Introduction
Pulmonary arterial hypertension is a fatal disease with a 3-year mortality rate of 20–40 % for which no cure is available [
12,
18,
43]. During disease progression, the RV undergoes compensatory hypertrophy to maintain physiological blood pressure and flow. As PAH progresses the RV becomes fibrotic and dilates and ultimately undergoes functional failure [
8,
38]. In recent years, the outlook for PAH patients has improved due to earlier recognition and new therapies. However, despite these interventions vascular pulmonary resistance was high and increased over time leading eventually still to RV failure [
8,
41]. Thus, it is important to elucidate mechanisms driving RV remodeling and transition to RV dysfunction to identify new targets for therapeutic intervention [
47].
Previously, the receptor Fn14 and its ligand TWEAK have been associated with LV remodeling after myocardial infarction (MI) [
6,
33]. Systemic overexpression of TWEAK induced via Fn14 progressive dilated cardiomyopathy and heart failure affecting both the LV and RV [
20]. This phenotype was associated with cardiomyocyte elongation and cardiac fibrosis. Fn14 is also expressed in human cardiomyocytes and circulating TWEAK levels were correlated with idiopathic dilated cardiomyopathy [
20]. In contrast, TWEAK levels were inversely correlated with the severity of PAH in patients suggesting that the TWEAK/Fn14 axis might play no role in the RV failure [
11]. However, correlations of cytokine blood levels to heart disease can be misleading. For example, TNF overexpression leads to heart failure and its endogenous expression is positively correlated with heart failure [
23]. However, clinical trials with anti-TNF therapies were disappointing [
27]. Moreover, as with PAH, TWEAK blood plasma levels are decreased in patients and animal models with chronic kidney disease (CKD). In contrast, Fn14 is upregulated in the kidney in animal models as well as patients with CKD [
19,
52]. Deletion of Fn14 in the animal models protects against kidney fibrosis and failure [
19]. Thus, it is important to study the endogenous role of Fn14 in heart disease.
TWEAK plays an important role in several biological processes including inflammation, angiogenesis, cell growth, cell death and fibrosis. Depending on cell type and context TWEAK mediates these activities via Fn14 by activating a variety of downstream signaling cascades. Our previous data have suggested that TWEAK-induced cardiomyocyte proliferation is mediated through activation of ERK and PI3K as well as inhibition of GSK-3beta [
35]. Recently, NFκB has been identified as a major downstream target of the TWEAK/Fn14 axis through functional validation in numerous cellular and invivo contexts [
31]. With regard to signaling in heart cell types, the TWEAK/Fn14 axis was shown to induce NFκB signaling in cardiomyocytes [
6] as well as fibroblasts [
5].
Here we examined Fn14 expression in models of pressure overload (3 weeks after PAB in mice [
2,
3] or monocrotaline (MCT) treatment of rats [
7,
26]) and found Fn14 markedly upregulated in the RV. Fn14
−/− mice exhibited reduced fibrosis and improved RV function following PAB. Our data show that Fn14 activation regulates fibroblast proliferation, differentiation and collagen expression. Finally, our data suggest a novel TWEAK/Fn14/RhoA/MAL pathway downstream of ET-1 signaling in cardiac fibroblasts (CFs). Collectively, our data demonstrate that Fn14 is an important endogenous mediator of RV remodeling and failure.
Methods
An expanded version of methods is available in Supplementary material online.
Animal studies
The investigation conforms with the Guide for the Care and Use of Laboratory Animals published by the Directive 2010/63/EU of the European Parliament. In vivo procedures were approved by a local Animal Ethics Committee in accordance to governmental and international guidelines on animal experimentation. Fn14
−/− mice were previously generated at Biogen (Biogen Idec, Inc, Cambridge) [
21]. PAB and/or SHAM operation of mice (20–23 g) was performed under isoflurane anesthesia (1.5 % v/v) and 0.03 mg/kg buprenorphine hydrochloride (s.c.). Analgetic therapy post operation was achieved by buprenorphine (0.03 mg/kg, 48 h) and carprofen (s.c., 4 mg/kg, 3–7 days). Pulmonary hypertension in Sprague–Dawley rats (300–350 g) was induced with 60 mg/kg MCT (s.c.) [
42]. Animals were daily controlled for signs of pain. Functional analyses were performed by MRI and hemodynamic measurements with a Millar microtip catheter under inhalation of isoflurane (1.5–2.0 % v/v). For organ/tissue sampling, animals were anesthetized (120 mg/kg ketamine + 16 mg/kg xylazine) and subsequently euthanized through exsanguination. Blood was intracardially collected and analyzed with a RayBio_Mouse TWEAK ELISA Kit.
