- Email: [email protected]
The Influence of Tubular Phenotypic Changes on the Development of Diffuse Interstitial Fibrosis in Renal Allografts
The Influence of Tubular Phenotypic Changes on the Development of Diffuse Interstitial Fibrosis in Renal Allografts B.H. Özdemir, A.A. Özdemir, T. Colak, S. Sezer, and M. Haberal ABSTRACT It has been reported that myofibroblasts are the major cells in the development of interstitial fibrosis (IF) and, therefore, chronic dysfunction in renal allografts. Our aim was two fold: first, to understand the key markers controlling tubular and glomerular epithelial-to-mesenchymal transition (EMT); second, to show the role of tubular EMT on the development of interstitial fibrosis (IF) and the role of glomerular EMT on the development of proteinuria and, therefore, graft survival. For this purpose we evaluated the vinculin- and paxillin-containing adhesion complexes and ␣-smooth muscle actin (␣-SMA) expression on tubular cells and glomerular podocytes in first year renal allograft biopsy specimens of 74 patients. We established the proteinuric state at the time of the biopsy of all patients. Follow-up biopsy specimens of all patients were evaluated for the development of diffuse IF (ⱖ50% of the biopsy specimen). Among 74 patients, 21 showed grade 1 and 9 showed grade 2 tubular EMT. Only 25/74 cases showed glomerular EMT. The incidence of the development of diffuse IF at 18 and 24 months after the initial biopsy and the graft loss at 5 years were higher among subjects with tubules and glomerular podocytes showing EMT (P ⬍ .001 for all). The development of IF and graft loss was significantly earlier in cases with grade 2 compared with grade 1 or no tubular EMT (P ⬍ .001 for all). The proteinuric state at the time of the biopsy showed a significant positive correlation with the glomerular EMT (P ⬍ .001). In conclusion, our results showed that renal tubular cells and glomerular podocytes can undergo epithelial-to-mesenchymal differentiation. The transformed cells with reorganized cytoskeletal features have an important role in renal allograft survival by inducing the development of diffuse IF and proteinuria. OSS of nephrons together with the development of tubulointerstitial fibrosis are processes that lead to renal failure, which is common in kidney allografts. Nephron loss and development of interstitial fibrosis (IF) can be linked directly to the aggravated transition of tubular epithelial cells to myofibroblasts. Myofibroblasts are activated fibroblasts identified in humans by the expression of ␣-smooth muscle actin (␣SMA).1 The expression of ␣-SMA and the formation of stress fibers and paxillin- and vincullin-containing adhesion complexes are the primary criteria determining the differentiation of nonmuscle cells into contractile myofibroblasts.2 Myofibroblasts have the ability to proliferate and to synthesize collagen. Recent studies have reported that myofibroblasts may be derived from tubular epithelial cells by epithelialto-mesenchymal transition (EMT); however, the details
L
© 2011 by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710 Transplantation Proceedings, 43, 527–529 (2011)
regarding the conversion between these 2 cell types are poorly understood. Previous reports have suggested that at the cellular level tubular EMT is likely a step-wise process involving several key events.3 Proteinuria of any cause has been suggested to be associated with changes in glomerular intercellular adhesion and in addition to eventual EMT of glomerular cells inducing glomerulosclerosis.4 Mesangial cells, tubulointerstitial fibroblasts, and perhaps tubular epithelial cells unFrom the Departments of Pathology (B.H.Ö.), Public Health (A.A.Ö.), Nephrology (T.C., S.S.), and General Surgery and Transplant Surgery (M.H.), Baskent University, Ankara, Turkey. Address reprint requests to Binnaz Handan Özdemir, MD, Professor, Department of Pathology, Bas¸kent University, 12. sokak 7/4, Bahcelievler, 06490, Ankara, Turkey. E-mail: handan27@ hotmail.com 0041-1345/–see front matter doi:10.1016/j.transproceed.2011.01.017 527
528
dergo myofibroblast differentiation, a process characterized by ␣-actin expression, formation of stress fibers, and synthesis of interstitial collagens. Inhibition of myofibroblast differentiation, which might prevent progression of kidney disease, may be difficult to achieve because it may require inhibition of multiple signaling pathways. In the present study our aim was two fold: first, to understand the key markers controlling tubular and glomerular EMT. For this purpose we evaluated the vinculin- and paxillin-containing adhesion complexes and ␣-SMA expression on tubular and glomerular cells seeking to define tubular and glomerular EMT. Second, we sought to show the role of tubular EMT on the development of IF and the role of glomerular EMT on the development of proteinuria and, therefore, graft survival.
