Abstract
Fibroblasts isolated from strictures in Crohn's disease (CD) exhibit reduced responsiveness to stimulation with transforming growth factor (TGF) β1. TGF-β1, acting through the smad pathway, is critical to fibroblast-mediated intestinal fibrosis. The membrane glycoprotein, endoglin, is a negative regulator of TGF-β1.
Intestinal fibroblasts were cultured from seromuscular biopsies of patients undergoing intestinal resection for CD strictures or from control patients. Endoglin expression was assessed using confocal microscopy, flow cytometry and western blot. The effect of small interfering (si) RNA-mediated knockdown and plasmid-mediated overexpression of endoglin on fibroblast responsiveness to TGF-β1 was assessed by examining smad phosphorylation, smad binding element (SBE) promoter activity, connective tissue growth factor (CTGF) expression and ability to contract collagen.
Crohn's stricture fibroblasts expressed increased constitutive cell-surface and whole-cell endoglin relative to control cells. Endoglin co-localized with filamentous actin. Fibroblasts treated with siRNA directed against endoglin exhibited enhanced TGF-β1-mediated smad-3 phosphorylation, and collagen contraction. Cells transfected with an endoglin plasmid did not respond to TGF-β1 by exhibiting SBE promoter activity or producing CTGF.
Fibroblasts from strictures in CD express increased constitutive endoglin. Endoglin is a negative regulator of TGF-β1 signalling in the intestinal fibroblast, modulating smad-3 phosphorylation, SBE promoter activity, CTGF production and collagen contraction.
Introduction
Crohn's disease (CD) results from a complex interplay between genetic, immunological and microbial factors1. More than one-third of patients with CD will develop a distinct fibrostenosing phenotype that manifests in recurrent intestinal stricture formation2. Strictures result from chronic transmural inflammation and dysregulated wound healing, leading to excessive and aberrant deposition of extracellular matrix3. Abnormal contraction of this extracellular matrix leads to scar formation, tissue distortion and ultimately intestinal obstruction.
Intestinal fibroblasts mediate stricture formation in CD4. The cytokine, transforming growth factor (TGF) β1, is critical to this process4. Fibroblasts isolated from CD strictures express increased constitutive growth factors (connective tissue growth factor (CTGF), vascular endothelial growth factor), produce excess collagen and exhibit an enhanced ability to contract and reorganize collagen in vitro5–8. Both TGF-β and its receptors are overexpressed in the intestine of patients with CD9,10. Intestinal fibroblast expression of TGF-β isoforms varies according to the nature of the native tissue. Fibroblasts from normal and inflamed CD mucosa express both TGF-β1 and TGF-β3 isoforms. Fibroblasts from fibrotic tissue show reduced expression of TGF-β3 with enhanced TGF-β1 and TGF-β2 production11. However, CD stricture fibroblasts exhibit a decreased magnitude of response to TGF-β1 relative to control cells, a finding that is contrary to expectation5,7.
Endoglin (cluster of differentiation105), a homodimeric membrane glycoprotein, is a type III TGF-β receptor that negatively regulates certain aspects of cellular responsiveness to exogenously added TGF-β112–14. Endoglin is expressed in stromal tissues in a variety of fibrotic tissue types and may itself modulate fibrosis15. It is overexpressed in biopsies of renal and hepatic fibrotic tissue16–18, as well as in fibroblasts isolated from the fibrotic skin of patients with scleroderma15, a syndrome characterized by systemic fibrosis. The role of endoglin in CD-associated intestinal fibrosis remains to be determined.
The aims of this study were to characterize endoglin expression in fibroblasts present in areas of strictured bowel in patients with CD and to explore the function of endoglin on TGF-β1-mediated intestinal fibroblast activation in vitro by using plasmid-mediated overexpression and interference RNA-based approaches.
Methods
The study protocol was approved by the Mater Misericordiae University Hospital Ethics Committee and written informed consent was obtained from all patients. Ten patients with CD (median age 37·5 (interquartile range (i.q.r.) 24·0–41·8) years) who were undergoing resection of ileal or ileocolonic strictures were recruited. All patients exhibited a fibrostenosing phenotype, were undergoing resection for obstruction and were infliximab naive at the time of surgery. Eight patients who were having intestinal resection for colorectal cancer (median age 68 (i.q.r. 55·0–81·3) years) were also recruited.
