Matrix metalloproteinase-2, matrix metalloproteinase-9 ... - Springer Link

2 downloads 0 Views 2MB Size Report
stage renal failure due to an accumulation of the mesangial matrix (MM) and a thickening of the glomerular basement membrane (GBM). Both the MM and GBM ...
Clin Exp Nephrol (2006) 10:253–261 DOI 10.1007/s10157-006-0438-3

© Japanese Society of Nephrology 2006

ORIGINAL ARTICLE Tetsuya Endo · Kimimasa Nakabayashi · Makiho Sekiuchi Tadahide Kuroda · Akinori Soejima · Akira Yamada

Matrix metalloproteinase-2, matrix metalloproteinase-9, and tissue inhibitor of metalloproteinase-1 in the peripheral blood of patients with various glomerular diseases and their implication in pathogenetic lesions: study based on an enzyme-linked assay and immunohistochemical staining Received: March 6, 2006 / Accepted: August 31, 2006

Abstract Background. Various glomerular diseases progress to endstage renal failure due to an accumulation of the mesangial matrix (MM) and a thickening of the glomerular basement membrane (GBM). Both the MM and GBM are consistently metabolized through the synthesis and destruction of the matrix. Such synthesis is influenced by transforming growth factor-β (TGF-β) and other factors, whereas the destruction is presumed to be mediated by both matrix metalloproteinases (MMPs) and inhibitors of matrix metalloproteinases (TIMPs). Based on such evidence, we tried to detect MMP-2, MMP-9, and TIMP-1 in the peripheral blood of patients with various glomerular diseases. Methods. Serum was used to detect MMP-2 and TIMP-1, while plasma was used to detect MMP-9. These enzymes were detected using an enzyme-linked assay. Results. The findings showed an increased level of MMP-2 in patients with a alteration of GBM, typically membranous nephropathy (MN), regardless of the differences in their etiological processes. In contrast, MMP-9 did not show a strong association with any specific glomerular abnormalities. However, it mainly tended to increase in patients with MM accumulation. In addition, the localization of MMP-2, MMP-9, and TGF-β1 was studied using immunohistochemical staining. MMP-2 was demonstrated to exist in the glomerular capillary loop (GCL) as well as in the mesangial cells and the mesangial matrix. MMP-9 was found to exist in mesangial cells and the matrix, GCL, infiltrated neutrophils, and some tubular epithelial cells. Positive staining for TGFβ1 in GCL was found to be associated with an increased level of MMP-2 in patients with MN, whereas in MM such positive staining was not necessarily associated with an increased level of MMP-9.

T. Endo · K. Nakabayashi (*) · M. Sekiuchi · T. Kuroda · A. Soejima · A. Yamada First Department of Internal Medicine, Kyorin University School of Medicine, 6-20-2 Shinkawa, Mitaka, Tokyo 181-8611, Japan Tel. +81-422-47-5511, Ext. 5915; Fax +81-422-42-9607 e-mail: [email protected]

Conclusions. These results therefore suggest that MMP-2 plays an important role in the degradation of GBM, while MMP-9 only moderately affects the degradation of MM. Key words Metalloproteinase · Tissue inhibitor of metalloproteinase · Glomerulonephritis

Introduction Advanced glomerular diseases are characterized by the accumulation of type IV collagen, proteoglycan, laminin, etc., in the mesangial matrix (MM) and glomerular basement membrane (GBM).1,2 These components of the matrix are metabolized by matrix metalloproteinases (MMPs) as well as tissue inhibitors of metalloproteinase (TIMPs).3,4 The main constituent of MM and GBM is type IV collagen, which is degraded by MMP-2 and MMP-9, whereas the degradation of this collagen is mainly inhibited by TIMP-1.3,4 MMP-2 synthesis has been shown to be produced by fibroblasts and mesangial cells, while MMP-9 synthesis is produced by glomerular epithelial cells and mesangial cells.5–8 These synthesized MMP-2, MMP-9, and TIMP-1 are reported to be restored in the MM and GBM under the preenzymatic status, becoming active forms in local tissues, and thereby playing important roles in the progression and resolution of Thyl.1 nephritis and Heymann nephritis, as well as human glomerulonephritis.3,4,9–14 These MMPs and TIMPs are also released into the blood stream, and their concentrations have recently been determined by enzymelinked assay systems.15 However, only a few studies have so far documented their blood concentrations in human kidney diseases.12–14 Therefore, we carried out this study in order to detect MMP-2, MMP-9, and TIMP-1 in the peripheral blood of patients with various glomerulonephritides, and herein we discuss the implications of our findings in the pathogenesis of glomerulonephritis.

