Introduction
The matrix metalloproteinases (MMPs) are zinc-dependent endopeptidases within the metzincin subfamily, composed of 25 structurally related peptides [1], classified according to their presynthetic region on chromosomes, their various substrate specificities, and partly to their cellular localization. The MMPs share a common functional domain structure. Three common highly conserved domains constitute the general MMP architecture: the pro-domain, the catalytic domain containing the highly conserved zinc-binding site, and a hemopexin C terminal domain [2]. The MMPs are also classified according to fibronectin-like repeats and a flexible hinge region.
Based on their structural features, MMPs can be divided into six major groups, namely: collagenases, gelatinases, stromelysins, matrilysins, membrane-bound MMPs and “other MMPs.” Number designations, MMP-1 to MMP-28, are used for classification [3]. MMPs contain the motif His-Glu-X-X-His (X represents any amino acid) that binds zinc in the catalytic site, as well as another zinc molecule and two calcium molecules structurally. The matrixin subfamily are EC designated 3.4.24.x. This group also contains astacin, reprolysin, and serralysin, as well as other more divergent metalloproteinases.
All MMPs are initially synthesized as inactive proenzymes. The pro-domain removal of MMP enables the activation of the proteolytic abilities of these enzymes. The activation of proenzymes occurs either intracellularly by pro-protein convertases, or extracellularly by peptidases such as other MMPs, or plasmin [2, 4]. This tight regulation of MMP activity at the transcriptional and posttranscriptional levels leads to the constitutive expression of some MMPs or to the expression of other MMPs only after transcriptional activation [5].
The processes which lead to the activation of zymogens are crucial for the extracellular matrix (ECM) breakdown. MMPs cleave several ECM components including collagens, gelatins, elastin, glycoproteins, proteoglycans, and casein [6, 7]. Moreover, modern biochemical techniques, e.g. the usage of the degradomic approach, high-throughput and high-resolution analysis, allow the uncovering of not only other ECM substrates, but also novel non-ECM substrates of MMPs that modify cellular functions [2, 7, 8]. MMPs cleave cell surface receptors, process diverse bioactive molecules including chemokines and growth factors [8]. They are also implicated in the release of apoptotic signals and the activation or inactivation of many cytokines [8]. The extracellular MMP activity ranges from changes in cell-to-cell interactions, cell to ECM adhesions, cell survival, and proliferation to tissue composition and cell motility [8, 9].
The proteolytic activities of the MMPs are further controlled mainly by specific endogenous inhibitors comprising tissue inhibitors of metalloproteinases (TIMPs), α2-macroglobulin, netrins, tissue factor inhibitor 2, type I collagen C-proteinase enhancer protein, cell surface inhibitor, and the reversion-inducing cysteine-rich protein with Kazal motifs (RECK) [2, 10]. In general, all MMPs are inhibited by TIMPs. The mechanism of TIMP inhibition encompasses the binding of N-terminal domain to the catalytic site of the MMPs, where a cysteine-chelating group inactivates the catalytic zinc site with an N-terminal amino and carbonyl group [1, 11](Fig. 1). TIMPs specifically inhibit MMPs and regulate ECM turnover and tissue remodeling by forming tight-binding inhibitory complexes with MMPs, binding the zinc atom on the MMP active site [12]. Thus, the role of TIMPs involves the maintenance of the balance between ECM formation and ECM breakdown.
Fig. 1.
Schematic structure of the MMP-TIMP interaction. Modified from Gu et al. [301] with permission from ASPET.
MMPs have significant substrate overlap. In general, the collagenases MMP-1, MMP-8, and MMP-13 degrade fibrillar collagen, which is further cleaved by gelatinases MMP-2 and MMP-9. MMP production is tightly regulated, and its expression may be cell specific and depends on signals that are temporarily expressed [13]. MMPs play a pivotal role in embryonic development, tumor growth, and tissue fibrosis in the liver, lung, and kidney [13, 14].
Detailed information on immunohistochemical analysis and subcellular localization of MMP expression in human kidney tissues is available at www.proteinatlas.org [15, 16].
For better orientation in the following text, a short recapitulation of cells present in the kidney is provided.
In the kidney, the capillaries and the mesangium are covered with epithelial cells (podocytes), constituting the visceral epithelium of Bowman’s capsule. Visceral epithelial cells produce the components of the glomerular basement membrane. The parietal epithelium of Bowman’s capsule is composed of squamous epithelial cells located on a basement membrane. Mesangial cells are derived from a smooth muscle lineage and are adjacent to the endothelium and thus regulate signaling, recruitment of nonresident cells, and maintenance of the vascular tone in the kidney. There are other mesangial cells nested in the kidney originating from macrophages and monocytes. The mesangial cells are embedded in ECM matrix comprising collagen IV, V, fibronectin, laminin, and proteoglycans. The tubular epithelium contains a single layer of cells anchored to the basement membrane. The collecting duct contains two types of cells: principal cells and intercalated cells, these cells regulate the transport of ions and water. The juxtaglomerular apparatus contains renin-producing juxtaglomerular cells. The interstitium of the kidney is composed of the resident fibroblasts providing a scaffold frame for glomeruli, tubuli, and blood vessels. The interstitium also contains the migrating cells of the immune systems, mainly the dendritic cells. The space between cells is filled by ECM proteins, e.g. proteoglycans, glycoproteins, fibrils, and interstitial fluid.
In the kidney, glomerular intrinsic cells and tubular epithelial cells synthesize MMPs [16, 17]. In addition, macrophages also produce MMPs. Macrophages are cells which play an important role in most human glomerulonephritides [17, 18]. The expression profile of MMPs and TIMPs along the nephron is shown in Figure 2.
Fig. 2.
Known expression profile of MMPs and TIMPs along the nephron in healthy and disease states. Modified from Tan and Liu [20] with permission from The American Physiological Society. Data is accumulated from many original sources.
MMPs are dominant regulators of ECM formation and breakdown in the glomerulus. Differences in MMP expression or MMP activity determine intrarenal ECM composition [15, 19, 20]. Moreover, MMPs in line with TIMPs influence multiple processes, including: chemokine and cytokine production and release, chemotaxis, the release of apoptotic signals, the activity and proliferation of multiple cell lines [20].
The current knowledge of MMPs in the kidney and recent data from animal and human studies with reference to specific renal disease are further elucidated in this review (Table 1).
Collagenases (MMP-1, MMP-8, MMP-13, and MMP-18)
Collagenases: Structure and Function
The collagenases (MMP-1, MMP-8, MMP-13, and MMP-18) cleave fibrillar collagens into 3/4 and 1/4 fragments, which become available for degradation by other MMPs, specifically by the gelatinases MMP-2 and MMP-9 [5]. The other important substrates of collagenases include gelatin, aggrecan, casein, versican, laminin, tenascin [10]. While MMP-1 preferentially degrades collagen III, MMP-8 cleaves collagen I, MMP-13 preferentially cleaves collagen II over collagen I and III [21]. MMP-18 cleaves collagen I. MMP-18 is present in many tissues but is not detected in kidney [22].
