Effect of Smad pathway activation on podocyte cell cycle regulation: an immunohistochemical evaluation

Abstract

Background: Mature podocytes are in cell cycle arrest and their inability to proliferate successfully is a consequence of negative cell-cycle regulators’ expression, such as p57. Phosphorylated smad2/smad3 (pSmad2/3) is an intracellular heteromeric mediator of transforming growth factor beta (TGF-β) signals and, together with co-activators such as P300, regulates gene transcription, including cell cycle regulator proteins. Methods: In order to investigate Smad pathway activation and podocyte cell cycle regulation in glomerular injury, we studied the glomerular immunohistochemical expression of p57, pSmad2/3 and P300 in samples from 67 patients with various types of glomerulonephritis (GN) and 10 normal kidney tissue specimens. Results: pSmad2/3 and p300 expression were found significantly increased in all glomerular cell types in both proliferative and nonproliferative GN, while a significant reduction in p57-positive podocytes was observed when compared to controls. Staining for p57 was found to inversely correlate to pSmad2/3 suggesting that glomerular Smad pathway activation is related to down-regulation of p57 expression in proliferative glomerulonephritis. To our knowledge, this is the first study that indicates a relation between the TGF-beta/Smad signalling pathway and the cell cycle regulatory protein p57 in human GN. Conclusion: The increased pSmad2/3 staining together with the reduced p57 expression found in biopsy specimens with intense interstitial inflammation, indicate a possible relation between interstitial inflammation, glomerular Smad pathway activation and podocyte cell-cycle deregulation.

• Glomerulonephritis
• immunohistochemistry
• p57
• podocytes
• Smad

Introduction

Podocyte injury and glomerulosclerosis are the hallmarks of progressive glomerular damage. Podocytes are terminally differentiated epithelial cells with specialized morphology and molecular architecture that are main contributors to the glomerular filtration barrier. Mature podocytes are in cell cycle arrest and their inability to proliferate successfully is most likely the consequence of robust expression of Cyclin-dependent Kinase inhibitors (CKIs), which are negative regulators of cell-cycle progression.1 p57 is a member of the CKIs family associated with G1 cell cycle arrest.2 In glomerular diseases, p57 expression is affected and cell-cycle deregulation can occur. Specifically, in focal segmental glomerulosclerosis (FSGS) cellular lesions, damaged podocytes have reduced p57 protein expression that is related to unsuccessful podocyte proliferation.3 This podocyte cell-cycle deregulation is thought to be involved in the development of cellular FSGS and even the formation of crescents, although it has been argued that proliferating cells in these lesions mainly originate from parietal epithelial cells.4–7

It is well established that TGF-β has effects on proliferation, hypertrophy, and apoptosis in renal cells.8–10 The expression of TGF-β mRNA and protein in podocytes has been found to be increased in a variety of glomerulopathies, such as FSGS11 and Membranous Nephropathy (MN).12 In TGF-β1 transgenic mice, an acute and massive increase in plasma levels of TGF-β1 results in severe GN with crescent formation.13,14 TGF-β is also strongly expressed in the cellular crescents in the early stage of experimental crescentic GN.15 Furthermore, podocytes covering the cellular lesion of FSGS exhibit increased expression of TGF-β1 protein in glomerular diseases.11,16,17 TGF-β activates the Smad, Ras/extracellular signal-regulated kinase (ERK) and phosphatidyl inositol-3-kinase (PI3K) pathways in podocytes, of which the major TGF-β/Smad signalling pathway appears to override the minor TGF-β -induced Ras/ERK/PI3K pathways.18 The TGF-β/Smad signalling pathway is initiated by ligand-binding induced signal transduction through the TGF-β receptor complex which results in the intracellular mediator Phosphorylated smad2/smad3 (pSmad2/3).19 The activated pSmad2/3 heteromer binds with Smad4, translocates into the nucleus and, together with co-activators such as P300, induces transcription of genes such as the cell-cycle regulators p15 and p21.20–23 This process is negatively regulated by the inhibitory Smad7.24 P300 is a transcriptional co-activator that facilitates pSmad2/3 and Smad4 complexes binding to DNA and enables specific sequence targeting.25–27 Increased expression of Smad2/3 and their activated forms pSmad2/3 in glomeruli of both experimental and human glomerulonephritis has been reported in the literature.11,28 Moreover, it has been shown that Smad2/3 and Smad4 require the co-activator p300 to activate transcription.23,26,27 P300 is also involved in TGF-β/Smad mediated type I collagen gene transcription.29–31

Limited data exist on the glomerular expression patterns of p57, p300 and phosphorylated Smads in human glomerulonephritis (GN). It appears that there is interplay between the TGF-β/Smad signaling pathway and podocyte cell cycle regulation that could be involved in the pathogenesis and disease progression of GN. In order to investigate glomerular Smad pathway activation and p57 expression in podocytes, we studied the glomerular immunohistochemical expression of p57, pSmad2/3 and P300 in various types of human GN.

