2) regulates myofibroblast activation under hypoxic conditions in vitro. A** and B, EphrinB2 and α-SMA (α-smooth muscle actin) protein levels in cardiac fibroblasts under hypoxia conditions for 24 and 48 h were assessed via Western blotting. C and D, EphrinB2 and α-SMA expression levels in cardiac fibroblasts under hypoxic conditions after EphrinB2-shRNA lentiviruses infection were detected via Western blotting. Each experiment was repeated 3×, **P<0.01.

EphrinB2 Activates Stat3 and TGF-β/Smad3 Signaling to Regulate Fibrotic Response

EphrinB2 protein facilitates the phosphorylation of Stat3 by its upstream signal, Jak2, resulting in the nuclear translocation of Stat3, which is a critical nuclear signaling pathway for EphrinB family proteins.^15,16 Therefore, we examined Jak2/Stat3 signaling in an in vivo mouse MI model. We found that the phosphorylation levels of Jak2 and Stat3 were significantly increased in the Scramble MI group, whereas were significantly decreased in the EphrinB2-shRNA MI group (Figure 7A and 7B). In addition, TGF-β/Smad3 signaling has been shown to be a major pathway contributing to both reactive and replacement cardiac fibrosis.^17,18 As expected, our data confirmed that the mice in the Scramble MI group showed a greater induced expression of active TGF-β, which was cleaved from full-length and inactive TGF-β, compared with the sham groups. However, the production of active TGF-β in the EphrinB2-shRNA MI group was significantly lower than that in the Scramble MI group. Consistently, the phosphorylated Smad3 levels were highly consistent with those of active TGF-β, suggesting the activation of TGF-β/Smad3 in EphrinB2-mediated cardiac fibrosis (Figure 7A and 7B). In vitro study further demonstrated that EphrinB2 activates phosphorylation of Stat3 and Smad3 in cultured cardiac fibroblasts (Figure 7C and 7D). Furthermore, the cultured cardiac fibroblasts infected with vector or EphrinB2 overexpression lentiviruses and specific antagonists against Stat3 (Stattic, Selleck) and Smad3 (SIS3, Selleck) were used. Our data showed that inhibition of the Stat3 using Stattic (2.5 μmol/L)¹⁹ or the Smad3 using SIS3 (1 μmol/L)²⁰ significantly reversed EphrinB2-mediated α-SMA expression, respectively. Inhibition of both Stat3 and Smad3 activation showed a further decrease in the expression level of α-SMA in EphrinB2-overexpressed cardiac fibroblasts than inhibition of Stat3 or Smad3 alone, indicating a synergistic effect of Stat3 and TGF-β/Smad3 signaling in cardiac fibrogenesis upon EphrinB2 activation (Figure 7E). In summary, these results indicated that EphrinB2 activated both Stat3 and TGF-β/Smad3 signaling to regulate the fibrotic response.

**Figure 7.** EphrinB2 (erythropoietin-producing hepatoma interactor B2) activates Jak2 (Janus kinase 2)/Stat3 (signal transducer and activator of transcription 3) and TGF-β (transforming growth factor-β)/Smad3 (mothers against decapentaplegic homolog 3) signaling to regulate the fibrotic response. A and B, Activation of Jak2/Stat3 and TGF-β/Smad3 signaling in vivo in a mouse model of myocardial infarction (MI) after lentivirus injection was detected and quantified via Western blotting. C and D, Activation of Stat3 and Smad3 in cultured cardiac fibroblasts overexpressing EphrinB2 or downregulating EphrinB2 was detected and quantified via Western blotting. E, Cultured cardiac fibroblasts infected with vector lentiviruses or EphrinB2-overexpressing lentiviruses were stimulated with specific antagonist against Stat3 (Stattic, 2.5 μmol/L) and Smad3 (SIS3, 1 μmol/L), respectively. The expression level of α-SMA was quantified via Western blotting. In vitro experiments repeat 3×, *P<0.05, **P<0.01.

