Up-regulation of endoglin, a TGF-b-binding protein, in rats with ...

Nephrol Dial Transplant (2001) 16 [Suppl 1]: 34±39

Up-regulation of endoglin, a TGF-b-binding protein, in rats with experimental renal ®brosis induced by renal mass reduction Ana RodrõÂguez-PenÄ a, Marta Prieto, Annette Duwel, Juan V. Rivas, Ne lida Eleno, Fernando Pe rez-Barriocanal, Miguel Are valo1, Joshua D. Smith2, Calvin P. H. Vary2, Carmelo Bernabeu3 and Jose M. Lo pez-Novoa Instituto `Reina SofõÂa' de InvestigacioÂn NefroloÂgica, Departamento de FisiologõÂa y FarmacologõÂa, 1 Departamento de AnatomõÂa e HistologõÂa Humanas, Universidad de Salamanca, Salamanca, 3Centro de Investigaciones BioloÂgicas, Consejo Superior de Investigaciones Cientõ®cas, Madrid, Spain and 2The Center for Molecular Medicine, Maine Medical Center Research Institute, South Portland, Maine, USA

Abstract Background. The central process in chronic renal failure is the progressive accumulation of extracellular matrix in the glomeruli and in the tubulo-interstitial space, resulting in renal ®brosis. Transforming growth factor-b1 (TGF-b1) up-regulation plays a major role in the genesis of renal ®brosis. Endoglin is a membrane glycoprotein that binds TGF-b1 and TGF-b3 with high af®nity. An increased level of endoglin immunostaining has been demonstrated previously in biopsies from patients with chronic progressive renal disease. We have assessed the expression of endoglin in the rat 5u6th renal mass reduction (RMR) model. Methods. One, 3 and 5 months after RMR, mean arterial pressure and renal function were measured, animals were sacri®ced, renal ®brosis was evaluated quantitatively and the expression of endoglin was assessed by western blot, northern blot and immunohistochemistry. Results. RMR induced a progressive increase in mean arterial pressure and urinary protein excretion. Renal corpuscular area, and mesangial and interstitial ®brosis increased with time after RMR. Immunohistochemical staining for endoglin demonstrated its expression mainly on the endothelial surface of major vessels. In kidneys 1 and 3 months after RMR, the expression of endoglin in renal corpuscles was limited to Bowman's parietal epithelium. In rats 5 months after RMR, the immunoexpression in glomerular endothelium was more marked. Northern blot analysis revealed that rats with RMR showed an increase in the expression of mRNA for endoglin, only at 5 months after RMR. Western blot analysis gave a different time

Correspondence and offprint requests to: J. M. LoÂpez-Novoa, Departamento de FisiologõÂa y FarmacologõÂa, Edi®cio Departamental, Campus Miguel de Unamuno, 37007 Salamanca, Spain. A. RodrõÂguez-PenÄa and M. Prieto contributed equally to this work. #

course: a marked increase in the ®rst month, a decrease in the 3rd month and a further increase in the 5th month after RMR. Conclusions. The present study demonstrates increased endoglin expression in rats with severe hypertension and renal damage. This increased endoglin expression coincides with the period of higher renal damage and renal dysfunction. Keywords: chronic renal failure; endoglin; glomerulosclerosis; renal mass reduction; transforming growth factor-b; tubulo-interstitial ®brosis

Introduction The central process in chronic renal failure is the progressive accumulation of extracellular matrix in the glomeruli and in the tubulo-interstitial space. Extracellular matrix is subjected to continued remodelling, the rate of synthesis and degradation being similar under normal conditions. Accumulation of extracellular matrix occurs as a result of increased synthesis, decreased degradation or some combination of both processes. Transforming growth factor-b (TGF-b) is a member of a large family of proteins that has many biological effects including regulation of cellular proliferation, differentiation and migration, extracellular matrix formation, and modulation of the immune response w1,2x. TGF-b seems to play a major role in the genesis of renal ®brosis. First, the up-regulation of TGF-b expression in several models of experimental renal ®brosis, as well as the expression of both type I and type II receptor mRNA has been reported w3x. In addition, renal ®brosis has been improved by the administration of antibodies against TGF-b, by the administration of antisense oligonucleotides or by transfection with decorin, a

