Bridging the gap
Spiekman, Maroesjka
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from
it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date:
2018
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Spiekman, M. (2018). Bridging the gap: Adipose tissue-based therapy for dermal scarring. Rijksuniversiteit
Groningen.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
MicroRNA-15b is decreased during
cardiac fibrosis and inhibits cardiac
fibroblast activation through targeting
of the small GTPase intermediates
Growth Factor Receptor-Bound 2 (Grb2)
and Son-of- Sevenless homologue
(SOS) 1 and SOS2
Maroesjka Spiekman
1, Guido Krenning
1CHAPTER 8
1. Department of Pathology and Medical Biology, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands
ABSTRACT
BackgroundIn the heart, chronic injury results in the accumulation of extracellular matrix (ECM). The accumulation of fibrotic tissue might culminate in the progressive decline in cardiac function and, ultimately, heart failure. A pivotal process in cardiac fibrogenesis is fibroblast activation, entailing fibroblast proliferation, myofibroblast differentiation and excessive ECM production. Although canonical Transforming Growth Factor β (TGF-β) signaling is well-established as an inducer of cardiac fibroblasts activation, non-canonical TGF-β signaling via small GTPases has emerged as a second regulatory mechanism of fibroblast activation. We found that the expression of microRNA (miR)-15b is decreased in fibrotic hearts and identified the small GTPase intermediates Growth Factor Receptor-Bound 2 (Grb2) and Son-of-Sevenless homologue (SOS) 1 and SOS2 as putative targets of miR-15b. Therefore, we hypothesize that miR-15b might inhibit cardiac fibroblast activation by the inhibition of non-canonical TGF-β signaling.
Methods
Cardiac fibrosis was induced in mice by transverse aortic constriction and miR-15b expression was analyzed in the hearts of mice with and without aortic constriction. Binding of miR-15b to the 3’-UTR region of its putative gene targets was investigated using dual luciferase reporter assays.
In vitro, cardiac fibroblasts transfected with miR-15b mimics or scrambled control sequences were
stimulated with TGF-β1 and the gene expression of small GTPase intermediates (Grb2, SOS1 and SOS2), ECM components (COL1A1 and COL3A1) and myofibroblast markers (ACTA2 and CNN1) were analyzed. At the functional level, gel contraction assays were performed and fibroblast proliferation was assessed. Finally, the activity of the small GTPase Ras was measured in TGF-β-stimulated cardiac fibroblasts that were transfected with miR-15b mimics or scrambled control sequences.
Results
Aortic constriction induced left ventricular fibrosis in mice. In the fibrotic mouse hearts, miR-15b expression was lower compared to hearts of sham-operated mice. In TGF-β1-stimulated cardiac fibroblasts, the expression of miR-15b was lower compared to non-stimulated cardiac fibroblasts. Grb2, SOS1 and SOS2 were confirmed as miR-15b targets, as binding of miR-15b to the 3’-UTR region of Grb2, SOS1 and SOS2 reduced reporter activity. TGF-β1 stimulation, Ras inhibition or miR-15b transfection did not alter fibroblast proliferation, however, transfection of cardiac fibroblasts with miR-15b mimics precluded myofibroblast differentiation and ECM production, as indicated by a decrease in the expression of ACTA2, CNN1, COL1A1, and COL3A1. Moreover, in cardiac fibroblasts that were transfected with miR-15b mimics, collagen gel contraction was inhibited.
Conclusion
During cardiac fibrosis, miR-15b expression is decreased, which culminates in the activation of cardiac fibroblasts. Maintaining the expression of miR-15b, via the exogenous delivery of miR-15b mimics precludes cardiac fibroblast activation by TGF-β1, which might be due to the decreased
expression of the small GTPase intermediates Grb2, SOS1 and SOS2. Therefore, the decrease of miR-15b expression during cardiac fibrosis might contribute to the progression of heart failure and could thus be used as a therapeutic target.
INTRODUCTION
Chronic fibroproliferative disease, leading to organ failure, is the cause of nearly 45% of deaths in the developed world1. A hallmark feature of fibrosis in any organ is fibroblast activation, i.e. fibroblast proliferation, myofibroblast differentiation and ECM deposition1,2. Chronic injury leads to a damage repair response by the resident tissue cells1. However, if this repair response becomes dysregulated and fibroblast activation persists, it might culminate in organ fibrosis2.
In the healthy heart, ECM production and degradation are in homeostasis. In response to chronic injury, ECM accumulates in the cardiac tissue, which results in stiffening of the heart3 and reduced cardiac contractility. Furthermore, cardiac ECM accumulation can impair the electrophysiological coupling of cardiomyocytes 4 and might result in compensatory cardiac hypertrophy. Combined, these processes might result in the progressive loss of cardiac function and finally heart failure5. Transforming Growth Factor (TGF) β1 and β2, play a pivotal role in the development of heart failure through the induction of cardiac fibroblast activation5,6.
TGF-β induces a pro-fibrotic response through the activation of its canonical and non-canonical signal transduction pathways. The binding of TGF-β to its canonical receptor complex composed of TGF-β receptor (TGFBR) 1 and TGFBR27, results in the activation of Smad2/3 complexes, which is extensively studied in the context of cardiac fibrosis6. Yet, non-canonical TGF-β signaling also contributes to the development and progression of cardiac fibrosis4,6. Recently, non-canonical TGF-β signal transduction via TGFBR1/2 and the associated small GTPases has emerged as regulatory mechanism of fibroblast activation and fibrosis8,9. Small GTPases are cell membrane-bound G-proteins, which alternate between an active GTP and an inactive GDP state10. When TGF-β binds to its receptor complex, the small GTPase intermediates Growth Factor Receptor-Bound 2 (Grb2) and Son-of-Sevenless homologue (SOS) 1 and SOS2 are recruited to the receptor complex, which activate the downstream GTPases 9,11. Ultimately, the TGF-β-induced activation small GTPases might culminate in fibroblast activation12,13 and the inhibition the small GTPase Ras can limit fibrogenesis14,15.
MicroRNAs (miRs) are small (20-25 nucleotides), non-coding RNAs that regulate gene expression at the post-transcriptional level through their interaction with their target mRNA16. MiRs interact with their target mRNAs through imperfect base pairing to the 3’ untranslated region (3’UTR) of its target mRNA. Upon the formation of a miR-mRNA duplex, translational repression of the mRNA occurs by either the degradation of the mRNA or by blocking of translation initiation or elongation17. In cardiac physiology, miRs are one of the regulatory mechanisms that orchestrate
proliferation, differentiation and angiogenesis18, whereas in cardiac pathology, dysregulated miR expression might induce or facilitate the pathogenesis19,20.
We found that miR-15b expression is decreased during cardiac fibrosis and identified miR-15b as a putative modulator of non-canonical TGF-β signaling via the small GTPases. Using online bioinformatics tools (i.e. TARGETScan21 and the miRANDA algorithm22), we uncovered that miR-15b putatively targets the small GTPase intermediates Grb2, SOS1 and SOS2. We hypothesize that miR-15b might inhibit cardiac fibroblast activation by targeting these small GTPase intermediates. Subsequently, we investigated the role of miR-15b on non-canonical TGF-β signaling during cardiac fibroblast activation.