MATERIALS AND METHODS Renal allograft biopsy specimens from 74 patients were obtained during the first year. Those showing IF in ⬎10% of the biopsy specimen were not included in the study. Renal grafts had been obtained from cadaveric (n ⫽ 16) or living (n ⫽ 58) donors. Mean follow-up time was 59.6 ⫾ 32 months. None of the patients showed humoral rejection. We established the proteinuric status of all patients at the time of the biopsy. A 3-drug immunosuppressive protocol was used in all patients. The biopsy specimens were evaluated for the development of diffuse IF (ⱖ50% of the biopsy specimen). Patients were grouped into 2 groups regarding the development of IF during 18 or 24 months after initial biopsy. All specimens were immunostained for epithelial, cytoskeleton, and mesenchymal markers. Antibody against cytokeratin (mouse monoclonal antibody; Dako, Denmark) was used to detect epithelial markers. Mesenchymal and cytoskeletal markers were stained with ␣-SMA (mouse monoclonal antibody; Dako, Denmark), vinculin and paxillin (mouse monoclonal, clone 5H11; Neomarkers, Calif, United States). We assessed the expression of cytokeratin, ␣-SMA, vinculin, and paxillin both on glomeruli and in tubules using the following: Grade 0, proximal and/or distal tubule cells and glomerular podocytes expressing only cytokeratin without ␣-SMA, vinculin, or paxillin; Grade 1, proximal and/or distal tubule cells and glomerular podocytes co-expressing both cytokeratin and all of the mesenchymal and/or cytoskeletal markers (␣-SMA, vinculin, and paxillin) (This stage represented the transition between epithelia and mesenchyme.); and Grade 2, proximal and/or distal tubule cells and glomerular podocytes expressing all mesenchymal and/or cytoskeletal markers (␣-SMA, vinculin, and paxillin) but not cytokeratin.
Statistical Analysis Statistical analysis was performed using SPSS software (Statistical Package for the Social Sciences, version 9.05, SSPS Inc, Chicago, Ill, United States). For quantitative variables, results are given as mean values ⫾ SD, which were compared using Student t test for normal or Kruskal-Wallis test for non-normal distribution. Categorical data were compared using either the chi-square test or Fisher exact probability test, when the expected cells contained 5 or fewer subjects. Results were considered statistically significant at P ⬍ .05.
ÖZDEMIR, ÖZDEMIR, COLAK ET AL
RESULTS
Of 74 patients, 21 (28.4%) showed grade 1 and 9 (12.2%) grade 2 tubular EMT. Only 25/74 (33.8%) cases showed grade 2 glomerular EMT. The incidence of graft loss during 5 years and the development of diffuse IF at 18 and 24 months after initial biopsy were higher among cases with tubule and glomerular podocytes showing EMT (Table 1). The development of IF and graft loss over 5 years were increased with greater tubular EMT grade (Table 1). The development of diffuse IF occurred at 11 ⫾ 1.8 months in cases with grade 2 versus 25.6 ⫾ 6.2 months for grade 1 or 31.8 ⫾ 11 months for no tubular EMT (P ⬍ .001). In addition the development of IF showed a significant difference between grade 1 and grade 2 tubular EMT (P ⬍ .001). Graft loss occurred significantly earlier in cases with grade 2 (19.6 ⫾ 8.4 months) than grade 1 (47 ⫾ 24.8 months) or no tubular EMT (73.7 ⫾ 30 months; P ⬍ .001). The proteinuric state at the time of biopsy showed a significant positive correlation with the glomerular EMT (P ⬍ .001). Of 25 cases with glomerular EMT, 13 (52%) displayed proteinuria. Whereas only 11/49 (22.4%) cases without glomerular EMT showed proteinuria (Table 2). The development of IF and graft loss were significantly earlier in cases with compared with without glomerular EMT (29.3 ⫾ 9.7 versus 18 ⫾ 10.3 months for IF and 70.6 ⫾ 31 versus 37.8 ⫾ 22 months for graft loss, respectively; P ⬍ .001 for all). DISCUSSION
Tubular EMT, by definition, is a process in which renal tubular cells lose epithelial phenotype, acquiring new characteristic features of mesenchyme. The hypothesis of EMT in the development of IF5 has been widely supported by several outstanding studies 6 – 8 Our results also confirmed this hypothesis: patients with tubular EMT showed significantly earlier development of IF than those without tubular EMT. In previous studies the potential role of tubular epithelial cells in renal fibrosis has been overlooked. The possible reason may have been due to using only ␣-SMA antibody as Table 1. Comparison of Tubular and Glomerular EMT With the Development of IF During 18 and 24 Months and With Graft Loss During 5 Years
Tubular EMT Grade 0 Grade 1 Grade 2 Glomerular EMT Grade 0 Grade 1 Grade 2 *P ⬍ .001. † P ⫽ .001.