Collagenases and Kidney Disease
In glomerular disease, mesangial cells dedifferentiate in reaction to injury, from a mature cell to a myofibroblastic one capable of contraction and proliferation. Mesangial cells are known to be able to produce ECM constituents including type IV collagen. On the other hand, myofibroblastic cells can produce collagens I and III, which normally are not found in glomeruli [23]. In fact, glomerular sclerosis is triggered by the transforming growth factor-β1 (TGF-β1), resulting in an accumulation of type I and III collagen instead of type IV collagen [23, 24].
The known inducers of MMP-1 are ECM MMP inducer [5, 25], oxidized low-density lipoprotein [5, 26], and such cytokines as tumor necrosis factor-α (TNF-α) and interleukin-1β (IL-1β) [27]. IL-1, TNF-α and also to some extent IL-6, induce MMP-13 expression [5, 28]. In contrast, inflammatory cytokines IL-1β, TNF-α, or interferon-γ stimulate the MMP-8 production in monocyte-derived human macrophages to a much lesser extent [5, 29].
In an experimental study with adriamycin-induced nephropathy, spontaneously hypertensive rats increased kidney MMP-1 overexpression predominantly in the glomerular basement membranes and ECM, and subsequent glomerulosclerosis was most successfully reversed with the use of tempol, most probably due to a blockage of peroxynitrite caused by a combination of oxidative stress and deterioration of matrix proteins [30].
Another experimental study with alloxan-induced diabetes in rodents was associated with low serum MMP-1 levels and increased levels of TIMP-1 expression in promoting kidney damage [31].
The only study of end-stage renal disease (ESRD) patients caused by hypertension failed to detect any differences in circulating MMP-1 levels between the patients with essential hypertension, hypertensive ESRD patients, and normotensive subjects [32]. In patients with polycystic kidney disease (PKD), serum MMP-1 concentrations were elevated compared to the controls [33]. Serum MMP-1 levels were found to be significantly increased in the induced pluripotent stem cell-derived endothelial cells from patients with ADPKD and intracranial aneurysms [34].
MMP-1 cleaves interstitial collagens and participates in the ECM turnover, which is a hallmark of interstitial kidney fibrosis [35]. In acute rejection, the proMMP-1 concentrations increase in response to multiple stimuli of inflammatory cytokines [36]. In contrast to that, another study outlined the role of atorvastatin in the changes of MMP expression and signaling within abdominal aortic aneurysms. This study showed significantly reduced MMP-13 expression, a tendency toward reduced MMP-8 expression, and also a trend toward diminished TIMP-1 expression within aortic aneurysms [9]. Furthermore, MMP-1 concentrations tend to rise in more advanced coronary lesions, especially in patients with complex lesions [37], including patients with multivessel disease [5, 38]. MMP-8 levels also rise in patients with coronary artery disease (CAD) compared to those without it [5, 39]. Interestingly, MMP-13 levels were not detected in most of both healthy subjects and CAD patients [5, 29]. Corticosteroids were shown to repress the expression of MMP-8 [40]. Corticosteroids decreased MMP-8-positive eosinophils and macrophages and at the same time upregulated TIMP-1 from macrophages protecting the proteolytic effect of MMP-8 in the airways of asthmatic children [40]. Glucocorticoid treatment downregulates the inflammatory genes through the activation of the glucocorticoid receptor, inhibits the cell expression of MMPs and activates the cellular expression of TIMPs. Experimental studies provide evidence that glucocorticoids mediated the transcriptional suppression of cytokine-induced MMP-9 expression in rat mesangial cells [41, 42]. A recent article exploring the effect of biological sealants and adhesive treatments on MMP expression during kidney damage healing found that the expression of collagenases (MMP-1, MMP-8, and MMP-13) and subsequently the expression of gelatinases (MMP-2 and MMP-9) occur at different tested time periods [43]. This overexpression of metalloproteinases at the early healing period obviously does not influence the final kidney repair [43].
Among collagenases with marked glomerular expression, MMP-13 shows a particular propensity towards collagen I, III, and specifically for collagen type II [44-46]. Regarding the role of MMP-13 in kidney injury, the published data is scant. The absence of the MMP-13 in one study resulted in protection from adriamycin-induced kidney injury in mice [47]. The protective role of a decreased MMP-13/TIMP-1 ratio was defined in a study with cardiotrophin 1 infusion in inducing vascular, cardiac, and renal fibrosis [48]. Another murine study suggests that MMP-13, together with MMP-2 and MMP-9, is capable of regulating ECM turnover in hypertension [46]. A specific pattern of changes in the ECM proteolytic system, including decreased MMP-13 values in patients with left ventricular hypertrophy (LVH), was linked to structural, functional, and clinical signs of hypertensive heart disease, with further reduction in patients with LVH and chronic heart failure [49, 50].
The MMP-1 promotor polymorphism 1G/2G together with MMP-3 promotor polymorphisms 5A/6A carries an increased risk of developing renal cell carcinoma as well as various other cancers [51].
Although circulating levels of collagenases in patients with atherosclerosis lesions and hypertensive heart disease have been determined, the situation in patients with renal diseases is vague. Therefore, future studies are needed to define the significance of collagenases in kidney disease, as well as the possible links between renal disease and increased cardiovascular risk in these patients.
Gelatinases (MMP-2, MMP-9)
Gelatinases: Structure and Function
Gelatinases MMP-2 (gelatinase A, 72 kDa) and MMP-9 (gelatinase B, 92 kDa), both abundantly expressed in the kidney, cleave components in the glomerular basement membrane and degrade other ECM structures of the kidney tissue [52]. All TIMPs can inhibit the active forms of MMPs, but the gelatinases in their latent forms are able to organize complexes with TIMPs. TIMP-1 inhibits the latent and active form of MMP-9, but the role of the pro-MMP-9/TIMP-1 complex is not clear. TIMP-2 mostly inhibits pro-MMP-2 [53]. Although TIMP-2 is a major MMP-2 inhibitor, TIMP-2 also facilitates the activation of proMMP-2 via a cell membrane anchored membrane type 1 metalloproteinase (MT-1-MMP) [11, 54]. Increased expressions of both TIMPs have been associated with glomerulosclerosis [1, 55].