MATERIALS AND METHODS
Transverse aortic constriction cardiac fibrosis mouse model
Male C57Bl/6j mice, aged 10-12 weeks, were purchased (Harlan, Horst, The Netherlands). Animals were randomly assigned to the experimental groups. Aortic banding was performed to induce cardiac fibrosis (n=5) as described previously23. In brief, under anesthesia, the ascending aorta was visualized through an incision in the second intercostal space. A suture was placed around the aorta and a blunted 27G needle was placed onto the aorta. The suture was tightened around the aorta and needle to constrict the aorta after which the needle was removed to allow a standardized opening of the aorta. In sham animals (n=6), the same operative procedure was performed, except for placement of the constricting suture. Two weeks after the aortic banding, animals were sacrificed by exsanguination under general anesthesia. Hearts were collected for histochemistry and gene expression analyses. The local Committee for Animal Experimentation (University of Groningen, University Medical Centre Groningen) approved all experimental procedures.
Cardiac fibroblast culture, microRNA transfection and stimulation
Experiments were conducted with mouse or human cardiac fibroblasts (CF). Mouse CF were isolated from adult mouse hearts. Hearts were minced into small (<1mm3) pieces and subjected to 0.1% collagenase (Roche Diagnostics, Mannheim, Germany) digestion. Digested tissue was filtered through a 70µm cell strainer to obtain a single cell suspension. Cells were pelleted by centrifugation and plated at a density of 5,000 cells/cm2 in Dulbecco’s Modified Eagle’s/Ham’s F12 Medium (DMEM/F12; Lonza, Breda, The Netherlands) with 10% fetal bovine serum (FBS; GE Lifesciences, Piscataway, NJ) with 2mM L-glutamine (Sigma-Aldrich, St. Louis, MO) and 1% v/v penicillin-streptomycin (Sigma-Aldrich, St. Louis, MO).
Human CF were purchased (Promocell, Heidelberg, Germany) and culture expanded in DMEM with 10% FBS, 2mM L-glutamine, 1% penicillin/streptomycin, 10ng/mL recombinant human fibroblast growth factor 2 (FGF-2; Peprotech, Rocky Hill, NJ) and 1x Insulin-Transferrin-Selenium (ITS; Gibco,
Grand Island, NY). Approximately a week before transfection and stimulation, CF were switched to culture medium without FGF-2 and ITS.
CF were transfected with pre-miR-15b (cat# PM10904, Ambion Life Technologies, Carlsbad, CA) or scrambled miR control sequences (cat# AM171110, Ambion Life Technologies, Carlsbad, CA) with the siRNA Reagent System (SantaCruz Biotechnology, Heidelberg, Germany) in accordance with the manufacturer’s instructions. For lentiviral expression of miR-15b or scrambled control sequences, pGreenPuro-miR-15b and pGreenPuro-SCR were generated by unidirectional cloning of oligonucleotides (Biolegio, Leiden, The Netherlands) encoding the pre-miR-15b or a scrambled sequence into the BamH1/EcoR1 sites of the pGreenPuro vector (Systembio). Vectors were enumerated in competent Top10 E.coli (Invitrogen, Carlsbad, CA) and isolated using the Qiagen miniprep kit. Correct insertion of the oligonucleotides was validated by sequencing (Baseclear, Leiden, The Netherlands). HEK293T cells were transfected with pGP-miR-15b or pGP-SCR, pVSVG (envelope plasmid) and pCMV-R8.91 (gag-pol 2nd generation packaging plasmid) using the Endofectin-Lenti (Gene Copoeia, Rockville, MD) transfection reagent. 24-72 hours post-transfection, viral supernatants were collected in DMEM/F-12, centrifuged at 500xg for 10 min and filtered through a 0.45μm PES filter to remove any remaining cell debris. Viral supernatants were supplemented with polybrene (6µg/ml; Sigma, St. Louis, MO) and applied on 30% confluent CF for three consecutive rounds of 24h exposure. Transduced CF were passaged once and transduced cells were selected by puromycin (6µg/ml; Invitrogen, Carlsbad, CA).
After transfections and prior to stimulation, cells were synchronized for 24h in DMEM with 0.5% FBS, 2mM L-glutamine and 1%penicillin/streptomycin. For experiments, CF were stimulated with 5ng/mL TGF-β1 (R&D Biosystems, Abingdon, UK) and/or 10μM of the small GTPase inhibitor farnesyl thiosalicylic acid (FTS; Cayman Chemical, Ann Arbor, MI) and/or 100ng/mL Bone morphogenetic protein (BMP7; R&D Biosystems, Abingdon, UK) in DMEM with 0.5% FBS, 2mM L-glutamine and 1% penicillin/streptomycin. DMSO was used as vehicle control. Cells were cultured under experimental conditions for 24 hours (proliferation assessment by Ki67 immunofluorescence) or 72h (other assays).
Histochemistry
Mouse hearts were formalin fixed and paraffin embedded. Four μm transverse sections were deparaffinnized and stained with Masson’s Trichrome solutions: samples were fixed for 5 minutes with Bouins Fixative at 51°C and were consecutively stained for 20 minutes with Weighert’s Iron Hematoxilin, for 20 minutes with Biebrich Scarlet-Acid Fuchsine, for 12 minutes with Phosphomolybdic-Phosphotungstic Acid and for 7 minutes with Aniline Blue (all at room temperature). Between each staining solution, slides were extensively washed with demi-water. Finally, slides were incubated with 1% glacial Acetic Acid, were mounted in toluene solution and visualized by light microscopy.
Gene and microRNA transcript analysis
Total RNA was isolated using Trizol Reagent (Invitrogen, Carlsbad, CA) in accordance with the manufacturer’s instructions. RNA quantity was assessed spectrophotometrically (Nanodrop, Wilmington, DE) and RNA integrity was assessed by gel electrophoresis. For gene expression analysis, 500ng total RNA was reverse transcribed into cDNA using the First Strand cDNA synthesis kit and random hexamer primers (Applied Biosystems, Carlsbad, CA) according to the manufacturer’s instructions. Quantitative polymerase chain reaction (qPCR) was performed using 10ng cDNA, human or mouse gene specific primers (all from Biolegio, Leiden, the Netherlands; Primer sequences, see Table 1) and Fast Start SYBR green (Roche, Almere, the Netherlands). For miR expression analysis, RNA was reverse transcribed using the TaqMan microRNA reverse transcription kit (Applied Biosystems, Carlsbad, CA) and specific miR-15b and RNU-6B stem loop primers. For qPCR analysis, 20ng cDNA and specific primers for miR-15b and RNU-6B and a common antisense primer (all from Biolegio, Leiden, the Netherlands; Primer, see Table 2) and Fast Start SYBR green were used. For mRNA and miR expression analyses, reactions were run on a Viia7 real time PCR machine (Applied Biosystems, Carlsbad, CA) and analyzed using the Viia7 software. All reactions were performed in duplicate. Gene expression was calculated using the ΔΔCt method. GAPDH (for mRNA) and RNU-6B (for miRs) were used as reference genes. Data are presented as fold change compared to control.