254

Materials and methods

used for the staining of MMP-2, MMP-9, TIMP-1, and transforming growth factor-1 (TGF-β1).

Materials One hundred and fifty-nine patients with various glomerular diseases proven by renal biopsy were included in the study. The main enzymatic activities in the blood were examined at the time of renal biopsy. Some cases were also examined at different times, but the results were similar to the amounts of urinary protein and the renal function obtained at the time of renal biopsy. The patients investigated with various glomerulopathies, included 50 with IgA nephritis (IgAN), 17 with membranous nephropathy (MN), 14 with membranoproliferative glomerulonephritis (MPGN), 13 with minimal-change nephrotic syndrome (MCNS), 10 with nephrosclerosis, 28 with lupus nephritis (LN), and 27 with other types of glomerular disease. At the time of blood sampling for the experiment, the patients with primary glomerulonephritides were not on steroid therapy, whereas all the patients with lupus nephritis were already on steroid treatment, including large doses of steroid and pulse therapy. These patients were not associated with any other manifestations or unusual laboratory data except for the presence of renal diseases. Nineteen healthy people were selected as controls. The patients with IgAN were categorized into four groups according to the prognostic criteria for IgA nephropathy which had been established by the Research Project for Progressive Glomerular Disease, which is supported by the Ministry of Heath, Labour, and Welfare of Japan.16 The four groups were those with a good prognosis, a relatively good prognosis, a relatively poor prognosis, and a poor prognosis. MN patients were also divided into four stages according to Churg’s classification.17 The four stages included stages I, II, III, and IV. Methods The determination of the MMP-2, MMP-9, and TIMP-1 levels in the peripheral blood Serum or heparinized plasma was used for the experiment. The serum was used for MMP-2 and TIMP-1 determination, while the plasma was used for MMP-9 determination. The assay system is an enzyme-linked and one-step sandwich method, which has been described elsewhere in the literature.15 These assay systems were purchased from Daiichi Fine Chemical (Takaoka, Japan). Renal biopsy specimens Routine examination The renal tissue specimens obtained by biopsy were divided into three sections for light, immunofluorescence, and electron microscopic examinations, and their histopathological diagnoses were made. Several tissue specimens which were from either embedded paraffin or frozen tissue were also

Immunohistochemical study for MMP-2, MMP-9, and TIMP-1 Staining was performed with mouse monoclonal antibodies to the human enzymes described above, and they were also purchased from Daiichi Fine Chemical. The staining was mainly carried out using paraffin-embedded tissues, and the PAP staining method was used. A brief description of the staining procedure follows. Mouse monoclonal antibody was diluted to 100× and then applied to renal tissue specimens for the first antibody under 4°C and an overnight reaction. After washing with phosphate-buffered saline (PBS), goat anti-mouse IgG antibody labeled with horseradish peroxidase (Envision System, Dako, Kyoto, Japan) was reacted on the renal tissue as a second antibody for 60 min. After washing with PBS, 3-diaminobenzidine tetrahydrochloride (DAB), including 0.01% hydrogen peroxide, was used for enzymatic visualization. The stained tissue specimens were then examined. The patients investigated included 13 with IgAN, 3 with MN, 1 with diabetic nephropathy, and 3 with LN. An immunohistochemical study for transforming growth factor-b1 (TGF-b1) TGF-β is known to have subsets such as TGF-β1, TGF-β2, and TGF-β3,18 and TGF-β1 has a strong influence on the production of the extracellular matrix from various cells.18 Therefore, the localization of TGF-β1 was investigated in the renal specimens. The staining procedure was almost the same as that used for MMPs and TIMP-1, namely, rabbit anti-human TGF-β1 antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA) was diluted to 100× and was then applied to renal tissue specimens for the first antibody under 4°C and an overnight reaction. Next, the goat antirabbit IgG antibody labeled with horseradish peroxidase (Dako) was used as the second antibody. Enzymatic visualization with DAB was finally performed in the standard manner and the results were examined. The patients investigated included 5 with IgAN, 3 with MN, and 2 with MPGN. The patients with IgAN and MN were different from the patients examined using MMP-2, MMP-9, and TIMP-1 staining. MMP-2, MMP-9, and TIMP-1 levels in the peripheral blood and clinical data Each enzymatic level and clinical parameter, such as the systolic and diastolic blood pressure, urinary protein, serum creatinine, and creatinine clearance, were analyzed to identify any correlations. The level of urinary protein was calculated according to the fractional urinary protein level (mg/dl) divided by its urinary creatinine concentration (mg/dl).