Gelatinases are composed of 3 domains, which are distinguished by the presence of type II additional fibronectin domain inserted into the catalytic domain [5]. The main substrates of gelatinases include: collagen IV and V, aggrecan [56], elastin [57], and denatured collagens (gelatins) [19, 58]. Gelatinases also can degrade collagen types I, VII, X, and XI [59], fibronectin and laminin [60], and vitronectin [61]. Furthermore, MMP-2 has the ability to cleave brevican [62], neurocan [63], and decorin [63, 64].
Both gelatinases MMP-2 and MMP-9 have proinflammatory and anti-inflammatory impacts on numerous tissues [65]. They also process various bioactive molecules significant in kidney diseases, including growth factors such as stromal cell-derived factor-1 (becomes inactive after cleavage) [66], cytokines such as proTGF-β1 [67], proTNF-α [68], and pro IL-1β (activates after cleavage) [68], chemokines (MMP-9 shortens IL-8 at the N-terminus, thus enhancing its potency) [69], and the vasoconstrictor endothelin-1 (ET-1) (MMP-2 participates in the cleavage of big ET-1 into ET-1) [70].
Both MMP-2 and MMP-9 can activate the latent forms of MMPs (pro-MMPs). While proMMP-2, proMMP-9, and proMMP-13 can be cleaved by MMP-9, the proMMP-1, proMMP-2, and proMMP-13 are cleaved by MMP-2 [5].
MMP-2 synthesis and storage occur in the form of a stable, latent pro-enzyme (proMMP-2) form under physiological conditions [5, 71]. Secreted MMP-2 is bound to the tissue inhibitor of MMP 2 (TIMP-2) [72]. This binding occurs between the hemopexin domain of MMP-2 and the noninhibitory C terminal domain of TIMP-2 [11]. Thereafter, the activation of proMMP-2 can be induced by gelatinases MMP-2 or MMP-9 [73, 74]. The inactivation of the enzyme occurs through binding of TIMP-1 (and particularly TIMP-2) to the catalytic site of MMP-2. However, under certain conditions TIMP-2 can positively induce MMP-2 enzymatic activity [5, 75]. TIMPs inhibit all active MMPs, but different TIMPs inhibit different MMPs better than other TIMPs. For example, TIMP-1 inhibits MMP-1, MMP-3, MMP-7, and MMP-9 better than TIMP-2, and TIMP-2 inhibits MMP-2 more effectively than the other TIMPs [76].
The synthesis of MMP-9 also occurs as an inactive proenzyme, which is immediately bound to TIMP-1 [77]. In contrast to MMP-2, MMP-9 can be stored not only in its latent form, but also in an active form [71]. The processes leading to the activation of MMP-9 are rather complex and are finely regulated by other MMPs and TIMPs [5, 78].
The activation of proMMP-9 to the active enzyme form occurs in two major pathways, namely by the exposure to nitric oxide (NO) or proteolytically. The numerous inducers of MMP-9 include MMP-2, MMP-3, collagenase I [78], tissue kallikrein [79], leukocyte elastase, and trypsin [80]. In addition, growth factors, cell surface molecules, cytokines, lipoproteins, and matrix-derived proteins also induce MMP-9 expression [5, 81-83]. Another important novel molecule which can modulate MMP-9 activity is the neutrophil gelatinase associated lipocalin (NGAL), a biomarker of kidney injury. By binding NGAL to MMP-9, the enzyme degradation is inhibited, thereby prolonging MMP-9 peptidase activity [84].
MMP-9 interacts with plasminogen and generates angiostatin. MMP-9 is important for interaction with intracellular adhesion molecule-1 and for the enhanced affinity to collagen. The anti-inflammatory effects of MMP-9 include a change of IL-1β from its precursor and a reduction of IL-2 response. MMP-9 is self-regulated by its own secretion and expression. Furthermore, increased mRNA transcript of MMP-9 is associated with the activation of inflammatory pathways involving nuclear binding of transcription factors nuclear factor-κB (NF-κB) and activator protein-1 complex [83, 85].
TIMPs regulate the effects of MMP-9. The activity of TIMPs facilitates ECM turnover and inflammatory pathways in kidney tissue, thus modifying renal anatomy and physiology [86]. While TIMP-1 inhibits MMP-9, TIMP-2 is the main inhibitor of MMP-2. Enhanced expressions of both TIMPs have been linked with glomerulosclerosis [1, 11, 55]. TIMP-2 in interaction with MT-1-MMP facilitates the activation of pro-MMP-2 [11, 54]. TIMP-2 and TIMP-3 play differential and contrasting roles in renal injury in rodent models: TIMP-3 protects from damage, whereas TIMP-2 promotes injury through MMP-2 activation [87]. TIMP-3, an ECM-bound protein that is expressed ubiquitously, inhibits tumor growth and angiogenesis. TIMP-3 might regulate VEGF-mediated angiogenesis [88]. TIMP-4’s structure shows a closer homology to TIMP-2 and TIMP-3, but much less homology to TIMP-1. TIMP-4 expression in tissues is limited, with prevailing tissue expression in the heart, but with little expression in the kidney. TIMP-4 specifically binds to progelatinase A (pro-MMP-2) [89].
MMP-2 and MMP-9 are especially involved in the cross-talk directly between the cells and ECM, through remodeling of the tubular basement membrane. Indeed, MMP-2- and MMP-9-associated proteolytic processes of type IV collagen are crucial for renal basement membrane remodeling, as well as embryonic growth, tissue remodeling and repair, and certain disease states [90, 91].
Furthermore, a number of drug groups including angiotensin II receptor antagonists [92], diuretics [93], calcium channel blockers [94], statins [95, 96], acetylsalicylic acid [97], and thrombin inhibitors [98] have also been linked to modulate gelatinase activity.
Gelatinases and Kidney Disease
In the kidney, MMP expression varies according to their localization across the entire nephron. Mesangial and epithelial cells along with proximal tubular cells are sites where MMP-2, MMP-9, and TIMPs are expressed the most in the kidney [99, 100]. Due to the unique modifying ability of MMP-2 and MMP-9 to cleave plasminogen to angiostatin, MMP-2 can be mapped either near the tubular basement membrane or near the interstitial space. The presence of MMP-9 in tubular and interstitial cells is increased after acute renal failure due to ischemia in rats [101]. MMP-9 is also produced by neutrophils, macrophages, and T-lymphocytes [102]. In addition, the activity of MMP-3, MMP-12, and MMP-7 leads to the generation of angiostatin [103-106].
MMP-2 and MMP-9 are produced at the first stage of kidney embryogenesis in vivo, but only MMP-9 is absolutely required for kidney organogenesis in vitro [107]. MMP-9 protects the mesenchymal cells from apoptosis and stimulates branching morphogenesis during kidney development through the release of stem cell factor (SCF), suggesting the necessity of MMP-9 in normal kidney development [107-109]. The proposed mechanism in the limited proteolysis of ECM proteins includes synthesis of MMP-2 by the activated mesangial cells, with immediate accumulation of MMP-2 at these sites and their involvement in protease cleavage [100]. The subsequent change of mesangial cells to their inflammatory phenotype cells occurs as a consequence of interactions between ECM and integrins, or as a response to ambient growth factors [110].
