TY - JOUR
AU - Egido, Jesús
AB - Abstract Background. Recent evidence in vitro and in vivo suggests that gremlin, a bone morphogenetic protein antagonist, is participating in tubular epithelial mesenchymal transition (EMT) in diabetic nephropathy as a downstream mediator of TGF-β. Since EMT also occurs in parietal epithelial glomerular cells (PECs) leading to crescent formation, we hypothesized that gremlin could participate in this process. With this aim we studied its expression in 30 renal biopsies of patients with pauci-immune crescentic nephritis. Methods. Gremlin was detected by in situ hybridization (ISH) and immunohistochemistry (IMH) and TGF-β by ISH and Smads by southwestern histochemistry (SWH). Phosphorylated Smad2, CTGF, BMP-7, PCNA, α-SMA, synaptopodin, CD-68, and phenotypic markers of PECs (cytokeratin, E-cadherin), were detected by IMH. In cultured human monocytes, gremlin and CTGF induction by TGF-β was studied by western blot. Results. We observed strong expression of gremlin mRNA and protein in cellular and fibrocellular crescents corresponding to proliferating PECs and monocytes, in co-localization with TGF-β. A marked over-expression of gremlin was also observed in tubular and infiltrating interstitial cells, correlating with tubulointerstitial fibrosis ( r = 0.59; P < 0.01). A nuclear Smad activation in the same tubular cells, that are expressing TGF-β and gremlin, was detected. In human cultured monocytes, TGF-β induced gremlin production while CTGF expression was not detected. Conclusion. We postulate that gremlin may play a role in the fibrous process in crescentic nephritis, both in glomerular crescentic and tubular epithelial cells. The co-localization of gremlin and TGF-β expression found in glomeruli and tubular cells suggest that gremlin may be important in mediating some of the pathological effects of TGF-β. BMP-7 antagonist, crescentic glomerulonephritis, epithelial mesenchymal transition (EMT), gremlin, TGF-β, Smad Introduction Gremlin, a member of the cysteine knot protein super family 1, is a highly conserved glycosylated and phosphorylated secreted protein present both on the external cell surface and within the ER-Golgi compartment of a variety of cell types [ 1 ]. Gremlin, as bone morphogenetic protein (BMP) antagonist, has been reported to influence diverse processes in growth, differentiation and development, in many cases by heterodimerization with BMP-2, -4 and -7, thereby inhibiting the ability of these ligands to bind to their receptors. Recent work has established that gremlin mediates its action via induction of epithelial to mesenchymal feedback signalling. Metanephric renal (and limb bud) organogenesis occurs via BMP antagonism and thus gremlin is confirmed as the essential extracellular signal which initiates renal development [ 2 ]. The induction of gremlin in cultured human mesangial cells exposed to high glucose and transforming growth factor-β (TGF-β) in vitro and in kidneys from diabetic rats in vivo , has been reported [ 3 ]. Recently we have shown that gremlin mRNA and protein are highly expressed in the tubular compartment of advanced human diabetic nephropathy [ 4 ]; kidney biopsies from patients with diabetic nephropathy had significantly increased gremlin expression when compared with normal kidney or biopsies from patients with non-scarring renal disorders such as minimal change disease. Gremlin expression was most pronounced in areas of interstitial fibrosis and it is co-localized with TGF-β [ 4 ]. Since gremlin expression has been induced by TGF-β1 in renal proximal tubular cells in vitro undergoing transdifferentiation to a fibroblast phenotype, a role of gremlin in the epithelial mesenchymal transition (EMT) process, as BMP-7 antagonist has been proposed [ 2 , 3 ]. On the other hand, the discovery of TGF-β/Smad-signalling pathway has allowed to study the intracellular mechanisms of TGF-β and BMP-7 at the EMT [ 5 ]. TGF-β signalling is mediated by nuclear translocation of phosphorylated Smad 2 and Smad 3 and BMP-7 by Smad 1, Smad 5 and Smad 8 [ 6 ]. Recently, Bowman's epithelial-mesenchymal transdifferentiation, identified by alpha-smooth muscle actin (α-SMA) expression in the evolution of glomerular crescent formation, has been reported [ 7–9 ]; so far we hypothesize a role for gremlin in the scar formation of crescentic glomerulonephritis. The purpose of this study is to examine the presence of gremlin in the glomerular crescents formation in 30 renal biopsies from patients with pauci-immune crescentic glomerulonephritis and correlate its expression with TGF-β and markers of EMT, trying to define whether gremlin is a phenotypic modulator in the formation of glomerular crescents. Materials and methods Assessment of the pattern of gremlin expression by in situ hybridization (ISH) Kidney samples were obtained by percutaneous renal biopsy from patients undergoing diagnostic evaluation at the Division of Nephrology, Austral University, Valdivia, Chile. The samples were studied after obtaining patient consent and the project was approved by the local hospital ethics committee. Renal biopsies from 30 patients with pauci-immune crescentic nephritis before specific treatment were studied. Control human kidney specimens ( n = 5) were taken from normal portions of renal tissue from patients who underwent surgery because of localized renal tumours. ISH was performed as described previously for antisense TGF-β probe (R&D Systems, Minneapolis, MN, USA) [ 10 ], and with the following modifications for biotin-labelled human gremlin