Material and methods

Patients

Percutaneous renal biopsies were obtained from 67 patients (28 male/39 female) with various types of GN. Renal biopsies were performed for diagnostic purposes and informed consent for the use of excess renal tissue was obtained from the patients. GNs were divided into proliferative (n = 40) and non-proliferative (n = 27). The non-proliferative group of primary GNs included Minimal Change Disease (MCD, n = 3), MN (n = 9), FSGS (n = 5), and cases with WHO class II and V Lupus Nephritis (LN, n = 10). The proliferative group of GNs included IgA nephropathy (IGAN, n = 10), pauci-immune GN (Pauci-Im GN, n = 10) as well as cases with WHO class III and IV LN (n = 20). The summary of the classification and the clinical parameters of patients with GN are given in Table 1. Ten samples of normal kidney tissue that were surgically removed from patients with renal cell carcinoma served as controls.

Table 1. Summary of the classification and the clinical parameters of patients with glomerulonephritis.

Disease	Number of patients	Age (Years)	Gender (M/F)	Creatinine (mg/dl)	Proteinuria (g/24 h)
FSGS	5	42 ± 16.6	4/1	1.87 ± 0.92	2.82 ± 1.54
MN	9	59.1 ± 20.1	5/4	1.05 ± 0.52	4.1 ± 1.83
MCD	3	30.7 ± 7.4	3/0	0.93 ± 0.15	5.75 ± 3.61
pauci-IM-GN	10	61.6 ± 12.3	4/6	3.26 ± 2.32	0.72 ± 0.87
Proliferative LN	20	33.8 ± 2.8	3/17	1.41 ± 0.73	3.56 ± 2.84
Non-proliferative  LN	10	32.2 ± 6.2	1/9	0.8 ± 1.41	2.06 ± 2.84
IGAN	10	40.7 ± 12.8	8/2	1.08 ± 0.46	1.9 ± 1.8

FSGS: Focal segmental glomerulosclerosis, MN: Membranous nephropathy, MCD: Minimal Change Disease, pauci-IM-GN: pauci-immune glomerulonephritis, LN: Lupus nephritis, IGAN: IgA nephropathy. Values expressed in mean ± std. dev.

Immunohistochemistry

Immunohistochemical staining was performed on 4 mm thick formalin-fixed paraffin embedded sections, using a biotin free, two-step, HRP-labelled detection system (DACO EnVision™, Glostrup, Denmark).

Sections were deparaffinised in xylene and rehydrated through graded alcohols. After quenching of endogenous peroxidase activity using a methanol/hydrogen peroxide solution (0.3% in TBS for 30 min), we proceeded to microwave heat-mediated antigen retrieval in 10 mM, pH = 6 citrate buffer and blockage of non-specific binding by incubation in 10% normal horse serum in TBS (Vector Lab, Burlingame, CA) for 30 min at room temperature. Subsequently, sections were incubated for 40 min at 4 °C with the primary antibodies.

The primary antibodies used in our study were: (1) anti-p57 rabbit polyclonal antibody(C20/sc-1040) (Santa Cruz Biotechnology Inc., Dallas, TX) at a dilution 1:80, (2) anti-pSmad2/3 rabbit polyclonal antibody (#3101) (Cell Signalling Technology Inc., Danvers, MA) at a dilution 1:500 and (3) anti-p300 rabbit polyclonal antibody (sc-585) (Santa Cruz Biotechnology Inc.) at a dilution 1:500.

Sections were incubated with a biotin free, two-step, HRP-labelled detection system (DACO EnVision™) for 30 min at room temperature followed by the addition of 3,3′-diaminobenzidine tetrachlorohydrate (DAB), in order to achieve visualization of the requested antigens. The biotin-avidin/streptavidin system is often used as an amplification step to increase sensitivity. However, in some tissues such as kidney, “non-specific” interactions may be a problem due to high levels of endogenous biotin-containing proteins.32 Thus, a biotin free detection system was used in our experiments. Finally, sections were rinsed, counterstained and mounted.