Diverse glomerular diseases are associated with profound imbalance in ECM turnover. On one hand, increased formation of matrix proteins accompanies scarring processes; on the other, enhanced degradation facilitates glomerular damage in inflammatory disease states [111, 112]. MMP might play a dual role in primary nephropathies: in the early phases, MMP activity contributes to acute damage, while in the later period MMP protects renal parenchyma against excessive matrix deposition [90, 113].
Both the mesangial matrix and glomerular basement membrane are consistently processed via matrix formation and matrix degradation. The main constituent of the mesangial matrix and glomerular basal membrane is type IV collagen, which is the main substrate of the gelatinases MMP-2 and MMP-9, whereas TIMP-1 is the main inhibitor of this degradation process. Various signals including TGF-β, IL-1, TNF-α, and thrombin influence matrix turnover in the kidney.
These processes are specifically present in chronic kidney disease (CKD). CKD results from a process that causes persistent kidney parenchymal loss or damage. The irreversible reduction in the number of functioning nephrons leads to a gradual inability of renal maintenance of homeostasis over the span of weeks to years. The rate of progression varies according to the underlying pathology, e.g. glomerular disease progresses faster than tubulointerstitial disease, and between individual patients. Progression of CKD is more due to the secondary maladaptive hemodynamic and metabolic changes than due to the underlying disease. In general, CKD comprises complex changes with progressive interstitial scarring, glomerular sclerosis, vessel stiffening, and calcification [114].
The transformation of tubular epithelial cells into profibrotic, mesenchymal scarring cells is linked with fibroblast and cytokine activation leading to ECM remodeling [114]. Indeed, due to the MMPs activities, the imbalanced equilibrium is shifted towards increased matrix synthesis and decreased matrix degradation [115], which promotes the vascular, glomerular, and tubular changes found in CKD [90, 116-118]. An upregulation of MMP-2 and downregulation of MMP-9 activities have been demonstrated in different stages of CKD [58]. In addition, the dysregulation of MMP/TIMP activity and matrix deposition in CKD plays a role in the progressive stages of the disease [20, 114]. Gelatinases impact the initial and progressive changes of CKD in various ways. In children with CKD, MMP-2 levels kept increasing from the beginning of renal failure progression, and the levels of TIMP-1 and TIMP-2 correlated with TGF-β1 [119], suggesting that these markers indicate increased cell damage, inflammation, and aggravation of proteolytic processes in CKD children [119]. A detailed review of MMP-2 and MMP-9 in CKD has already been published, emphasizing the role of the gelatinases in cell structural membrane changes, the activation of various cell signaling pathways, and the processes associated with hypoxia [120].
Moreover, progressive interstitial kidney scarring and tubular atrophy represent the final histological stages of all forms of kidney diseases that develop into ESRD.
Kidney fibrosis accompanies chronic renal diseases. Nowadays, the concept of the progressive glomerular and tubular injury resulting in fibrosis is accepted [121].
Glomerular sclerosis starts with the formation of an adhesion of tuft to Bowman’s capsule, which in degenerative conditions leads to focal segmental glomerulosclerosis while in inflammatory processes to crescentic glomerulonephritis. In both conditions, the outcome is similar, leading to the progression of glomerular sclerosis [122]. Both processes of glomerular sclerosis encroach on other lobules and on the glomerular tubular junction resulting in the obstruction of the urinary orifice. This deprives the tubule from filtrate and leads to atrophy. Altogether, these processes result in the degeneration of the corresponding tubule and the loss of the entire nephron [121, 122].
Kidney fibrosis associated with primary tubular disease comes from the regeneration of the tubular cells. Peritubular inflammation and fibrosis support the tubular cells in recovery. When tubular cells fail to fully regenerate and differentiate, the peritubular fibrosis persists. The progressing tubular disease finally involves the glomerulus causing the progression of the CKD [121].
MMPs actively participate in kidney remodeling by altering matrix degradation, specifically changing the glomerular, tubulointerstitial, and vascular spaces of the kidney. The consequences of such ECM alterations include the accumulation of collagens type I and III, and a higher collagen to elastin ratio leading to enhanced volume flow resistance [46, 123-125]. MMPs also cleave bioactive molecules which may therefore stimulate increased ECM protein deposition [67, 126].
The sources of the interstitial collagen deposition are myofibroblastic cells. Interestingly, the organized tubular epithelium is able to transform into the myofibroblastic phenotype under the influence of TGF-β1 [117]. The proteins of the TGF-β subfamily comprise a group of multifunctional cytokines with an impact on the differentiation of the cells, their organization, and proliferation. The role of TGFs in the development of vascular pathologies has already been described [127]. MMP-2, MMP-9, and MMP-13 are capable of releasing latent TGF-β1 [128]. The released TGF-β1 forms tight, large latent protein complexes, and it is only after the proteolytic degradation that the active TGF-β1 becomes available to its receptors TGF-β RI or TGF-β RII and induces signaling. TGF-β1 stimulates the synthesis of MMP-2 through its activator membrane type metalloproteinase MT1-MMP (MMP-14) [117]. Active MMP-2 is necessary for epithelial mesenchymal transformation in the absence of exogenous TGF-β1. Another interesting ability of MMP-2 includes the mediation of the epithelial mesenchymal transformation by the production of TGF-β1 peptide due to proteolysis in a paracrine manner. Another mechanism of importance in kidney fibrosis encompasses the involvement of MMP-9-dependent Notch pathway activation in CKD [129] through TGF-β endothelial-mesenchymal transition in the murine model of kidney scarring [130]. Moreover, MMP-2 alone is responsible for causing the degradation and disruption of the underlying IV collagen-rich basal lamina which is a critical constituent of renal tubular epithelial mesenchymal transformation [117]. Taken together, this data marks that the active MMP-2 is essential to induce the complex downstream genetic changes resulting in tubular epithelial mesenchymal transdifferentiation.
Autophagy is another important process significant for ECM turnover, where cytosolic proteins in autosomes are presented to lysozyme vacuoles for further degradation [131, 132]. Autophagy has been connected to such renal diseases as AKI [133], PKD [134], diabetic nephropathy [135], cisplatin-induced renal injury [136], tubular injury following sepsis [137], and acute ischemic kidney injury [138].