3’UTR reporter assays
3’UTR reporter assays were performed as described previously24. In brief, 3′UTR fragments were isolated from a cDNA pool derived from various human tissues using specific primers (Biolegio,
Table 1 | Primer sequences for mRNA qPCR
Gene Forward sequence Reverse sequence
ACTA2 CTGTTCCAGCCATCCTTCAT TCATGATGCTGTTGTAGGTGGT CNN1 CCAACCATACACAGGTGCAG TCACCTTGTTTCCTTTCGTCTT COL1A1 GGGATTCCCTGGACCTAAAG GGAACACCTCGCTCTCCA COL3A1 CTGGACCCCAGGGTCTTC CATCTGATCCAGGGTTTCCA Grb2 CCAAGAACTACATAGAAATGAAACCA CCGCTGTTTGCTAAGCATTT SOS1 GCAATGATACCGTCTTTATCCAA CACTTCATCAGTGCCTTTGGT SOS2 GCTGGCAGCTGTTTTGAAG GATAATGTTTCATAAGGATCAAATGC Table 2 | Primer sequences for miR cDNA synthesis and qPCR
miRNA Stemloop Forward sequence Reverse sequence miR-15b GTCGTATCCAGTGCAGG- GTCCGAGGTATTCGCACTG-GATACGACTGTAAACC TGCGGTAGCAGCACATCAT CCAGTGCAGGGTCCGAG-GTCCG RNU6B GTCGTATCCAGTGCAGG- GTCCGAGGTATTCGCACTG-GATACGACAAAAATATGG TGCGGCTGCGCAAGGATGA CCAGTGCAGGGTCCGAG-GTCCG
Leiden, The Netherlands). Primer sequences are found in Table 3. Primers were extended with restriction sequences and amplification was performed. The amplicon size was validated by agarose gel electrophoresis. The purified amplicons were cloned into a dual luciferase reporter vector, psiCHECK-2 (Promega, Madison, WI). COS7 cells were transfected with the UTR reporter plasmids and 50nM pre-miR-15b or scrambled control (Ambion Life Technologies, Carlsbad, CA). After 48h, luciferase activity was assayed using the Dual-Glo Luciferase Assay kit (Promega, Madison, WI) to assess binding of miR-15b to the reporter plasmids.
Immunoblot analyses
Protein levels of Grb2 and SOS1 were determined by use of immunoblot analyses. The immunoblot procedures were conducted as previously described25,26. The primary antibodies rabbit anti-Grb2 (1:1,000; cat# sc-255, Santa Cruz Biotechnologies, Santa Cruz, CA), rabbit anti-SOS1 (1:2,000; cat# D3T7T, Cell Signaling Technology, Beverly, MA), rabbit anti-β-actin (1:1,000; cat# 4967L, Cell Signaling Technology, Beverly, MA) and rabbit anti-GAPDH (1:1,000; cat# ab9485, Abcam, Cambridge, UK) were used. The secondary antibody goat anti-rabbit IRdye680 (1:10,000; cat# 925-68021, Li-Cor Biosciences, Bad Homburg, Germany) was used. Proteins were visualizes using an Odyssey infrared imaging system (Li-Cor Biosciences, Bad Homburg, Germany). Densitometry analysis was performed using the TotalLab software (Nonlinear Dynamics, Newcastle-upon-Tyne, UK).
Ras activation assays
Ras activation was assessed using a Ras Activation ELISA assay kit (Millipore, Temecula, CA) in accordance with the manufacturer’s instructions. Cells were lysed in the supplied lysis buffer with a protease inhibitor (Sigma Aldrich, St. Louis, MO). The samples were sonicated and the protein concentrations were measured using the Quick Start Bradford protein assay (BioRad Laboratories, Hercules, CA). 100μg protein per experimental condition was used. 5 minutes after addition of the chemiluminescent substrates, the signal was recorded for 1 second on a Luminoskan ASCENT (Thermo Scientific, Waltham, MA). Data are presented as fold change compared to control.
Immunofluorescence staining
CF were treated with TGF-β1 and/or FTS for 24 hours (Ki67 staining) or 72 hours (αSMA staining). Next, the cells were washed and fixed in 2% paraformaldehyde (Sigma Aldrich, St. Louis, MO). The cells were permeabilized using 0.5% Triton X-100 (Sigma Aldrich, St. Louis, MO), blocked with 10% donkey serum in phosphate buffered saline (PBS). Next, the cells were incubated with rabbit anti-Ki67 (1:250; cat# PSX1028, Monosan, Uden, The Netherlands) or mouse anti-αSMA (1:100; cat#
Table 3 | Primer sequences 3’UTR reporter assays / 3′UTR fragments:
Gene Forward sequence Reverse sequence
Grb2 GAGTCAAGAAGCAATTATTTAAAG GAAGAATTCATTGTGTATTTATTATTC SOS1 GTTCCTCTGTACTGGGATGTATATT GTGGGCTATGTAAGGCATTT SOS2 CCTTAGCCATATGTAGTCATT CTTTTAAAAATTAAGACTGGGGTG
ab7817, Abcam, Cambridge, UK) primary antibodies. The secondary antibodies, donkey anti-rabbit AlexaFluor 594 and donkey anti-mouse AlexaFluor 594 (both 1:500; both from Molecular Probes, Carlsbad, CA) were used for signal development. Images were acquired using a Zeiss Axio Observer. Z1 microscope (Carl Zeiss, Mainz, Germany) in fluorescence mode. The images were analyzed using the TissueFAXS Immunofluorescence Cell Analysis system (TissueGnostics USA Inc., Tarzana, CA).
Gel contraction assay
For gel contraction assays, CF transfected with pre-miR15b or a scrambled control sequence were treated with TGF-β1 and/or FTS or vehicle controls for 3 days. Afterwards, CF were detached and the cell number was determined using a hemocytometer (Beckman Coulter, Palo Alto, CA). CF were resuspended in DMEM with 0.5% FBS and mixed with high concentration rat tail collagen (Corning, New York, NY) and the pH of this mixture was adjusted using NaOH (Sigma-Aldrich, St. Louis, MO), HEPES (Gibco, Grand Island, NY) and 10xPBS according to the manufacturer’s instructions. 100μl of this cell-collagen mixture was pipetted into the wells of a 96 wells plate. The gels were made in triplicate for each experimental condition. The final gels consisted of 3x105 cells/mL and 2.4mg/ mL collagen. The collagens gels were allowed to polymerize for one hour after which the gels were circumferentially detached with a 25G needle. Appropriate media with TGF-β1 and/or FTS or vehicle controls were added and the gels were allowed to contract. 72h after detachment, the collagens gels were scanned with a flatbed scanner. The gel surface area was determined using ImageJ (NIH, Bethesda, ML).
Statistical analyses
The data are derived from at least three independent experiments, unless stated otherwise. For statistical analyses, GraphPad Prism version 5 (GraphPad Software Inc., La Jolla, CA) was used. Data are presented as mean±SEM, unless stated otherwise. Data were analyzed using a one-way analysis of variance (ANOVA) with a Bonferroni post-hoc test, unless stated otherwise. P-values <0.05 were considered statistically significant.