probes (Invitrogen, Carlsbad, CA, USA). The gremlin probes were as follows: 478 antisense 5′-TGAAAGGAACCTTCCTCCTTCC-3′,2416 antisense 5′-ATGGGAGAGCACTGGATCAAAA-3′ and 3553 antisense 5′-CAGGCACTGACTCAGGAAGACA-3′ For gremlin analysis, pretreatment with endogenous biotin blocking system (Dako Co, Carpinteria, CA, USA) was performed prior to proteinase K digestion. The sections were incubated with a pre-hybridization solution (Dako, mRNA ISH Solution) for 60 min at 37°C and with the antisense probe overnight at 37°C. The slides for gremlin were washed with 2 × SSC and 1 × SSC for 10 min at room temperature and then with 0.5 × SCC for 20 min at 37°C. Detection was performed with avidin-alkaline phosphatase conjugate (Dako) for 30 min at room temperature, washed 5 min with 1 × TBS and using NBT-BCIP as the enzyme substrate for 120 min at 37°C (R&D Systems). Tissues were then dehydrated in ethanol series and mounted in Canadian balsam (Polysciences Inc., Warrington, PA, USA). The specificity of the reaction was confirmed: (i) by demonstrating the disappearance of hybridization signal when RNAse (100 μg/ml) (Sigma Chemicals Co., St. Louis, MO, USA) was added in 0.05 M Tris after the digestion with proteinase K; (ii) by the use of a sense probe (R&D Systems); (iii) with a negative control (Plasmid DNA) (Dako) and (iv) without probe. For gremlin ISH slides, Dako nuclear fast red was used for 10 min. Immunohistochemistry (IMH) For light microscopy, kidney tissues were fixed in 4% buffered formalin, or Bouin, dehydrated and embedded in paraffin by conventional techniques. Sections were stained with haematoxylin and eosin (HE), periodic acid-Schiff (PAS), and silver methenamine. Paraffin embedded biopsy specimens were used for detection of gremlin, BMP-7, CTGF, pSmad2, α-SMA, PCNA, e-cadherin, cytokeratin, synaptopodin, and macrophage marker (CD68). The following primary antibodies were employed: rabbit polyclonal anti-gremlin (ABGENT, AP6133a, San Diego CA, USA); goat polyclonal anti-human BMP-7 (Santa Cruz Biotechnology, CA, USA); rabbit polyclonal anti-human CTGF (ABCAM, ab6992, Cambridge, UK); rabbit polyclonal anti-Smad2 (phospho S465) ABCAM ab5490; mouse anti-PCNA clone PC10 (Dako); mouse anti-human alpha-smooth muscle actin clone 1A4 (Dako); mouse anti-human e-cadherin clone 36B5 (Novocastra, Newcastle, UK); mouse anti-multi-cytokeratin NCL-AE1/AE3 clone AE1 y AE3 (Novocastra); mouse anti-human synaptopodin clone G1D4 (Progen, Heidelberg, Germany); mouse anti-CD68 (Dako). Briefly, 5 μm thick renal sections, Bouin- or formalin-fixed were deparaffinized through xylene, alcohol and distilled water. Endogenous peroxidase was blocked by 3% H 2 O 2 for 15 min and then the sections were treated in a microwave oven in a solution of 0.1 mM citrate buffer pH 6.0 for 10 min or EDTA buffer 1 mM pH 8.0 for synaptopodin detection. After blocking, the sections were incubated overnight at 4°C with the specific primary antibody. The sections were then incubated with the correspondent biotinylated secondary antibodies for 30 min at 22°C. After three rinses in Tris saline buffer, they were incubated with streptavidin-peroxidase (Dako) 1/1000 for 30 min. Color was developed with substrate (Dako) and then counterstained with haematoxylin, dehydrated, and mounted with Canadian balsam (Polysciences, Inc.). The specificity was checked by omission of primary antibodies and use of non-immune sera. Southwestern histochemistry (SWH) This technique has been described by Isono et al . [ 11 ] and recently reported in our laboratory for NF-κB detection [ 12 ]. Briefly, complementary oligonucleotides containing a Smad binding consensus sequence (21) were synthesized by Invitrogen as follows: 5′-GAGTATGTCTAGACTGACAATGTAC-3′. After annealing with their complementary DNA (80°C during 2 min), the probe was labelled with digoxigenin (DIG oligonucleotide three-end labelling kit), (Boehringer Mannheim, Mannheim, Germany). Paraffin-embedded kidney sections were dewaxed, rehydrated and incubated with 5 mM levamisole (Sigma Chemical Co.) to inhibit endogenous alkaline phosphatase, and fixed with 0.2% p-formaldehyde for 30 min at 28°C. Sections were subsequently digested with pepsin A (433 U/mg; Sigma), washed twice with buffer 1 (10 mM HEPES, 40 mM NaCl, 10 mM MgCl 2 , 1 mM DTT, 1 mM EDTA, 0.25% BSA, pH 7.4), and then with 0.1 mg/ml DNAse I, washed once with buffer 2 (10 mM HEPES, 40 mM NaCl, 1 mM DTT, 10 mM EDTA, 0.25% BSA, pH 7.4) to stop the reaction. The labelled probe (100 pM) diluted in buffer 1 containing 0.5 mg/ml poly (dI-dC) (Pharmacia LKB, Piscataway, NJ, USA) was applied overnight at 37°C. After washing, sections were incubated for 1 h in blocking solution (0.01 × SSC, 0.01% SDS, 0.03% Tween 20, 0.1 M maleic acid, 0.15 M NaCl, pH 7.5), and with anti-digoxigenin antibody conjugated with alkaline phosphatase (1:250 in blocking solution; Boehringer Mannheim) overnight at 4°C. The color reaction was developed using (NBT/BCIP) (Dako). Immunohistochemistry quantification The surface area labelled was evaluated by quantitative image analysis using a KS 300 imaging system 3.0 (Zeiss, München-Hallbergmoos, Germany). For each sample, the mean staining area was obtained by analysis of 20 different fields (20×). Quantification was done twice, independently, and interassay variations were not significant. The staining score is expressed as percentage/mm 2 . Tubulointerstitial cell infiltration and interstitial fibrosis was classified into four groups according to the extent of them and the presence of tubular atrophy and