It has been suggested that in a murine model, renovascular fibrosis due to hypertension can be partly dependent on the enhanced expression and activity of MMP-9, MMP-2, and MMP-13 [46]. In addition, autophagy plays a role in hypertension-induced renal remodeling [46]. Although elevated MMP-9 levels correlated with systolic blood pressure, the increase in plasma MMP-2 and MMP-10 correlated better with renal damage induced by hypertension [32]. MMP-2 degrades calponin-1, which leads to vascular smooth muscle cell proliferation and contributes to hypertrophic remodeling in the aorta. In contrast, the MMP-2 induced increased calponin-1 expression in mesenteric arteries in two-kidney, one-clip hypertensive male rats leading to vessel hypercontractility [139]. The divergent effect of calponin 1, regulated by MMP-2, represents one of the possible mechanisms of vessel maladaptation in hypertension [139]. A recent systematic review focused on blood and urine markers found that MMP-2, MCP-1, and TGF-β might identify patients at risk for the finding of renal scarring in biopsy, and subsequent poor renal outcomes [140].
The most common cause of AKI is acute tubular necrosis (ATN), better termed acute tubular injury after ischemia and reperfusion. Endothelial injury together with epithelial cell injury and inflammation are the main pathophysiological mechanisms involved in ATN. An altered expression of cellular adhesion molecules, cytokine and chemokine release, activation of apoptotic stimuli, leucocyte activation, influence of proinflammatory mediators, intracellular calcium influx, reactive oxygen species lead to the dysregulated immune response in ATN [141].
In the kidney, gelatinases MMP-2 and MMP-9 are upregulated after ischemia/reperfusion injury, and the MMP activation modulates renal microvascular permeability and increased microvascular density [103, 142]. MMP-9 has been found coupled in disulphide-linked complexes with NGAL [143]. NGAL has been suggested as an early predictive biomarker of ischemic AKI in the adult and pediatric population after cardiac surgery [144, 145]. Urinary MMP-9, together with other markers, has also been proposed as a novel biomarker of AKI [146]. A cross-sectional study has demonstrated significantly higher urinary MMP-9 levels in AKI in comparison to urinary MMP-9 concentrations found in CKD patients and healthy control subjects [146]. However, urinary MMP-9 levels in AKI proved to be a poor predictor in distinguishing those patients with urinary tract infection [146].
Moreover, in the rodent model, MMP-9 and SCF have been shown to be important regulators of renal function in AKI. The evidence that MMP-9 protects mice from apoptosis in AKI comes from the ability of MMP-9 to induce the release a soluble form of SCF in kidney tissue. The inhibitory effect of MMP-9 on apoptosis and the rescue by SCF were shown in the ischemia reperfusion model of AKI. This process where MMP-9 facilitated the mobilization of SCF might represent a survival mechanism in AKI [109]. Gelatinases MMP-2 and MMP-9 cleave the tight junction protein zonula occludens-1 in the glomeruli. Moreover, MMP-9 specifically degrades the endothelium-occluding cells, leading to an enhanced vessel permeability [116, 142].
The role of gelatinases in diabetic nephropathy has been illustrated in a number of clinical studies. Indeed, gradual renal remodeling in diabetic nephropathy leads to a deterioration of kidney function, where enhanced storage of ECM plays a crucial role [147, 148]. In diabetic nephropathy, the cardinal histological changes include mesangial expansion, glomerular and tubular membrane thickening and nodular glomerulosclerosis. Further vascular changes, tubular atrophy and interstitial fibrosis appear as diabetic nephropathy progresses.
Among the factors that are involved in ECM breakdown and turnover are reactive oxygen species, plasminogen inhibitor, plasmin, TGF-β1, and MMPs [147-150]. The protein kinase C and hexosamine biochemical pathway activation in diabetic kidney disease leads to the enhanced expression of these profibrotic agents and to induced collagen synthesis and ECM deposition [151].
Mostly, the elevated levels of MMP-2 and MMP-9 in the serum and urine were observed in patients with diabetic nephropathy [152-155]. Elevated MMP-9 levels in plasma even forgo albuminuria development in diabetic nephropathy [154]. Moreover, urinary MMP-9 concentrations correlated with albuminuria as well [156]. Interestingly, an increased MMP-9/TIMP-1 ratio has also been observed in diabetic nephropathy [157]. The enhanced arterial expression of MMP-9 has been noted in diabetic CKD patients, where it is believed to exacerbate the stiffening of the arteries, leading to endothelial dysfunction and impaired angiogenesis in this population [158]. Many factors including MMPs, advanced glycation end products (AGEs), and chemokines contribute to the inflammatory changes in diabetic nephropathy. Using monocyte human renal mesangial cell co-culture system, the complex interconnection between MMP systems and chemokines was identified [159]. Firstly, fractalkine inhibits the mRNA expression and activity of MMP-2, which is important in ECM breakdown. The ECM accumulation is a typical characteristic of diabetic nephropathy, and ECM presence is fundamental for the extravasation of monocytes, and their subsequent migration into the tissues [160]. Secondly, s-fractalkine tempers the intrinsic activity of MMP-2 produced by monocytes and restricts monocyte migration in reaction to MCP-1, illustrating s-fractalkine participation in the regulation of the inflammatory process [161]. Finally, AGEs not only inhibit MMP-2 activity directly, but they also indirectly influence the MMP-2 expression by increasing fractalkine activity [159].
Increased MMP-2 activity and the presence of MMP-2 protein were highly expressed in kidney tissue samples from human diabetic patients [162]. An elevation in urinary MMP-2 levels and MMP-2 enzyme activity has also been observed in type 1 DM patients with albuminuria, in comparison to nonalbuminuric diabetic patients or control subjects [163, 164]. In addition, podocytes were detected in the urinary sediment of albuminuric diabetic patients, but not in patients with type 1 DM without albuminuria or control subjects [165]. Plasma MMP-9 levels correlated with podocyte number in the urine in this diabetic population [165]. A panel of nine biomarkers including MMP-2, MMP-7, MMP-8, and MMP-13, all involved in collagen metabolic turnover and scarring, were measured in 1,765 type 2 DM and CKD patients of various stages in order to predict the progression of kidney function loss [166]. The encompassing role of matrix gelatinases in CKD, cardiovascular disease, and diabetic nephropathy has been recently emphasized in a detailed review [167]. Taken together, these studies demonstrated deregulated activity of gelatinases and TIMPs in clinical trials of human type 1 or type 2 diabetes, providing a link between MMP-2 and MMP-9 expression and the development of diabetic nephropathy.
Increased MMP activity in glomerulopathies participates in the disease process, and their elevated glomerular expression contributes to the extent of structural glomerular damage. In the animal model of MMP-9 deficient mice, the crescentic proliferative glomerulonephritis and AKI were more pronounced because MMP-9 can degrade fibrin [168] and is capable of releasing soluble SCF from renal cells [109].