255

Statistical analysis A Mann–Whitney U-test19 was performed to find any significant differences between each group and the healthy controls, as well as between two different groups. Spearman’s correlation coefficient20 was used to elucidate the relationships between the MMP-2, MMP-9, and TIMP1 levels in the peripheral blood and the clinical data. Ethics The experimental design was approved by the Ethics Committee of Kyorin University. The blood and renal tissue specimens from the patients were obtained only after informed consent had been received, and then they were used for the study.

Results The MMP-2, MMP-9, and TIMP-1 levels in the peripheral blood and various glomerulonephritides Serum MMP-2 level The mean concentrations (M) and standard deviations (SD) in the healthy controls and for various glomerulonephritides are shown in Fig. 1. The M ± SD in each patient group with glomerulonephritis was higher than that of healthy controls. Patients with an M + 2SD higher than that of the

Fig. 1. Serum matrix metalloproteinase-2 (MMP-2) titers versus various glomerulonephritides. In Figs. 1–7, the solid lines indicate the mean +2 standard deviations (M + 2SD) level. MCNS, minimalchange nephritic syndrome; IgAN, IgA nephritis; MN, membranous nephropathy; MPGN, membranoproliferative glomerulonephritis; NPS, nephrosclerosis; LN, lupus nephritis

healthy controls belonged mainly to the MCNS, MN, MPGN, and LN groups. In contrast, patients with IgAN and nephrosclerosis did not show such a high level except for a few cases. These data suggest that the glomerular diseases involved in the capillary loop structure showed a high concentration. Plasma MMP-9 level All M ± SDs in the healthy controls and in various glomerulonephritides (unclear) are shown in Fig. 2. The M + SDs in each patient group with glomerulonephritis did not show any significant differences from those of the healthy controls. However, markedly higher levels than M + 2SD of the healthy controls were observed in a considerable number of patients with IgAN and LN, but in only a few patients with MPGN, MN, and nephrosclerosis. MCNS did not show any significantly high levels except in a few patients. These data imply that the glomerular diseases manifested with the expansion of the mesangial matrix tended to reveal a high concentration. Serum TIMP-1 level All M ± SDs in the healthy controls and in various glomerulonephritides are shown in Fig. 3. The M ± SD in each patient group with glomerulonephritis was higher than that of the healthy controls. Levels higher than that of M + 2SD in the healthy controls were observed in all glomerulonephritides. These data indicate that the glomerular diseases involved in the capillary loop as well as the mesangial matrix showed a high concentration.