EphrinB2 Regulates the Interaction of Stat3 and Smad3

Several studies have reported the association between Stat3 and Smad3 in mediating tumor fibrotic processes.^11,21 But the interplay between Stat3 and Smad3 in heart was not studied. Coimmunoprecipitation results showed a physical interaction of Stat3 and Smad3 under normoxic condition, which was enhanced under hypoxic condition in cultured cardiac fibroblasts (Figure 8A). Next, fibroblasts were used to overexpress EphrinB2, and the results showed that upregulated EphrinB2 enhanced the interaction of Stat3 and Smad3 (Online Figure I). To clarify the specific molecular mechanism of Stat3/Smad3 interaction, tyrosine at 705 was mutated to glutamate to constitutively activate Stat3 (Y705E), which enhanced the Stat3/Smad3 interaction (Figure 8B), suggesting that phosphorylation of Stat3 promotes the interaction of Stat3/Smad3. Furthermore, to determine the structural features underlying Stat3/Smad3 interaction, we mapped the regions in Smad3 and in Stat3 that mediated their interaction. Smad3 contains MAD homology 1 and MAD homology 2 domains (Figure 8C). The interaction of Stat3 with mutant Smad3 where MAD homology 1 or MAD homology 2 domain of Smad3 was deleted was assessed via co-IP assay. Stat3 bound to the MAD homology 2 domain (176-425aa) of Smad3 (Figure 8D). Stat3 contains several protein–protein interaction domains including a coil–coil domain, a DNA-binding domain, and an Src homology 2 domain (Figure 8E). Co-IP results elucidated that Smad3 bound to the coil–coil domain and DNA-binding domain (141-290aa) of Stat3 (Figure 8F). Therefore, EphrinB2 regulates cardiac fibrosis through the interaction of Stat3 and Smad3.

**Figure 8.** EphrinB2 (erythropoietin-producing hepatoma interactor B2) regulates the interaction of Stat3 (signal transducer and activator of transcription 3) and Smad3 (mothers against decapentaplegic homolog 3). A, The interaction between Stat3 and Smad3 in cultured cardiac fibroblasts under normoxic or hypoxic condition was detected via coimmunoprecipitation analysis. B, Constitutive activation of Stat3 (Y705E) promoted the interaction between Stat3 and Smad3 in HEK293T cells. C, Schematic diagram of Smad3 and its deletion mutants. D, Stat3 binds with the MAD homology 2 (MH2) domain (176-425aa) of Smad3. E, Schematic diagram of Stat3 and its deletion mutants. F, Smad3 binds with the coil–coil and DNA-binding domain (141-290aa) of Stat3. G, Schematic figure demonstrating that EphrinB2 regulates cardiac fibrosis through the interaction of Stat3 and TGF-β/Smad3 signaling. FL indicates full length; and WCL, whole cell lysates.

Discussion

In the present study, we have uncovered a previously unrecognized role of EphrinB2 in cardiac fibrosis. We observed that EphrinB2 was increased during cardiac remodeling process which was associated with a progressive cardiac fibrosis process. The role of upregulated EphrinB2 expression was then investigated by EphrinB2-shRNA approach, and we showed that knockdown of EphrinB2 by intramyocardial injection of lentiviruses containing EphrinB2-shRNA can ameliorate cardiac fibrosis and improve cardiac function. Mechanistically, EphrinB2 activated Stat3 and TGF-β/Smad3 signaling, both of which synergistically promoted the fibrotic process in heart. Furthermore, EphrinB2 regulated the interaction of Stat3 and Smad3, which was enhanced by Stat3 phosphorylation.