2001 European Renal Association±European Dialysis and Transplant Association

Endoglin and renal ®brosis

soluble molecule that can bind and inactivate TGF-b1 w4±6x. Endoglin is a 180 kDa homodimeric membrane glycoprotein expressed by human endothelial cells w7x, macrophages w8x, vascular smooth muscle cells and other cell types w9x. The gene encoding endoglin has been identi®ed as the target gene for the autosomal dominant vascular disorder known as hereditary haemorrhagic telangiectasia type 1 w10x. Endoglin binds TGF-b1 and TGF-b3 with high af®nity in human endothelial cells w11x. Recently it has been demonstrated that endoglin is up-regulated in biopsies from patients with chronic progressive renal disease w12x. However, in that study, endoglin was only detected by immunohistochemistry, and no detailed study of protein and mRNA expression or the time course of up-regulation was performed. The purpose of the present study was to assess the expression of endoglin by western blot, northern blot and immunohistochemistry in a model of renal ®brosis associated with hypertension and renal mass reduction (RMR), the rat 5u6th nephrectomy model.

Materials and methods Disease model and experimental design Experiments were performed in male Wistar rats weighing ;250 g. Rats were housed in standard cages in a daylight(12 h), temperature- (208C) and humidity- (60%) controlled room. Rats were allowed free access to food (standard rat chow containing 20% protein by weight) and water. Rats were anaesthetized with 2-2-2 tribromoethanol and subjected to 5u6th RMR by ligation of two or three branches of the left renal artery and right uninephrectomy, as previously reported w13x. After 2 days, the survival rate was 60%. Veri®cation of RMR was con®rmed by elevated blood pressure, increased values for urinary protein excretion and visual inspection of the remnant kidney at the end of the study. Sham-operated animals were used as controls. Each month after surgery, systolic and diastolic arterial pressure were measured in awake animals with a tail-cuff method (Electro-sphygmomanometer, Letica LE 5000, Letica Barcelona, Spain). Then, body weight was measured and rats were placed into metabolic cages with free access to food and water during four equilibration days. On the two following days, urine was collected free of food and faeces. A sample of blood (0.15 ml) was also obtained from a cut in the tail tip collected into heparinized capillaries. Plasma and urine creatinine levels were measured by a colorimetric method based on the Jaffe reaction. Endogenous creatinine clearance was used to estimate the glomerular ®ltration rate (GFR). Urinary protein excretion was determined by Bradford's method. Surviving RMR animals were distributed randomly into three groups: group 1, sacri®ced 1 month after surgery; group 2, sacri®ced 3 months after surgery; and group 3, sacri®ced 5 months after surgery.

Morphological and morphometric studies For this purpose, animals were anaesthetized with ether and the kidney was perfused with heparinized isotonic saline

35

to clear the kidney of blood. A piece of the kidney, including cortex and medulla, was trimmed and ®xed by immersion in 4% buffered formalin for 24 h. The blocks thus obtained were dehydrated in a graded series of ethanol washes and embedded in paraf®n. Sections 3 mm thick were cut, mounted on glass slides and counterstained with either haematoxylin±eosin or Masson's trichrome for light microscopy analysis. For quantitative determination of glomerular and interstitial ®brosis, 5 mm sections were stained with syrius red. Images were captured using a high-resolution videocamera (SONY ccd-iris) connected to a light microscope (Leitz Laborlux S) using the 20 3 objective and a green optical ®lter (IF 550). The evaluation and image analysis procedures were performed with speci®c software (Fibrosis HR1 Master Diagnostica, Granada, Spain) as previously reported w13x. To increase the precision of this method further, the same person captured the images in a blind manner. A total of 25 glomerular and 10 interstitial images for each slide were captured and processed, in six rats per group. For each image, the values, obtained in mm2 were: (i) glomerular ®brosis area; (ii) glomerular section area; and (iii) tubulo-interstitial ®brosis area. To assess the variability of the technique, a single glomerular image was captured and processed 20 times. Coef®cients of variation for the parameters analysed were: glomerular area, 0.19%; intraglomerular ®brosis, 1.52%; and tubulo-interstitial ®brosis, 0.78%. Renal endoglin expression was detected in 3 mm coated sections using a rabbit polyclonal anti-endoglin antibody w14x, followed by a standard avidin±biotin complex method (Santa Cruz Biotechnology, CA), as previously described w15x.