n
IF 18 Months (⫹)
IF 24 Months (⫹)
Graft Loss (%)
44 21 9
4 (9.1%)* 3 (14.3%) 9 (100%)
8 (18.2%)* 12 (57.1%) 9 (100%)
11 (25%)* 15 (71.4%) 9 (100%)
49 0 25
5 (10.2%)† 0 11 (44%)
11 (22.4%)* 0 18 (72%)
15 (30.6%) 0 20 (80%)
DIFFUSE INTERSTITIAL FIBROSIS
529
Table 2. The Relationship Between the Glomerular EMT and the Presence of Proteinuria
Glomerular EMT Grade 0 Grade 1 Grade 2
n
Proteinuria No
Proteinuria Yes
49 0 25
38 (77.6%) 0 12 (48%)
11 (22.4%)* 0 13 (52%)
*P ⫽ .01.
a marker of epithelial-to-mesenchymal transdifferentiation. Tubular ␣-SMA expression alone is not sufficient to define tubular EMT. Because even in a homogenous cell population that originates from a single cell clone only a small fraction of cultured tubular cells express ␣-SMA in response to TGF-1 stimulation.9 Nevertheless, ␣-SMA cannot be construed as the gold standard for all collagenexpressing cells during fibrogenesis. For this reason, in addition to ␣-SMA we stained biopsy specimens with other cytoskeletal markers—vinculin and paxillin. Only cells that co-expressed all mesenchymal and cytoskeletal markers were accepted as displaying EMT. In the present study we noted co-expression of both cytokeratin and all mesenchymal markers in some tubules, indicating that they were in a transitional stage between epithelia and mesenchyme. We found a significant difference between tubules showing pure EMT (grade 2) and those in a transitional stage (grade 1) in regard to the development of IF; grade 2 EMT cases showed significantly earlier evolution to IF compared with transitional stage EMT (grade 1). This finding suggested that tubules must show complete EMT by loss of cytokeratin expression, for the a complete fibrogenesis process of the kidney. One important feature of EMT for the development of IF during chronic renal allograft dysfunction is the potential for transforming epithelial cells to migrate across the basement membrane from tubules into the interstitium.10 The reorganization of the cytoskeleton of tubular cells and induction of ␣-SMA may provide a structural foundation, not only to define the morphology of the transformed cells but also for them to migrate, invade, and even acquire the capacity for contractility.3 We noted a significant relationship between ␣-SMA expression by glomerular podocytes and the development of proteinuria. Prior to and during the development of proteinuria, visceral epithelial cells (podocytes) expressed ␣-SMA and lost expression of the epithelial marker cyto-
keratin. In principle, proteinuria may be caused by podocyte defects. Thus we suggest that mesenchymal differentiation of podocytes alters their actin cytoskeleton and, therefore, induces the podocyte injury in proteinuria cases. Changes in the podocyte cytoskeleton due to transdifferentiation could provide a route by which the glomerular filtration rate is modified, or the matrix synthesis is augmented. An active role of glomerular EMT has been also reported in anti-glomerular basement membrane glomerulonephritis, diabetic nephropathy, and cresentic glomerulonephritis.11,12 In conclusion, our results showed that renal tubular cells and glomerular podocytes can undergo epithelialto-mesenchymal differentiation. The transformed cells with reorganized cytoskeletal features have an important role in renal allograft survival by inducing the development of diffuse IF and proteinuria.