Patients with various forms of glomerulonephritis including focal segmental glomerulosclerosis, minimal change disease, human immunodeficiency-associated nephropathy and antineutrophil cytoplasmatic antibody-associated vasculitis each had distinct expression of gelatinases [169-174]. Plasma concentrations of MMP-2, MMP-9, and TIMP-1 are remarkably changed in patients with primary glomerulonephritis including IgA nephropathy, membranous glomerulonephritis, minimal change nephrotic disease, and segmental glomerular sclerosis [169, 171]. Plasma profiles of MMPs and TIMPs differ significantly between various histopathological types of primary glomerulonephritides, thus implying diverse mechanisms in the development of the glomerulus and tubulointerstitial fibrosis [169, 171]. Urinary MMP-9/NGAL ratio has been proposed as a potential distinguishing biomarker between minimal change disease and focal segmental glomerulosclerosis in nephrotic children [175].
Changes in plasma concentrations and glomerular expression of gelatinases were found in lupus nephritis, IgA nephropathy, IgA vasculitis, and the postinfectious nephritides [171, 176, 177].
Some studies [178, 179] reported constitutive and unchanged levels of MMP-2 in the serums of systemic lupus erythematosus (SLE) patients. Conversely, MMP-9 serum levels were increased in SLE patients when compared with healthy subjects [179]. Elevated MMP-9 concentrations were found to be correlated with pneumonitis, Raynaud phenomenon, discoid rash, neurological disorders, and the presence of APLA antibodies [179]. In this study, MMP-9 activity levels were not correlated with SLE disease activity indices (SLEDAI, BILAG) in females, but correlations were found with the SLE activity in the group of male patients [179]. It is possible that the differences between disease activity indices and MMP-9 activity levels might be attributed to the influence of sex hormones. Indeed, progesterone was shown to regulate the MMP-9 generation and activity [180].
However, in other studies, the total concentrations of MMP-9 and active MMP-9 were unexpectedly lower in patients with SLE in comparison with control groups [181, 182]. Moreover, lower levels of total MMP-9 were detected in patients with active disease [182]. Similarly, TIMP-1 concentrations were lower in SLE patients than in healthy volunteers [182]. The levels of these markers did not correlate with the activity of SLE, and the presence of clinical and laboratory symptoms of SLE. Lower levels of MMP-9 and TIMP-1 can result from the accumulation of these enzymes in the inflamed vessels and tissues.
The failure of normal polycystin formation stimulates production of growth factors that lead to the dysregulated cell turnover and uncontrolled proliferation of tubular cells in autosomal PKD in response to the abnormal response to elevated cyclic adenosine monophosphate production. Proliferating tubular cells form tubular cysts that close off the associated nephron and expand in size over time. Although only 1% of nephrons undergo such cystic change, the associated profibrotic cytokine release and local ischemic surrounding promote epithelial mesenchymal transformation and kidney scarring and progression of CKD. In addition, the enlarging cysts become walled off from the rest of the collecting duct; the cysts might also impinge the normal blood flow to nephrons [183].
PKD has also been associated with MMP dysregulation. Increased levels of MMP-1, MMP-9, TIMP-1, and type IV collagen have been found in the serum of PKD patients when compared with controls [184]. Elevated levels of MMP-2, MMP-3, and MMP-9 have been noted in patients with Alport syndrome [184]. Increased levels of gelatinases, MMP-2, MMP-9, as well as their inhibitors, TIMP-1 and TIMP-2, have also been demonstrated in allograft rejection [185-187]. MMP-2 and MMP-9 gene polymorphisms have been linked with enhanced allograft survival in renal transplant recipients [188], and the immunosuppressive drug rapamycin downregulates MMP-9 expression while upregulating TIMP-1 [189].
To date, multiple studies exploring potential reliable, noninvasive tumor marker for renal cell carcinoma have not been successful. The metzincin family of MMPs and their inhibitors represent feasible marker candidates for diagnosis and surveillance in the follow-up of cancer. The imbalance of MMPs over TIMPs is accompanied by tissue destruction in malignancies. In renal cell carcinoma using a zymography technique, it has been shown that MMP-9 levels were higher in the serums from clear cell carcinoma in comparison to levels from oncocytoma patients. The prevailing enhanced MMP-9 lytic activity was abundantly present in clear cell carcinoma, whereas MMP-2 activity values were detected in far lesser concentrations [190]. These results were further confirmed by the same author in a study where MMP-9 levels were increased in the clear cell carcinoma patients versus the oncocytoma patients. On the contrary, the mean values of MMP-2 and TIMP-1 were comparable in both tumor groups [191]. The results of the discussed studies imply the potential usage of MMP-9, but not MMP-2 in diagnosing renal cancer. Nevertheless, in spite of the findings in these tissues, the measurement of serum and urine MMPs and TIMPs can so far not be used to diagnose kidney cancer.
Gelatinases are immensely important in the development of cardiovascular alterations associated with cardiovascular disease in patients with renal disease [158, 192]. MMP-9 strongly correlated with carotid atherosclerotic burden, irrespectively of other contributing factors in the early, moderate, and advanced stages of CKD [193]. Similarly, the circulating MMP-2 levels were found to be strongly linked to intima thickness in ESRD patients on hemodialysis [194]. MMP upregulation probably results from the activation of NF-κB proinflammatory pathways [195]. Progressive CKD increased the circulating levels of MMP-2, MMP-9, and TIMP-1 expression in the aorta [192]. Moreover, the initiation of higher MMP-2 activity starts at the predialysis CKD stage. Patients with ESRD on peritoneal dialysis and hemodialysis are characterized by further, gradual MMP-2 activation. The increased expression of MMP-2 could represent one of the presumably nontraditional mechanisms that facilitate multiple adverse processes including structural remodeling, vascular dysfunction, and oxidative stress in increased cardiovascular burden in CKD patients receiving dialysis. The TT allele of the C1562T polymorphism in the MMP-9 gene has been linked to the enhanced oxidative stress in patients with nephrolithiasis, in addition to imposing cardiovascular risk in patients with the homozygous TT genotype of MMP-9 [196].
Furthermore, MMP-2 expression associated with the phenotype changes of vascular smooth muscle cells is typical for elastocalcinosis, and even precedes cell loss and the development of arterial medial calcifications in the early stages of CKD [197]. Vascular calcification comprises a number of events beginning with the apoptosis of smooth muscle cells, change in their phenotype and the release of a multitude of proteases, including MMPs [198]. Activation of the Wnt/β-catenin pathway in vascular smooth muscle cells (VSMC) in the precalcifying (increased calcium and phosphorus concentrations) culture conditions, associated with further enhanced overexpression of gelatinases (MMP-2 and MMP-9), facilitates the CKD-linked vascular calcification in the process of arteriosclerosis [30, 199-201]. Further reciprocal mechanisms between NO bioavailability and MMP activation [3] strengthens the association between MMP activity and endothelium-dependent relaxation in the dialysis patients [158]. Thus, the proteolytic activity of MMP-2 leads to the loss of arterial structural and functional integrity. The activity of gelatinases, mainly MMP-2, could aid in understanding the pathogenesis of the higher prevalence of vascular disease in CKD.