Cell culture experiments
Cardiac fibroblasts were isolated with Liberase TH enzyme mix (Roche) from Fn14−/− mice and wild-type littermates. All experiments were performed with primary CFs after one or two passages. As an immortalized fibroblast cell line the Rat2 fibroblast cell line was used; a normal, non-tumorigenic and highly transfectable cell line derived from the Fischer rat fibroblast 3T3 like cell line Rat1. HEK293T cells were utilized for luciferase reporter assays. NIH3T3 cells were used as control cells for MAL nuclear translocation assays. Cells were cultured under standard conditions or as indicated.
Construction of pEGFP-TWEAK
TWEAK cDNA was amplified from HEK293T cells, ligated into pGEM-T-easy and a HindIII/SalI TWEAK fragment was subcloned into pEGFP-N1.
Transfection and luciferase promoter assays
Transient transfections were performed with Fugene 6 (Roche) with a total of 50 ng of constructs. Luciferase activity was measured by LightSwitch Luciferase Assay Reagent after 24 h of stimulation or 30–48 h after overexpression.
Western blot
Tissues were lysed and homogenized in RIPA buffer. Protein extracts were resolved on SDS gels, blotted on nitrocellulose membranes and incubated with the following primary antibodies: rabbit anti-TWEAK/Fn14 receptor, rabbit anti-pan-actin, monoclonal rabbit anti-RhoA (1:1,000) (all cell signaling), goat anti-DDR2 (Santa Cruz), rabbit anti-Collagen Type 1 (1:500) (Rockland Immunochemicals), mouse anti-GAPDH (1:2,000) (Sigma) and polyclonal rabbit anti-MAL (1:200) (from G. Posern). Antigen–antibody complexes were visualized using horseradish peroxidase-conjugated secondary antibodies. Western blots were quantified by Image J (NIH) software.
Cardiac fibroblasts from wild-type mice were split in serum-free medium and 24 h later transfected with 40 pM of siRNA perfectly matching the sequence 5′-ATGGAGCTGGTGGAGAAGAA-3′ of both murine MAL and MRTF-B and 40 pM of scrambled siRNA (Qiagen) with lipofectamine [
30]. 24 h later cells were stimulated with 100 ng/ml of TWEAK.
Determination of activated RhoA
Activated RhoA was determined by immunoprecipitation with the fusion protein GST-Rho-binding domain of Rhotekin (GST-RBD) [
39].
RNA extraction and Real-Time PCR
RNA was isolated with an RNeasy Fibrous Tissue Kit. RT reaction was performed using oligo (dT) primer. For Real-Time PCR analysis, cDNA was amplified with IQTM SYBR® Green SuperMix (Biorad) and Bio-Rad iCYCLER iQ5. Real-Time PCR was performed in triplicates and relative gene expression was calculated on the basis of ΔCt values to gapdh.
Immunohistochemistry
Mouse hearts were isolated, dissected in RV and LV + S, washed in PBS, fixated overnight in 10 % PFA, embedded in paraffin and sectioned longitudinally (5 μm). Sections were deparaffinized in xylene and rehydrated in ethanol. Heat-mediated antigen retrieval was performed in 0.05 M EDTA buffer (pH 8.0). Tissues for cryosections (5 μm) were frozen in OCT, fixated in acetone and stained as indicated. To quantify fibrosis RV sections were stained with 0.1 % Sirius Red F3B in picric acid and analyzed using a QWin V3 computer-assisted image analysis software (Leica) [
15].