Quantification of pSmad2/3 and p300 immunostaining and histologic injury in human renal biopsies

Serial sections were immunostained for each molecule. The intensity of glomerular staining of pSmad2/3 p300 and p57 was evaluated according to the following 0 to 3 scale: 0: <10% cells with nuclear staining; (1) weak nuclear staining intensity or 11–50% of cells with nuclear staining; (2) moderate nuclear staining intensity and >50% of cells with nuclear staining and (3) strong nuclear staining intensity and >50% of cells with nuclear staining. The staining status of glomerular lesions was also assessed, with immunopositivity being recorded when some degree of immunostaining was noticeable. When no staining was observed, the respective lesions were characterized as negative. Sections were scored independently by two investigators (A.N. and L.N.) blinded to the patients’ clinical profile. The human renal biopsy specimens have been reviewed and analyzed by an independent pathologist, who was blinded to the quantification of immunostaining. The severity of glomerulosclerosis was expressed as the percentage of globally sclerosed glomeruli out of the total number of glomeruli in the renal biopsy. The non-globally sclerosed glomeruli were analyzed for the presence of proliferative segmental lesions, crescents, microadhesions, and segmental sclerotic lesions. The degree of interstitial inflammation, interstitial fibrosis and tubular atrophy was scored between 0 and 3 according to the following: 0, 0%, 1, 1 to 25%; 2, 26 to 50%; and 3, >50% of the biopsy area demonstrating inflammation, fibrosis or atrophy, respectively.

Data are presented as mean ±1 standard deviation. Comparisons between the groups were made with the Kruskal–Wallis test. Mann–Whitney U test was used to compare the molecules’ expression between controls and every form of GN. The Spearman correlation analysis was used to detect possible correlations between staining scores, clinical data and indices of histologic injury. Statistical significance was set at 5%.

Results

Normal human kidney

Within normal human kidney tissue, pSmad2/3 was expressed in the nuclei of a few glomerular cells (less than 10%) (Figure 1a), whereas less that 30% of glomerular cells exhibited nuclear positivity for p300 (Figure 1b).35 The expression of both molecules was not restricted in any particular glomerular cell type. p57 immunostaining in the glomerulus was mainly restricted to normal podocytes and was not present in glomerular mesangial or endothelial cells (Figure 1c).

Figure 1. Expression pSmad2/3, p300 and p57 in normal human kidney: A few glomerular cells, including all cell types, were pSmad2/3 positive (a) (arrows). Less than 30% of glomerular cells, including all cell types, were p300 positive (b). Podocyte staining for p57 in the normal kidney was abundant with a nuclear staining pattern (c). Immunopositivity for psmad2/3 and p300 (d, e,) but not for p57 (f) in a cellular crescent (arrows) (Wegener granulomatosis). Immunopositivity for psmad2/3 and p300 (g, h) but not for p57 (i) in hyperplastic lesions and microadhesions (arrows) (proliferative lupus nephritis, WHO class IV). (Figure 1a, b, d, e, g and h are reproduced from Kassimatis et al.35). Original magnification ×400.

Glomerular pSmad2/3, p300 and p57 expression in human glomerulonephritis

pSmad2/3 and p300 expression was significantly increased (p < 0.001) in all glomerular cell types in both proliferative and nonproliferative GN, when compared to controls. In contrast, a significant reduction in p57-positive podocytes was observed in samples from patients with GN (p = 0.006). In proliferative GN there was a significant general increase in pSmad2/3 (p < 0.001) and p300 (p < 0.001) expression and a decrease in the intensity of p57 podocyte staining (p < 0.001). In areas of proliferation, there was a segmental absence of podocyte immunostaining for p57. The absence was not due to cell loss, as nuclei were easily detected by the hematoxylin counterstain. In these areas, on the other hand, an increase in staining for pSmad2/3 and p300 was observed. In proliferative lupus nephritis, pSmad2/3 and p300, but not p57, were overexpressed in hyperplastic lesions and microadhesions and the same pattern of expression was observed in cellular crescents (Figure 1g, h and i). Comparing the non-proliferative GN group with the control group resulted in similar results regarding pSmad2/3 (p = 0.011) and p300 (p = 0.003), while podocyte expression of p57 was found to be preserved (p = 0.418).