256 Fig. 2. Plasma MMP-9 titers versus various glomerulonephritides

Fig. 3. Serum tissue inhibitor of matrix metalloproteinase-1 (TIMP-1) titers versus various glomerulonephritides

Relationship of the serum MMP-2, plasma MMP-9, and serum TIMP-1 levels in the peripheral blood to the prognostic criteria for IgAN (Figs. 4–6) Serum MMP-2 level In Fig. 4, the M ± SDs in the group of patients with a relatively poor prognosis and in those with a good prognosis show a statistically higher level than those of the healthy controls. In contrast, the M ± SDs in the patient groups showing a relatively good prognosis and in those with a

poor prognosis did not show any statistical difference in comparison to the healthy controls. The patients with a level higher than M + 2SD of the healthy controls belonged mainly to the groups with a relatively poor prognosis or a poor prognosis, except for one patient who demonstrated a relatively good prognosis. Plasma MMP-9 level The M ± SDs in each patient group shown in Fig. 5 did not reveal any statistically significant differences from those of

257 Fig. 4. Serum MMP-2 titers versus prognostic groups categorized from lightmicroscopy findings in IgA nephropathy

Fig. 5. Serum MMP-9 titers versus prognostic groups categorized from lightmicroscopy findings in IgA nephropathy

the healthy controls. However, several patients in the group with a relatively poor prognosis, as well as a few patients in the group of patients with a relatively good prognosis and those with a poor prognosis, showed a level higher than M + 2SD of those observed in the healthy controls. The proportion of patients showing a level higher than M + 2SD in each group was 39% in the groups with a relatively poor prognosis, 28.6% in the group with a relatively good prognosis, 14.3% in the group with a poor prognosis, and 0% in the group with a good prognosis. Serum TIMP-1 level The M ± SDs in each group of patients shown in Fig. 6 showed a higher level in the groups with a relatively poor prognosis and those with a poor prognosis than that of the healthy controls. The proportion of patients showing a level

higher than M + 2SD of the healthy controls belonged mainly to the group of patients with a relatively poor prognosis, except for a few patients in the groups with a relatively good prognosis or a poor prognosis. These data suggest that a high proportion of the patients with IgAN in the relatively poor prognosis group showed remarkable abnormalities of MMP-2, TIMP-1, and MMP-9.

Relationship of serum MMP-2, plasma MMP-9, and serum TIMP-1 levels in the peripheral blood to the stage for membranous nephropathy Serum MMP-2 level The M ± SD in the stage II and III patients shown in Fig. 7 showed a higher level than that of the healthy controls. The

258 Fig. 6. Serum TIMP-1 titers versus prognostic groups categorized from lightmicroscopy findings in IgA nephropathy

Fig. 7. Serum MMP-2 titers versus histopathological stages of membranous nephropathy

stage II patients in particular showed markedly high levels, as well as the stage I and IV patients, although a few patients, tended to show high levels. Plasma MMP-9 level The M ± SD for the patients at each stage (data not shown) did not demonstrate any correlation with the stages of MN.

The relationship of serum MMP-2, plasma MMP-9, and serum TIMP1 levels in peripheral blood to the WHO classification of lupus nephritis was also studied, but the data are complex because the plasma MMP-9 levels were greatly influenced by the preceding large doses of steroid treatment. Therefore, these data will be described in a future publication.

Serum TIMP-1 level The M ± SD in the stage II patients (data not shown) was a higher level than that of the healthy controls. Five patients in I, II, and IV excluding the other 2 patients, showed a high level. These data were very similar to the serum MMP-2 level. These data suggest that in MN, which demonstrates a thickening of basement membrane, both the serum MMP-2 and TIMP-1 levels increased to high levels, whereas the plasma MMP-9 remained within the normal levels. The localization of MMP-2, MMP-9, and TIMP-1 in renal tissue A preliminary study for these enzymes performed with immunofluorescence staining did not show any positive

259

Fig. 8. MMP-2 staining (×400). Membranous glomerulonephritis (MGN) (56-year-old man). The positive findings are shown in the capillary loop (arrow heads), as well as some mesangial cells (long arrows) and matrices (short arrows). Inset. Another area in the same patient

Fig. 10. Transforming growth factor-1 (TGF-β1) staining (×400). IgA nephritis (IgAN) (53-year-old man). Positive findings were found in the mesangial cells (long arrow) and matrices (short arrow) as well as in the capillary loop (arrow heads)

The TGF-β1 existence in renal tissue specimens and the serum MMP-2, plasma MMP-9, and serum TIMP-1 levels