Immunofluorescence
Heart sections were blocked in 5 % goat serum/0.2 % Tween-20/PBS for 1 h at RT. Cells were fixated in 3.7 % paraformaldehyde and permeabilized in PBS/0.5 % Triton. Samples were stained as indicated. F-actin was detected by rhodamine–phalloidin and cell membranes by WGA staining (Molecular Probes, Invitrogen). Cell size was determined using the ImageJ (NIH) software.
Endothelin stimulation
After serum starvation cells were stimulated for 24 h with 100 nM ET-1 (R and D systems).
Collagen assay
Serum-starved Rat2 fibroblasts were stimulated with TWEAK (100 ng/ml) for 48 h. l-ascorbic acid (0.25 mM) was added to the medium daily. Cells were lysed in RIPA buffer and total collagen (Types 1–5) was assessed using a Sircol soluble collagen assay kit (Biocolor Ltd).
MAL translocation
MAL translocation was determined with the Count software (by B. Waclaw). Nuclei number was assessed based on DAPI staining. MAL translocation was considered positive if the pixel number in the purple channel (MAL/DAPI overlap) was greater or equal compared to the threshold defined by control experiments.
Proliferation assay
Proliferation was determined with a Countess™ cell counter (Invitrogen) or CellTiter 96 Aqueous Cell Proliferation Assay (Promega).
Statistical analyses
Data were analyzed with GraphPad Prism. Data are presented as mean ± standard error of the mean (SEM). Statistical significance was determined using Student’s t test or for multiple comparisons One-way ANOVA followed by Bonferroni’s post hoc test. Values of p < 0.05 were considered statistically significant.
Discussion
The development of RV failure involves complex pathological mechanisms whereas identification of underlying causes and successful medical treatment remain a major challenge. Here, we show that Fn14 deletion reduces markedly PAB-induced right heart fibrosis and dysfunction. Our data provide strong evidence that inhibition of endogenous TWEAK/Fn14 pathway may potentially be clinically beneficial in treating right heart disease due to pressure overload. This is important as right heart disease is poorly characterized with limited treatment options. Furthermore, our study suggests that Fn14 is linked at the mechanistic level with the MAL-collagen pathway and provides evidence that Fn14 itself is regulated by ET-1.
It has previously been shown that inhibition of Fn14 signaling reduces fibrotic processes in several organs in disease models [
19,
51,
53,
54]. In the heart, this pathway has been associated with LV remodeling. Systemic overexpression of TWEAK induced heart failure affecting LV and RV [
20]. However, it remained unclear whether endogenous TWEAK/Fn14 signaling plays an active role in cardiac remodeling. Decreased TWEAK blood levels in patients with PAH argue against a role in the RV [
11]. However, reduced TWEAK blood levels might be due to sequestration of circulating TWEAK by the upregulated Fn14 receptor or might be a compensatory mechanism to protect from the consequences of Fn14 activation. This is supported by our findings that TWEAK blood plasma levels were reduced in MCT-treated rats while Fn14 was upregulated in the heart. In fact, opposite to most other signaling principles, in vivo TWEAK/FN14 pathway activation is regulated by changing the concentration of the receptor, not of the ligand, [
1]. Moreover, high Fn14 levels are reported to activate downstream signaling with few or none TWEAK present [
50]. Finally, correlations of cytokine blood levels to heart disease can be misleading as exemplified by TNF [
27]. Interestingly, TWEAK blood plasma levels are also decreased in patients and animal models with CKD whereas Fn14 was upregulated in the kidney. Importantly, deletion of Fn14 in the animal models protects against kidney fibrosis and failure [
19,
52].
Fibrosis is characterized by fibroblast accumulation and excess deposition of extracellular matrix proteins, which leads to tissue remodeling and dysfunction. The traditional view is that the underlying mechanism is induction of resident fibroblast proliferation [
24]. Our data suggest that reduced RV fibrosis upon PAB in Fn14
−/− mice is partially due to inhibition of this process. This is in accordance with reports that TWEAK has pro-mitogenic effects on cardiac cells including fibroblasts [
5,
28,
35]. Thus, reduced numbers of myofibroblasts and fibrosis might be due to reduced fibroblast proliferation. However, as the activation of TWEAK/Fn14 signaling was sufficient to induce myofibroblast differentiation and collagen expression in vitro, it appears likely that induced Fn14 re-expression also promotes myofibroblast differentiation driving RV fibrosis.