Northern and western blot Renal endoglin expression was assessed by northern blot analysis. Total RNA was isolated from rat homogenates using the guanidinium thiocyanate±phenol±chloroform method. The probe used was a 360 bp SacI±SacII fragment of rat endoglin cDNA in pGEM-T. A probe for glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal loading control. Endoglin expression was also measured by western blot in renal plasma membranes obtained by differential centrifugation. The crude membrane preparation was resuspended in lysis buffer (140 mM NaCl, 10 mM EDTA, 10% glycerol, 20 mM Tris, pH 8.0 containing 100 Uuml aprotinin, 2 mM phenylmethylsulfonyl ¯uoride, 60 mg of soy bean tripsin inhibitor and 1% NP-40). A sample containing 150 mg of protein was mixed with the same volume of 2 3 sample buffer (125 mM Tris, pH 6.8, 4% SDS, 20% glycerol, 0.01% bromophenol blue). Proteins were separated on a 10% SDS±polyacrylamide gel and transferred to a nitrocellulose membrane (0.45 mm pore) at 400 mA for 2 h in a buffer containing 20 mM Tris and 190 mM glycine, pH 8.3. Non-speci®c binding blockade was performed overnight in blocking buffer (20 mM Tris containing 150 mM NaCl and 0.1% Tween-20) to which 8% dry milk was added. The membrane was then rinsed with blocking buffer, and incubated for 1 h in standard solution containing rabbit antiendoglin polyclonal antiserum w14x at 1 : 1000 dilution. The membrane was washed four times for 10 min in blocking buffer and incubated for 30 min in standard buffer containing horseradish peroxidase-labelled goat anti-rabbit IgG (1 : 10 000 dilution). The membrane was washed a further four times for 10 min with the standard incubation buffer.

36

A. RodrõÂguez-PenÄa et al.

The detection of rat endoglin was performed with the ECL chemiluminiscence western blotting system from Amersham (Buckinghamshire, UK). For quanti®cation, ®lms were scanned and relative optical densities of each lane were obtained from the digitalized image using an image analysis program (MacBAS, version 2.2, Fuji, Japan).

Statistical methods The Kolmogorov±Smirnov test was used to assess normality of the data distribution. One-way analysis of variance was used to compare differences between groups. ScheffeÂ's t-test for parametric data and the Kruskal±Wallis multiple comparison Z-value test for non-parametric data were used. A P-value -0.05 was considered statistically signi®cant. All the data are expressed as means"SEM.

Results RMR induced a progressive increase in mean arterial pressure and urinary protein excretion, as shown in Table 1. Light microscopy analysis of haematoxilin± eosin- and Masson's trichrome-stained slides showed that most glomeruli in group 1 had a preserved architecture with only a slight mesangial matrix increment observed (Figure 1a). No alterations were detected in tubule interstitial area in these animals. Renal tissues from group 2 showed a moderate increment in mesangial matrix. In addition, in¯ammatory in®ltration and low-grade interstitial ®brosis were observed in the interstitial tissue (Figure 1b). Group 3 showed evident glomerulosclerosis in a large number of renal corpuscles, with tuft adhesions to Bowman's capsule. In¯ammatory in®ltration and focal tubular atrophy and dilatation with PAS-positive casts and ®brosis were observed in the interstitium (Figure 1c). As shown in Table 2, computer-assisted quantitative detection of ®brous tissue revealed that renal corpuscular area, and mesangial and interstitial ®brosis were increased with the time of evolution. Immunohistochemical staining for endoglin demonstrated expression mainly on the endothelial surface of large vessels in all groups of the study. The immunoexpression in renal corpuscles was limited to Bowman's parietal epithelium in group 1, while glomerulus showed very slight staining in this group

and in group 2 (Figure 1d and e). The immunoexpression in conserved glomerular endothelium from group 3 was more evident (Figure 1f). Northern blot experiments revealed that rats with RMR showed an increase in the expression of mRNA for endoglin, only at 5 months after RMR (Figure 2A). Western blot analysis (Figure 2B) produced a different time course: a marked increase in the ®rst month, a decrease in the 3rd month and a further increase in the 5th month after RMR.