degeneration: (i) normal, (ii) involvement up to 25% of the cortex, (iii) involvement of 26 to 50% of cortex, and (iv) extensive damage involving more than 50% of the cortex. In vitro studies Human monocyte cells (THPs cell line) were used. Cells were grown in RPMI with 10% FCS. Cells were serum starved for 24 h before the experiments. Cell culture reagents were from Life Technologies, Inc. Recombinant TGF-β was from Preprotech. Protein levels were evaluated by western blot. Cell samples were homogenized in lysis buffer and protein content was determined by the BCA method (Pierce, Rockford, IL, USA). Proteins (20 μg/lane) were separated on 15% SDS-PAGE gels, transferred, blocked and incubated with Gremlin antibody overnight at 4°C. Detection was made with corresponding peroxidase-conjugated secondary antibody and developed using an ECL chemiluminiscence kit (Amersham). The autoradiographs were scanned using the GS-800 Calibrated Densitometer (Quantity One, Bio-Rad, Spain). Results are expressed as n-fold increase over control as mean ± SEM of experiments made. Statistical analysis The statistical analysis was performed with the GraphPad Instat, GraphPad Software, San Diego, CA, USA. The results of the clinical data are expressed as the mean ± SD. A Spearman correlation was used to correlate tubule interstitial gremlin expression and tubule interstitial cell infiltration and fibrosis. A Pearson correlation was used to correlate the gremlin mRNA and protein expression. Results The clinical data of the patients studied are presented in Table 1 . Biopsies from 30 patients with pauci-immune crescentic glomerulonephritis, 80% of them with rapidly progressive GN (mean s. creat. 7.8 ± 4.5 mg/dl), and with different range of proteinuria, were included in these studies. The mean age of the patients was 47 ± 20 years old (median 50) and 18 were females. The percentage of glomerular crescents was over 50%, the immunofluorescence microscopy was negative or with rare immune deposits in all the cases (pauci-immune); 43% of the patients were P-ANCA positive and 20% were C-ANCA positive. Table 1. Clinical and biological data from the 30 patients with pauci-immune crescentic glomerulonephritis Patient  Age (years)/ Gender  Initial S Cr (mg/dl)   Urinary findings  Proteinuria (g/24 h)  Clinical features  ANCA  Follow-up  1  57/F  16.7  P: 75 mg/dl; RBC: 50–62  N.A.  RPGN Pulmonary haemorrhage  c-ANCA  On chronic dialysis  2  50/F  8.2  P: (–); RBC: 5–10  N.A.  RPGN  c-ANCA  After 3 years, stable serum cretainine (3 mg/dl)  3  22/M  8.8  P: 500 m/dl,RBC: 6–8; hyaline, granular, RBC casts; oval fat bodies  3.7  RPGN Uraemic syndrome  (–)  Transplanted, after 6 years on chronic dialysis  4  73/F  4.9  P: 100 mg/dl; RBC: >100  N.A.  RPGN  p-ANCA  Death on chronic dialysis  5  66/F  5.7  P: 25 mg/dl; RBC: 30–40; WBC: 6–8  0.42  RPGN Pulmonary haemorrhage  c-ANCA  Death on chronic dialysis (CNS vasculitis)  6  72/M  8.2  P: 75 mg/dl; RBC: >100  N.A.  RPGN  c-ANCA  After 5 years, stable serum creatinine (1.4 mg/dl)  7  47/M  7.8  N.A.  N.A.  RPGN  (–)  Unknown  8  50/F  7.7  P: (–); RBC: 6–8  N.A.  RPGN  c-ANCA  On chronic dialysis  9  22/F  10.0  P: 500 mg/dl; RBC: >100; granular, waxy, broad, hyaline, fatty, RBC casts  2.4  RPGN Pulmonary haemorrhage  p-ANCA  Transplanted, after 2 year on chronic dialysis  10  58/M  8.0  P: 100 mg/dl; RBC: >100; granular, hyaline, fatty, RBC casts  6.4  RPGN  (–)  Death on chronic dialysis  11  21/M  12.0  P: 300 mg/dl; RBC: 180–200; WBC: 8–20  N.A.  RPGN  (–)  After 10 years, stable serum creatinine (1.9 mg/dl)  12  34/M  6.3  P: 500 mg/dl; RBC: >100  8.1  RPGN  (–)  On chronic dialysis  13  62/F  4.9  P: 75 mg/dl; RBC: 0–4  N.A.  RPGN  c-ANCA  After 5 years, stable serum creatinine (1.6 mg/dl)  14  41/M  2.9  P: 150 mg/dl; RBC: >100; granular, epithelial, waxy, RBC casts; oval fat bodies  5.6  Nephrotic syndrome, pulmonary haemorrhage  p-ANCA  Transplanted, after chronic dialysis  15  10/F  13.6  P: 500 mg/dl; RBC: 5–8; WBC: 15–20  1.7  Uraemic syndrome  N.A.  Transplanted, after chronic dialysis  16  44/M  1.3  P: 220 mg/dl; RBC: 3–5; hyaline casts  2.7  Pulmonary haemorrhage, non–nephrotic proteinuria  p-ANCA  After 2 years, stable serum creatinine (1.0 mg/dl)  17  66/M  6.9  P: 500 mg/dl; RBC: >100; granular, waxy, hyaline, WBC, RBC casts; oval fat bodies  1.6  RPGN Uraemic syndrome  (–)  Unknown  18  75/F  3.3  P: 300 mg/dl; RBC: >100  0.5  RPGN Pulmonary haemorrhage  p-ANCA  Unknown  19  55/F  9.9  P: 110 mg/dl; RBC: >100  1.7  RPGN  p-ANCA  Unknown  20  43/M  4.7  P: 500 mg/dl; RBC: >100; granular casts; oval fat bodies  9.0  Nephrotic syndrome  N.A.  Unknown  21  34/F  0.9  P: 310 mg/dl; RBC: >100  6.0  Nephritic syndrome  N.A.  Unknown  22  14/F  3.2  P: 150 mg/dl; RBC: >100  1.7  RPGN  p-ANCA  Unknown  23  62/F  3.2  P: 240 mg/dl; RBC: >100  1.9  RPGN Pulmonary haemorrhage  p-ANCA  Unknown  24  38/F  11.1  P: 60 mg/dl; RBC: >100; WBC: 3–5; granular, broad, waxy, RBC casts  1.0  Uraemic syndrome  p-ANCA  After 2 years, on chronic dialysis  25  27/M  18.6  P: 300 mg/dl; RBC: 21–23; WBC: 2–3; granular, hyaline, RBC casts  3.2  RPGN  (–)  On chronic dialysis  26  65/F  5.6  P: 25 mg/dl; RBC: 2–4  1.0  RPGN  p-ANCA  On chronic dialysis  27  51/F  9.2  P: (+); RBC: (+)  4.3  RPGN Oligoarthritis  p-ANCA  On chronic dialysis  28  69/F  4.9  P: 60 mg/dl; RBC: (+)  1.2  RPGN Pulmonary haemorrhage  p-ANCA  On chronic dialysis  29  18/F  7.3  P: 200 mg/dl; RBC: >100  2.8  RPGN Nephrotic proteinuria  N.A.  