Materials
RPMI cell culture medium, penicillin, streptomycin, l-glutamine and fetal bovine calf serum (FBS) were purchased from Gibco Life Technologies (Paisley, UK). Human TGF-β1 was purchased from R&D Systems (Oxford, UK). Protein concentrations were determined using the Bio-Rad DC protein assay kit (Blessington, Ireland). Goat antihuman CTGF was purchased from Santa Cruz Biotechnology (Heidelberg, Germany). Rabbit antihuman smad-2 and -3, and phosphorylated (phospho-) smad-2 and -3 were purchased from Cell Signaling Technology (Boston, Massachusetts, USA). Mouse antihuman endoglin was purchased from BD Biosciences (Erembodegem, Belgium). A phycoerythrin-labelled antihuman endoglin antibody used for flow cytometry was purchased from eBioscience (Hatfield, UK). Rat-tail collagen type I was purchased from BD Biosciences. The chemiluminescent detection kit was from Amersham Biosciences (Amersham, UK). The AlexiFluor 488 FITC-conjugated antimouse IgG and Texas red-labelled phalloidin were purchased from Alexis Biochemicals (Lausen, Switzerland) and Lipofectamine™ 2000 transfection reagent from Invitrogen Life Technologies (Paisley, UK). Scrambled and endoglin-directed small interfering (si) RNA products were purchased from Ambion (Cambridge, UK). The pcEXV-Endo-L endoglin plasmid was a generous gift from Professor C. Bernabeu (Centro de Investigaciones Biologicas, Madrid, Spain). The smad binding element (SBE)–luciferase reporter, which contains four repeats of the CAGACA sequence and has been identified as a SBE in the Jun-B promoter19, was a generous gift from Professor B. Vogelstein (The Howard Hughes Medical Institute, The Johns Hopkins Kimmel Cancer Center, Baltimore, Maryland, USA). The dual luciferase reporter assay was from Promega (Southampton, UK). All other reagents were purchased from Sigma-Aldrich (Poole, UK).
Fibroblast isolation and culture
Intestinal fibroblast cultures were established using a primary explant technique as described previously10,20. Patients with CD had 1-cm2 serosal biopsies taken from sites of stricture and from adjacent, macroscopically normal, bowel (non-strictured) within the resection specimen. As a control, single serosal biopsies were taken from macroscopically normal colon at the edge of the resection specimen in patients undergoing resection for colorectal cancer. Confluent cells were characterized by their morphological features and immunohistochemical staining properties for vimentin, α-smooth muscle actin and desmin, as described previously21,22. For all experiments, fibroblasts were used between the second and fifth passages, and maintained in low-serum medium. The human fibroblast/myofibroblast cell line (CCD-18Co) was obtained from American Type Culture Collection (Rockville, Maryland, USA) and used between passages eight and 15 for plasmid transfection experiments. Cells were maintained in Dulbecco's modified Eagle's medium F12 containing 10 per cent FBS, 2 mmol/l glutamine and 1 per cent penicillin–streptomycin.
Cell-surface endoglin expression
Once confluent, fibroblasts were isolated by trypsinization, and expression of endoglin was assessed using an Epics XL-mcl cytofluorometer (Beckman Coulter, Galway, Ireland). Fibroblasts (5 × 103) in 100 µl medium were incubated with 5 µl phycoerythrin-labelled antihuman endoglin antibody at 4 °C for 30 min, then washed with RPMI and analysed. Results are expressed as the log of mean channel fluorescence.
Protein extraction and western blot analysis
Total protein was isolated from 2 × 106 fibroblasts using NP-40 isolation solution (0·5 per cent NP-40, 10 mmol/l Tris, pH 8·0, 60 mmol/l potassium chloride, 1 mmol/l EDTA, pH 8·0, 1 mmol/l dithiothreitol plus protease inhibitor mix) and western blot analysis was performed as described previously20. Some 30 µg was resolved on a 12 per cent sodium dodecyl sulphate gel (75 min at 140 V) before transfer at 100 V for 80 min. Blots were incubated with the appropriate primary antibody (concentrations: CTGF, 1 : 100; endoglin, 1 : 200; smad-2 and -3 and phospho-smad-2 and -3, 1 : 500). Equal protein loading was confirmed by β-actin or unphosphorylated protein expression. Protein density on scanned western blots was determined using the Un-Scan-It™ gel automated digitizing system, version 5.1 (Silk Scientific, Orem, Utah, USA).