Although EphrinB2 is pivotal in angiogenesis and development, the role of EphrinB2 in fibrosis is not fully understood. Kida et al⁷ have reported an inhibitory role of EphrinB2 in renal fibrosis after kidney injury, whereas in skin and retinal tissues EphrinB2 was found to promote fibrogenesis.^22,23 These studies suggested different effects of EphrinB2 on fibrosis with organ specificity. Cardiac fibrosis plays a vital role in ventricular remodeling and heart failure. Thus far, the role of EphrinB2 in cardiac fibrosis remains unclear. Our data supported EphrinB2 as a critical profibrotic regulator in cardiac fibrosis, which differed from the finding that EphrinB2 plays an antifibrotic role in renal fibrosis. Furthermore, we clearly illustrated that EphrinB2 promoted cardiac fibrosis via a new mechanism mediated the activation and interaction of Stat3 and TGF-β/Smad3 signaling.

TGF-β/Smad3 signaling is a primary pathway in fibrogenesis, majorly contributing to both reactive fibrosis in transverse aortic constriction-induced cardiac hypertrophy and replacement fibrosis in MI in mice.^17,18 Smad3 has been identified to bind to the promoter of α-SMA in mouse fibroblasts by the approach of using the chromatin setting.²⁴ We found that EphrinB2 promoted the activation of intrinsic TGF-β and subsequent phosphorylation of Smad3, which was associated with cardiac fibrosis characterized by myofibroblast activation. Inhibition of Smad3 activation significantly reversed EphrinB2-induced α-SMA expression level, suggesting that EphrinB2 regulates cardiac fibrotic process by activating TGF-β/Smad3 signaling.

Stat3 is a critical transcription factor transducing the signal of EphrinB family proteins to the nuclei.¹⁶ The role of Stat3 in organic fibrogenesis is complicated. Aggravating cardiac fibrosis was observed in aged mice with cardiomyocyte-specific knockout of Stat3.²⁵ Conversely, selective inhibition of Stat3 phosphorylation via the antagonist S3I-201 attenuated left atrial fibrosis in an in vivo mouse MI model and suppressed collagen synthesis of cardiac fibroblasts in vitro.²⁶ In a rat model of cardiac hypertrophy generated by renal artery ligation, inhibition of Stat3 activation via specific antagonist resulted in significant decrease in collagen synthesis and regression of cardiac hypertrophy.²⁷ These studies implied that exact effects of Stat3 on cardiac fibrosis need to be further clarified. We observed that EphrinB2 activated the phosphorylation of Stat3 in an in vivo mouse MI model and in cultured cardiac fibroblasts in parallel with the expression level of α-SMA, suggesting a potential involvement of Stat3 in EphrinB2-mediated cardiac fibrosis. We further confirmed that EphrinB2-activated Stat3 promoted myofibroblast activation using the specific antagonist against the phosphorylation of Stat3. These data supported the notion that EphrinB2-activated Stat3 signaling regulates cardiac fibrosis.

In addition, our data showed that inhibition of both Stat3 and Smad3 activation synergistically reduced α-SMA expression, which suggested a potential interplay between Stat3 and Smad3 in heart. Existing studies have reported the association between Stat3 and Smad3 in liver fibrosis and tumor fibrogenesis. Yamamoto et al²⁸ proved that in hepatoma cells the activity of Stat3 was augmented by TGF-β/Smad3 signaling through the interaction between Stat3 and Smad3 bridged by p300. Tang et al²¹ reported a cooperative role of Stat3 with TGF-β/Smad3 in liver fibrosis. In contrast, Wang et al¹¹ illustrated a physical direct interaction of Stat3 and Smad3, which inhibited TGF-β–induced fibrogenesis in tumor. However, the interaction of Stat3 and Smad3 in cardiac fibrosis is not clarified. Using cultured cardiac fibroblasts, we discovered that Stat3 physiologically interacted with Smad3. Both EphrinB2- and hypoxia-activated Stat3 enhanced the interaction of Stat3 and Smad3. Constitutive activation of Stat3 (tyrosine to glutamate mutation at 705) further promoted the interaction of Stat3/Smad3 complex. These data confirmed that Stat3 activation is critical in enhancing Stat3 and Smad3 interaction. Furthermore, we identified that the MAD homology 2 domain of Smad3 and the coil–coil and DNA-binding domain of Stat3 were attributed as the binding sites of Stat3/Smad3 complex, which advanced our knowledge on the structural feature of Stat3/Smad3 interplaying complex.