Unknown  30  77/M  17.7  P: 500 mg/dl; RBC: >100  3.0  RPGN Pulmonary haemorrhage  p-ANCA  Death on chronic dialysis  Patient  Age (years)/ Gender  Initial S Cr (mg/dl)   Urinary findings  Proteinuria (g/24 h)  Clinical features  ANCA  Follow-up  1  57/F  16.7  P: 75 mg/dl; RBC: 50–62  N.A.  RPGN Pulmonary haemorrhage  c-ANCA  On chronic dialysis  2  50/F  8.2  P: (–); RBC: 5–10  N.A.  RPGN  c-ANCA  After 3 years, stable serum cretainine (3 mg/dl)  3  22/M  8.8  P: 500 m/dl,RBC: 6–8; hyaline, granular, RBC casts; oval fat bodies  3.7  RPGN Uraemic syndrome  (–)  Transplanted, after 6 years on chronic dialysis  4  73/F  4.9  P: 100 mg/dl; RBC: >100  N.A.  RPGN  p-ANCA  Death on chronic dialysis  5  66/F  5.7  P: 25 mg/dl; RBC: 30–40; WBC: 6–8  0.42  RPGN Pulmonary haemorrhage  c-ANCA  Death on chronic dialysis (CNS vasculitis)  6  72/M  8.2  P: 75 mg/dl; RBC: >100  N.A.  RPGN  c-ANCA  After 5 years, stable serum creatinine (1.4 mg/dl)  7  47/M  7.8  N.A.  N.A.  RPGN  (–)  Unknown  8  50/F  7.7  P: (–); RBC: 6–8  N.A.  RPGN  c-ANCA  On chronic dialysis  9  22/F  10.0  P: 500 mg/dl; RBC: >100; granular, waxy, broad, hyaline, fatty, RBC casts  2.4  RPGN Pulmonary haemorrhage  p-ANCA  Transplanted, after 2 year on chronic dialysis  10  58/M  8.0  P: 100 mg/dl; RBC: >100; granular, hyaline, fatty, RBC casts  6.4  RPGN  (–)  Death on chronic dialysis  11  21/M  12.0  P: 300 mg/dl; RBC: 180–200; WBC: 8–20  N.A.  RPGN  (–)  After 10 years, stable serum creatinine (1.9 mg/dl)  12  34/M  6.3  P: 500 mg/dl; RBC: >100  8.1  RPGN  (–)  On chronic dialysis  13  62/F  4.9  P: 75 mg/dl; RBC: 0–4  N.A.  RPGN  c-ANCA  After 5 years, stable serum creatinine (1.6 mg/dl)  14  41/M  2.9  P: 150 mg/dl; RBC: >100; granular, epithelial, waxy, RBC casts; oval fat bodies  5.6  Nephrotic syndrome, pulmonary haemorrhage  p-ANCA  Transplanted, after chronic dialysis  15  10/F  13.6  P: 500 mg/dl; RBC: 5–8; WBC: 15–20  1.7  Uraemic syndrome  N.A.  Transplanted, after chronic dialysis  16  44/M  1.3  P: 220 mg/dl; RBC: 3–5; hyaline casts  2.7  Pulmonary haemorrhage, non–nephrotic proteinuria  p-ANCA  After 2 years, stable serum creatinine (1.0 mg/dl)  17  66/M  6.9  P: 500 mg/dl; RBC: >100; granular, waxy, hyaline, WBC, RBC casts; oval fat bodies  1.6  RPGN Uraemic syndrome  (–)  Unknown  18  75/F  3.3  P: 300 mg/dl; RBC: >100  0.5  RPGN Pulmonary haemorrhage  p-ANCA  Unknown  19  55/F  9.9  P: 110 mg/dl; RBC: >100  1.7  RPGN  p-ANCA  Unknown  20  43/M  4.7  P: 500 mg/dl; RBC: >100; granular casts; oval fat bodies  9.0  Nephrotic syndrome  N.A.  Unknown  21  34/F  0.9  P: 310 mg/dl; RBC: >100  6.0  Nephritic syndrome  N.A.  Unknown  22  14/F  3.2  P: 150 mg/dl; RBC: >100  1.7  RPGN  p-ANCA  Unknown  23  62/F  3.2  P: 240 mg/dl; RBC: >100  1.9  RPGN Pulmonary haemorrhage  p-ANCA  Unknown  24  38/F  11.1  P: 60 mg/dl; RBC: >100; WBC: 3–5; granular, broad, waxy, RBC casts  1.0  Uraemic syndrome  p-ANCA  After 2 years, on chronic dialysis  25  27/M  18.6  P: 300 mg/dl; RBC: 21–23; WBC: 2–3; granular, hyaline, RBC casts  3.2  RPGN  (–)  On chronic dialysis  26  65/F  5.6  P: 25 mg/dl; RBC: 2–4  1.0  RPGN  p-ANCA  On chronic dialysis  27  51/F  9.2  P: (+); RBC: (+)  4.3  RPGN Oligoarthritis  p-ANCA  On chronic dialysis  28  69/F  4.9  P: 60 mg/dl; RBC: (+)  1.2  RPGN Pulmonary haemorrhage  p-ANCA  On chronic dialysis  29  18/F  7.3  P: 200 mg/dl; RBC: >100  2.8  RPGN Nephrotic proteinuria  N.A.  Unknown  30  77/M  17.7  P: 500 mg/dl; RBC: >100  3.0  RPGN Pulmonary haemorrhage  p-ANCA  Death on chronic dialysis  S Cr : serum creatinine, M: males, F: females, P: proteinuria, RBC: red blood cells, WBC: white blood cells, N.A.: not available. View Large The cellular components of these crescents are mainly parietal epithelial cells, cytokeratin-positive ( Figure 1 B) and dysregulated parietal epithelial cells negative for cytokeratin, with different participation of infiltrating monocyte/macrophage cells (CD68-positive cells) ( Figure 1 C). All these crescentic cells are mainly PCNA-positive cells as it is shown in Figure 1 A. PCNA was also expressed in numerous tubular and interstitial cells. On the other hand, these crescentic cells, are negative for synaptopodin ( Figure 1 F), a phenotypic marker of podocytes, and negative for E-cadherin ( Figure 1 E), a tubular epithelial marker, that is lost during the glomerular epithelial-myofibroblast transition process. In glomerular crescents in which a fibrous transition to fibrocellular or fibrous crescent is occurring, the cellular components are replaced by proliferating α-SMA (+) myofibroblasts, within a collagen-rich sclerotic crescent ( Figure 1 D). Fig. 1. View largeDownload slide Immunohistochemistry (IMH) for PCNA, cytokeratin (CK), CD68, α-SMA, E-cadherine and synaptopodin, in glomerular crescents of patients with pauci-immune rapidly progressive GN. Many crescentic parietal epithelial cells (PECs) strongly expressed PCNA ( A ). Cytokeratine (CK) was strongly expressed in the majority of these proliferating crescentic PECs, as well as in the tubular epithelial cells. ( B ). Macrophages (CD68 (+) cells) are identified in these cellular crescents and also as interstitial infiltrating cells ( C ). α-SMA (+) cells were found in crescentic lesions, as well as in periglomerular interstitial fibrosis ( D ). Most of these proliferating crescentic PECs are negative for E-cadherin, that is preserved in some of the epithelial tubular cells (Case 20) ( E ). In this fibrocellular crescent, no cell expressed synaptopodin, a podocyte marker (Case 29) ( F ). (magnification ×200). Fig. 1. View largeDownload slide Immunohistochemistry (IMH) for