Expression and functional analysis of endoglin in isolated liver cells and its involvement in fibrogenic Smad signalling

doi:10.1016/j.cellsig.2010.12.002

Cellular Signalling

Volume 23, Issue 4, April 2011, Pages 683-699

https://doi.org/10.1016/j.cellsig.2010.12.002 Get rights and content

Abstract

Endoglin is an accessory component of the TGF-β-binding receptor complex that differentially modulates TGF-β and BMP responses. The existence of two splice variants L- and S-endoglin which differ in their cytoplasmic domain has already been shown in human and mice. Endoglin is located on the cell surfaces of cultured hepatic stellate cells and transdifferentiated myofibroblasts suggesting that this receptor might be associated with the profibrogenic attributes of these liver cell subpopulations. We now show that endoglin expression is increased in transdifferentiating hepatic stellate cells and in two models of liver fibrosis (i.e. bile duct ligation and carbon tetrachloride model) and further detectable in cultured portal fibroblasts representing another important fibrogenic cell type but not in hepatocytes. In respect to TGF-β1-signalling, we demonstrate that endoglin interacts with and is phosphorylated by TβRII. In hepatic stellate cells, TGF-β1 upregulates endoglin expression most likely via the ALK5 pathway and requires the SP1 transcription factor. We further identified a novel rat splice variant that is structurally and functionally different from that identified in human and mouse. Transient overexpression of endoglin resulted in a strong increase of TGF-β1-driven Smad1/5 phosphorylation and α-smooth muscle actin expression in a hepatic stellate cell line. In supernatants of respective cultures, we could detect the ectodomain of endoglin suggesting that shedding is a further key process involved in the regulation of this surface receptor.

Introduction

Endoglin (CD105) is a type III TGF-β co-receptor that is expressed as a disulfide-linked 180-kD homodimer in the plasma membrane of diverse cells. Of fundamental interest is the expression of endoglin in profibrogenic cells including mesangial cells, cardiac fibroblasts, scleroderma fibroblasts and hepatic stellate cells (HSC) [1], [2], [3], [4]. The glycosylated extracellular domain contains a putative protease cleavage site in human, mouse and rat [5], [6], which upon proteolysis by matrix metalloproteinase-14 (MT1-MMP) gives rise to a soluble, shedded form of endoglin [7]. This process is triggered under pathological conditions [8], [9], [10]. The single membrane spanning domain is connected to a short cytoplasmic domain (CD) that lacks catalytic activity [5], [6] but comprises several serine/threonine residues and a PDZ-domain [11] both of which regulate protein–protein interactions [12]. The retention of an intron in the CD results in a longer transcript, which is translated into a shorter protein, i.e. S-endoglin. This variant misses most of the CD domain including phosphorylation sites and protein–protein interaction domains [13]. Due to this modification, S-endoglin has a different role in cellular responses in comparison to the full-length receptor [14]. The occurrence of S-endoglin has been described in human and mouse but not in rat [13], [15].

The expression of endoglin is up-regulated under conditions of cellular activation, e.g. in endothelial cells (angiogenesis) and monocytes/macrophages [16], [17], and under hypoxic conditions [18]. In addition, angiotensin II and TGF-β1 are capable of inducing endoglin expression [2], [19]. Under basal conditions, endoglin expression is mainly influenced by the transcription factor specificity protein 1 (SP1) [20], whereas a network of SP1, KLF-6, Smad3 and Hif-1α is involved under hypoxic or TGF-β1-stimulated conditions [18], [21], [22]. The fact that TGF-β/Smad3 is involved in regulating endoglin promoter activity implies a regulatory feedback loop due to the ligand-binding spectrum of endoglin encompassing TGF-β1 and TGF-β3 in addition to the BMP-ligands BMP-2, BMP-7, and BMP-9 [23], [24], [25].