Discussion The present study demonstrates increased expression of endoglin in rats with severe hypertension and renal damage. The up-regulation of endoglin has been demonstrated by northern blot, western blot and immunohistochemistry, and this increased endoglin expression coincides with the major renal damage and renal dysfunction. The increase in endoglin expression in the kidneys after RMR can be explained by the increase in TGF-b1 expression in remnant rat kidneys after subtotal nephrectomy w16,17x. TGF-b1 itself can stimulate the expression of endoglin in cultured human monocytes and in the U-937 monocytic line w18x, mouse ®broblasts (Eleno et al., unpublished data) and human arterial smooth muscle cells w19x in vitro. The biological meaning of the increased endoglin expression in glomeruli and interstitium of diseased kidneys remains to be determined. Since endoglin is a component of the TGF-b receptor system and the pro-matrix effects of TGF-b1 are recognized to be a key factor in the glomerulosclerosis and interstitial ®brosis that characterize chronic progressive renal disease w3,20x, it is tempting to speculate that the increased expression of endoglin by interstitial ®broblasts or glomerular mesangial cells modulates the stimulation of extracellular matrix production. Endoglin is an accessory protein that requires the presence of the signalling receptor type II for ligand binding w21,22x. The association of endoglin with type I and II receptors, as well as the modulation of cellular TGF-b responses by endoglin overexpression have been demonstrated w18,21x. Among the cellular TGF-b responses

Table 1. Effect of renal mass reduction on renal function in rats Sham

Mean arterial pressure (mmHg) Creatinine clearance (mlumin) Urinary protein excretion (mgu24 h)

105"3 1.7"0.2 13.7"1.1

Data are the mean"SEM of nine animals per group. *Signi®cant difference (P-0.05) vs sham-operated rat values.

Months after RMR 1

3

5

115"4 0.9"0.1* 21.7"2.3

134"4* 1.4"0.2* 119.0"21.2*

146"9* 1.2"0.2* 439"96*

Endoglin and renal ®brosis

37

Fig. 1. Morphology and immunohistochemical localization of endoglin in the three different periods of the study. (a±c) Masson's trichrome staining light microscopy pictures. (a) One month after RMR, a mild mesangial increase only is detected. (b) At 3 months after RMR, mononuclear in¯ammatory in®ltrate and light interstitial ®brosis with mild mesangial expansion is seen. (c) At 5 months after RMR, glomerulosclerosis (note the renal corpuscle in the centre of the image), in¯ammatory in®ltrate and more pronounced interstitial ®brosis are observed. (d±f) Immunohistochemical staining against endoglin. (d) Endoglin is detected in the Bowman's parietal epithelium; however, no evident expression is seen in the glomerulus 1 month after RMR. (e) After 2 months of RMR, expression is still seen in the Bowman's parietal epithelium, and a mild glomerular expression is detected mainly in podocytes. (f) More evident immunostaining in glomerulus podocytes and capillaries was observed after 3 months of RMR. Magni®cation: (a) and (b) 3 144; (c) 3 72; (d), (e) and (f) 3 288.

modulated by endoglin is the synthesis of the extracellular matrix components ®bronectin and plasminogen activator inhibitor type 1. Thus, the presence of endoglin as a part of the TFG-b receptor system could be important in determining the extent of extracellular matrix protein production by renal cells following the binding of TGF-b1. It is unclear whether TGF-b1 binding to endoglin in the kidney would lead to increased or decreased TGF-b signalling. In this

regard, we have already obtained some preliminary evidence showing that endoglin overexpression can diminish rather than enhance the effect of TGF-b on extracellular matrix synthesis and cell proliferation w18,21, Eleno et al., unpublished results, Smith et al., unpublished resultsx. Thus, endoglin hyperexpression could protect the kidney against the continuous pro®brogenic effect of TGF-b1, thus ®tting into the concept of `glomerular self-defence' developed by