In summary, we have discovered a profibrotic role of EphrinB2 in heart via the interaction of Stat3 and TGF-β/Smad3 signaling. Our work has provided a potential promising therapeutic target for treating fibrotic diseases and heart failure.

Nonstandard Abbreviations and Acronyms

α-SMA	α-smooth muscle actin
EphrinB2	erythropoietin-producing hepatoma interactor B2
Jak2	Janus kinase 2
MI	myocardial infarction
Smad3	mothers against decapentaplegic homolog 3
Stat3	signal transducer and activator of transcription 3
TGF-β	transforming growth factor-β

Sources of Funding

This work was supported by the National Natural Science Foundation of China (81270179 and 81470384 to M. Xiang).

Disclosures

None.

Footnotes

In June 2017, the average time from submission to first decision for all original research papers submitted to Circulation Research was 12.45 days.

*These authors contributed equally to this article.

The online-only Data Supplement is available with this article at http://circres.ahajournals.org/lookup/suppl/doi:10.1161/CIRCRESAHA.117.311045/-/DC1.

Correspondence to Meixiang Xiang, MD, PhD, Department of Cardiology, The Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang 310009, China. E-mail xiangmxhz@163.com

References

1. Rai V, Sharma P, Agrawal S, Agrawal DK. Relevance of mouse models of cardiac fibrosis and hypertrophy in cardiac research.Mol Cell Biochem. 2017; 424:123–145. doi: 10.1007/s11010-016-2849-0.Crossref Medline Google Scholar
2. Kis K, Liu X, Hagood JS. Myofibroblast differentiation and survival in fibrotic disease.Expert Rev Mol Med. 2011; 13:e27. doi: 10.1017/S1462399411001967.Crossref Medline Google Scholar
3. Kuijper S, Turner CJ, Adams RH. Regulation of angiogenesis by Eph-ephrin interactions.Trends Cardiovasc Med. 2007; 17:145–151. doi: 10.1016/j.tcm.2007.03.003.Crossref Medline Google Scholar
4. Adams RH, Diella F, Hennig S, Helmbacher F, Deutsch U, Klein R. The cytoplasmic domain of the ligand ephrinB2 is required for vascular morphogenesis but not cranial neural crest migration.Cell. 2001; 104:57–69.Crossref Medline Google Scholar
5. Gerety SS, Wang HU, Chen ZF, Anderson DJ. Symmetrical mutant phenotypes of the receptor EphB4 and its specific transmembrane ligand ephrin-B2 in cardiovascular development.Mol Cell. 1999; 4:403–414.Crossref Medline Google Scholar
6. Foo SS, Turner CJ, Adams S, Compagni A, Aubyn D, Kogata N, Lindblom P, Shani M, Zicha D, Adams RH. Ephrin-B2 controls cell motility and adhesion during blood-vessel-wall assembly.Cell. 2006; 124:161–173. doi: 10.1016/j.cell.2005.10.034.Crossref Medline Google Scholar
7. Kida Y, Ieronimakis N, Schrimpf C, Reyes M, Duffield JS. EphrinB2 reverse signaling protects against capillary rarefaction and fibrosis after kidney injury.J Am Soc Nephrol. 2013; 24:559–572. doi: 10.1681/ASN.2012080871.Crossref Medline Google Scholar