In contrast to betaglycan, binding of TGF-β1 to endoglin requires the presence of the corresponding type I and II receptors [23]. As a consequence, endoglin interacts with the TGF-β signalling receptor type II (TβRII) and the two type I receptors ALK5 and ALK1 [26], [27], [28]. Interaction with the signalling receptors involves specific N- and C-terminal domains of endoglin and results in the phosphorylation of its CD [27], [28], [29], [30]. TβRII phosphorylates distinct serine residues in endoglin [11], a modification which has been postulated to be lost upon internalization in endothelial cells [31]. Interestingly, ALK1 and ALK5 are known to act on endoglin-CD serine and threonine residues [11], [32] but since endoglin binds BMP-ligands (e.g. BMP-2 and BMP-9) it is also substrate for BMP-type I receptors (e.g. ALK2) [32], [33]. In turn, phosphorylation of the CD of endoglin influences the association with ALK1 and ALK5 and interaction with other signalling intermediates [11], [34], [35].

In the context of the afore-mentioned signalling pathways, the role of endoglin is extremely versatile involving both Smad-independent and -dependent mechanisms [34], [36], [37], [38]. In HSC, TGF-β1 activates two different Smad pathways involving Smad2/3 or Smad1/5. The Smad3 branch is clearly linked to fibrogenic signalling [39]. On the other hand it has been shown that activation and transcriptional activity of Smad1 increases during in vitro activation of HSC and in HSC isolated from carbon tetrachloride-intoxicated rats [40], [41]. The pivotal role of Smad1 in the process of cellular activation is implied by the finding that BMP-2 which is a genuine activator of Smad1/5 forces α-SMA expression in activated HSC [40]. Nevertheless, α-SMA expression is preserved on a Smad3 background [42], pointing on a vital role for Smad1/5, which are prominently activated by TGF-β1 [43]. In addition, the TGF-β1/ALK1/Smad1-mediated Id1 expression was shown to promote HSC activation via re-organization of α-SMA-fibres [43].

It has been shown that endoglin is up-regulated by and negatively regulates profibrogenic responses of TGF-β1 in several cell types [1], [2], [3], [44]. Since endoglin is highly expressed in HSC [4] representing the major profibrogenic cell type in the liver, we sought to get a more comprehensive view on the expression of endoglin in liver cells and to get a first glimpse on endoglin function and involvement in TGF-β1-signalling in culture-activated, transdifferentiating HSC.

We show here that endoglin is expressed in all liver cell types except hepatocytes and that endoglin is transiently up-regulated during transdifferentiation. Likewise we found an increased expression of endoglin in two animal models of liver fibrogenesis. In line, we could show an increased expression of endoglin mediated by TGF-β1 that involves SP1 and activated Smad3. Moreover, we identified a splice variant of L-endoglin which differs from the previously identified short form of mouse and rat. Both variants interact with and are phosphorylated by TβRII. In CFSC, which represents an immortalized activated HSC cell line that show a low endogenous endoglin expression, we could demonstrate that overexpression of endoglin results in an overall stronger TGF-β1-mediated activation of Smad1/Smad5 and increased expression of α-SMA. Finally we could demonstrate that endoglin is shedded from the surface of CFSC giving rise to a soluble receptor.

Section snippets

Animal treatment

To induce liver fibrosis in male Sprague–Dawley rats, the common bile duct was double ligated (BDL) or CCl₄ was repeatedly administrated as described in detail elsewhere [45], [46].

Cell culture

HSC, Kupffer cells (KC) and liver sinusoidal endothelial cells (LSEC) were isolated from male Sprague–Dawley rats by the pronase–collagenase method followed by a density centrifugation on a Nycodenz gradient as described before [47], [48]. HSC were passaged once to obtain myofibroblast-like cells (MFB). Primary

Endoglin expression in isolated liver cells

HSC and MFB express the two TβRIII receptors betaglycan and endoglin (CD105), both promiscuously detected by the antibody sc-6199 directed against an epitope of the betaglycan C-terminus [4]. To analyze endoglin in the liver in more detail, we generated two independent antibodies, i.e. PPabE1 and PPabE2 that are directed against different epitopes within the extracellular domain of rat endoglin (Fig. 2A) and are not cross-reactive with other TGF-β-receptors (Fig. 2B) and are highly specific for

Discussion

In the present study we demonstrate that endoglin is expressed in all isolated liver cells except hepatocytes. Endoglin expression is transiently up-regulated during HSC transdifferentiation and as a mutual consequence in two models of liver fibrogenesis, i.e. BDL and CCl₄ administration. The increase in receptor expression in HSC is at least partly regulated by TGF-β1 itself, since this ligand induces endoglin expression in a time- and concentration-dependent manner. This regulation involves

