Metalloproteinases in the Pathogenesis of Diabetic Nephropathy

Diabetic nephropathy is one of the major complications of diabetes mellitus and an important cause of increased morbidity and mortality in these patients. The increase in the number of patients with chronic renal failure due to diabetic nephropathy is a major problem in several countries. The pathogenesis of diabetic nephropathy therefore requires elucidation in order to develop effective treatment modalities. Several pathogenic mechanisms have been proposed, such as the presence of glomerular hyperfiltration and/or hyperperfusion [1], activation of the polyol pathway, myoinositol metabolism abnormalities [2, 3], high activity of protein kinase C [4, 5], increased nonenzymatic glycated proteins [6, 7, 8, 9], hypercoagulability [10], involvement of cytokines and growth factors [11, 12, 13, 14]and hereditary factors [15]. However, the precise mechanism remains to be clarified (table 1). The pathological hallmarks of diabetic nephropathy are increased thickness of the glomerular basement membrane (GBM) and mesangial expansion [16, 17, 18, 19, 20]. The latter is considered more significant than thickening of the GBM as expansion of the mesangium has been reported to ultimately lead to renal insufficiency [19]. These changes in the GBM and mesangium are characterized by expansion of all intrinsic components of extracellular glycoproteins, such as collagen types IV, V and VI, laminin, and fibronectin, as well as other intrinsic components of the normal renal extracellular matrix (ECM) that have not yet been identified [21, 22, 23, 24, 25, 26, 27]. Although the balance between ECM production and degradation is important, these relationships are not fully understood in diabetic nephropathy. Recent developments in cellular and molecular biology have allowed a better understanding of ECM production and degradation in this disease. This article reviews recent studies on the balance between ECM production and degradation and its role in the pathogenesis of diabetic nephropathy. The pathological change in mesangial expansion is characterized by an increase in all intrinsic components of mesangial matrix, such as collagen types IV, V and VI, laminin and fibronectin, as well as other components of ECM that have not yet been identified in normal glomeruli. Several immunohistological studies have demonstrated that intrinsic components of the glomeruli stain positive to certain antibodies in diabetic nephropathy (table 2) [21, 22, 23, 24, 25, 26, 27]. Falk et al. [21]showed that collagen types IV and V, laminin and fibronectin were increased in the mesangial matrix in early and moderate stages of diabetic nephropathy. However, the mesangial levels of all these components, with the exception of type V collagen, are diminished with disease progression. Nerlich and Schleichert [22]also reported increased staining for collagens type IV and V, laminin and fibronectin in enlarged mesangial matrix of diabetic nephropathy, whereas staining of heparan sulfate proteoglycan, which maintains a negative charge in GBM, was markedly reduced. We have previously demonstrated that the proportions of areas positive for periodic acid Schiff (PAS), collagens type III, IV, V and VI, or fibronectin staining in mild-to-moderate specimens of diabetic nephropathy were significantly greater than in diffuse proliferative glomerulonephritis [23]. Moreover, in patients with diabetic nephropathy the percentages of PAS-positive areas correlated significantly with both type IV collagen-positive areas and fibronectin- positive areas. The percentages of PAS-positive areas also correlated significantly with serum creatinine levels and the degree of proteinuria in these patients. These results are summarized in table 3. Increased collagen, fibronectin and laminin synthesis may be an important factor in the expansion of ECM in diabetic nephropathy [28, 29, 30, 31, 32]. In vitro studies indicate that glomerular resident cells, such as glomerular mesangial, epithelial and endothelial cells, are capable of producing ECM [33, 34, 35]. In this regard, increased type IV collagen synthesis is one of the major causes of mesangial expansion in diabetic nephropathy. Cultured glomerular cells overproduce type IV collagen in the presence of high concentrations of glucose [36, 37]. Haneda et al. [36]reported that type IV collagen secreted from, and associated with, glomerular mesangial cells increased significantly when the cells were cultured in high concentrations of glucose. Furthermore, Danne et al. [37]demonstrated that all three glomerular resident cells – epithelial, endothelial and mesangial – were able to synthesize type IV collagen in culture in the presence of high glucose concentrations. Moreover, the level of mRNA for type IV collagen was significantly increased in isolated glomeruli from diabetic rats [38]and in the cortical kidney from diabetic mice [39]. However, we reported that the expression of type IV collagen mRNA tended to correlate inversely with the degree of mesangial expansion in diabetic nephropathy [40, 41]. Furthermore, immunofluorescence studies using specific antibodies against α1–4 isoforms of type IV collagen have shown a tendency for a decrease in the intensity of staining for some isoforms with deteriorating histological features in patients with non-insulin-dependent diabetes mellitus (NIDDM) associated with diabetic nephropathy [24]. The discrepancy between human studies and the experimental model suggests that glomerular synthesis of type IV collagen increases during the early to middle phase of mesangial expansion in human diabetic nephropathy. Moreover, the results of short-term in vitro studies may not always correspond with the long-term situation in vivo. In fact, mesangial cells, both in long-term cultures and when cultured on glycated matrices, show reduced collagen production [42]. Further human studies are necessary to investigate these processes. The ECM comprises many components, including glycoproteins such as collagens, laminin and fibronectin, and proteoglycans such as heparan sulfate proteoglycan and chondroitin sulfate proteoglycan. For this reason, several enzymes are known to degrade ECM. The four main types are aspartic, cysteine, serine proteinases and matrix metalloproteinases (MMPs) [43]. Of these enzymes, MMPs are thought to play an important role in the pathogenesis of diabetic nephropathy. MMPs possess the following characteristics: (1) they degrade one or more components of the ECM; (2) they contain a zinc ion at the active center; (3) they are secreted in a latent form and are activated by proteinases or organomercurial compounds, and (4) the activities of MMPs are inhibited by tissue inhibitors of metalloproteinases (TIMPs) [44, 45, 46]. MMPs are divided into three main subgroups by virtue of their substrate specificity – collagenases, gelatinases and stromelysin – and ten types of MMPs have been defined (table 4) [47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57]. MMP-1 and MMP-8, which are collagenases, mainly degrade connective tissue collagens including types I, II and III. On the other hand, MMP-2 and MMP-9, which are gelatinases, degrade basement membrane collagens as well as having gelatinolytic activity. MMPs of the stromelysin class – MMP-10 and especially MMP-3 – have a broad substrate specificity, including type IV collagen, laminin, fibronectin and proteoglycans. Conversely, TIMPs are known to inhibit the activity of MMPs and expression of TIMP mRNA in human glomeruli was recently described by Carome et al. [58]. TIMPs form a 1:1 stoichiometric complex with MMPs and TIMP-1, TIMP-2, TIMP-3 and TIMP-4 have been described [59, 60, 61, 62, 63]. Cultured mesangial and glomerular epithelial cells are able to synthesize not only ECM, but also MMPs and TIMPs [43, 64, 65, 66, 67, 68, 69]. Reduced degradation of ECM leads to accumulation of ECM and increased production of ECM components. The high level of ECM production under conditions of low levels of MMP and high levels of TIMP easily explains mesangial expansion in diabetic models. For instance, in in vitro studies, it has been reported that a high glucose concentration suppresses MMP mRNA expression and enhances type IV collagen and TIMP mRNA by cultured mesangial cells [65, 70]. Moreover, several investigators have examined MMP activity in diabetic models. Reckelhoff et al. [71]reported that glomerular metalloproteinase activity was decreased in diabetic rats and Teschner et al. [72]showed that glomerular proteinase activity was decreased in the diabetic rat. Nakamura et al. [73]demonstrated diminishing levels of MMP-3 with age in glomeruli of streptozotocin-induced diabetic rats. Furthermore, Schaefer et al. [74]showed that ECM expansion in glomeruli from obese Zucker rats (an autosomal recessive model of obesity and hyperlipidemia with many similarities to human NIDDM) is due to both enhanced synthesis of matrix components and reduced degradation by MMPs. Previous studies examining the balance between synthesis and degradation of ECM have largely been limited to animal models and in vitro experiments, and results cannot be readily extrapolated to humans. Moreover, most animal studies do not permit assessment of the specific renal cells involved and only a few studies have examined ECM production and degradation in human diabetic nephropathy. To determine the balance between turnover and degradation of ECM in renal tissue of patients with NIDDM with diabetic nephropathy, we examined the expression of MMP-3, TIMP-1 and type IV collagen mRNAs using high-resolution in situ hybridization [40]. Patients were divided into three grades based on the light microscopic findings: mild (grade I), moderate (grade II) and severe (grade III) mesangial expansion and tubulointerstitial injury. We evaluated the degree of expression of the above mRNAs in the glomeruli by scoring the intensity of immunoperoxidase staining, from 0 (– = no staining) to 3 (+++ = very strong staining). We also examined the relationship between the expression of MMP-3, TIMP-1 and type IV collagen mRNAs and the degree of glomerular mesangial expansion and interstitial injury. Positive staining for each mRNA was observed in glomerular resident cells, including glomerular mesangial, epithelial and endothelial cells and cells of Bowman’s capsule. Expression of MMP-3 and TIMP-1 mRNAs was strongest in glomeruli of grade I disease and inversely correlated with mesangial expansion. Furthermore, we undertook an additional experiment using an MMP-2 probe in another group of NIDDM patients with diabetic nephropathy. Positive staining for MMP-2 mRNA was observed in glomerular resident cells along with positive staining for type IV collagen, MMP-3 and TIMP-1 mRNA (fig. 1). Expression of MMP-2 mRNA was strongest in glomeruli of grade