RESULTS
MiR-15b expression is decreased in cardiac fibrosis and in TGF-ß1-activated cardiac fibroblasts
Aortic banding in mice caused a distortion of the cardiac architecture and the deposition of high amounts of ECM between the muscle fibers, compared to the compact myocardial tissue in the sham control hearts (Fig. 1A). On average, aortic banding for 14 days increased the ECM amount in the left ventricle from ~3 to ~8% (Fig. 1B). In non-fibrotic hearts, miR-15b expression could be readily detected, whereas in the hearts of banded mice, miR-15b expression was decreased (Fig. 1C). We questioned if the decrease in miR-15b expression levels would also occur in TGF-β-activated cardiac fibroblasts. CF readily expressed miR-15b, however, in response to TGF-β1 stimulation, CF decreased the expression of miR-15b (Fig. 1D).
MiR-15b mimics target the small GTPase intermediates Grb2, SOS1 and SOS2
Online bioinformatics tools (TARGETScan21 and the miRANDA algorithm22) imply that the downstream mediators of non-canonical TGF-β signaling, i.e. the small GTPase intermediates Grb2, SOS1 and SOS2 are putative targets of miR-15b (Fig. 2A). The co-transfection of Grb2, SOS1 and SOS2 reporter constructs with miR-15b mimics into COS7 cells reduced the reporter activity compared to the scrambled controls (Fig. 2B), indicating a direct repression of miR-15b on the 3’-UTRs of these genes. Thus, we established that Grb2, SOS1 and SOS2 are genuine miR-15b targets.
A.
Sham TACB.
C.
D.
β Figure 1 | MicroRNA-15b levels are decreased in fibrotic mouse hearts and in TGF-β1-stimulated
mouse cardiac fibroblasts. (A) Masson’s trichrome staining of a healthy (sham-operated) mouse heart and a mouse heart after transverse aortic constriction (TAC). (B) Quantification of the fibrotic area in hearts of normal (sham operated) mice and of mice that underwent TAC. MiR-15b levels in (C) normal hearts and TAC hearts and in (D) unstimulated and TGF-β-activated mouse cardiac fibroblasts. **P<0.01 (Mann-Whitney Test)
In TGF-β1-stimulated CF, the expression of Grb2, SOS1 and SOS2 increased compared to the non-stimulated CF (Fig. 3A-C). Next, we questioned if miR-15b might preclude the increase in expression of Grb2, SOS1 and SOS2 in TGF-β-stimulated CF. CF transfected with miR-15b mimics showed reduced expression of Grb2, SOS1 and SOS2 compared to the unstimulated control cells and to the TGF-β1-stimulated CF (Fig 3A-C). Based on preliminary experiments (n=1), TGF-β1 stimulation seems to decrease the Grb2 protein level (Fig. 3D,E), whereas the SOS1 protein level seems to be slightly increased (Fig. 3F,G) compared to unstimulated control cells. In TGF-β1-stimulated miR-15b transfected cells, the Grb2 and SOS protein levels decreased (Fig 3D-G) compared to TGF-β1-stimulated cells. In combination, these data suggest that miR-15b regulates the expression of the small GTPase intermediates Grb2, SOS1 and SOS2 in TGF-β-stimulated CF.
Concurrently, after aortic banding, the Grb2 expression level tended to be increased (Fig. 3H) and SOS1 and SOS2 expression levels are higher in the banded hearts compared to the non-banded hearts (Fig. 3I,J).
A.
B.
RNA Nucleotide Nucleotide of position Sequence interest in 3’UTR
miR15b 3' ACAUUUGGUACUACACGACGAU 5' GRB2 220: 5' UACAAAUUUUCACUGCUGCUC 3' SOS1 3224: 5' UUUAUUUCUUGGUAUUCUGCUGCUG 3' SOS2 512: 5' CUUAAAUGCCGGUAUUUGCUGCUG 3'
Figure 2 | MiR-15b directly targets small GTPase intermediates Grb2, SOS1 and SOS2. (A) In silico
analysis identified Grb2, SOS1 and SOS2 as putative miR-15b targets, with a single binding site in the 3’UTR region of these genes. (B) Luciferase reporter assays in miR-15b transfected COS7 cells with Grb2, SOS1 and SOS2 reporter constructs. Data are mean±SEM. ***P<0.001 (one-way ANOVA and Bonferroni post-hoc test)
Control TGF-β1FTS miR-15bmiR-15b + TGF-β1 FTS + TGF-β1 SOS1 GAPDH 152 kDa 37 kDa TGF-β1 - - + + - + miR-15b - + - + - -FTS - - - - + + Control TGF-β1FTS miR-15bmiR-15b
+ TGF-β1 FTS + TGF-β1 Grb2 β-actin 25 kDa 42 kDa TGF-β1 - - + + - + miR-15b - + - + - -FTS - - - - + +
A.
B.
C.
H.
I.
J.
D.
E.
F.
G.
TGF-β1 - - + +miR-15b - + - + TGF-β1 - - + + miR-15b - + - + TGF-β1 - - + + miR-15b - + - +
Figure 3 | Grb2, SOS1 and SOS2 levels are decreased by miR-15b. Gene expression levels of (A)
Grb2, (B) SOS1 and (C) SOS2 in mouse cardiac fibroblasts (one-way ANOVA and Bonferroni post-hoc test). Protein analysis by Western blot of (D,E) Grb2 and (F,G) SOS1 in human cardiac fibroblasts (n=1). Gene expression levels of (H) Grb2, (I) SOS1 and (J) SOS2 in in hearts of mice that underwent transverse aortic constriction (TAC) or a sham operation (Mann-Whitney test). *P<0.05, **P<0.01
miR-15b decreases the expression of ECM and cytoskeleton components and contractility
Since miR-15b levels are decreased during cardiac fibroblast activation, we questioned if maintaining miR-15b expression could inhibit fibroblast activation, i.e. fibroblast proliferation, myofibroblast differentiation and ECM deposition. Therefore, we firstly investigated the proliferation in TGF-β1-stimulated CF (Fig. 4A). The addition of TGF-β1 to CF did not increase the cell proliferation as compared to unstimulated CF (Fig. 4B). Transfection of CF with miR-15b mimics or a pharmacological inhibitor of small GTPases (FTS), did not alter the proliferation as compared to unstimulated cells (Fig. 4B). However, in TGF-β1-stimulated CF, miR-15b transfection decreased CF proliferation in all three independent experiments, albeit these results did not reach statistical difference due to the large interexperimental variability (Fig. 4B).
We next investigated if miR-15b mimics or FTS could inhibit the TGF-β1-induced expression of ECM genes. TGF-β1 stimulation of CF increased the gene expression of COL1A1 and COL3A1 (Fig. 5). Addition of FTS did not alter COL1A1 (Fig. 4A) or COL3A1 (Fig. 4B) gene expression, compared to unstimulated control CF, however, in FTS prevented induction of COL1A1 and COL3A1 gene expression by TGF-β1. In CF transfected with miR-15b mimics, the induction of COL1A1 (Fig. 5A) and COL3A1 expression (Fig. 5B) following TGF-β1 stimulation was precluded.