Most of the reviewed studies demonstrated elevated levels of MMP-2 in ESRD on dialysis in comparison to controls [32, 202], but two studies did not [203, 204]. MMP-9 levels were comparable in two studies [194, 202], increased in two other studies [32, 205], and decreased in a different study [203]. These differences are likely due to multiple factors including the diversities in age, ethnicity, and specificities of dialysis membranes, causes of renal diseases or other clinical states. Moreover, MMP-9 genetic polymorphism influences MMP-9 expression in hemodialysis patients [206]. In addition, a single hemodialysis session does not influence or diminish MMP-2 and MMP-9 levels, or TIMP-1 and TIMP-2 concentrations [204, 207]. While in the previously mentioned study neither MMP-2 serum levels, nor MMP-9 concentrations were shown to be mortality predictors in hemodialysis [208], in a 5-year recent cohort study it was shown that MMP-2 serum concentrations in the fifth year predict mortality in hemodialysis patients [209].
It has been shown that treatment with exogenous MMP-9 demonstrated the protective role of MMP-9 in chronic kidney failure in an animal model [210]. It is feasible that the protective impact of MMP-9 is reached either due to the MMP-9 degradative ability of ECM, or via the inhibitory effect on TIMP-1 and the subsequent proinflammatory response [210].
Interestingly, in the rodent model of renal traumatic injuries, the MMP expression, namely MMP-1, MMP-2, MMP-8, MMP-9, and MMP-13 was different. Different time manner was specifically prominent with maximal MMP-8 and MMP-9 expression after 2 days of injury in healing of the wound. In contrast, the MMP-2 expression peaked after 6 days of kidney traumatic injury [43].
Recently, it has been demonstrated that the astacin proteases meprin-α and -β might serve as targets of treatment in kidney fibrosis and scarring. The acidic modification of meprin-β exhibits a specific selectivity for matrix proteases, such as MMP-2 [211].
Taken together, the measurement of plasma and serum levels of gelatinases may aid in the determination of a variety of renal diseases and also in ESRD patients on dialysis, or patients with a transplanted kidney. To date, ambiguous data exists in relation to the potential ability of blood levels of the gelatinases MMP-2 and MMP-9 to serve as markers for kidney disease. Therefore, to unravel these contradictory issues, the exact determination of the possible transient levels of MMP-2 and MMP-9 (and their inhibitors) in the course of renal disease in prospective studies is required.
Stromelysins (MMP-3, MMP-10, and MMP-11)
Stromelysins: Structure and Function
Stromelysins, namely MMP-3, MMP-10, and MMP-11, are capable of degrading many of the constituents of basement membranes and connective tissues. The inability to degrade type I collagen distinguishes stromelysins from collagenases [5, 212]. MMP-3 degrades such substrates as type II, IV, and IX collagens, as well as fibronectin, laminin, proteoglycans, elastin, and gelatins [5, 213]. MMP-3 and MMP-10 can serve as activators of other MMPs, specifically the precursor pro-forms of MMP-1, MMP-7, MMP-8, and MMP-9 [214], and also other proinflammatory mediators, such as pro-TNF-α [215] and osteopontin [216]. Among other targets, MMP-10 degrades collagen types III, IV and V, gelatin, and elastin in vitro. MMP-10 is best known for enhancing tissue plasminogen activator’s fibrinolytic activity [217]. The role of MMP-10 in atherosclerotic vascular remodeling has also been described [218]. In contrast, MMP-11 does not cleave gelatins, collagens, or other major matrix components [219], but is capable of the degradation of serine proteinases like α-1-proteinase inhibitor [220], and also cleaves insulin-like growth factor-binding protein [5, 221].
MMP-11 (stromelysin 3) expression appears to be associated with normal tissue remodeling [17, 222]. MMP-11 is particularly present around a variety of invasive carcinomas, along with stromal fibroblasts [17]. MMP-11, in contrast to other MMPs, presents with a distinct molecular structure, divergent enzyme activity, and different gene arrangement and regulation [17, 223]. The different role of MMP-11 is marked by its weak activity toward matrix substrates, and probably indirect participation in remodeling only through activation of serpins [17].
Stromelysins and Kidney Disease
MMP-3 and TIMP-1 expression are found in the glomerulus (specifically in the mesangial and epithelial cells) and some tubular cells (mostly in the proximal tubular cells) in humans [224]. A diverse group of MMPs, namely MMP-1, MMP-2, and MMP-3, is produced mainly by fibroblasts and smooth muscle cells [225, 226].
Animal studies in rodents highlighted the potential role of stromelysins in the pathogenesis of diabetic nephropathy. The expression of kidney MMP-3 mRNA is diminished in streptozotocin-induced diabetic rats [227]. Kidney tissue analysis from patients with diabetic nephropathy using in situ hybridization techniques revealed that MMP-3mRNA expression was most abundantly present in those with the mildest mesangial expansion, in contrast with the far less mesangial presence in moderate or severe histopathological samples [228]. High MMP-10 levels were associated with nephropathy and proliferative retinopathy in type 1 diabetic patients. The knockout of the Mmp10reduces these disorders in a diabetes murine model [229].
Significant progress was reached in evaluating the importance of MMPs including MMP-3 in the dynamic changes of kidney disease in Alport’s syndrome, using a mouse model of the disease which is deficient in the α3-chain of type IV collagen [α3(IV)] [90, 184]. MMP-3, similar to MMP-2 and MMP-9, appeared in Alport’s syndrome in both human and animal models. Gelatinases and stromelysin MMP-3 are able to cleave the components of the basement membrane in Alport’s syndrome in vitro [230]. The significance of MMPs in Alport’s syndrome differs in relation to the duration of the kidney disease. While in the early stages, the MMPs inhibit the development of kidney lesions, in the later (proteinuric) stage MMPs actually enhance the disease progression [184].
Both MMP-3 and MMP-9 could cause epithelial mesenchymal transition by the degradation of epithelial cell marker E-cadherin in kidney cell cultures of tubular cells, resulting in the activation of β-catenin [20, 231].
Serum MMP-3 concentrations are increased in mesangial proliferative glomerulonephritides such as IgA nephropathy and lupus nephritis [112]. In addition, MMP-3 levels correlate with the clinical parameters of lupus nephritis, including a decreased C4 complement, circulating immune complexes, and decreased creatinine clearance [232]. An enhanced expression of MMP-3 has been observed in IgA vasculitis and ANCA vasculitis [174, 233]. Vasculitis refers to an autoimmune condition with both inflammation and leukocyte infiltration of the blood vessel walls. The damage of the vessel wall encompasses aneurysm formation, vessel occlusion leading to ischemia and infarction.