I disease and inversely correlated with mesangial expansion and MMP-3 and TIMP-1 (results are summarized in tables 5 and 6). We also examined the expression of type IV collagen, MMP-2, MMP-3 and TIMP-1 mRNA using high-resolution in situ hybridization in renal tissues from patients with insulin-dependent diabetes mellitus (IDDM) associated with diabetic nephropathy, and in normal human kidney (NHK) [41]. IDDM patients were divided into 2 groups: ‘slow track’ with very slowly developing structural changes of diabetic nephropathy, and ‘fast track’ with very rapidly developing structural changes of diabetic nephropathy. All glomerular nuclei as well as nuclei with surrounding positive cytoplasm were counted and the results expressed as percent positive cells. Positive staining for each mRNA was observed in glomerular resident cells as in the NIDDM study. The percent positive cells for each mRNA in ‘fast track’ disease was significantly lower than that in NHK. Also, the percent positive cells for MMP-2 and MMP-3 mRNA was significantly lower in ‘fast track’ than in ‘slow track’ disease. The percent positive cells for MMP-2 and MMP-3 mRNA in all IDDM patients was significantly lower than that in NHK. Moreover, the percent positive cells for each mRNA correlated inversely with the degree of mesangial expansion. Considered together, these results suggest that reduced activity of glomerular MMPs may contribute to the accumulation of ECM and mesangial expansion in diabetic nephropathy. Recently, several cytokines and growth factors have been identified that may stimulate matrix production or reduce matrix degradation [11, 12, 13, 14]. In particular, transforming growth factor-β (TGF-β), is an important regulator of glomerular ECM degradation [75]. Several investigators have demonstrated that TGF-β mRNA and protein expression is increased in glomeruli and renal cortex of streptozotocin-induced diabetic rats and that insulin treatment ameliorates these changes [76, 77]. Moreover, it is known that TGF-β increases the synthesis of ECM proteins [78, 79, 80, 81]. Park et al. [82]showed that mRNA expression of both TGF-β and type IV collagen was increased in glomeruli during the early phase of rapid renal growth in diabetic rats and that insulin treatment inhibited the increased expression of both mRNAs. Yamamoto et al. [83]reported that a key action of TGF-β was induction of ECM production, and that matrix proteins including fibronectin, tenascin and proteoglycan biglycan were increased by TGF-β in diabetic rats. Furthermore, in our in situ hybridization study using a TGF-β probe, TGF-β signal was observed in glomerular epithelial cells, mesangial cells and epithelial cells of Bowman’s capsule in diabetic nephropathy [84](fig. 2). The signal intensity was very strong in tissues from mild and moderate mesangial expansion. These data suggest that TGF-β plays an important role in the regulation of ECM production and degradation in diabetic nephropathy. It has been reported that advanced glycation end products (AGEs), which are non-enzymatically glycated proteins, contribute to the pathogenesis of diabetic nephropathy and may be associated with the balance between turnover and degradation of ECM [6, 7, 8, 85, 86, 87, 88]. We have reported that a monoclonal anti-AGE antibody reacted with the mesangial area in patients with diabetic nephropathy and that the intensity of AGE staining gradually increased in parallel with the degree of mesangial expansion [7]. Moreover, we showed that a monoclonal antibody to reduced glycated lysine reacted with GBM in the glomeruli of diabetic nephropathy patients [8]. AGEs stimulate matrix production and inhibit mesangial proliferation [85]and glycated GBM is more resistant to digestion by protease [86]. Doi et al. [87]showed that mouse mesangial cells exhibited an increase in mRNA and peptide synthesis of type IV collagen, laminin and heparan sulfate proteoglycan after exposure to AGEs. Yang et al. [88]demonstrated that AGEs induced an increase in mRNA expression for ECM proteins such as type IV collagen, laminin, and TGF-β. Conversely, the AGE inhibitor, aminoguanidine, reduced all these changes. These results indicate that AGEs may contribute to both the production and degradation of ECM proteins in diabetic nephropathy. Although many studies have been performed to investigate the balance between production and degradation of ECM in diabetic nephropathy, the precise mechanism is not fully understood. As shown in figure 3, the high level of ECM production and TIMPs expression and low level of MMPs easily explain ECM expansion in diabetic nephropathy. However, a low production of MMPs together with reduced ECM and TIMPs production may also result in ECM expansion. The latter may also occur in the presence of high levels of ECM production with high MMPs production. Moreover, several factors such as cytokines, growth factors and AGEs are implicated in ECM production and degradation. In the balance between synthesis and degradation of ECM, previous studies have largely been limited to animal models and in vitro experiments; however, these studies cannot be readily extrapolated to humans. Diabetic nephropathy is an insidious process that occurs over a long period and accurate models of diabetic nephropathy are therefore difficult to establish. Further human studies examining the balance between ECM production and degradation are therefore warranted.