Finally, we assessed if miR-15b mimics or FTS could also inhibit myofibroblast differentiation and examined the expression of ACTA2 and CNN1 and CF contraction at the functional level. TGF-β1 stimulation of CF resulted in the increased gene expression of ACTA2 and CNN1 (Fig. 6A,B), which was abolished by the addition of FTS or the transfection of miR-15b mimics (Fig. 6A,B). On protein level, the percentage of α-SMA positive cells decreased in TGF-β1-activated CF compared to unstimulated CF (Fig. 6C,D). However, in TGF-β1-activated, miR-15b transfected CF, αSMA levels seemed to decrease even further compared to TGF-β1-stimulated scrambled control CF (Fig. 6C,D). We questioned if the reduced expression of ACTA2 and CNN1 in miR-15b-expressing fibroblasts would culminate in a decreased contractile capacity (Fig. 7). CF have a baseline contractile capacity (Fig. 7B), which is elevated by the addition of TGF-β1 (Fig. 7B). In miR-15b-expressing CF, this enlargement by TGF-β1 stimulation was largely abrogated (Fig. 7B) and the contractile capacity remained at baseline levels.
TGF-β1-induced activation of cardiac fibroblasts is dependent on small GTPase activation
As the transfection of miR-15b mimics in CF decreased the expression of the small GTPase intermediates Grb2 and SOS2 and precluded fibroblast activation, we questioned if non-canonical TGF-β signaling via the small GTPases result in fibroblast activation. We inhibited small GTPase activation using the inhibitor FTS (Fig. 5 and 6). When small GTPase signaling was inhibited during TGF-β1-induced CF activation, the increase in the expression of COL1A1 and COL3A1 (Fig. 5A,B) and ACTA2 and CNN1 (Fig. 6A,B) was precluded as compared to TGF-β1-stimulated CF. These data suggest that non-canonical TGF-β signaling via the small GTPases is required for fibroblast activation.
A.
Control TGF-β1 miR15b + TGF-β1 miR-15b FTS + TGF-β1 FTSB.
Ki67
DAPI
Ki67
DAPI
TGF-β1
- - +
+
-
+
miR-15b
- + -
+
-
-FTS
- -
- -
+
+
Figure 4 | Proliferation of cardiac fibroblasts is not altered by TGF-β1 stimulation, however miR-15b
transfection tended to decrease proliferation. (A) Representative pictures of immunofluorescent staining for proliferation marker Ki67 in mouse cardiac fibroblasts (B) Percentage of Ki67+ cells (one-way ANOVA and Bonferroni post-hoc test). Scale bar represents 100µm.
A.
B.
TGF-β1 - - + + - + miR-15b - + - + - -FTS - - - - + + TGF-β1 - - + + - + miR-15b - + - + - -FTS - - - - + + Figure 5 | miR-15b transfection inhibits TGF-β1-induced ECM production in mouse cardiacfibroblasts. Gene expression levels of (A) COL1A1 and (B) COL3A1. Data are mean±SEM. **P<0.01, ***P<0.001, ****P<0.0001 (one-way ANOVA and Bonferroni post-hoc test).
Because miR-15b reduces the expression the small GTPase mediators Grb2, SOS1 and SOS2, we questioned if this would ultimately also reduce the activation of the small GTPase Ras. TGF-β1 stimulation or FTS inhibition did not change Ras activation, compared to unstimulated CF (Fig. 8). Also, miR-15b transfection decreased Ras activation as compared to unstimulated scrambled control cells (Fig. 8), however, this difference was not statistically significant due to the large interexperimental variability. In TGF-β1-stimulated cells miR-15b transfection did not alter Ras activation compared to the TGF-β1-stimulated scrambled control cells (Fig. 8).
BMP-7 treatment restores miR-15b expression levels, reduces Grb2, SOS1 and SOS2 expres-sion and inhibits fibroblast activation
BMP7 is an antagonists of TGF-β1 signaling27. Therefore, we questioned if BMP7 could increase miR-15b expression levels to antagonize fibroblast activation. BMP7 alone did not significantly alter the miR-15b expression (Figure 9A) in unstimulated CF, but in TGF-β1- stimulated CF, the miR-15b expression levels were maintained at baseline level by the addition of BMP7 (Fig 9A). Also, BMP7 treatment inhibited the increase of SOS1 and SOS2 expression, compared to TGF-β1-stimulated CF (Fig. 9B,C). BMP7 stimulation did not alter the expression of ACTA2 and CNN1 in presence or absence of TGF-β1 (p>0.05, data not shown). The COL1A1 and COL3A1 gene expression was increased by TGF-β1 stimulation (Fig. 9D,E). The increase in COL1A1 expression by TGF-β1 stimulation could be partially inhibited by BMP7 addition compared to TGF-β1-stimulated CF (Fig 9D), whereas the TGF-β1-induced increase in COL3A1 expression was inhibited almost completely compared to TGF-β1-stimulated CF (Fig. 9E).
TGF-β1 - - + + - + miR-15b - + - + - -FTS - - - - + + TGF-β1 - - + + - + miR-15b - + - + - -FTS - - - - + +
C.
D.
Control miR-15b TGF-β1 miR15b + TGF-β1
TGF-β1 - - + + miR-15b - + - +
A.
B.
αSMA DAPI
Figure 6 | Mesenchymal marker expression is decreased in miR-15b transfected, TGF-β1-stimulated
cardiac fibroblasts. Gene expression of (A) ACTA2 or αSMA and (B) CNN1 or Calponin in mouse cardiac fibroblasts. (D) Representative pictures of immunofluorescent stainings for αSMA and (C) quantification of the percentage of αSMA positive human cardiac fibroblasts (n=1). Data are mean±SEM. **P<0.01, ***P<0.001, ****P<0.0001 (one-way ANOVA and Bonferroni post-hoc test).
A.
B.
TGF-β1 - - + + miR-15b - + - +
Control miR-15b TGF-β1 miR15b + TGF-β1
Figure 7 | MiR-15b decreases contractility of human cardiac fibroblasts. (A) Photographs
of representative collagen gels and (B) quantification of the percentage of gel contraction by human cardiac fibroblasts. *P<0.05, **P<0.01. (one-way ANOVA with Fisher’s LSD test). Scale bar represents 200µm.
DISCUSSION
Here, we show that miR-15b levels are decreased during cardiac fibrosis and during TGF-β1-induced cardiac fibroblast activation. We demonstrate that Grb2, SOS1 and SOS2 are targets of miR-15b. During TGF-β1-induced cardiac fibroblast activation, the expression of the small GTPase intermediates (Grb2, SOS1 and SOS2), ECM components (COL1A1 and COL3A1) and the myofibroblast markers (ACTA2 and CCN1) is readily induced at the gene level. Transfection of cardiac fibroblasts with miR-15b mimics decreases the expression of the small GTPase intermediates Grb2, SOS1 and SOS2 in TGF-β1-stimulated CF, putatively leading to the observed inhibition of TGF-β1-induced cardiac fibroblast activation (summarized in Fig. 10). Particularly, we demonstrate that miR-15b is a novel regulator of cardiac fibroblast activation and might thus represent a new possibility for therapeutic intervention.