PCNA, cytokeratin (CK), CD68, α-SMA, E-cadherine and synaptopodin, in glomerular crescents of patients with pauci-immune rapidly progressive GN. Many crescentic parietal epithelial cells (PECs) strongly expressed PCNA ( A ). Cytokeratine (CK) was strongly expressed in the majority of these proliferating crescentic PECs, as well as in the tubular epithelial cells. ( B ). Macrophages (CD68 (+) cells) are identified in these cellular crescents and also as interstitial infiltrating cells ( C ). α-SMA (+) cells were found in crescentic lesions, as well as in periglomerular interstitial fibrosis ( D ). Most of these proliferating crescentic PECs are negative for E-cadherin, that is preserved in some of the epithelial tubular cells (Case 20) ( E ). In this fibrocellular crescent, no cell expressed synaptopodin, a podocyte marker (Case 29) ( F ). (magnification ×200). The gremlin expression was studied by in situ hybridization and by immunohistochemistry, as it is illustrated in Figure 2 . Gremlin was not expressed in normal human kidney ( Figure 2 C). Conversely, abundant gremlin mRNA expression ( Figure 2 A–D) and gremlin protein staining ( Figure 2 B) were observed in cellular glomerular crescents. As it is shown in Figure 2 B, the gremlin protein was detected in proliferating cellular crescents, and this expression was not observed in those fibrous crescents. Unexpectedly, abundant gremlin expression was also observed in tubular epithelial cells, as it is illustrated in Figure 2 A, and a strong correlation between the gremlin mRNA and protein expression was observed in the samples studied ( r = 0.8; P < 0.01). Fig. 2. View largeDownload slide In-situ hybridization (ISH) and IMH demonstrating gremlin mRNA and gremlin protein in crescentic PECs and tubular epithelial cells of patients with pauci-immune crescentic GN. In comparison with normal renal tissue in which there is no expression of gremlin mRNA ( C ), proliferating PECs of glomerular crescents show a strong expression of gremlin mRNA by ISH ( A – D ) (case 7) (magnification ×200 and ×400); and gremlin protein expression by IMH (case 7) ( B ) (magnification ×200); immune competent infiltrating interstitial cells are also strongly positive for gremlin staining ( E ); and CTGF was also expressed in these glomerular crescentic cells. (case 20) ( F ). Fig. 2. View largeDownload slide In-situ hybridization (ISH) and IMH demonstrating gremlin mRNA and gremlin protein in crescentic PECs and tubular epithelial cells of patients with pauci-immune crescentic GN. In comparison with normal renal tissue in which there is no expression of gremlin mRNA ( C ), proliferating PECs of glomerular crescents show a strong expression of gremlin mRNA by ISH ( A – D ) (case 7) (magnification ×200 and ×400); and gremlin protein expression by IMH (case 7) ( B ) (magnification ×200); immune competent infiltrating interstitial cells are also strongly positive for gremlin staining ( E ); and CTGF was also expressed in these glomerular crescentic cells. (case 20) ( F ). Moreover, gremlin expression was most prominent in areas of tubulointerstitial fibrosis, and was also observed in interstitial inflammatory cells, as it is illustrated in Figure 2 E, with a strong correlation between the tubular and interstitial gremlin expression and the tubulointerstitial fibrosis score ( r = 0.59; P < 0.01) ( Figure 5 ). Since gremlin induction has been reported in tubular epithelial cells exposed to TGF-β, we studied the TGF-β and gremlin mRNA expression by ISH in serial sections of some biopsies. TGF-β mRNA was not detected by ISH in normal renal tissue (data not shown), however, there was a marked increase in association with glomerular crescent formation. As it is illustrated in Figure 3 A and B (case 7), we observed a strong co-localization of TGF-β and gremlin expression in the crescentic cells. Moreover, we also observed a strong co-localization of TGF-β and gremlin on the tubular epithelial cells as it is shown in Figures 3 D and E (case 24). Fig. 3. View largeDownload slide Co-expression of TGF-β mRNA, gremlin mRNA and Smads in glomerular crescentic cells and tubular cells of patients with RPGN. Strong expression of TGF-β mRNA in proliferating crescentic cells by ISH ( A ) and gremlin mRNA in sections of the same biopsy ( B ). Activated pSmad2 was also observed by IMH in the crescentic glomerular, tubular and infiltrating interstitial cells ( C ) (case 7). On serial sections we observed tubular TGF-β mRNA transcription ( D ) and gremlin mRNA expression ( E ), on the same tubuli (case 16). Smad3 translocated to the nucleus is demonstrated by SWH on tubular cells ( F ) (case 24) (magnification ×200). Fig. 3. View largeDownload slide Co-expression of TGF-β mRNA, gremlin mRNA and Smads in glomerular crescentic cells and tubular cells of patients with RPGN. Strong expression of TGF-β mRNA in proliferating crescentic cells by ISH ( A ) and gremlin mRNA in sections of the same biopsy ( B ). Activated pSmad2 was also observed by IMH in the crescentic glomerular, tubular and infiltrating interstitial cells ( C ) (case 7). On serial sections we observed tubular TGF-β mRNA transcription ( D ) and gremlin mRNA expression ( E ), on the same tubuli (case 16). Smad3 translocated to the nucleus is demonstrated by SWH on tubular cells ( F ) (case 24) (magnification ×200). In addition, in cultured human monocyte cells, we have observed that treatment with TGF-β increased gremlin production in a dose-dependent way after 18 h (data not shown) remaining elevated until 24 h ( Figure 6 ). However, in these cells TGF-β did not increase CTGF production (data not shown). This result suggests that the main inducer of gremlin in crescentic, and interstitial cells could be TGF-β, moreover, considering the participation of monocytic cells in the crescents formation and into the interstitial cells. As TGF-β activate a unique signal transduction pathway acting through the Smad family of proteins, we studied the expression of activated Smad by southwestern histochemistry and IMH. In Figure 3 C, we observed a definitive nuclear pSmad2 activation by IMH in the crescentic glomerular cells. In addition, we observed the same tubular cells on serial sections, positive for gremlin mRNA ( Figure 4 A), and activated pSmad2 by IMH ( Figure 4 B) (case 24). Fig. 4. View largeDownload slide Co-localization of gremlin and Smad in tubular cells and BMP-7 expression in pauci-immune GN. In situ hybridization and immunohistochemistry demonstrating co-localization of tubular gremlin mRNA ( A ) and pSmad2 ( B ) on the same tubular cells. Serial sections of case 24 (magnification 400×). BMP-7 normally expressed in the luminal side of distal tubules ( D ), was significantly decreased in renal sections of patients with pauci-immune RPGN (C). Fig. 4. View largeDownload slide Co-localization of gremlin and Smad in tubular cells and BMP-7 expression in pauci-immune GN. In situ hybridization and immunohistochemistry demonstrating co-localization of tubular gremlin mRNA ( A ) and pSmad2 ( B ) on the same tubular cells. Serial sections of case 24 (magnification 400×). BMP-7 normally expressed in the luminal side of distal tubules ( D ), was significantly decreased in renal sections of patients with pauci-immune RPGN (C). Fig. 5. View largeDownload slide Correlation between tubule interstitial gremlin mRNA expression and tubule interstitial fibrosis. Gremlin expression correlates with fibrosis score ( r = 0.59; P < 0.01). Points represent individual values for patients with pauci immune crescentic GN. Fig. 5. View largeDownload slide Correlation between tubule interstitial gremlin mRNA expression and tubule interstitial fibrosis. Gremlin expression correlates with fibrosis score ( r = 0.59; P < 0.01). Points represent individual values for patients with pauci immune crescentic GN. Fig. 6. View largeDownload slide In human monocyte cells TGF-β increases gremlin expression in a dose dependent way. Cultured human monocyte cells were stimulated with TGF-β (range 0, 1–10 ng/ml) for 24 h. Results of total gremlin production were obtained from densitometric analysis and expressed as ratio Gremlin/Tubuline as n -fold over control. Figure shows in the top panel a representative western blot and in the bottom panel, data of gremlin production as mean ± SEM of 4–7 independent experiments. * P < 0.05 vs control. Fig. 6. View largeDownload slide In human monocyte cells TGF-β increases gremlin expression in a dose dependent way. Cultured human monocyte cells were stimulated with TGF-β (range 0, 1–10 ng/ml) for 24 h. Results of total gremlin production were obtained from densitometric analysis and expressed as ratio Gremlin/Tubuline as n -fold over control. Figure shows in the top panel a representative western blot and in the bottom panel, data of gremlin production as mean ± SEM of 4–7 independent experiments. * P < 0.05 vs control. Since the fibrogenic effects of TGF-β are partially mediated by CTGF and this fibrogenic factor participates in the scar formation of crescentic glomerulonephritis, we decided to study the expression of CTGF in these samples. As we denoted in Figure 2 F, there was a strong up-regulation of CTGF in crescentic glomerular cells, capsular adhesions and periglomerular fibrosis, as well as in tubular epithelial cells. In addition, since gremlin is a BMP-7 antagonist, we studied their expression by IMH ( Figure 4 ); BMP-7, predominantly expressed in the luminal side of the collecting duct and distal tubular epithelial cells in the normal kidney, was significantly decreased in tubular cells of patients with crescentic GN, and in close relation with the tubular-interstitial involvement (data not shown). Discussion Glomerular cellular crescents consist of parietal epithelial cells (PECs) and macrophages, which can undergo an irreversible process of fibrous organization. Epithelial-mesenchymal transition (EMT) is the term to describe this conversion [ 5 ] and plays an important role in the genesis of fibroblasts in interstitial renal fibrosis [ 13 ]. A more specific Bowman's epithelial-mesenchymal transition participating in the formation and evolution of glomerular crescents, has been also proposed [ 7–9 ]. This study provides the first phenotypic and morphological evidence that gremlin could participate in glomerular crescent formation. In aggregate, these results indicate that the developmental gene gremlin re-emerges in the context of tubulointerstitial fibrosis and suggests a role for TFG-β as an inducer of gremlin expression in this context. Tubular EMT is an orchestrated, highly regulated process involving four key steps: (i) loss of epithelial cell adhesion by suppression of E-cadherin expression, (ii) de novo α-smooth muscle actin expression and actin reorganization, (iii) disruption of the tubular basement membrane and (iv) enhanced cell migration and invasion. TGF-β as a sole factor is capable of inducing epithelial cells to undergo all four steps [ 13 ]. In this study, we have observed that PECs lost expression of the epithelial marker, E-cadherin and some cells co-expressed CK and α-SMA