Comparisons between the different types of GN showed a statistically significant difference in the glomerular expression of pSmad2/3, p300 and p57 (Figure 2). In particular, when compared to controls, a significant decrease in p57 expression was found in pauci-immune GN, SLE and IGA (p = 0.004, p = 0.019, p = 0.01, respectively), but not in FSGS, MN and MCD (p = 0.733, p = 0.4, p = 0.573, respectively). Similarly, increased pSmad2/3 staining was found in pauci- immune GN, SLE and IGA (p < 0.001, p < 0.001, p = 0.03, respectively) but not in FSGS, MN and MCD (p = 0.198, p = 0.095, p = 0.364, respectively). P300 staining was found to be increased in all GN, compared to controls.

Figure 2. Staining scores for the glomerular expression of pSmad2/3, p300 and p57 in different types of GN. (% patients). (FSGS: Focal segmental glomerulosclerosis, MN: Membranous nephropathy, MCD: Minimal Change Disease, pauci-IM-GN: pauci-immune glomerulonephritis, LN: Lupus nephritis, IGAN: IgA nephropathy.)

Correlation of p57 with pSmad2/3 and p300 expression – Correlations between immunohistochemical, clinical, and histologic parameters

Taking into account glomerular staining in all GN, pSmad2/3 expression correlated positively to P300 (r = 0.603, p < 0.01), while an inverse correlation between p57 expression, and pSmad2/3 (r = −0.4, p < 0.01), as well as P300 (r = −0.45, p < 0.01) staining was observed. The expression of pSmad2/3 in diseased glomeruli correlated positively with serum creatinine values (r = 0.36, p < 0.05) and interstitial inflammation scores (r = 0.29, p < 0.05). A negative correlation between p57 and interstitial inflammation was also found. (r = −0.3, p < 0.05), while p300 was found to be inversely correlated to glomerulosclerosis (r = 0.35, p < 0.01). No correlation was found between proteinuria or interstitial fibrosis and the expression of the studied proteins.

Stratification of the GN biopsy samples to proliferative and non-proliferative revealed that the correlations between pSmad2/3, p300 and p57 were significant only for the proliferative GN group. In particular, in the proliferative GN group pSmad2/3 expression correlated positively to P300 (r = 0.35, p = 0.04) and inversely to p57 (r = −0.37, p = 0.03) while in the non-proliferative there were no significant correlations. It should be noted that in the non-proliferative group the significant negative correlation between p300 and glomerulosclerosis persisted (r = −0.49, p < 0.05).

Discussion

To our knowledge, this is the first study that indicates a relation between the TGF-beta/Smad signalling pathway and the cell cycle regulatory protein p57 in human GN.

In contrast to the findings of the present study, no expression of pSmad2/3 in normal human renal tissue was reported by another group of researchers.11 This divergence most likely reflects differences in the immunostaining protocol or the usage of a different primary antibody.33,34 Indeed, in a previous study from our laboratory, similar results of positive immunostaining in normal human renal tissue were found using the same biotin free, two-step, HRP-labelled detection system for antibody localization with increased sensitivity.35 Moreover, Smad2 and 3 have been found to be weakly expressed in the glomeruli and tubules of the mouse embryo and adult kidney.29,36,37 p300 integrates transcriptional signals thus regulating various cellular processes such as DNA repair, cell growth, differentiation, and apoptosis.38 It is therefore possible that its expression in normal glomeruli, detected in our study, implies a role for this molecule in maintaining normal glomerular function. Podocyte staining for p57 in the normal kidney was abundant with a nuclear staining pattern, and reduced in samples from patients with GN, as expected.39 It is known that p57 and p27 are decreased in damaged podocytes from patients with FSGS in comparison with controls.3,40 Our findings are also in agreement with previous studies in experimental models in mice and rat showing that p57 is suppressed in glomerular diseases.41