38

A. RodrõÂguez-PenÄa et al.

Table 2. Quantitative morphometric study of renal ®brosis in rats with renal mass reduction Sham

Corpuscular section area Glomerular ®brosis Tubulo-interstitial ®brosis

10 558"205 874"64 3225"197

Months after RMR 1

3

5

11 349"310 959"82 3811"293

16 969"439* 1888"104* 4708"241*

17 198"585* 2008"113* 9075"357*#

Data are shown in mm2 as median"SEM of six animals per group. No statistical differences were found between sham-operated and 1 month after RMR in the parameters observed. *Statistically signi®cant difference (P-0.01; Kruskal±Wallis test) vs sham-operated or 1 month after RMR. #Statistically signi®cant difference (P-0.01; Kruskal±Wallis test) vs 3 months after RMR.

Fig. 2. (A) Representative northern blot analysis for endoglin (upper blot) and GAPDH (lower blot) using kidney extracts from rats 1 month (lanes 1±3), 3 months (lanes 4±7) and 5 months (lanes 8±11) after RMR and sham-operated rats (lanes 12±15). This is a representative northern blot of three different experiments performed in the same conditions. The average endoglinuGAPDH signal (phosphorimager; ns9 animals per group) is 1.44"0.03 for sham-operated rats, 1.49"0.11 for 1 month RMR, 1.60"0.18 for 3 months RMR and 1.99"0.05 for 5 months RMR (P-0.05 vs sham-operated rats and 1 month RMR). (B) Representative western blot for endoglin using kidney extracts from rats 1 month (lanes 1 and 2), 3 months (lanes 3 and 4) and 5 months after RMR (lanes 5 and 6) and sham-operated rats (lanes 7 and 8). The ®gure shows one of three different but equally performed experiments. The average signal (densitometry; ns6 animals per group) is 9.6"1.8 for sham-operated rats, 15.5"0.9 for 1 month RMR (P-0.05 vs sham-operated rats), 11.5"1.1 for 3 months RMR and 14.4"1.1 for 5 months RMR (P-0.05 vs sham operated rats and 1 month RMR).

Kitamura and Fine w23x. In support of this hypothesis is the ®nding that the time course of endoglin induction is slightly delayed as compared with that of the extracellular matrix deposition in the kidney upon RMR. Endoglin also contains the RGD sequence in an exposed region of the extracellular domain w9x. This is an important recognition motif for a number of extracellular matrix proteins such as ®bronectin, collagen, laminin and vitronectin for their speci®c integrin receptors. Thus, the presence of endoglin in the kidney could also serve to organize extracellular matrix in the sites of extracellular matrix deposition. In conclusion, this is the ®rst study demonstrating that endoglin, a TGF-b-binding protein, is

up-regulated in the kidneys of rats with renal ®brosis induced by RMR. The mechanism of endoglin up-regulation and the elucidation of its pathological or physiological signi®cance deserve to be investigated further.

Acknowledgements. The authors wish to thank Dr Neil Docherty for correcting the language of the manuscript. This study was partially supported by a grant from DireccioÂn General de EnsenÄanza Superior e InvestigacioÂn Cientõ®ca (grant SAF19880095), Comision Interministerial de Ciencia y Tecnologia (CICYT), Junta de Castilla y Leon (grant SA70u00B) and Comunidad de Madrid (CAM). A.R.-P. is a fellow of Fondo de Investigaciones Sanitarias (FIS). M.P. is a Fellow of Junta de Castilla y Leon and Fondo Social Europeo.