Crystal violet assay assessment of cell number
Control intestinal fibroblasts were grown in 24-well plates to 70 per cent confluence, serum starved for 24 h, and then incubated with TGF-β1 (1 ng/ml) for 24 h. Cells were then fixed in 1 per cent gluteraldehyde solution, washed in phosphate-buffered saline (PBS), incubated with 1 per cent crystal violet, washed, solubilized with 1 per cent Triton X-100, and analysed at an absorbance of 590 nm as described previously23.
Confocal laser microscopic analysis
Cells were grown on two-chamber slides, fixed with 4 per cent paraformaldehyde in PBS for 30 min and permeabilized with 0·1 per cent Triton X-100 in PBS for 5 min. Cells were then blocked with 5 per cent normal goat serum in PBS for 1 h and left at room temperature in mouse antihuman endoglin (diluted 1 : 500). Cells were washed with PBS and left at room temperature for 40 min with Alexa Fluor 488 (green) goat antimouse IgG (diluted 1 : 1000). Confocal analysis and digital imaging was carried out using a Carl Zeiss LSM 510 UVMETA system (Welwyn Garden City, UK).
siRNA-mediated inhibition and plasmid-mediated upregulation of endoglin expression
Incorporation of siRNA (targeted against endoglin or scrambled control; optimal dose following dose response experiments was 5 nmol/l) was achieved using Lipofectamine™ 2000 (5 µl). The cells were plated in antibiotic-free medium for 24 h, followed by transfection for 6 h on the evening before experimentation. Endoglin overexpression was achieved using the pcEXV-Endo-L plasmid. Following dose–response experiments, CCD-18Co cells were treated with 5 µg plasmid DNA for 6 h on the evening before experimentation using Lipofectamine™ 2000. All transfections were carried out at a confluence of 80 per cent in six-well plates.
Smad binding element promoter luciferase assay
CCD-18Co cells diluted in 10 per cent FBS were adhered to the wells of a six-well plate for 24 h. The promoter–reporter SBE was transfected as described above. The pRL-CMV–Renilla luciferase reporter plasmid was co-transfected and its activity was used to correct for variation in transfection efficiency. Cells were treated with TGF-β1 for 48 h, then lysed; luciferase assays were performed as per the manufacturer's instructions.
Fibroblast-populated collagen lattice
Fibrosis progresses because of both increased extracellular matrix deposition and reorganization of matrix4. TGF-β1 augments fibroblast-mediated collagen gel contraction, an in vitro model of connective tissue remodelling, which has been shown to be a smad-3-dependent response24,25. Free-floating fibroblast-populated collagen lattices (FPCLs) were prepared in 24-well plates as described previously7. Fibroblasts were mixed with rat-tail collagen type I and RPMI containing 1 per cent FBS to give final concentrations of 6·25 × 104 fibroblasts/ml and 1·25 mg/ml collagen in a total volume of 0·5 ml. Incubation at 37 °C in 5 per cent carbon dioxide for 25 min caused the mixture to set, and the resultant gel was freed from the well by gentle manipulation with a pipette tip and refloated in 0·5 ml RPMI culture medium containing 1 per cent FBS and test substances. Medium was replaced on a daily basis. Contracting lattices were photographed at 24 h over a lightbox from a fixed height using a mounted digital camera (Powershot A70; Canon, Dublin, Ireland). Analysis of calibrated images allowed area determination in square centimetres (Image J Software, version 1.26t; National Institutes of Health, Bethesda, Maryland, USA).
Statistical analysis
Unless otherwise stated, data are expressed as mean(s.e.m.). For western blot experiments, densitometry values for protein bands were adjusted for β-actin expression, and statistical analysis was performed on the resultant values. For contraction studies, the raw data were analysed. The statistical software program SPSS® version 12.0 (SPSS, Chicago, Illinois, USA) was used. Statistical analysis between groups was determined using Student's t test or ANOVA as appropriate; P < 0·050 was considered statistically significant.
Results
Endoglin expression is upregulated in Crohn's stricture fibroblasts
Endoglin expression was determined in Crohn's stricture fibroblasts, in fibroblasts from adjacent non-stricturing areas and in fibroblasts isolated from controls. Whole-cell endoglin protein levels, as detected by western blotting, were increased in Crohn's stricture fibroblasts relative to control fibroblasts (32·77(11·69) versus 0·80(0·17) relative density (RD) respectively; P = 0·021) (Fig. 1a,b). Cell-surface endoglin levels followed a similar pattern. Surface endoglin was present at greater levels in non-stricture Crohn's fibroblasts relative to control fibroblasts (1·56(0·14) versus 0·87(0·11) respectively; P = 0·009), and was markedly overexpressed in Crohn's stricture fibroblasts relative to fibroblasts derived from adjacent, macroscopically normal, bowel (2·59(0·09) versus 1·56(0·14); P = 0·001) (Fig. 1c). Using confocal microscopy, endoglin was seen to co-localize to areas of filamentous actin accumulation within the fibroblast, and there was a greater expression of endoglin in Crohn's stricture fibroblasts relative to that in control fibroblasts (Fig. 2).