The traditional view of fibrosis has been challenged during the last years as it became clear that the fibroblast population exhibits a large phenotypic heterogeneity [
24]. It is now clear that fibroblasts can be derived from endothelial cells, pericytes, bone marrow-derived progenitor cells, monocytes, and fibrocytes. Thus, it is possible that inhibition of TWEAK/Fn14 signaling is not only regulating fibroblasts proliferation and collagen expression but also fibroblast precursor recruitment. Whether this plays in PAB-induced fibrosis a major role is unknown. However, it is well known that the TWEAK/Fn14 axis affects the immune response upon tissue injury [
4] and it has been hypothesized that it induces the recruiting of proinflammatory mediators during the acute phase of MI while at later time points, it participates in extracellular matrix remodeling and fibrosis. Therefore, it will be interesting to determine in future experiments whether TWEAK/Fn14 controls fibroblast precursor recruitment in the PAB model.
Although our in vitro data indicate that TWEAK can modulate directly fibroblast proliferation, differentiation and collagen synthesis, we cannot conclude from our data whether the observed protection is due to a direct myocardial effect or due to an unknown systemic, e.g. immune-mediated effect as we have utilized a general knockout model. Future studies utilizing conditional knockout mice will elucidate the underlying cellular mechanism.
SM22 and SMA both contain CArG boxes in their transcription control region, which are known targets of actin-MAL/MRTF-SRF signaling [
9,
44,
49]. MAL directly regulates
Col1a2 gene expression [
44]. It is a downstream target of Rho GTPase-actin signaling, which targets ROCK and MLC kinases [
32]. Nuclear MAL translocation links reorganization of the actin cytoskeleton to SRF-dependent gene transcription and myofibroblast differentiation [
16,
25,
44]. Finally, ROCK inhibition can reduce cardiac fibrosis [
16,
17,
40]. In accordance with these data, TWEAK stimulation led to rapid activation of RhoA kinase in CFs. Blockage of the Rho/SRF or ROCK activity abolished nuclear MAL translocation and reduced collagen expression. Our data suggest that TWEAK activates via Fn14 the RhoA-ROCK-dependent nuclear translocation of MAL to trigger SRF-dependent transcription. As Chen and co-workers have recently demonstrated that TWEAK/Fn14 signaling promotes proliferation and collagen synthesis of rat CFs via the NF-кB pathway it will be interesting to determine in the future how these pathways interact [
5,
48,
55].
Fn14 expression itself can be induced in CFs by ET-1, one of the key regulators implicated in the pathogenesis of PAH [
45]. Importantly, ET-1 also facilitated TWEAK/Fn14 signaling towards MAL translocation. Reduced ET-1 levels in Fn14
−/− mice are probably due to reduced fibrosis [
10,
29].
RV failure is the leading cause of death in PAH [
8,
38]. However, available therapies all target pulmonary vasoconstriction, but not the remodeling of the right heart [
12]. Fn14 combines several features that make it a good therapeutic target: Fn14
−/− mice are viable and show no obvious phenotype under physiological conditions. Fn14 is specifically upregulated in the RV after PAB, and its genetic ablation protected from right heart disease. Thus, global inhibition during therapy appears not prone to cause side effects. However, it is possible that Fn14
−/− mice activate compensatory pathways. In that case, therapeutic inhibition might have negative effects. For example, it has recently been suggested, in contrast to previous reports, that the human heart is a highly dynamic organ with the ability to regenerate cardiomyocytes [
22]. Thus, as TWEAK has recently been shown to be a positive regulator of cardiomyocytes proliferation [
35], a therapy targeting TWEAK/Fn14 signaling might interfere with cardiac homeostasis and cause over adverse effect.
Anti-TWEAK blocking antibodies and Fn14-Fc decoy receptor are available and have successfully been tested in other disease models. Given our findings, blocking the TWEAK/Fn14 axis may be a useful therapy for protecting patients from right heart disease and therefore, warrants further preclinical investigation.