8. Lagares D, Ghassemi-Kakroodi P, Tremblay C, Baron M, Blati M, Zhan Y, Pelletier J, Martel-Pelletier J, Tager AM, Kapoor M. Fibroblast-derived ephrin-b2 regulates lung fibrosis in vivo by promoting fibroblast recruitment and myofibroblast differentiation.Am J Resp Crit Care. 2014; 189:A1242.Google Scholar
9. Bassat E, Mutlak YE, Genzelinakh A, et al.. The extracellular matrix protein agrin promotes heart regeneration in mice.Nature. 2017; 547:179–184. doi: 10.1038/nature22978.Crossref Medline Google Scholar
10. Seth M, Zhang ZS, Mao L, Graham V, Burch J, Stiber J, Tsiokas L, Winn M, Abramowitz J, Rockman HA, Birnbaumer L, Rosenberg P. TRPC1 channels are critical for hypertrophic signaling in the heart.Circ Res. 2009; 105:1023–1030. doi: 10.1161/CIRCRESAHA.109.206581.Link Google Scholar
11. Wang G, Yu Y, Sun C, Liu T, Liang T, Zhan L, Lin X, Feng XH. STAT3 selectively interacts with Smad3 to antagonize TGF-β.Oncogene. 2016; 35:4388–4398. doi: 10.1038/onc.2015.446.Crossref Medline Google Scholar
12. Su SA, Yang D, Zhu W, Cai Z, Zhang N, Zhao L, Wang JA, Xiang M. Interleukin-17A mediates cardiomyocyte apoptosis through Stat3-iNOS pathway.Biochim Biophys Acta. 2016; 1863:2784–2794. doi: 10.1016/j.bbamcr.2016.08.013.Crossref Medline Google Scholar
13. Zhang N, Ye F, Zhu W, Hu D, Xiao C, Nan J, Su S, Wang Y, Liu M, Gao K, Hu X, Chen J, Yu H, Xie X, Wang J. Cardiac ankyrin repeat protein attenuates cardiomyocyte apoptosis by upregulation of Bcl-2 expression.Biochim Biophys Acta. 2016; 1863:3040–3049. doi: 10.1016/j.bbamcr.2016.09.024.Crossref Medline Google Scholar
14. Cai Z, Shen L, Ma H, Yang J, Yang D, Chen H, Wei J, Lu Q, Wang DW, Xiang M, Wang J. Involvement of endoplasmic reticulum stress-mediated C/EBP homologous protein activation in coxsackievirus B3-induced acute viral myocarditis.Circ Heart Fail. 2015; 8:809–818. doi: 10.1161/CIRCHEARTFAILURE.114.001244.Link Google Scholar
15. Salvucci O, Ohnuki H, Maric D, Hou X, Li X, Yoon SO, Segarra M, Eberhart CG, Acker-Palmer A, Tosato G. EphrinB2 controls vessel pruning through STAT1-JNK3 signalling.Nat Commun. 2015; 6:6576. doi: 10.1038/ncomms7576.Crossref Medline Google Scholar
16. Bong YS, Lee HS, Carim-Todd L, Mood K, Nishanian TG, Tessarollo L, Daar IO. ephrinB1 signals from the cell surface to the nucleus by recruitment of STAT3.Proc Natl Acad Sci USA. 2007; 104:17305–17310. doi: 10.1073/pnas.0702337104.Crossref Medline Google Scholar
17. Dobaczewski M, Chen W, Frangogiannis NG. Transforming growth factor (TGF)-β signaling in cardiac remodeling.J Mol Cell Cardiol. 2011; 51:600–606. doi: 10.1016/j.yjmcc.2010.10.033.Crossref Medline Google Scholar
18. Zeisberg EM, Tarnavski O, Zeisberg M, Dorfman AL, McMullen JR, Gustafsson E, Chandraker A, Yuan X, Pu WT, Roberts AB, Neilson EG, Sayegh MH, Izumo S, Kalluri R. Endothelial-to-mesenchymal transition contributes to cardiac fibrosis.Nat Med. 2007; 13:952–961. doi: 10.1038/nm1613.Crossref Medline Google Scholar