The reduced glomerular p57 expression together with the increased pSmad2/3 and P300 staining that we observed in the glomeruli of patients with GN indicate that these proteins may be involved in the progression of glomerular injury. The finding that staining for p57 inversely correlated to pSmad2/3 could suggest that glomerular Smad pathway activation is related to down-regulation of p57 expression. Our findings, which are based solely on immunohistochemistry, cannot illustrate a causal relationship between Smad signalling pathway activation and cell cycle deregulation, but there are data in the literature suggesting that this relation might exist. It has been suggested that TGF-β, which is overexpressed by hyperplastic podocytes, may be involved in the regulation of podocyte proliferation.18 Moreover, p57 is thought to be involved in the control of TGF-β-induced proliferation and differentiation of osteoblastic cells.42 In osteoblastic cells on cell cycle arrest, p57 is rapidly degraded upon TGF-β stimulation.43 One proposed mechanism of this p57 degradation is that it is mediated through TGFβ1-activated, Smad-dependent transcription of the gene for the F-box protein FBL12, a protein expressed in the limb bud of developing embryos42,44 FBL12 forms a complex that binds to and ubiquitinates phosphorylated mouse p57. Inhibition of FBL12 suppresses TGFβ-induced degradation of p57, increases the steady-state level of p57, and promotes the differentiation of primary osteoblasts.44 In addition; the interaction between TGF-beta and p57 has been demonstrated to affect proliferative processes of both renal and non-renal cells. TGF-β-induced cell cycle arrest of hematopoietic cells requires p57 up-regulation,45 while the targeted inhibition of p57 blocks TGF-β-inhibited proliferation of primary cultured human limbal epithelial cells.46 In an animal model that involved unilaterally nephrectomised male Sprague-Dawley rats, TGF-β-induced compensatory tubular cell hypertrophy that was found to be regulated by p57.47 Liu et al. demonstrated that TGF-β-mediated hypertrophy in rat renal epithelial cells is regulated by p57, and does not involve regulation by the CKIs p21 and/or p27.48 The finding that staining for p57 inversely correlated to P300 expression suggests that glomerular p300 could also serve as a mediator of podocyte proliferation. Research in cancer cells has shown that p300 plays a crucial role in cell cycle progression and its activity is required for G1/S transition, while p300 inhibition induces block of progression into the S-phase of the cell cycle and apoptosis.49–51 Our finding suggest that p300 over-expression accompanied by p57 down-regulation might induce podocyte cell cycle deregulation and be involved in disease progression, especially in proliferative GNs.

The increased expression of pSmad2/3 and p300 in the glomeruli of proliferative GNs may be related to the inflammation that accompanies disease progression. This is supported by the positive correlation of pSmad2/3 and p300 glomerular immunostaining with interstitial inflammation. It is well accepted that TGF-beta/Smad signalling is a major pathway for renal fibrosis. In experimental and human kidney diseases such as diabetic nephropathy, obstructive kidney disease, 5/6 nephrectomy, and hypertensive nephropathy, Smad2 and Smad3 are strongly activated, in the context of renal fibrosis.52,53 It should be noted, though, that except for its pro-fibrotic role, TGF-β also exerts anti-inflammatory effects and possesses immunoregulatory properties. Deletion of TGF-β1 in mice results in lethal multi-organ inflammation and death at 3 weeks of age. Similarly, conditional deletion of theTGF-β1 gene from T cells in mice develops autoimmune diseases.54,55 These findings suggest a vital role for TGF-β as an anti-inflammatory mediator. This is confirmed further by the findings that application of TGF-β improves autoimmune diseases, including allergic encephalomyelitis, arthritis, and experimental colitis.56–59 In the normal kidney, TGF-β induces Smad7 expression to exert its negative feedback mechanism by causing degradation, and prevent phosphorylation of Smad2 and Smad3.60,61 Recent data have shown that in progressive kidney disease, renal Smad7 protein is reduced, resulting in over activation of TGF-β signalling and progressive renal fibrosis as shown in various rodent models of kidney disease, such as an anti-Thy1 model of glomerulonephritis, obstructive kidney disease, remnant kidney disease, and autoimmune crescentic kidney disease.62–65 The loss of renal Smad7 not only enhances TGF-β Smad3-mediated renal fibrosis, but also enhances renal inflammation through Smad2 and Smad3 phosphorylation and activation of the NF-kB–dependent inflammatory response.66,67 This mechanism could explain our findings of pSmad2/3 and p300 over-expression in the glomeruli of proliferative GNs with an increased inflammation score.

Our results indicate an interaction between the TGF-β/SMAD signaling pathway, the co-activator p300 and the cell cycle regulatory protein p57 in the glomerulus that could be involved in the pathogenesis and disease progression of GN. The increased pSmad2/3 staining together with the reduced p57 expression found in biopsy specimens with intense interstitial inflammation, indicating a possible relation between interstitial inflammation, glomerular Smad pathway activation and podocyte cell-cycle deregulation.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.