Endoglin and renal ®brosis

References 1. Massague J, Attisano SL, Wrana SL. The TGF-b family and its composite receptors. Trends Cell Biol 1994; 4: 172±178 2. Roberts AB, Sporn MB. Physiological actions and clinical applications of transforming growth factor-b (TGF-b). Growth Factors 1993; 8: 1±9 3. Border WA, Noble N. Cytokines in kidney disease: the role of transforming growth factor beta. Am J Kidney Dis 1993; 22: 105±113 4. Border WA, Okuda S, Languino LR, Sporn MB, Ruoslahti E. Suppression of experimental glomerulonephritis by antiserum against transforming growth factor b1. Nature 1990; 346: 371±374 5. Border WA, Noble NA, Yamamoto T et al. Natural inhibitor of transforming growth factor-b protects against scarring in experimental kidney disease. Nature 1992; 360: 361±364 6. Isaka Y, Brees DK, Ikegaya K et al. Gene therapy by skeletal muscle expression of decorin prevents ®brotic disease in rat kidney. Nature Med 1996; 2: 418±423 7. Gougos A, Letarte M. Identi®cation of a human endothelial cell antigen with monoclonal antibody 44G4 produced against a pre-B leukemic cell line. J Immunol 1988; 141: 1925±1933 8. Lastres P, BelloÂn T, CabanÄas C et al. Regulated expression on human macrophages of endoglin, an Arg-Gly-Asp-containing surface antigen. Eur J Immunol 1992; 22: 393±397 9. Gougos A, St Jacques S, Creaves A et al. Identi®cation of distinct epitopes of endoglin, an RGD-containing glycoprotein of endothelial cells leukemic cells and syncytiotrophoblasts. Int Immunol 1992; 4: 83±92 10. MacAllister KA, Grogg KM, Johnson DW et al. Endoglin, a TGF-b binding protein of endothelial cells, is the gene for hereditary haemorragic telangiectasia type 1. Nature Genet 1994; 8: 345±351 11. Cheifetz S, BelloÂn T, CaleÂs C et al. Endoglin is a component of the transforming growth factor-b receptor system in human endothelial cells. J Biol Chem 1992; 267: 19027±19030.

39 12. Roy-Chaudhury P, Simpson JG, Power DA. Endoglin, a transforming growth factor-beta-binding protein, is upregulated in chronic progressive renal disease. Exp Nephrol 1997; 5: 55±60 13. Flores O, Gallego B, AreÂvalo M, LoÂpez-Novoa JM. Effect of trandolapril and verapamil on renal damage progression associated to 5u6 nephrectomy in rats. Exp Nephrol 1998; 6: 39±49 14. Raab U, Velasco, B, Lastres P et al. Expression of normal and truncated forms of human endoglin. Biochem J 1999; 339: 579±588 15. Raab U, Lastres P, AreÂvalo MA et al. Endoglin is expressed in the chicken vasculature and is involved in angiogenesis. FEBS Lett 1999; 459: 249±254 16. Wu LL, Cox A, Roe CJ, Dziadek M, Cooper ME, Gilbert RE. Transforming growth factor beta 1 and renal injury following subtotal nephrectomy in the rat. Role of the renin±angiotensin system. Kidney Int 1997; 51: 1553±1567 17. Gilbert RE, Wu LL, Kelly DJ et al. Pathological expression of renin and angiotensin II in the renal tubule after subtotal nephrectomy. Implications for the pathogenesis of tubulointerstitial ®brosis. Am J Pathol 1999; 155: 429±440 18. Lastres P, LetamendõÂa A, Zhang H et al. Endoglin modulates cellular responses to TGF-b1. J Cell Biol 1996; 13: 1109±1121 19. Conley BA, Smith JD, Guerrero-Esteo M, Bernabeu C, Vary CPH. Endoglin, a TGF-beta binding protein, is expressed by smooth muscle cells in human atherosclerotic plaques. Atherosclerosis 2000; 153: 323±335 20. Border WA, Noble N. Transforming growth factor beta in glomerular injury. Exp Nephrol 1994; 2: 13±17 21. LetamendõÂa A, Lastres A, Botella LM et al. Role of endoglin in cellular responses to transforming growth factor b. A comparative study with betaglycan. J Biol Chem 1998; 273: 33011±33019 22. Barbara NP, Wrana JL, Letarte M. Endoglin is an accessory protein that interacts with the signaling receptor complex of multiple members of the transforming growth factor-b superfamily. J Biol Chem 1999; 274: 584±594 23. Kitamura M, Fine LG. The concept of glomerular self-defense. Kidney Int 1999; 55: 1639±1671