Conclusions

Our study shows that endoglin is expressed in all profibrogenic liver cells and that its expression and surface exposure is increased during activation and transdifferentiation of HSC. TGF-β1 upregulates endoglin expression most likely via the ALK5 pathway and requires the SP1 transcription factor. Endoglin interacts with TβRII and enhances Smad1/5 phosphorylation and strongly increases the expression of α-SMA representing a marker of HSC activation. We further identified a novel rat splice

Acknowledgements

The authors thank Andreas Ludwig (Institute of Pharmacology and Toxicology, RWTH University Hospital Aachen, Germany) for kindly providing HUVEC cells, M. Rojkind (Marion Bessin Liver Research Center, Albert Einstein College of Medicine, Bronx, New York, USA) for CFSC-2G, Scott L. Friedman (Division of Liver Diseases, Mount Sinai School of Medicine, New York, NY 10029, USA) for LX-2 cells, M. Cristina Cardoso (Max Delbrück Center for Molecular Medicine, Berlin, Germany) for L₆E₉ cells, and

References (88)

L. Diez-Marques et al.
Biochim. Biophys. Acta
(2002)
S.K. Meurer et al.
J. Biol. Chem.
(2005)
A. Gougos et al.
J. Biol. Chem.
(1990)
O. Llorca et al.
J. Mol. Biol.
(2007)
R.I. Koleva et al.
J. Biol. Chem.
(2006)
B.A. Conley et al.
J. Biol. Chem.
(2004)
J.R. Westphal et al.
J. Invest. Dermatol.
(1993)
T. Sánchez-Elsner et al.
J. Biol. Chem.
(2002)
L.M. Botella et al.
J. Biol. Chem.
(2001)
L.M. Botella et al.
Blood
(2002)

N.P. Barbara et al.

J. Biol. Chem.

(1999)

A. Letamendía et al.

J. Biol. Chem.

(1998)

S. Cheifetz et al.

J. Biol. Chem.

(1992)

M. Guerrero-Esteo et al.

J. Biol. Chem.

(2002)

H. Yamashita et al.

J. Biol. Chem.

(1994)

H. Tang et al.

Blood

(2005)

N.Y. Lee et al.

J. Biol. Chem.

(2007)

N.Y. Lee et al.

J. Biol. Chem.

(2008)

O. Scherner et al.

J. Biol. Chem.

(2007)

B. Schnabl et al.

Hepatology

(2001)

E. Borkham-Kamphorst et al.

Lab. Invest.

(2004)

E. Borkham-Kamphorst et al.

Lab. Invest.

(2008)

H. Fehrenbach et al.

Hepatology

(2001)

P.O. Seglen

Meth. Cell Biol.

(1976)

S. Vogel et al.

J. Lipid Res.

(2000)

B. Nadal-Ginard

Cell

(1978)

H.Y. Lin et al.

Cell

(1992)

R. Weiskirchen et al.

Biochem. Biophys. Res. Commun.

(2000)

S.L. Friedman

Gastroenterology

(2008)

Y.Q. Wang et al.

Biochem. Biophys. Res. Commun.

(2004)

K. Miao et al.

Arch. Biochem. Biophys.

(1997)

S. Gaidarova et al.

J. Biol. Chem.

(2002)

C. Ríus et al.

Blood

(1998)

F. Sanz-Rodriguez et al.

J. Biol. Chem.

(2004)

Q. Meng et al.

J. Biol. Chem.

(2006)

S. Dooley et al.

FEBS Lett.

(2001)

K.M. Koli et al.

J. Biol. Chem.

(1997)

R.G. Wells et al.

J. Biol. Chem.

(1997)

R.G. Wells et al.

J. Biol. Chem.

(1999)

J. Pannu et al.

J. Biol. Chem.

(2007)

K.W. Finnson et al.

Osteoarthritis Cartilage

(2010)

K. Chen et al.

Circ. Res.

(2004)

A. Leask et al.

Arthritis Rheum.

(2002)

L.J. Hawinkels et al.

Cancer Res.