suggesting a transitional phase in the dynamic phenomenon of EMT. In addition, cellular crescents contained small numbers of α-SMA-positive myofibroblasts, that become the dominant population in fibrocellular crescents, in which CTGF appears. The findings showing PCNA-positive crescentic cells, the majority of them dysregulated PECs that become negative for CK, followed by some PECs still positive for CK, macrophagic cells and myofibroblasts, extend and confirm the observations of EMT in pauci-immune crescentic GN reported by Bariety [ 9 ]. Recently, we have reported that gremlin mRNA levels by real-time polymerase chain reaction, were significantly elevated in biopsy specimens with diabetic nephropathy compared with normal tissue [ 4 ]. Gremlin expression was not increased in specimens of either minimal change disease or IgA nephropathy. Interestingly, gremlin mRNA was increased in some samples of crescentic nephritis (data not shown), and in diabetic specimens there was a direct correlation between gremlin mRNA levels and tubulointerstitial fibrosis [ 4 ], supporting the present data in which we have observed a strong gremlin mRNA and protein expression in the glomerular and interstitial compartment of specimens with crescentic GN. Co-localization of gremlin and TGF-β1 expression in crescentic cells suggests that TGF-β1, may also modulate gremlin expression in the glomeruli; this hypothesis is strengthened further because co-localization of expression is also evident in tubular cells, particularly in areas where tubulointerstitial fibrosis has developed. In addition, we have shown that gremlin expression is induced in vitro by TGF-β in human monocytic cells, extending those observations performed on renal proximal tubule cells undergoing transdifferentiation to a fibroblast phenotype [ 3 ], and supporting a role for gremlin in the pathogenesis of tubulointerstitial fibrosis. As noted, gremlin is a BMP antagonist, by heterodimerization with BMP 2, 4 and 7 and thereby prevents receptor binding [ 2 ]. Recently, it has been reported that BMP-7 reverses TGF-β1-induced epithelial to mesenchymal transition by reinduction of E-cadherin, a key epithelial cell adhesion molecule [ 14 ]. In this context it is very intriguing that administration of the gremlin ligand BMP-7 is protective in models of progressive renal disease [ 15 ], raising the possibility that gremlin may even be a therapeutic target. Actually, we have observed that the tubular BMP-7 expression is significantly decreased in biopsies of patients with progressive crescentic GN in comparison with normal renal tissue, and these findings are coincident with the loss of tubular BMP-7 and the increase in gremlin observed in experimental diabetic nephropathy [ 15 ]. TGF-β is known to mediate its fibrotic effects by activating the receptor-associated Smads and the discovery of the TGF-β/Smad signalling pathway has allowed to study the intracellular mechanisms of TGF-β in EMT [ 16 ]. The Smads are divided into three categories. The first category consists of the receptor-activated or pathway-restricted Smads (R-Smads), that include Smad2 and Smad3, and are activated by relatively specific TGF-β receptor ligands. The phosphorylated R-Smads associate to form a heteromultimer that includes the second type of Smad, the common-partner Smad 4. The complex is then translocated to the nucleus, where it can regulate target gene transcription. A third category comprises of the inhibitory Smads (Smad 6, Smad 7) [ 17 ] that appear to function as competitive inhibitors of Smad activation. The complementary technique of SWH performed in this study using a labelled Smad-binding element demonstrated increased binding of nuclear proteins to the Smad-binding element, indicating active signalling downstream of the TGF-β stimulus. The observed co-localization of TGF-β, and activated Smad2 and Smad3 translocated into the nucleus, confirm the TGF-β/Smad-signalling pathway in the EMT. Furthermore, the co-localization of the BMP antagonist, gremlin, on the same glomerular and tubular cells suggests that the gremlin induction by TGF-β could be occurring through the TGF-β/Smad-signalling pathway. Recent results suggest that CTGF may play a crucial role in the renal EMT and the subsequent deposition/degradation process of ECM during tubulointerstitial fibrosis [ 18 ]. In this context, the present study confirms observations of Ito et al . [ 19 ] that first demonstrated that CTGF is expressed in the crescents of human GN, and observations of Kanemoto et al . [ 20 ], that showed the participation of CTGF in scar formation of crescentic GN. Interestingly, CTGF mRNA was localized in the PECs, but not in macrophages, suggesting that inflammatory cells are not a major source of CTGF in cellular crescents [ 20 ]. Our in vitro results confirm that monocytes are not a source of CTGF when stimulated by TGF-β. Instead, gremlin is induced. The regulatory mechanisms for CTGF and gremlin in cellular crescents need to be determined, although it is well known that CTGF play an important role in mediating the profibrotic effects of TGF-β [ 18 ]. Of interest, on the other hand, in vitro studies performed in cultured human proximal tubular epithelial cells have shown that TGF-β1 upregulated CTGF gene expression, preceding that of α-SMA and fibronectin. The α-SMA was significantly inhibited by the CTGF antisense oligodeoxynucleotide [ 18 ]. In conclusion, we have reported the presence of gremlin in crescentic proliferating cells, the majority PECs, which are evolving to a myofibroblast phenotype, that suggests gremlin as a downstream mediator of TGF-β either acting as an inhibitory trap protein for BMP-7 creating a profibrotic positive loop or directly by promoting the transdifferentiation of epithelial cells. Studies are in progress to demonstrate the exact role of gremlin in the EMT. Acknowledgement Supported by grant FONDECYT, Chile, 1040163, 7060074 and Ministerio de Educación y Ciencia, España SAF:2005-03378. Conflict of interest statement. None declared. References 1 Topol LZ,  Bardot B,  Zhang Q, et al.  