REFERENCES 1. Gabbiani G: The biology of the myofibroblast. Kidney Int 41:530, 1992 2. Brown MC, Turner CE: Paxillin: adapting to change. Physiol Rev 84:1315, 2004 3. Yang J, Liu Y: Dissection of key events in tubular epithelial to myofibroblast transition and its implications in renal interstitial fibrosis. Am J Pathol 159:1465, 2001 4. Bains R, Furness PN, Critchley R: A quantitative immunofluorescence study of glomerular cell adhesion proteins in proteinuric states. J Pathol 183:272, 1997 5. Strutz F, Muller GA: Transdifferentiation: a new angle on renal fibrosis. Exp Nephrol 4:267, 1996 6. Okada H, Inoue T, Suzuki H, et al: Epithelial-mesenchymal transformation of renal tubular cells in vitro and in vivo. Nephrol Dial Transplant 15(suppl 6):44, 2000 7. El Nahas AM: Plasticity of kidney cells: role in kidney remodeling and scarring. Kidney Int 64:1553, 2003 8. Lan HY: Tubular epithelial-myofibroblast transdifferentiation mechanisms in proximal tubule cells. Curr Opin Nephrol Hypertens 12:25, 2003 9. Iwano M, Plieth D, Danoff TM, et al: Evidence that fibroblasts derive from epithelium during tissue fibrosis. J Clin Invest 110:341, 2002 10. Robertson H, Ali S, Mc Donnel BJ, et al: Chronic renal allograft dysfunction: the role of T cell-mediated tubular epithelial to mesenchymal cell transition. J Am Soc Nephrol 15:390, 2004 11. Bariety J, Hill GS, Mandet C, et al: Glomerular epithelialmesenchymal transdifferentiation in pauci-immune cresentic glomerulonephritis. Nephrol Dial Transplant 18:1777, 2003 12. Zeisberg M, Hanai J, Sugimoto H, et al: BMP-7 counteracts TGF-beta1-induced epithelial-to-mesenchymal transition and reverses chronic renal injury. Nat Med 9:964, 2003
L
© 2011 by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710 Transplantation Proceedings, 43, 527–529 (2011)
regarding the conversion between these 2 cell types are poorly understood. Previous reports have suggested that at the cellular level tubular EMT is likely a step-wise process involving several key events.3 Proteinuria of any cause has been suggested to be associated with changes in glomerular intercellular adhesion and in addition to eventual EMT of glomerular cells inducing glomerulosclerosis.4 Mesangial cells, tubulointerstitial fibroblasts, and perhaps tubular epithelial cells unFrom the Departments of Pathology (B.H.Ö.), Public Health (A.A.Ö.), Nephrology (T.C., S.S.), and General Surgery and Transplant Surgery (M.H.), Baskent University, Ankara, Turkey. Address reprint requests to Binnaz Handan Özdemir, MD, Professor, Department of Pathology, Bas¸kent University, 12. sokak 7/4, Bahcelievler, 06490, Ankara, Turkey. E-mail: handan27@ hotmail.com 0041-1345/–see front matter doi:10.1016/j.transproceed.2011.01.017 527
528
dergo myofibroblast differentiation, a process characterized by ␣-actin expression, formation of stress fibers, and synthesis of interstitial collagens. Inhibition of myofibroblast differentiation, which might prevent progression of kidney disease, may be difficult to achieve because it may require inhibition of multiple signaling pathways. In the present study our aim was two fold: first, to understand the key markers controlling tubular and glomerular EMT. For this purpose we evaluated the vinculin- and paxillin-containing adhesion complexes and ␣-SMA expression on tubular and glomerular cells seeking to define tubular and glomerular EMT. Second, we sought to show the role of tubular EMT on the development of IF and the role of glomerular EMT on the development of proteinuria and, therefore, graft survival.
MATERIALS AND METHODS Renal allograft biopsy specimens from 74 patients were obtained during the first year. Those showing IF in ⬎10% of the biopsy specimen were not included in the study. Renal grafts had been obtained from cadaveric (n ⫽ 16) or living (n ⫽ 58) donors. Mean follow-up time was 59.6 ⫾ 32 months. None of the patients showed humoral rejection. We established the proteinuric status of all patients at the time of the biopsy. A 3-drug immunosuppressive protocol was used in all patients. The biopsy specimens were evaluated for the development of diffuse IF (ⱖ50% of the biopsy specimen). Patients were grouped into 2 groups regarding the development of IF during 18 or 24 months after initial biopsy. All specimens were immunostained for epithelial, cytoskeleton, and mesenchymal markers. Antibody against cytokeratin (mouse monoclonal antibody; Dako, Denmark) was used to detect epithelial markers. Mesenchymal and cytoskeletal markers were stained with ␣-SMA (mouse monoclonal antibody; Dako, Denmark), vinculin and paxillin (mouse monoclonal, clone 5H11; Neomarkers, Calif, United States). We assessed the expression of cytokeratin, ␣-SMA, vinculin, and paxillin both on glomeruli and in tubules using the following: Grade 0, proximal and/or distal tubule cells and glomerular podocytes expressing only cytokeratin without ␣-SMA, vinculin, or paxillin; Grade 1, proximal and/or distal tubule cells and glomerular podocytes co-expressing both cytokeratin and all of the mesenchymal and/or cytoskeletal markers (␣-SMA, vinculin, and paxillin) (This stage represented the transition between epithelia and mesenchyme.); and Grade 2, proximal and/or distal tubule cells and glomerular podocytes expressing all mesenchymal and/or cytoskeletal markers (␣-SMA, vinculin, and paxillin) but not cytokeratin.