(2010)

Cited by (53)

Endoglin and soluble endoglin in liver sinusoidal endothelial dysfunction in vivo
2024, Biochimica et Biophysica Acta - Molecular Basis of Disease
Liver sinusoidal endothelial cells (LSECs) play a crucial role in regulating the hepatic function. Endoglin (ENG), a transmembrane glycoprotein, was shown to be related to the development of endothelial dysfunction. In this study, we hypothesized the relationship between changes in ENG expression and markers of liver sinusoidal endothelial dysfunction (LSED) during liver impairment.
Male C57BL/6J mice aged 9–12 weeks were fed with 3,5-diethoxycarbonyl-1,4-dihydrocollidine (DDC) diet (intrahepatic cholestasis) or choline-deficient l-amino acid defined high-fat diet (CDAA-HFD) (non-alcoholic steatohepatitis (NASH)).
Significant increases in liver enzymes, fibrosis, and inflammation biomarkers were observed in both cholestasis and NASH. Decreased p-eNOS/eNOS and VE-cadherin protein expression and a significant increase in VCAM-1 and ICAM-1 expression were detected, indicating LSED in both mouse models of liver damage. A significant reduction of ENG in the DDC-fed mice, while a significant increase of ENG in the CDAA-HFD group was observed. Both DDC and CDAA-HFD-fed mice showed a significant increase in MMP-14 protein expression, which is related to significantly increased levels of soluble endoglin (sENG) in the plasma.
In conclusion, we demonstrated that intrahepatic cholestasis and NASH result in an altered ENG expression, predominantly in LSECs, suggesting a critical role of ENG expression for the proper function of liver sinusoids. Both pathologies resulted in elevated sENG levels, cleaved by MMP-14 expressed predominantly from LSECs, indicating sENG as a liver injury biomarker.
High soluble endoglin levels regulate cholesterol homeostasis and bile acids turnover in the liver of transgenic mice
2019, Life Sciences
Citation Excerpt :
Soluble endoglin (sEng) is free form of membrane bound Endoglin (CD 105) that is a transmembrane glycoprotein acting as coreceptor of the transforming growth factor-β (TGF-β) superfamily, which is expressed in the plasma membrane of diverse cells (e.g in endothelial cells, macrophages, fibroblasts or hepatic stellate cells) [1,2].
Increased plasma soluble endoglin concentrations (sEng) are frequently detected in metabolic disorders accompanied with hypercholesterolemia in serum, but effect of sEng on the cholesterol biochemistry is unknown. Cholesterol and bile acids (BA) are important products of liver metabolism with numerous functions within the organism. Turnover of these substances requires precise regulation due to potential toxicities during their cumulation. In this study, we hypothesized that high sEng levels affect cholesterol homeostasis and BA turnover in mice liver.
Nine-month-old transgenic male mice overexpressing human sEng and wild-type mice underwent plasma, bile, stool, and organ samples analysis by analytical, qRT-PCT and Western blot methods.
sEng mice demonstrated decreased plasma total and LDL cholesterol concentrations due to upregulation of hepatic Sr-b1 and Ldlr receptors, increased liver cholesterol content, and increased Abcg8-mediated cholesterol efflux into bile. sEng also increased conversion of cholesterol into bile acids (BA) via upregulation of Cyp7a1 and increased Mdr1 expression. Plasma concentrations of BA were increased in sEng mice due to their enhanced reabsorption via ileum. Increased hepatic disposition of BA led to their increased biliary excretion coupled with choleretic activity.
For the first time, we have shown that high sEng plasma levels affect cholesterol and BA homeostasis on the basis of complex liver and intestinal effects. The significance of these findings for pathophysiology of diseases associated with increased sEng concentrations remains to be elucidated in prospective studies.
Identification of bone morphogenetic protein 9 (BMP9) as a novel profibrotic factor in vitro
2016, Cellular Signalling
Citation Excerpt :
This pathway plays an anti-fibrotic role in different experimental models of renal fibrosis [8,10,21,22] or in in vitro experiments [23,24]. However, in other biological contexts, Smad1 activation promotes ECM protein synthesis in liver cells [25,26], glomerular mesangial cells [27,28] or dermal fibroblasts [29–31]. BMP9 also inhibits angiogenesis in different conditions [18,19,32].
Upregulated synthesis of extracellular matrix (ECM) proteins by myofibroblasts is a common phenomenon in the development of fibrosis. Although the role of TGF-β in fibrosis development has been extensively studied, the involvement of other members of this superfamily of cytokines, the bone morphogenetic proteins (BMPs) in organ fibrosis has given contradictory results. BMP9 is the main ligand for activin receptor-like kinase-1 (ALK1) TGF-β1 type I receptor and its effect on fibrosis development is unknown. Our purpose was to study the effect of BMP9 in ECM protein synthesis in fibroblasts, as well as the involved receptors and signaling pathways.
In cultured mice fibroblasts, BMP9 induces an increase in collagen, fibronectin and connective tissue growth factor expression, associated with Smad1/5/8, Smad2/3 and Erk1/2 activation. ALK5 inhibition with SB431542 or ALK1/2/3/6 with dorsomorphin-1, inhibition of Smad3 activation with SIS3, and inhibition of the MAPK/Erk1/2 with U0126, demonstrates the involvement of these pathways in BMP9-induced ECM synthesis in MEFs. Whereas BMP9 induced Smad1/5/8 phosphorylation through ALK1, it also induces Smad2/3 phosphorylation through ALK5 but only in the presence of ALK1.
Summarizing, this is the first study that accurately identifies BMP9 as a profibrotic factor in fibroblasts that promotes ECM protein expression through ALK1 and ALK5 receptors.
Overexpression of the short endoglin isoform reduces renal fibrosis and inflammation after unilateral ureteral obstruction