Fig. 9. MMP-9 staining (×400). Membranoproliferative glomerulonephritis (MPGN) (63-year-old man). Positive findings were found in the mesangial cells (long arrows) and matrices (short arrows), the capillary loop (arrow heads), and infiltrated neutrophils

results. Therefore, the immunohistochemical staining by the PAP method was used for this study. A positive MMP2 finding was demonstrated in the capillary loop as well as in some mesangial cells and matrices (Fig. 8). This MMP-2 positivity was observed in 3 patients with MN, 4 with IgAN, and 3 with LN. All these positive patients were characterized by a thickening of the glomerular basement membrane. In contrast, a positive MMP-9 finding was found in the mesangial cells and matrices, capillary loops, infiltrated neutrophils, and epithelial cells of the segmental tubules (Fig. 9). This MMP-9 positivity was observed in 4 patients with IgAN, 3 with LN, and 1 with diabetic nephropathy. As far as we could determine, all these positive patients were associated with cellular proliferation in the glomerulus and the expansion of the mesangial matrix. In TIMP-1 staining, no confirmed positivity was found based on this study.

TGF-β1 was demonstrated in the mesangial cells and matrices as well as in the capillary loops (Fig 10). In IgAN, positivity was observed in the mesangial matrix, and its intensity was remarkable in two patients with a relatively good prognosis, and in one patient with a relatively poor prognosis. In contrast, its intensity was not marked in one patient with a poor prognosis and one patient with a good prognosis. In MN, TGF-β1 was disclosed in the capillary loop and it was remarkable in two stage II patients, but it was faintly positive in one stage I patient. In MPGN, TGFβ1 was revealed in the mesangial cells and the matrices as well as the capillary loops in two patients, and positivity was mainly found in the peripheral area of the glomerulus. The relationship between the existence of TGFβ-1 in renal tissue and the plasma MMP-9 level of IgAN was as follows. The strong staining of TGF-β1 was found in four patients, and they were categorized into two groups, namely, an exceedingly high MMP-9 concentration level in two patients, and a normal MMP-9 concentration in two patients. A similarly high level of plasma MMP-9 was found in one patient with a relatively poor prognosis and one patient with a relatively good prognosis. However, a normal level of plasma MMP-9 was found in one patient with a relatively poor prognosis and in one patient with a relatively good prognosis. The serum MMP-2 and TIMP-1 levels in these four patients were not abnormal, and were fairly similar to the level of the healthy controls. These relationships were also studied in three patients with MN (2 with stage II and 1 with stage I disease) and two patients with MPGN, who were all positive for TGF-β1. These three MN patients had very high levels of MMP-2 and a high level of TIMP-1, but a normal level of MMP-9. Two MPGN patients had very high levels of MMP2 and TIMP-1, but one had a very high MMP-9 level, while the other patient showed a normal level.

260

These experimental results indicate that TGF-β1 was demonstrated in capillary loops, and was associated with a high level of MMP-2 in patients with MN and MPGN. In contrast, TGF-β1 was revealed in the mesangial matrix, but it was not necessarily accompanied by an increased level of MMP-9 in the patients with IgAN. Relationship between the MMP-2, MMP-9, and TIMP-1 levels in the peripheral blood and the clinical data The clinical parameters studied were mentioned in the Methods section, and the only positive correlation found in this study was between the serum MMP-2 and urinary protein (P < 0.0001).