Statistical Analysis Statistical analysis was performed using SPSS software (Statistical Package for the Social Sciences, version 9.05, SSPS Inc, Chicago, Ill, United States). For quantitative variables, results are given as mean values ⫾ SD, which were compared using Student t test for normal or Kruskal-Wallis test for non-normal distribution. Categorical data were compared using either the chi-square test or Fisher exact probability test, when the expected cells contained 5 or fewer subjects. Results were considered statistically significant at P ⬍ .05.
ÖZDEMIR, ÖZDEMIR, COLAK ET AL
RESULTS
Of 74 patients, 21 (28.4%) showed grade 1 and 9 (12.2%) grade 2 tubular EMT. Only 25/74 (33.8%) cases showed grade 2 glomerular EMT. The incidence of graft loss during 5 years and the development of diffuse IF at 18 and 24 months after initial biopsy were higher among cases with tubule and glomerular podocytes showing EMT (Table 1). The development of IF and graft loss over 5 years were increased with greater tubular EMT grade (Table 1). The development of diffuse IF occurred at 11 ⫾ 1.8 months in cases with grade 2 versus 25.6 ⫾ 6.2 months for grade 1 or 31.8 ⫾ 11 months for no tubular EMT (P ⬍ .001). In addition the development of IF showed a significant difference between grade 1 and grade 2 tubular EMT (P ⬍ .001). Graft loss occurred significantly earlier in cases with grade 2 (19.6 ⫾ 8.4 months) than grade 1 (47 ⫾ 24.8 months) or no tubular EMT (73.7 ⫾ 30 months; P ⬍ .001). The proteinuric state at the time of biopsy showed a significant positive correlation with the glomerular EMT (P ⬍ .001). Of 25 cases with glomerular EMT, 13 (52%) displayed proteinuria. Whereas only 11/49 (22.4%) cases without glomerular EMT showed proteinuria (Table 2). The development of IF and graft loss were significantly earlier in cases with compared with without glomerular EMT (29.3 ⫾ 9.7 versus 18 ⫾ 10.3 months for IF and 70.6 ⫾ 31 versus 37.8 ⫾ 22 months for graft loss, respectively; P ⬍ .001 for all). DISCUSSION
Tubular EMT, by definition, is a process in which renal tubular cells lose epithelial phenotype, acquiring new characteristic features of mesenchyme. The hypothesis of EMT in the development of IF5 has been widely supported by several outstanding studies 6 – 8 Our results also confirmed this hypothesis: patients with tubular EMT showed significantly earlier development of IF than those without tubular EMT. In previous studies the potential role of tubular epithelial cells in renal fibrosis has been overlooked. The possible reason may have been due to using only ␣-SMA antibody as Table 1. Comparison of Tubular and Glomerular EMT With the Development of IF During 18 and 24 Months and With Graft Loss During 5 Years
Tubular EMT Grade 0 Grade 1 Grade 2 Glomerular EMT Grade 0 Grade 1 Grade 2 *P ⬍ .001. † P ⫽ .001.
n
IF 18 Months (⫹)
IF 24 Months (⫹)
Graft Loss (%)
44 21 9
4 (9.1%)* 3 (14.3%) 9 (100%)
8 (18.2%)* 12 (57.1%) 9 (100%)
11 (25%)* 15 (71.4%) 9 (100%)
49 0 25
5 (10.2%)† 0 11 (44%)
11 (22.4%)* 0 18 (72%)
15 (30.6%) 0 20 (80%)
DIFFUSE INTERSTITIAL FIBROSIS
529
Table 2. The Relationship Between the Glomerular EMT and the Presence of Proteinuria
Glomerular EMT Grade 0 Grade 1 Grade 2
n
Proteinuria No
Proteinuria Yes
49 0 25
38 (77.6%) 0 12 (48%)
11 (22.4%)* 0 13 (52%)
*P ⫽ .01.