2016, Biochimica et Biophysica Acta - Molecular Basis of Disease
Citation Excerpt :
Surprisingly, total Smad1 levels were also lower in O than in NO kidneys in both S-ENG+ and WT mice (Fig. 8b). Eng is expressed in many profibrogenic cells as mesangial cells and renal fibroblasts of the kidney, as well as in skin, intestinal and cardiac fibroblasts, and in cultured portal fibroblasts and hepatic stellate cells of the liver [14,31]. Eng upregulation has been observed in several experimental models of renal fibrosis [12,32].
Transforming growth factor beta 1 (TGF-β1) is one of the most studied cytokines involved in renal tubulo-interstitial fibrosis, which is characterized by myofibroblast abundance and proliferation, and high buildup of extracellular matrix in the tubular interstitium leading to organ failure. Endoglin (Eng) is a 180-kDa homodimeric transmembrane protein that regulates a great number of TGF-β1 actions in different biological processes, including ECM synthesis. High levels of Eng have been observed in experimental models of renal fibrosis or in biopsies from patients with chronic kidney disease. In humans and mice, two Eng isoforms are generated by alternative splicing, L-Eng and S-Eng that differ in the length and composition of their cytoplasmic domains. We have previously described that L-Eng overexpression promotes renal fibrosis after unilateral ureteral obstruction (UUO). However, the role of S-Eng in renal fibrosis is unknown and its study would let us analyze the possible function of the cytoplasmic domain of Eng in this process. For this purpose, we have generated a mice strain that overexpresses S-Eng (S-ENG⁺) and we have performed an UUO in S-ENG⁺ and their wild type (WT) control mice. Our results indicate that obstructed kidney of S-ENG⁺ mice shows lower levels of tubulo-interstitial fibrosis, less inflammation and less interstitial cell proliferation than WT littermates. Moreover, S-ENG⁺ mice show less activation of Smad1 and Smad2/3 pathways. Thus, S-Eng overexpression reduces UUO-induced renal fibrosis and some associated mechanisms. As L-Eng overexpression provokes renal fibrosis we conclude that Eng-mediated induction of renal fibrosis in this model is dependent on its cytoplasmic domain.
TGF-β/BMP proteins as therapeutic targets in renal fibrosis. Where have we arrived after 25 years of trials and tribulations?
2015, Pharmacology and Therapeutics
The understanding of renal fibrosis in chronic kidney disease (CKD) remains as a challenge. More than 10% of the population of developed countries suffer from CKD. Proliferation and activation of myofibroblasts and accumulation of extracellular matrix proteins are the main features of kidney fibrosis, a process in which a large number of cytokines are involved. Targeting cytokines responsible for kidney fibrosis development might be an important strategy to face the problem of CKD.
The increasing knowledge of the signaling pathway network of the transforming growth factor beta (TGF-β) superfamily members, such as the profibrotic cytokine TGF-β1 or the bone morphogenetic proteins (BMPs), and their involvement in the regulation of kidney fibrosis, has stimulated numerous research teams to look for potential strategies to inhibit profibrotic cytokines or to enhance the anti-fibrotic actions of other cytokines. The consequence of all these studies is a better understanding of all these canonical (Smad-mediated) and non-canonical signaling pathways. In addition, the different receptors involved for signaling of each cytokine, the different combinations of type I–type II receptors, and the presence and function of co-receptors that can influence the biological response have been also described. However, are these studies leading to suitable strategies to block the appearance and progression of kidney fibrosis?
In this review, we offer a critical perspective analyzing the achievements using the most important strategies developed up till now: TGF-β antibodies, chemical inhibitors of TGF-β receptors, miRNAs and signaling pathways and BMP agonists with a potential role as therapeutic molecules against kidney fibrosis.
PDGF-D signaling in portal myofibroblasts and hepatic stellate cells proves identical to PDGF-B via both PDGF receptor type α and β
2015, Cellular Signalling
Citation Excerpt :
Thereafter, the medium was supplemented to 10% FCS. Upon culturing, pF differentiated into pMF that had spindle morphology and expressed fibulin-2 [15], α-SMA but were negative for desmin [16]. Primary HSC were isolated from male Sprague–Dawley rats through density gradient centrifugation technique and cultured as described [17,18].
Platelet-derived growth factor-D (PDGF-D) is one member of PDGF growth factors and known to signal by binding to and activating its cognate receptor type β (PDGFR-β). Beside PDGF-B, PDGF-D is a potent growth factor for stellate cell growth and proliferation and therefore potentiates the extracellular matrix deposition in liver fibrogenesis. We aimed to explore the signaling and molecular mechanisms of PDGF-D in liver fibrogenesis using the primary liver portal myofibroblasts and hepatic stellate cells. Unexpectedly we found PDGF-D to bind to PDGFR-α, thus inducing receptor endocytosis and decreasing the amount of PDGFR-α significantly. PDGF-D activates PDGFR-α specific tyrosine 754 and -1018 phosphorylation and CrkII, the adaptor protein that is specifically recruited by activated PDGFR-α. As a novel finding we could also demonstrate that recombinant PDGFR-α-Fc chimera homodimer is able to bind PDGF-D and thus prevent PDGF-D signaling. PDGF-D does induce individual PDGFR-β specific tyrosine phosphorylation similar to the PDGF-B. Additionally, PDGF-D enhances extracellular matrix accumulation comparable to the PDGF-B isoform.
PDGF-D signaling in pMF and HSC is identical to that of PDGF-B by binding to both PDGFR-α and -β.