Discussion Glomerulonephritis progresses to renal failure with an accumulation of the mesangial matrix (MM) and a thickening of the glomerular basement membrane (GBM). The accumulation and thickening of these components are controlled by the stimulation of TNFα, TGF-β, IL-1, and thrombin on the mesangial, epithelial, and endothelial cells of the glomerulus, while the degradation is conducted by MMPs and TIMPs.3,4,18,21 There are many articles which describe the existence and significance of these enzymes in renal tissues of experimental nephritis, and cultured cells from the glomerulus, and human glomerulus.4–6,11–14,21–24 However, to date only a few articles on the pathogenesis of human glomerulonephritis have reported the significance of these enzymes, not only in the peripheral blood, but also in the urine.12–14,24,25 We therefore measured these enzymes in the peripheral blood with various glomerulonephritides, at the same time as performing various staining procedures in the biopsy specimens from the kidney. The determination of the MMP2, MMP-9, and TIMP-1 levels in the peripheral blood was first performed in various glomerular diseases, and the implications of these enzymes are discussed in the pathogenesis of glomerulonephritides. Secondly, the influence of these enzymes on the progression of IgAN and MN was studied. Thirdly, the localization of MMP-2, MMP-9, and TIMP-1 in the renal tissue specimens was demonstrated, and we investigated their association with TGF-β1. Based on these results, a correlation of these enzymes in the peripheral blood with the renal tissues was suggested. An extremely high level of serum MMP-2 was observed in the patients with MN, MPGN, LN, and MCNS, and these glomerulonephritides were characterized by the abnormalities of the capillary loop, such as constituents, thickness, and electric charge. In contrast, almost normal or slightly increased levels of MMP-2 were detected in patients with IgAN and nephrosclerosis, and these glomerulonephritides were not commonly associated with any remarkable abnormalities of the capillary loop. However, a slightly increased level of MMP-2 was found in several patients with IgAN. These patients were shown to have some segmental loop

alterations, with thickening and partially extended depositions of IgA and C3 on the GBM, thus again suggesting a relationship between MMP-2 and the abnormalities of the capillary loop. A markedly high level of plasma MMP-9 was found in the patients with IgAN and LN, and also in some patients with MPGN, nephrosclerosis and MN. As far as we could tell, these glomerulonephritides were manifested by the expansion of the mesangial matrix. A normal MMP-9 level was found in patients with MCNS which was not associated with the accumulation of the mesangial matrix. The serum TIMP-1 increased in almost all patients who were investigated for glomerulonephritides. Although no direct evidence was found regarding the secretion of these enzymes from the glomerulus studied, these experimental data relating to the peripheral blood imply that the serum MMP-2 reflects the degradation of the GBM, while MMP-9 represents the degradation of the MM. The additional evidence is that these MMP-2 and MMP-9 are known to exist in joint tissues, neutrophils and macrophages,7,26 but the patients we studied did not show any other manifestations except for renal diseases. The second study on the association of serum MMP-2 and plasma MMP-9 with the prognostic classification of IgAN and MN was as follows. In IgAN, patients with a relatively poor prognosis associated with a widening of the MM were found to have high levels of both plasma MMP-9 and serum TIMP-1. In MN, the stage II and III patients who were shown to have a thickened GBM were revealed to have a high level of MMP-2 but a normal level of MMP-9. These experimental data also suggest that MMP-2 plays an important role in the metabolism of the GBM, whereas MMP-9 does not significantly act on the GBM metabolism, but it does act on the MM. The third study on the localization of MMP-2 and MMP-9 showed that MMP-9 was found in the mesangial cells and matrices as well as in the capillary loops and infiltrated neutrophils. As far as we could tell, such positivity was almost always associated with a widening of the MM. In addition, MMP-2 was found in the capillary loops as well as in some mesangial cells and matrices, while the positivity was generally associated with the thickening of the GBM. These results also support the data in the second study on MMP-9 and MMP-2 in the peripheral blood and their association of glomerular pathologies. The relationship between the glomerular localization of TGF-β1 and the MMP-2, MMP-9, and TIMP-1 levels in the peripheral blood showed that TGF-β1 was found in tissues of both the widened MM and the thickened GBM. In particular, TGF-β1 was found in the thickened GBM, and this was associated with the increased level of MMP-2. TGF-β1 was also found in the expanded MM area, but it was not necessarily associated with an increased MMP-9 level. These experimental data suggest that TGF-β1 stimulates the glomerular epithelial and mesangial cells to generate MMP-2, while factors other than TGF-β1 also play a role in the generation of MMP -9 by these cells. These experimental results for TGF-β1, which is known to stimulate various cells and to generate several cytokines,3,4,18,27 support the previous data showing it to play an important role in the repair and sclerosis of tissue through matrix proteinases,18,27