Intestinal fibroblast endoglin expression in Crohn's disease. a Western blot of endoglin concentration in fibroblasts cultured from normal colon (controls, lanes 1–6) and strictured bowel in patients with Crohn's disease (lanes 7–12). β-Actin concentration was measured to confirm equal protein loading. b Data from a expressed graphically as mean(s.e.m.) band densitometry of three independent experiments. *P = 0·021 (ANOVA). c Endoglin cell surface expression assessed by flow cytometry in fibroblasts cultured from strictured bowel, adjacent non-strictured, uninflamed bowel and controls. Data are expressed as mean(s.e.m.) log of the mean channel fluorescence (LnMCF) (n = 4 per group). *P = 0·009, †P = 0·001 (ANOVA)
Confocal microscopy images of endoglin cellular localization in fibroblasts from control and strictured bowel. Nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI), filamentous actin (F-actin) with Texas red, and endoglin with fluorescein isothiocyanate (FITC)
Suppression of endoglin expression enhances intestinal fibroblast sensitivity to TGF-β1
Endoglin levels were reduced using a targeted interference RNA-based approach. Endoglin-directed siRNA led to a reduction in whole-cell endoglin protein expression as detected using western blotting (Fig. 3a). In contrast, the same concentration of scrambled siRNA had no effect. TGF-β1-induced profibrotic responses were then compared between control and endoglin-deficient fibroblasts. TGF-β1 stimulation of scrambled siRNA-treated fibroblasts led to significant upregulation of CTGF protein (1·86(0·37) versus 0·08(0·01) RD in the absence of TGF- β1; P = 0·009) (Fig. 3b,c). CTGF production in endoglin-deficient fibroblasts exhibited a further, non-significant, increase following TGF-β1 stimulation (2·42(0·28) versus 1·86(0·37) RD; P = 0·294) (Fig. 3b,c). When FPCLs were stimulated with TGF-β1, contraction was enhanced (66·88(3·47) versus 88·83(2·30) per cent FPCL surface area at 24 h in controls; P = 0·003). However, endoglin-deficient fibroblasts exhibited enhanced basal (61·64(3·01) versus 88·83(2·30) per cent in controls; P = 0·001) and TGF-β1-stimulated (48·25(3·54) versus 66·88(3·47) per cent respectively; P = 0·048) collagen contraction relative to that in fibroblasts treated with a non-targeting siRNA (Fig. 3d). However, neither endoglin knockdown nor TGF-β1 stimulation promoted fibroblast proliferation (Fig. 3e).
Effect of targeted small interfering (si) RNA on endoglin expression and cellular responsiveness to transforming growth factor (TGF) β1. a Western blot of endoglin expression in intestinal fibroblasts transfected with siRNA directed against endoglin (si-Eng) or with scrambled siRNA (si-Sc). Equal loading was confirmed by analysis of β-actin. b Western blot of connective tissue growth factor (CTGF) expression in siRNA-transfected cells incubated with and without TGF-β1. c Data from b expressed graphically as mean(s.e.m.) band densitometry (n = 3 per group). *P = 0·009 (Student's t test). d Mean(s.e.m.) percentage reduction in surface area of fibroblast-populated collagen lattices prepared from siRNA-transfected cells and incubated with or without TGF-β1 for 24 h (n = 3 per group). *P = 0·001, †P = 0·003, ‡P = 0·048 (Student's t test). e Crystal violet assay of numbers of siRNA-transfected cells following incubation with or without TGF-β1. There were no differences between groups (ANOVA). f Western blot analysis of phosphorylated (phospho-) smad-2 and phospho-smad-3 in siRNA-transfected cells incubated with or without TGF-β1. Equal loading was confirmed by analysis of smad-2 and smad-3
When smad phosphorylation was examined in endoglin-deficient fibroblasts and scrambled siRNA-treated fibroblasts, stimulation with TGF-β1 led to an increase in the phosphorylation of both smads. However, smad-3 phosphorylation was increased even further in cells with reduced endoglin expression following stimulation with TGF-β1 (Fig. 3f). TGF-β1-induced smad-2 phosphorylation was unaltered in endoglin-deficient fibroblasts, relative to that observed in scrambled siRNA-treated fibroblasts.