View all citing articles on Scopus

View full text

Expression and functional analysis of endoglin in isolated liver cells and its involvement in fibrogenic Smad signalling

Abstract

Introduction

Section snippets

Animal treatment

Cell culture

Endoglin expression in isolated liver cells

Discussion

Conclusions

Acknowledgements

Biochim. Biophys. Acta

J. Biol. Chem.

J. Biol. Chem.

J. Mol. Biol.

J. Biol. Chem.

J. Biol. Chem.

J. Invest. Dermatol.

J. Biol. Chem.

J. Biol. Chem.

Blood

J. Biol. Chem.

J. Biol. Chem.

J. Biol. Chem.

J. Biol. Chem.

J. Biol. Chem.

Blood

J. Biol. Chem.

J. Biol. Chem.

J. Biol. Chem.

Hepatology

Lab. Invest.

Lab. Invest.

Hepatology

Meth. Cell Biol.

J. Lipid Res.

Cell

Cell

Biochem. Biophys. Res. Commun.

Gastroenterology

Biochem. Biophys. Res. Commun.

Arch. Biochem. Biophys.

J. Biol. Chem.

Blood

J. Biol. Chem.

J. Biol. Chem.

FEBS Lett.

J. Biol. Chem.

J. Biol. Chem.

J. Biol. Chem.

J. Biol. Chem.

Osteoarthritis Cartilage

Circ. Res.

Arthritis Rheum.

Cancer Res.