Unexpectedly, in our experiment we did not find an induction of cardiac fibroblast proliferation upon TGF-β1 stimulation. TGF-β is known to induce growth arrest or apoptosis in endothelial, epithelial and hematopoietic cells28,29. However, in certain types of mesenchymal cells, TGF-β can
TGF-β1
- - + + -
+
miR-15b
- + - + -
-FTS
- - - - +
+
Figure 8 | Ras activation tends to be reduced after miR-15b mimic transfection in unstimulated
mouse cardiac fibroblasts, as determined by Ras activation ELISAs. However, no difference in Ras activation is observed after miR-15b mimic transfection in TGF-β1-stimulated cardiac fibroblasts, as compared to TGF-β1-stimulated scrambled control cells. (one-way ANOVA and Bonferroni post-hoc test).
be pro-mitotic28. In fibroblasts, the effect of TGF-β seems to be context dependent, since it has been reported that TGF-β does not lead to proliferation directly, but induces the expression of pro-mitotic factors such as FGF-230 or CTGF31 and/or is dependent on the presence of mitogenic stimuli in the culture media, such as PDGF32 or EGF31. Under hypoxic conditions, the inhibition of TGF-β1 signaling by the overexpression of TGFBR3 even leads to a decrease in apoptosis33, suggesting that TGF-β1 can be pro-apoptotic under certain conditions. Thus, the traditional view that TGF-β1 induces fibroblast proliferation might be a biased view.
We found that miR-15b mimics seemed to decrease Ras activation in unstimulated CF, as compared to scrambled control CF. However, in TGF-β1-stimulated miR-15b mimics failed to decrease Ras activation compared to TGF-β1-stimulated scrambled control CF, suggesting compensatory activation of Ras by alternative upstream mediators in presence of TGF-β. However, the miR-15b mimics did decrease the Grb2, SOS1 and SOS2 and concurrently inhibited cardiac fibroblast activation. Thus, our data suggest that the inhibition of cardiac fibroblast activation by miR-15b only partially relies on Ras inactivation. This finding might be explained by the interaction of Grb2, SOS1 and SOS2 with small GTPases other than Ras. For example, Grb2 can interact with RAB1334, SOS1 with RAF35 and Rac36, and SOS2 with Rho and Rac36, thus creating alternative signaling routes. Activation of the small GTPases RAF, Rac and Rho might contribute to fibroblast activation37-39. Ultimately, the activation of the small GTPases leads to the induction of Mitogen-activated protein
A.
B.
TGF-β1 - - + + BMP7 - + - +C.
D.
E.
TGF-β1 - - + + BMP7 - + - + TGF-β1 - BMP7 - - + ++ - + TGF-β1 - - + + BMP7 - + - + TGF-β1 - BMP7 - - + ++ - +Figure 9 | BMP7 maintains miR-15b expression in TGF-β1-stimulated cardiac fibroblasts. Gene
expression levels of (A) miR15b, (B) SOS1, (C) SOS2, (D) COL1A1 and (E) COL3A1 in mouse cardiac fibroblasts. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001. (one-way ANOVA and Bonferroni post-hoc test)
kinase kinase/Extracellular signal-regulated kinase (MEK/ERK), phosphoinositide 3-kinase/Akt (PI3K/Akt) and p38/Mitogen activated protein kinase (p38/MAPK) signaling (Fig. 10), which are all implicated in fibroblast activation40-42. Moreover, the activation of p38/MAPK signaling by Grb2 has been reported during fibrogenesis43,44. For example, Grb+/- mice are protected against transverse aortic constriction induced cardiac fibrosis44.
It has been suggested that Grb2 can activate p38 directly, without the intermediation of small GTPases45,46 (Fig. 10). Thus, the observed inhibition of fibroblast activation by miR-15b might be mediated by a decrease in Grb2, SOS1 and SOS2, leading to less activation small GTPases such as RAF, Rac and Rho and/or by a decrease in Grb2-induced p38 activation. Thus, other small GTPases such as RAF, Rac and Rho should be investigated to further clarify how miR-15b can inhibit fibroblast activation.
Interestingly, gain-of-function mutations of SOS1 and SOS2 can cause Noonan syndrome 4 and 9, respectively47,48. Noonan syndrome is characterized by craniofacial anomalies, but patients affected by these syndromes also have a high risk (20-30%) of developing the fibroproliferative heart disease hypertrophic cardiomyopathy49. Additionally, gain-of-function mutations of SOS1 can lead to hereditary gingival fibromatosis, a fibroproliferative disease of the oral mucosa50.
Cell membrane miR-15b BMPR1 BMPR2 BMP-7 ? TGFBR1 TGFBR2
Pro-fibrotic gene expression Nucleus GTPase GTP Grb2 SOS1/2 MEK/
ERK1/2 PI3K/Akt GDP
p38/ MAPK TGF-β
Figure 10 | Schematic overview of the inhibitory role of miR-15b on non-canonical TGF-β signaling
via Grb2, SOS1, SOS2 and the small GTPases. Binding of TGF-β to the TGFBR1/2 complex results in the recruitment of Grb2 and SOS1/2 to the receptor tyrosine kinase, whereafter downstream GTPases are activated by phosphorylation of GDP, turning it into GTP. Ultimately, this leads to the induction of the fibrogenic Mitogen-activated protein kinase kinase/Extracellular signal-regulated kinase (MEK/ERK), phosphoinositide 3-kinase/Akt (PI3K/Akt) and p38/Mitogen activated protein kinase (p38/MAPK) signal transduction pathways. The BMP7-mediated induction of miR-15b decreases the levels of Grb2, SOS1 and SOS2 and might thus inhibit non-canonical TGF-β signaling via MEK/ERK, PI3K/Akt and p38/MAPK.
These pathologies underline the importance of SOS1 and SOS2 signaling in fibrogenesis, yet their molecular mechanisms need further elucidation.
We show here that BMP7 treatment of cardiac fibroblasts can maintain the expression of miR-15b and partially abrogates inhibits TGF-β1-induced fibroblast activation. Previously, BMP7 has been shown to reduce cardiac23 and kidney fibrosis51 by the induction of Smad1/5/8 phosphorylation52, which antagonizes the action of TGF-β1-induced Smad 2/3 phosphorylation23,51. Here, we demonstrate that BMP7 might limit fibroblast activation due to the decreased expression of the small GTPase intermediates Grb2, SOS1 and SOS2. We did not dissect the mechanism of BMP7-induced miR-15b maintenance. A better understanding of how miR-15b levels are regulated by BMP7, could elucidate new regulatory mechanisms of non-canonical TGF-β signaling, which might be used to interfere in fibroblast activation and fibrogenesis.