261

Furthermore, the clinical data which show a correlation between the serum MMP-2 level and the amount of urinary protein imply the degradation of the GBM by MMP-2, because the amount of proteinuria mainly depends on the abnormalities of the GBM, such as constituents, thickness, and electric charges. These experimental results correlate with the previous data on human glomerular histopathology, which showed the existence of MMP-2 in the GBM, the presence of MMP9 in mesangial cells, MMP-2 and MMP-9 generation by mesangial cells, and the enzymatic degradation of the MM by MMP-9.5,11,24,28 In addition, the serum MMP-2 level increased in patients with GCL abnormalities, including not only size-barrier diseases, but also a charge-barrier disease, when the MMP-2 level was examined in association with glomerular pathologies. This association closely correlates with the previous demonstration of MMP-2 in GBM,11 while also suggesting that MMP-2 plays important roles not only in GBM turnover, but also in GBM charge barrier loss. The plasma MMP-9 level tended to increase in patients with a widening of the MM, and these data also correlate with the findings of a previous report on the existence and production of MMP-9 in mesangial cells, specifically in association with mesangial proliferative glomerulonephritis.28 Therefore, the determination of the MMP-2 and MMP-9 levels as well as TIMP-1 in the peripheral blood of patients with glomerulonephritis is considered to a certain extent to be a helpful marker in speculations about the presence of GBM or mesangial lesions. Acknowledgments Valuable suggestions on the statistical analyses were given by Prof. Hideki Ohno, Chief, Department of Public Health, Kyorin University School of Medicine. We therefore express our sincere gratitude to him. The authors also thank Katuko Sano and Ayumi Sumiishi for their skillful assistance with the staining. This work was supported by grants (2003–2006) from the Ministry of Health, Labour, and Welfare of Japan.

References 1. Schnapper HW. Balance between matrix synthesis and degradation: a determinant of glomerulosclerosis. Pediatr Nephrol 1995; 9:104–11. 2. Weber M. Basement membrane proteins. Kidney Int 1992;41:620– 8. 3. Davis M, Martin J, Thomas GJ, Lovett DH. Proteinases and glomerular matrix turnover. Kidney Int 1992;41:671–8. 4. Lenz O, Elliot SJ, Stetler-Stevenson WG. Matrix metalloproteinases in renal development and disease. J Am Soc Nephrol 2000;11:574–81. 5. Osawa H, Yamabe H, Kaizuka M, Nakamura N, Shirato K, Sugawara T, et al. Platelet-derived growth factor stimulates matrix metalloproteinase-2 secretion in cultured human mesangial cells. Clin Exp Nephrol 2002;6:202–6. 6. Kuroda T, Yoshida Y, Kamiie J, Kovalenko P, Nammeta M, Fujinaka H, et al. Expression of MMP-9 in mesangial cells and its changes in anti-GBM glomerulonephritis in WKY rats. Clin Exp Nephrol 2004;8:206–15. 7. Sternlicht MD, Werb Z. Gelatinases (MMPs 2 and 9). In: Kreis T, Vale R, editors. Guidebook to the extracellalar matrix, anchor, and adhesion proteins. 2nd ed. Oxford: A Sambrook & Tooze Publication at Oxford University Press; 1999. p. 529–31.