a marker of epithelial-to-mesenchymal transdifferentiation. Tubular ␣-SMA expression alone is not sufficient to define tubular EMT. Because even in a homogenous cell population that originates from a single cell clone only a small fraction of cultured tubular cells express ␣-SMA in response to TGF-1 stimulation.9 Nevertheless, ␣-SMA cannot be construed as the gold standard for all collagenexpressing cells during fibrogenesis. For this reason, in addition to ␣-SMA we stained biopsy specimens with other cytoskeletal markers—vinculin and paxillin. Only cells that co-expressed all mesenchymal and cytoskeletal markers were accepted as displaying EMT. In the present study we noted co-expression of both cytokeratin and all mesenchymal markers in some tubules, indicating that they were in a transitional stage between epithelia and mesenchyme. We found a significant difference between tubules showing pure EMT (grade 2) and those in a transitional stage (grade 1) in regard to the development of IF; grade 2 EMT cases showed significantly earlier evolution to IF compared with transitional stage EMT (grade 1). This finding suggested that tubules must show complete EMT by loss of cytokeratin expression, for the a complete fibrogenesis process of the kidney. One important feature of EMT for the development of IF during chronic renal allograft dysfunction is the potential for transforming epithelial cells to migrate across the basement membrane from tubules into the interstitium.10 The reorganization of the cytoskeleton of tubular cells and induction of ␣-SMA may provide a structural foundation, not only to define the morphology of the transformed cells but also for them to migrate, invade, and even acquire the capacity for contractility.3 We noted a significant relationship between ␣-SMA expression by glomerular podocytes and the development of proteinuria. Prior to and during the development of proteinuria, visceral epithelial cells (podocytes) expressed ␣-SMA and lost expression of the epithelial marker cyto-
keratin. In principle, proteinuria may be caused by podocyte defects. Thus we suggest that mesenchymal differentiation of podocytes alters their actin cytoskeleton and, therefore, induces the podocyte injury in proteinuria cases. Changes in the podocyte cytoskeleton due to transdifferentiation could provide a route by which the glomerular filtration rate is modified, or the matrix synthesis is augmented. An active role of glomerular EMT has been also reported in anti-glomerular basement membrane glomerulonephritis, diabetic nephropathy, and cresentic glomerulonephritis.11,12 In conclusion, our results showed that renal tubular cells and glomerular podocytes can undergo epithelialto-mesenchymal differentiation. The transformed cells with reorganized cytoskeletal features have an important role in renal allograft survival by inducing the development of diffuse IF and proteinuria.
REFERENCES 1. Gabbiani G: The biology of the myofibroblast. Kidney Int 41:530, 1992 2. Brown MC, Turner CE: Paxillin: adapting to change. Physiol Rev 84:1315, 2004 3. Yang J, Liu Y: Dissection of key events in tubular epithelial to myofibroblast transition and its implications in renal interstitial fibrosis. Am J Pathol 159:1465, 2001 4. Bains R, Furness PN, Critchley R: A quantitative immunofluorescence study of glomerular cell adhesion proteins in proteinuric states. J Pathol 183:272, 1997 5. Strutz F, Muller GA: Transdifferentiation: a new angle on renal fibrosis. Exp Nephrol 4:267, 1996 6. Okada H, Inoue T, Suzuki H, et al: Epithelial-mesenchymal transformation of renal tubular cells in vitro and in vivo. Nephrol Dial Transplant 15(suppl 6):44, 2000 7. El Nahas AM: Plasticity of kidney cells: role in kidney remodeling and scarring. Kidney Int 64:1553, 2003 8. Lan HY: Tubular epithelial-myofibroblast transdifferentiation mechanisms in proximal tubule cells. Curr Opin Nephrol Hypertens 12:25, 2003 9. Iwano M, Plieth D, Danoff TM, et al: Evidence that fibroblasts derive from epithelium during tissue fibrosis. J Clin Invest 110:341, 2002 10. Robertson H, Ali S, Mc Donnel BJ, et al: Chronic renal allograft dysfunction: the role of T cell-mediated tubular epithelial to mesenchymal cell transition. J Am Soc Nephrol 15:390, 2004 11. Bariety J, Hill GS, Mandet C, et al: Glomerular epithelialmesenchymal transdifferentiation in pauci-immune cresentic glomerulonephritis. Nephrol Dial Transplant 18:1777, 2003 12. Zeisberg M, Hanai J, Sugimoto H, et al: BMP-7 counteracts TGF-beta1-induced epithelial-to-mesenchymal transition and reverses chronic renal injury. Nat Med 9:964, 2003