In our experiments, we focused on the miR-15b mediated inhibition of non-canonical TGF-β signaling via the small GTPase intermediates Grb2 and SOS1 and SOS2 in cardiac fibroblast activation. In accordance with our findings, a decrease in miR-15b levels has been found in hepatic
fibroblast activation, which culminated in a decrease in apoptosis, as Bcl2 is a target of miR15b53. Also, in vascular smooth muscle cell activation, a decrease in miR-15b expression levels coincides with a decrease in the expression of Yes-associated Protein (YAP)54. Moreover, the inhibition of miR-15b after transverse aortic constriction in mice leads to an increase in cardiac fibrosis: in this particular study miR-15b was shown to target TGFBR1 and TGFBR2 in cardiac fibroblasts55. Based on these findings and our results, miR-15b might inhibit TGF-β-induced activation of cardiac fibroblasts, since the TGFBR1/2 complex initiates activation of both canonical and non-canonical TGF-β signaling. Thus, this would explain our observation that miR-15b precludes TGF-β1-induced cardiac fibroblast activation, even though we did not observe differences in activation of the small GTPase Ras. Concluding, upon the inhibition of miR-15b, cardiac fibroblasts acquire the ability to be activated by TGF-β via its canonical and non-canonical signal transduction pathways. Hence, in cardiac pathologies where fibrotic tissue accumulation is present (e.g. pressure overload induced cardiac hypertrophy), the expression of miR-15b in fibroblasts could inhibit fibroblast activation and might prevent the loss of cardiac function and the development of heart failure in the long run. In other cardiac injury models (i.e. in Ren2 hypertrophic cardiomyopathy55 and after coronary artery ligation56,57), 15b levels were increased during disease. In most of these studies, miR-15b is studied in the cellular context of cardiomyocytes. MiR-miR-15b inhibition in cardiomyocytes leads to a higher number of viable cardiomyocytes and a decreased infarct size56 whereas miR-15b overexpression leads to increased cardiomyocyte apoptosis57. Thus, the function of miR-15b is cell type and context dependent. Hence, in pathology where the cardiomyocyte apoptosis is found (e.g. myocardial infarction), low miR-15b expression levels may be favorable, whereas in pathologies where fibroblast activation leads to ECM accumulation (e.g. cardiac fibrosis) high miR-15b levels might be protective.
In conclusion, we have shown that miR-15b is decreased in fibrotic hearts and during TGF-β-induced fibroblast activation. Transfection of miR-15b mimics in TGF-β-stimulated cardiac fibroblasts results in a decrease in the expression of the small GTPase intermediates Grb2, SOS1 and SOS2 and precludes fibroblast activation. The decrease of miR-15b in cardiac fibrosis might contribute to cardiac fibrogenesis and the progression of cardiac failure and restoration of miR-15b expression levels might thus be used as a novel antifibrogenic therapy.
ACKNOWLEDGEMENTS
We thank Marja Brinker for assistance in Western Blot analyses and Linda Brouwer for help with immunofluorescent stainings.
REFERENCES
1. Wynn, T. Cellular and molecular mechanisms of fibrosis. The Journal of pathology 214, 199-210 (2008).
2. Tomasek, J.J., Gabbiani, G., Hinz, B., Chaponnier, C. & Brown, R.A. Myofibroblasts and mechano-regulation of connective tissue remodelling. Nature reviews. Molecular cell biology
3, 349 (2002).
3. Chaturvedi, R.R., et al. Passive stiffness of myocardium from congenital heart disease and implications for diastole. Circulation 121, 979-988 (2010).
4. Vasquez, C., Benamer, N. & Morley, G.E. The cardiac fibroblast: functional and electrophysiological considerations in healthy and diseased hearts. Journal of cardiovascular
pharmacology 57, 380 (2011).
5. Krenning, G., Zeisberg, E.M. & Kalluri, R. The origin of fibroblasts and mechanism of cardiac fibrosis. Journal of cellular physiology 225, 631-637 (2010).
6. Leask, A. Getting to the Heart of the Matter. Circulation research 116, 1269-1276 (2015). 7. Biernacka, A., Dobaczewski, M. & Frangogiannis, N.G. TGF-β signaling in fibrosis. Growth
factors 29, 196-202 (2011).
8. Heineke, J. & Molkentin, J.D. Regulation of cardiac hypertrophy by intracellular signalling pathways. Nature reviews Molecular cell biology 7, 589-601 (2006).
9. Sugden, P.H. & Clerk, A. Activation of the small GTP-binding protein Ras in the heart by hypertrophic agonists. Trends in cardiovascular medicine 10, 1-8 (2000).
10. Malumbres, M. & Barbacid, M. RAS oncogenes: the first 30 years. Nature Reviews Cancer 3, 459-465 (2003).
11. Zhang, Y.E. Non-Smad pathways in TGF-β signaling. Cell research 19, 128 (2009).
12. Grande, M.T. & López-Novoa, J.M. Fibroblast activation and myofibroblast generation in obstructive nephropathy. Nature Reviews Nephrology 5, 319-328 (2009).
13. Leask, A. & Abraham, D.J. TGF-β signaling and the fibrotic response. The FASEB Journal 18, 816-827 (2004).
14. Bechtel, W., et al. Methylation determines fibroblast activation and fibrogenesis in the kidney.
Nature medicine 16, 544-550 (2010).
15. Grande, M.T., et al. Deletion of H-Ras decreases renal fibrosis and myofibroblast activation following ureteral obstruction in mice. Kidney international 77, 509-518 (2010).
16. Winter, J., Jung, S., Keller, S., Gregory, R.I. & Diederichs, S. Many roads to maturity: microRNA biogenesis pathways and their regulation. Nature cell biology 11, 228-234 (2009).
17. Filipowicz, W., Bhattacharyya, S.N. & Sonenberg, N. Mechanisms of post-transcriptional regulation by microRNAs: are the answers in sight? Nature reviews genetics 9(2008).
18. Chaturvedi, P. & Tyagi, S.C. Epigenetic mechanisms underlying cardiac degeneration and regeneration. International journal of cardiology 173, 1-11 (2014).
19. Orenes-Piñero, E., et al. Role of microRNAs in cardiac remodelling: new insights and future perspectives. International journal of cardiology 167, 1651-1659 (2013).
20. Port, J.D. & Sucharov, C. Role of microRNAs in cardiovascular disease: therapeutic challenges and potentials. Journal of cardiovascular pharmacology 56, 444-453 (2010).
21. Lewis, B.P., Shih, I.-h., Jones-Rhoades, M.W., Bartel, D.P. & Burge, C.B. Prediction of mammalian microRNA targets. Cell 115, 787-798 (2003).
22. Betel, D., Wilson, M., Gabow, A., Marks, D.S. & Sander, C. The microRNA. org resource: targets and expression. Nucleic acids research 36, D149-D153 (2008).
23. Zeisberg, E.M., et al. Endothelial-to-mesenchymal transition contributes to cardiac fibrosis.
Nature medicine 13(2007).
24. Correia, A.C., Moonen, J.-R.A., Brinker, M.G. & Krenning, G. FGF2 inhibits endothelial– mesenchymal transition through microRNA-20a-mediated repression of canonical TGF-β signaling. J Cell Sci 129, 569-579 (2016).