Endoglin overexpression attenuates TGF-β1 responsiveness in the intestinal fibroblast
Transfection of the intestinal fibroblast cell line CCD-18Co with the pcEXV-Endo-L plasmid resulted in an upregulation of endoglin expression relative to mock transfection with pcDNA3 (Fig. 4a). In pcDNA3-transfected fibroblasts, TGF-β1 stimulation led to an increase in CTGF production (1·88(0·48) versus 0·03(0·01) RD with no TGF-β1 stimulation; P = 0·003) (Fig. 4b,c). This response was abrogated in endoglin-transfected cells (0·04(0·02) versus 1·88(0·48) RD; P = 0·003). The profibrotic actions of TGF-β1 are mediated by its ability to bind to, and induce transcription at, the SBE. TGF-β1 stimulation of intestinal fibroblasts enhanced SBE promoter activity at 48 h (0·16(0·03) versus 0·03(0·01) with no TGF-β1 stimulation; P = 0·004). However, cells in which endoglin was overexpressed did not exhibit an increase in SBE activity to the same extent following TGF-β1 stimulation (0·06(0·01) versus 0·16(0·03); P = 0·029) (Fig. 4d).
Effect of pcEXV-Endo-L plasmid transfection on endoglin expression and cellular responsiveness to transforming growth factor (TGF) β1. a Western blot of CCD-18Co cells transfected with pcDNA3 (control) or pcEXV-Endo-L (p-Eng) plasmids. Equal loading was confirmed by analysis of β-actin. b Western blot of connective tissue growth factor (CTGF) expression in CCD-18Co-transfected cells incubated with and without TGF-β1. c Data from b expressed graphically as mean(s.e.m.) band densitomety (n = 6 per group). *P = 0·003, †P = 0·003 (ANOVA). d Smad binding element (SBE) promoter activity in fibroblasts co-transfected with a SBE promoter–luciferase reporter construct and pcDNA3 or p-Eng, and incubated with TGF-β1 (n = 3 per group). Values are mean(s.e.m.) luciferase units, normalized to co-transfected CMV–Renilla luciferase. *P = 0·004, †P = 0·029 (ANOVA)
Discussion
Endoglin is a 180-kDa homodimeric membrane glycoprotein expressed on endothelial cells, monocytes, macrophages and fibroblasts13–15,26. It binds TGF-β1 and TGF-β3 with high affinity, but does not itself transduce signals intracellularly15,27,28. Endoglin plays a role in angiogenesis. Antibodies directed against endoglin are antiangiogenic29, and mice heterozygous for endoglin exhibit reduced angiogenic responses30. Finally, loss-of-function mutations in the human endoglin gene are associated with the vascular dysplasia, hereditary haemorrhagic telangiectasia type 131.
Endoglin has recently been shown to inhibit TGF-β1-mediated profibrotic activation of renal and hepatic fibroblasts in vitro. TGF-β1 exerts its profibrotic actions primarily through activation of the smad pathway32. In this pathway, the receptor-associated smads (2 and 3) are activated by C-terminus phosphorylation. Phospho-smads then heterodimerize with smad-432. This complex translocates to the nucleus where it binds to and regulates the expression of target genes at motifs termed SBEs. A number of reports have identified a negative regulatory role for endoglin in TGF-β signalling15,33. This is believed to be due in part to its differential effect on smad-2/3 phosphorylation34.
Endoglin negatively regulates the TGF-β/activin receptor-like kinase (ALK) 5/smad-3 signalling pathway, resulting in increased activity along the TGF-β/ALK-1/smad-1 pathway35. Knockdown of endoglin in keratinocytes enhances smad-2 and smad-3 phosphorylation36, whereas its overexpression in prostate cancer cells potentiates the actions of TGF-β1 on smad-1 phosphorylation but has little effect on smad-3 phosphorylation37. Furthermore, endoglin overexpression reduces the phosphorylation of smad-3 and inhibits its nuclear translocation in rat myoblasts38 and NIH3T3 fibroblasts15. The present findings are in keeping with these observations.