8. McMillan JI, Riordan JW, Couser WG, Pollock AS, Lovett DH. Characterization of a glomerular epithelial cell metalloproteinase as matrix metalloproteinase-9 with enhanced expression in a model of membranous nephropathy. J Clin Invest 1966;47:1094–101. 9. Sato Y, Fujimoto S, Hamai K, Eto T. Serial alterations of glomerular matrix-degrading metalloproteinase activity in antithymocyte-induced glomerulonephritis in rats. Nephron 1998;78: 195–200. 10. Steinmann-Niggil K, Ziswiler R, Küng M, Marti HP. Inhibition of matrix metalloproteinases attenuates anti-Thy 1.1 nephritis. J Am Soc Nephrol 1998;9:397–407. 11. Jalalah SM, Furness PN, Barker G, Thomas M, Hall L, Bicknell GR, et al. Inactive matrix metalloproteinase 2 is a normal constituent of human glomerular basement membrane. An immunoelectron microscopic study. J Pathol 2000;191:61–6. 12. Kanauchi M, Nishioka H, Nakashima Y, Hashimoto T, Dohi K. Role of tissue inhibitors of metalloproteinase in diabetic nephropathy. Jpn J Nephrol 1996;38:124–8. 13. Akiyama K, Shikata K, Sugimoto H, Matsuda M, Shikata Y, Fujimoto N, et al. Changes in serum concentration of matrix metalloproteinases and type IV collagen in patients with various types of glomerulonephritis. Res Commun Mol Pathol Pharmacol 1997;95:115–28. 14. Waku A, Nakabayashi K, Karube M, Nagasawa T. Serum TIMP-1 level in various glomerulonephritides (in Japanese, with English abstract). Kyorin J Med Sci 1997;28:441–7. 15. Fujimoto N, Iwata K. Enzyme immunoassays for matrix metalloproteinases and their inhibitors. Connect Tissue 1994;26:237–44. 16. Tomino Y, Sakai H, Special Study Group (IgA nephropathy) on progressive glomerular disease: clinical guidelines for immunoglobulin A (IgA) nephropathy in Japan, 2nd version. Clin Exp Nephrol 2003;7:93–7. 17. Ehreinreich T, Churg J. Pathology of membranous nephropathy. In: Annual review of pathology 1968, Vol. 3. New York: Appleton– Century–Crofts; 1968. p. 145–8. 18. Border WA, Noble NA. Transforming growth factor β in tissue fibrosis. N Engl J Med 1994.;331:1286–92. 19. Daly LE, Bourke GJ. Comparisons of two independent medians. In: Interpretation and uses of medical statistics, 5th ed. Oxford: Blackwell Science; 2000. p. 218–21. 20. Daly LE, Bourke GJ. Associations between two quantitative variables: regression and correlation. In: Interpretation and uses of medical statistics, 5th ed. Oxford: Blackwell Science; 2000. p. 315– 38. 21. Kaizuka M, Yamabe H, Osawa H, Okumura K, Fujimoto N. Thrombin stimulates synthesis of type IV collagen and tissue inhibitor of metalloproteinases-1 by cultured human mesangial cells. J Am Soc Nephrol 1999;10:1516–23. 22. Martin J, Eynstone L, Davies M, Steadman R. Induction of metalloproteinases by glomerular mesangial cells stimulated by proteins of the extracellular matrix. J Am Soc Nephrol 2001;12:88– 96. 23. Martin J, Steadman R, Knowlden J, Williams J, Davies M. Differential regulation of matrix metalloproteinases and their inhibitors in human glomerular epithelial cells in vitro. J Am Soc Nephrol 1998;9:1629–37. 24. Fujimoto S, Hamai K, Sato Y, Yamamoto Y, Eto T. Neutral metalloproteinases in human urine from normal patients and renal disease patients. Nephrology 1996;2:329–37. 25. Koide H, Nakamura T, Ebihara I, Tomino Y. Increased mRNA expression of metalloproteinase-9 in peripheral blood monocytes from patients with immunoglobulin A nephropathy. Am J Kidney Dis 1996;28:32–9. 26. Okada Y, Morodomi T, Enghild JJ, Suzuki K, Yasui A, Nakanishi I, et al. Matrix metalloproteinase-2 from human rheumatoid synovial fibroblast. Purification and activation of the precursor and enzymic properties. Eur J Biochem 1990;194:721–30. 27. Yamamoto T, Noble NA, Cohen AH, Nast C, Hishida A, Gold LI, et al. Expression of transforming growth factor-β isoforms in human glomerular diseases. Kidney Int 1996;49:461–9. 28. Urushibara M, Kagami S, Kuhara T, Tamaki T, Kuroda Y. Glomerular distribution and gelatinolytic activity of matrix metalloproteinases in human glomerulonephritis. Nephrol Dial Transplant 2002;17:1189–96.