25. Przybyt, E., Krenning, G., Brinker, M.G. & Harmsen, M.C. Adipose stromal cells primed with hypoxia and inflammation enhance cardiomyocyte proliferation rate in vitro through STAT3 and Erk1/2. Journal of translational medicine 11, 39 (2013).
26. Spiekman, M., et al. Adipose Tissue–Derived Stromal Cells Inhibit TGF-β1–Induced Differentiation of Human Dermal Fibroblasts and Keloid Scar–Derived Fibroblasts in a Paracrine Fashion. Plastic and reconstructive surgery 134, 699-712 (2014).
27. Attisano, L. & Wrana, J.L. Signal transduction by the TGF-β superfamily. Science 296, 1646-1647 (2002).
28. Massagué, J., Blain, S.W. & Lo, R.S. TGFβ signaling in growth control, cancer, and heritable disorders. Cell 103, 295-309 (2000).
29. Schuster, N. & Krieglstein, K. Mechanisms of TGF-ß-mediated apoptosis. Cell and tissue
research 307, 1-14 (2002).
30. Strutz, F., et al. TGF-β1 induces proliferation in human renal fibroblasts via induction of basic fibroblast growth factor (FGF-2). Kidney international 59, 579-592 (2001).
31. Grotendorst, G.R., Rahmanie, H. & Duncan, M.R. Combinatorial signaling pathways determine fibroblast proliferation and myofibroblast differentiation. The FASEB Journal 18, 469-479 (2004).
32. Ishikawa, O., LeRoy, E.C. & Trojanowska, M. Mitogenic effect of transforming growth factor β1 on human fibroblasts involves the induction of platelet-derived growth factor α receptors.
Journal of cellular physiology 145, 181-186 (1990).
33. Chu, W., et al. TGFBR3, a potential negative regulator of TGF-β signaling, protects cardiac fibroblasts from hypoxia-induced apoptosis. Journal of cellular physiology 226, 2586-2594 (2011).
34. Zhang, L., Dai, F., Cui, L., Zhou, B. & Guo, Y. Up-regulation of the active form of small GTPase Rab13 promotes macroautophagy in vascular endothelial cells. Biochimica et Biophysica Acta
(BBA)-Molecular Cell Research 1864, 613-624 (2017).
35. Modzelewska, K., et al. An activating mutation in sos-1 identifies its Dbl domain as a critical inhibitor of the epidermal growth factor receptor pathway during Caenorhabditis elegans vulval development. Molecular and cellular biology 27, 3695-3707 (2007).
36. Loirand, G., Scalbert, E., Bril, A. & Pacaud, P. Rho exchange factors in the cardiovascular system. Current opinion in pharmacology 8, 174-180 (2008).
37. Akhmetshina, A., et al. Rho-associated kinases are crucial for myofibroblast differentiation and production of extracellular matrix in scleroderma fibroblasts. Arthritis & Rheumatology
58, 2553-2564 (2008).
38. Reimann, T., et al. Transforming growth factor-β1 induces activation of Ras, Raf-1, MEK and MAPK in rat hepatic stellate cells. FEBS letters 403, 57-60 (1997).
39. Shi-Wen, X., et al. Rac inhibition reverses the phenotype of fibrotic fibroblasts. PLoS One 4, e7438 (2009).
40. Mucsi, I., Skorecki, K.L. & Goldberg, H.J. Extracellular signal-regulated kinase and the small GTP-binding protein, Rac, contribute to the effects of transforming growth factor-β1 on gene expression. Journal of Biological Chemistry 271, 16567-16572 (1996).
41. Ricupero, D.A., et al. Phosphatidylinositol 3-kinase-dependent stabilization of α1 (I) collagen mRNA in human lung fibroblasts. American Journal of Physiology-Cell Physiology 281, C99-C105 (2001).
42. Sato, M., et al. Role of p38 MAPK in transforming growth factor β stimulation of collagen production by scleroderma and healthy dermal fibroblasts. Journal of investigative
dermatology 118, 704-711 (2002).
43. Thum, T., et al. MicroRNA-21 contributes to myocardial disease by stimulating MAP kinase signalling in fibroblasts. Nature 456, 980-984 (2008).
44. Zhang, S., et al. The role of the Grb2–p38 MAPK signaling pathway in cardiac hypertrophy and fibrosis. Journal of Clinical Investigation 111, 833 (2003).
45. Gutkind, J.S. The pathways connecting G protein-coupled receptors to the nucleus through divergent mitogen-activated protein kinase cascades. Journal of Biological Chemistry 273, 1839-1842 (1998).
46. Rausch, O. & Marshall, C.J. Cooperation of p38 and extracellular signal-regulated kinase mitogen-activated protein kinase pathways during granulocyte colony-stimulating factor-induced hemopoietic cell proliferation. Journal of Biological Chemistry 274, 4096-4105 (1999).
47. Roberts, A.E., et al. Germline gain-of-function mutations in SOS1 cause Noonan syndrome.
Nature genetics 39, 70-74 (2007).
48. Yamamoto, G.L., et al. Rare variants in SOS2 and LZTR1 are associated with Noonan syndrome.
Journal of medical genetics 52, 413-421 (2015).
49. Nishikawa, T., et al. Hypertrophic cardiomyopathy in Noonan syndrome. Pediatrics
International 38, 91-98 (1996).
50. Hart, T.C., et al. A mutation in the SOS1 gene causes hereditary gingival fibromatosis type 1.
The American Journal of Human Genetics 70, 943-954 (2002).
51. Zeisberg, M., et al. BMP-7 counteracts TGF-β1–induced epithelial-to-mesenchymal transition and reverses chronic renal injury. Nature medicine 9, 964-968 (2003).
52. Manson, S.R., Niederhoff, R.A., Hruska, K.A. & Austin, P.F. The BMP-7–Smad1/5/8 Pathway Promotes Kidney Repair After Obstruction Induced Renal Injury. The Journal of urology 185, 2523-2530 (2011).
53. Guo, C.-J., Pan, Q., Li, D.-G., Sun, H. & Liu, B.-W. miR-15b and miR-16 are implicated in activation of the rat hepatic stellate cell: An essential role for apoptosis. Journal of hepatology 50, 766-778 (2009).
54. Xu, F., et al. MicroRNA-15b/16 attenuates vascular neointima formation by promoting the contractile phenotype of vascular smooth muscle through targeting YAP. Arteriosclerosis,
thrombosis, and vascular biology, ATVBAHA. 115.305748 (2015).
55. Tijsen, A.J., et al. The microRNA-15 family inhibits the TGFβ-pathway in the heart.
Cardiovascular research 104, 61-71 (2014).
56. Hullinger, T.G., et al. Inhibition of miR-15 Protects Against Cardiac Ischemic Injury. Novelty and Significance. Circulation research 110, 71-81 (2012).
57. Liu, L., et al. MicroRNA-15b enhances hypoxia/reoxygenation-induced apoptosis of cardiomyocytes via a mitochondrial apoptotic pathway. Apoptosis 19, 19-29 (2014).