Endoglin expression was compared between fibroblasts from areas of stricture formation in CD, adjacent non-strictured bowel and normal colon. Whole-cell and cell-surface endoglin levels were greatest in stricture-associated fibroblasts. Endoglin levels were significantly increased in fibroblasts from adjacent non-strictured bowel, relative to levels observed in normal controls. Endoglin co-localized with filamentous actin, a pattern of distribution that has been observed in rat myoblasts, suggesting the cytoskeletal anchoring of endoglin39. This is a novel association between endoglin and CD-associated fibrogenesis. The upregulation of endoglin in the setting of CD is intriguing given that endoglin is a negative regulator of TGF-β1-mediated fibrotic responses in renal and hepatic models. As TGF-β1 plays a major role in CD-associated fibrogenesis, it is feasible that upregulated endoglin reflects an endogenous mechanism limiting the stimulatory effects of TGF-β1.
To investigate this association, endoglin levels were indirectly reduced using siRNA, after which responsiveness to TGF-β1 was assessed. Reduced endoglin expression alone led to enhanced collagen contraction, and stimulation of endoglin-deficient fibroblasts with TGF-β1 was followed by a further increase in contraction. The expression of CTGF following stimulation with TGF-β1 was not significantly different in cells treated with a scrambled control or a siRNA directed against endoglin. In contrast, in intestinal fibroblasts in which endoglin was overexpressed, TGF-β1-induced CTGF production was markedly reduced. Thus, very high levels of endoglin appear to have the potential to interrupt completely TGF-β1-mediated CTGF induction, but when endoglin expression is reduced the converse does not appear to hold true. These observations support the suggestion that endoglin inhibits TGF-β1-mediated signalling in intestinal fibroblasts.
The mechanism by which endoglin exerts the observed inhibitory effects was investigated. When compared with control fibroblasts, smad-3 phosphorylation was enhanced in endoglin-deficient fibroblasts following TGF-β1 stimulation. In contrast, smad-2 phosphorylation was unaltered. Phosphorylated smad complexes bind with and increase SBE transcriptional activity as part of the TGF-β1 intracellular signalling cascade. TGF-β1-mediated SBE promoter activity was abrogated completely when endoglin was overexpressed. Of note, smad-2 cannot bind to the SBE (owing to interference by a stretch of amino acids present immediately before the DNA-binding hairpin40,41), implying further specificity of endoglin for modulating smad-3 phosphorylation.
Smad-3 appears to be the key mediator of TGF-β1-mediated fibrogenesis4,20,32. Smad-3 null mice in which liver or intestinal fibrosis has been induced exhibit reduced tissue levels of CTGF and collagen relative to their wild-type controls42,43. In renal proximal tubule epithelial cells, TGF-β1-induced CTGF is a smad-3-dependent response44. In hepatocytes it has been demonstrated that CTGF is both a smad-2-dependent45 and a smad-3-dependent46 response. In the setting of intestinal fibroblasts, the effect of endoglin overexpression on SBE promoter activity and endoglin knockdown on smad-2/3 phosphorylation suggests that endoglin modulates TGF-β1-mediated smad-3 phosphorylation. This is highly significant given that smad-3 is regarded as the primary mediator of TGF-β1-associated fibrotic responses20.
Although fibroblasts isolated from CD strictures express increased constitutive growth factors such as CTGF, they also exhibit a reduced magnitude of response to stimulation with TGF-β15; endoglin overexpression may explain this phenomenon. However, although the regulatory effects of endoglin on TGF-β1 signalling may be beneficial in terms of CD-associated fibrosis, they could also be detrimental. TGF-β1 has important anti-inflammatory effects in CD47.
In summary, endoglin levels are increased in fibroblasts harvested from strictured intestine in CD. In fibroblasts deficient in endoglin, TGF-β1-mediated cellular responses are enhanced, but when endoglin is overexpressed TGF-β1 responsiveness is markedly reduced. These findings suggest that endoglin is a novel negative regulator of profibrotic TGF-β1 signalling in intestinal fibroblasts and may explain the reduced responsiveness of CD-associated fibroblasts to TGF-β1.
Acknowledgements
This work was supported in part by a grant from the Mater College for Postgraduate Education and Research and the Irish Research Council for Science, Engineering and Technology. The authors declare no conflict of interest.
References
Footnotes
Presented in part to the Sylvester O'Halloran Surgical Meeting, Limerick, Ireland, March 2006 and Digestive Disease Week, Washington, DC, USA, May 2007, and published in abstract form as Gastroenterology 2007; 132(Suppl 2): A239