19. Schust J, Sperl B, Hollis A, Mayer TU, Berg T. Stattic: a small-molecule inhibitor of STAT3 activation and dimerization.Chem Biol. 2006; 13:1235–1242. doi: 10.1016/j.chembiol.2006.09.018.Crossref Medline Google Scholar
20. Jinnin M, Ihn H, Tamaki K. Characterization of SIS3, a novel specific inhibitor of Smad3, and its effect on transforming growth factor-beta1-induced extracellular matrix expression.Mol Pharmacol. 2006; 69:597–607. doi: 10.1124/mol.105.017483.Crossref Medline Google Scholar
21. Tang LY, Heller M, Meng Z, Yu LR, Tang Y, Zhou M, Zhang YE. Transforming growth factor-β (TGF-β) directly activates the JAK1-STAT3 axis to induce hepatic fibrosis in coordination with the SMAD pathway.J Biol Chem. 2017; 292:4302–4312. doi: 10.1074/jbc.M116.773085.Crossref Medline Google Scholar
22. Umeda N, Ozaki H, Hayashi H, Oshima K. Expression of ephrinB2 and its receptors on fibroproliferative membranes in ocular angiogenic diseases.Am J Ophthalmol. 2004; 138:270–279. doi: 10.1016/j.ajo.2004.04.006.Crossref Medline Google Scholar
23. Avouac J, Clemessy M, Distler JH, Gasc JM, Ruiz B, Vacher-Lavenu MC, Wipff J, Kahan A, Boileau C, Corvol P, Allanore Y. Enhanced expression of ephrins and thrombospondins in the dermis of patients with early diffuse systemic sclerosis: potential contribution to perturbed angiogenesis and fibrosis.Rheumatology. 2011; 50:1494–1504. doi: 10.1093/rheumatology/keq448.Crossref Medline Google Scholar
24. Li Z, Xie WB, Escano CS, Asico LD, Xie Q, Jose PA, Chen SY. Response gene to complement 32 is essential for fibroblast activation in renal fibrosis.J Biol Chem. 2011; 286:41323–41330. doi: 10.1074/jbc.M111.259184.Crossref Medline Google Scholar
25. Jacoby JJ, Kalinowski A, Liu MG, Zhang SS, Gao Q, Chai GX, Ji L, Iwamoto Y, Li E, Schneider M, Russell KS, Fu XY. Cardiomyocyte-restricted knockout of STAT3 results in higher sensitivity to inflammation, cardiac fibrosis, and heart failure with advanced age.Proc Natl Acad Sci USA. 2003; 100:12929–12934. doi: 10.1073/pnas.2134694100.Crossref Medline Google Scholar
26. Chen Y, Surinkaew S, Naud P, Qi XY, Gillis MA, Shi YF, Tardif JC, Dobrev D, Nattel S. JAK-STAT signalling and the atrial fibrillation promoting fibrotic substrate.Cardiovasc Res. 2017; 113:310–320.Crossref Medline Google Scholar
27. Mir SA, Chatterjee A, Mitra A, Pathak K, Mahata SK, Sarkar S. Inhibition of signal transducer and activator of transcription 3 (STAT3) attenuates interleukin-6 (IL-6)-induced collagen synthesis and resultant hypertrophy in rat heart.J Biol Chem. 2012; 287:2666–2677. doi: 10.1074/jbc.M111.246173.Crossref Medline Google Scholar
28. Yamamoto T, Matsuda T, Muraguchi A, Miyazono K, Kawabata M. Cross-talk between IL-6 and TGF-beta signaling in hepatoma cells.FEBS Lett. 2001; 492:247–253.Crossref Medline Google Scholar

EphrinB2 Regulates Cardiac Fibrosis Through Modulating the Interaction of Stat3 and TGF