Biosynthesis, post-translation modification, and functional characterization of Drm/Gremlin,  J Biol Chem ,  2000, vol.  275 (pg.  8785- 8793) Google Scholar CrossRef Search ADS PubMed  2 Roxburgh SA,  Murphy M,  Pollock CA,  Brazil DP.  Recapitulation of embryological programmes in renal fibrosis. The importance of epithelial cell plasticity and developmental genes,  Nephron Physiol ,  2006, vol.  103 (pg.  139- 148) Google Scholar CrossRef Search ADS   3 McMahon R,  Murphy M,  Clarkson M, et al.  IHG-2, a mesangial cell gene induced by high glucose, is human gremlin. Regulation by extracellular glucose concentration, cyclic mechanical strain, and transforming growth factor-beta1,  J Biol Chem ,  2000, vol.  275 (pg.  9901- 9904) Google Scholar CrossRef Search ADS PubMed  4 Dolan V,  Murphy M,  Sadlier D, et al.  Expression of gremlin, a BMP antagonist, in human diabetic nephropathy,  Am J Kidney Dis ,  2005, vol.  45 (pg.  1034- 1039) Google Scholar CrossRef Search ADS PubMed  5 Kalluri R,  Neilson EG.  Epithelial-mesenchymal transition and its implications for fibrosis,  J Clin Invest ,  2003, vol.  112 (pg.  1776- 1784) Google Scholar CrossRef Search ADS PubMed  6 Zeisberg M.  Bone morphogenic protein-7 and the kidney: current concepts and open questions,  Nephrol Dial Transplant ,  2006, vol.  21 (pg.  568- 573) Google Scholar CrossRef Search ADS PubMed  7 Ng Y-Y,  Fan J-M,  Mu W, et al.  Glomerular epithelial-myofibroblast transdifferentiation in the evolution of glomerular crescent formation,  Nephrol Dial Transplant ,  1999, vol.  14 (pg.  2860- 2872) Google Scholar CrossRef Search ADS PubMed  8 Fujigaki Y,  Sun DF,  Fujimoto T, et al.  Mechanisms and kinetics of Bowman's epithelial-myofibroblast transdifferentiation in the formation of glomerular crescents,  Nephron ,  2002, vol.  92 (pg.  203- 212) Google Scholar CrossRef Search ADS PubMed  9 Bariéty J,  Hill GS,  Mandet Ch, et al.  Glomerular epithelial-mesenchymal transdifferentiation in Pauci-immune crescentic glomerunephritis,  Nephrol Dial Transplant ,  2003, vol.  18 (pg.  1777- 1784) Google Scholar CrossRef Search ADS PubMed  10 Mezzano S,  Droguett A,  Burgos ME, et al.  Overexpression of chemokines, fibrogenic cytokines and myofibroblasts in human membranous nephropathy,  Kidney Int ,  2000, vol.  57 (pg.  147- 158) Google Scholar CrossRef Search ADS PubMed  11 Isono M,  Chen Sh,  Hong SW, et al.  Smad pathway is activated in the diabetic mouse kidney and Smad3 mediates TGF-β-induced fibronectin in mesangial cells,  Bioch Bioph Res Comm ,  2002, vol.  296 (pg.  1356- 1365) Google Scholar CrossRef Search ADS   12 Mezzano S,  Barría M,  Droguett A, et al.  Tubular NFkβ and AP-1 activation in human proteinuric renal disease,  Kidney Int ,  2001, vol.  60 (pg.  1366- 1377) Google Scholar CrossRef Search ADS PubMed  13 Liu Y.  Epithelial to mesenchymal transition in renal fibrogenesis: pathologic significance, molecular mechanism and therapeutic intervention,  J Am Soc Nephrol ,  2004, vol.  15 (pg.  1- 12) Google Scholar CrossRef Search ADS PubMed  14 Zeisberg M,  Hanai J,  Sugimoto H, et al.  BMP-7 counteracts TGF-β1-induced epithelial-to-mesenchymal transition and reverses chronic renal injury,  Nat Med ,  2003, vol.  9 (pg.  964- 968) Google Scholar CrossRef Search ADS PubMed  15 Wang Sh-N,  Lapage J,  Hirschberg R.  Loss of tubular bone morphogenetic protein-7 in diabetic nephropathy,  J Am Soc Nephrol ,  2001, vol.  12 (pg.  2392- 2399) Google Scholar PubMed  16 Bottinger E,  Bitzer M.  TGF-β signaling in renal disease,  J Am Soc Nephrol ,  2002, vol.  13 (pg.  2600- 2610) Google Scholar CrossRef Search ADS PubMed  17 Schiffer M,  Schiffer LE,  Gupta A, et al.  Inhibitory Smads and TGF-β Signaling in glomerular cells,  J Am Soc Nephrol ,  2002, vol.  13 (pg.  2657- 2666) Google Scholar CrossRef Search ADS PubMed  18 Zhang Ch,  MenG X,  Zhu Zh, et al.  Role of connective tissue growth factor in renal tubular epithelial-myofibroblast transdifferentiation and extracellular matrix accumulation in vitro,  Life Sciences ,  2004, vol.  75 (pg.  367- 379) Google Scholar CrossRef Search ADS PubMed  19 Ito Y,  Aten J,  Bende R, et al.  Expression of connective tissue growth factor in human renal fibrosis,  Kidney Int ,  1998, vol.  53 (pg.  853- 861) Google Scholar CrossRef Search ADS PubMed  20 Kanemoto K,  Usui J,  Tomari SH, et al.  Connective tissue growth factor participates in scar formation of crescentic glomerulonephritis,  Lab Invest ,  2003, vol.  83 (pg.  1615- 1625) Google Scholar CrossRef Search ADS PubMed  © The Author [2007]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org
TI - Expression of gremlin, a bone morphogenetic protein antagonist, in glomerular crescents of pauci-immune glomerulonephritis
JF - Nephrology Dialysis Transplantation
DO - 10.1093/ndt/gfm145
DA - 2007-04-01
UR - https://www.deepdyve.com/lp/oxford-university-press/expression-of-gremlin-a-bone-morphogenetic-protein-antagonist-in-kOpqPh1dBl
SP - 1882
EP - 1890
VL - 22
IS - 7
DP - DeepDyve
ER -