In comparison to MMP-1 and MMP-9, MMP-3 levels distinguished between active AAV and a disease in remission better than C-reactive protein and erythrocyte sedimentation rate. Furthermore, patients with active ANCA vasculitis at month 6 showed elevated levels of MMP-3 in comparison to those whose disease was in remission [17]. Serum proforms of MMP-2 and MMP-3 illustrate the changes of ECM turnover in the glomeruli and interstitium in chronic transplant nephropathy, and highlight their possible pathogenetic role in the development of this condition [36].
MMP-10 expression has been detected in normal tubular cells [234]. The presence of MMP-10 in cancer cell cytoplasm of conventional renal cell carcinoma was significant and has been independently associated with tumor invasion [234].
MMP-11 can function as a potent mediator of inflammation and demonstrate a chemotactic impact on macrophages, which tend to accumulate in the mesangium and cause early fibrotic changes, participating in the origin of cellular and fibrocellular crescents in the secondary forms of glomerulonephritides [17]. No antifibrotic or mitogenic effects of MMP-11 have been observed within the glomeruli thus far [17].
MMPs influence cell proliferation indirectly. A programmed cell death, or apoptosis, is normally initiated via activation through the Fas receptors or other TNF superfamily receptors the proteolytic cascade of caspases and subsequent cleavage of subcellular structures and nucleus. Kidney tubular cells dispose such receptors [235]. MMPs also display other mechanisms in the regulation of apoptotic signals [236, 237]. Furthermore, the role of MMPs in apoptosis is augmented by TIMPs. The balance between MMPs and TIMPs mediated via various biologically active intracellular or extracellular molecules defines the upregulation or downregulation of apoptosis in relation to tissue characteristics, the levels of biologically active molecules and preservation of cellular homeostasis [238]. Excessive or impaired apoptosis induces tissue scarring and dysfunction [239]. Kidney impairment including ischemia, injury due to nephrotoxicity, radiation, and ureteral obstruction causes apoptosis in the kidney [240].
Cell to ECM adhesion is maintained through direct and indirect bonds to the actin cytoskeleton. Disruption of these connections to ECM has a harmful effect on cell preservation. MMPs induce a specific type of apoptosis known as anoikis altering these cell-to-matrix connections [237, 241]. MMPs promote cellular anoikis through the separation of various cadherins and integrins [242, 243]. The shedding of these adhesion molecules disables the unique assembly of cell-to-cell and cell-to-matrix interconnections in the apoptotically changing cell [244].
MMP-3 stimulates apoptosis by cleaving laminin [245], and MMP-11 suppresses tumor cell apoptosis in the murine model [246]. Furthermore MMP-11 can inhibit apoptosis by releasing IGFs, functioning as survival factors [247]; MMP-11 augments apoptosis in tissue remodeling and crossover [222].
Specifically, MMPs are located and translocated into the nucleus and through proteolytic cleavage control the regulation of cell cycle, nuclear matrix degradation, and apoptosis [248]. MMP-3, for example, stimulates apoptosis by MMP-3 translocation into the nuclei of mammalian cells. This effect on apoptosis is derived from the retention of its enzymatic action [86]. TIMP-1 also binds to the cell surface, and nuclear translocation occurs [249]. MMP-2 and MMP-9 are present in the nuclei and potentially able to cleave nuclear matrix proteins such as protein poly-ADP-ribose-polymerase [250], suggesting functioning of gelatinases similar to those of caspases in the nucleus [237, 251]. Taken together, these observations pinpoint the importance and multiple involvement of MMPs and TIMPs in various apoptotic mechanisms of cell biology. Further research is warranted to investigate the precise mechanisms elicited by stromelysins in the origin and progression of kidney diseases.
Matrilysins (MMP-7 and MMP-26)
Matrilysins: Structure and Function
Matrilysins have the structural peculiarity of lacking a hemopexin domain, having an atypical sixth exon instead [252]. The protease differs from other matrix MMPs by a comparatively smaller size. The enzyme synthesis starts from a 30-kDa inactive preform and continues through a stepwise activation to a definite 18-kDa active form [253, 254]. MMP-7 degrades ECM substrates such as fibronectin, laminin, elastin, entactin, and type IV collagen [5, 253, 254]. In addition, MMP-7 also cleaves proteoglycans, including versican [255]. The full activation of MMP-7 occurs through a series of various enzymes, including trypsin and MMP-3 [256]. MMP-7 could also be activated by aminophenylmercuric acetate in vitro and is partially activated by plasmin and leukocyte elastase [256]. MMP-26 selectively cleaves type I gelatin and α1-proteinase inhibitor. MMP-26 hydrolyzes the proteinous substrate of matrixins and the TNF-α-converting enzyme. Further selectivity of MMP-26 includes the cleaving of ECM proteins, inactivation of serpins, and the modulation of various cytokine effects [257].
Matrilysins and Kidney Disease
The expression of MMP-7 has been mostly associated with tissue remodeling [254, 258]. An overexpression of MMP-7 has been detected in pathological states such as unilateral ureteral obstruction, PKD, and acute folic nephropathy in mice, but no expression of MMP-7 has been observed in physiological human kidney tubular epithelium [254, 259].
In diabetic nephropathy, hyperglycemia causes alteration of matrix synthesis and degradation by MMP-7, which leads to changes of the physiological microarchitecture of the glomeruli. Moreover, MMPs modulate the enhancement of growth factors. Specifically, in early kidney hypertrophy in streptozotocin-treated diabetic rats, an increase in insulin-like growth factor binding protein (IGFBP)-1, IGFBP-3, and IGFGB-5 has been seen [260]. MMP-7 is capable of cleaving IGFBP-3 [261] and IGFBP-5 [262]. Degradation of IGFBPs by MMP-7 leads to the liberation of IGF-II [263], thus releasing the bioactive IGF, which results in the accumulation of ECM and mesangial expansion. TGF-β, a known profibrotic growth factor, becomes more available through MMP-7 enzymatic decorin processing, which is present in ECM and binds TGF-β [263]. Therefore, MMP-7 renal activity contributes in a direct or indirect way to the process of glomerular basement membrane thickening and the expansion of the mesangium, which constitute the initial changes present in the development of diabetic nephropathy.
An increase in MMP-7 expression (verified by renal biopsy) was found in kidney tissue from patients with hydronephrosis at hemodialysis [259, 264]. MMP-7 seems to act as a novel screening marker for renal disease in lupus nephritis [265], cardiovascular alterations in CKD patients [266], and as a possible predictive marker of kidney transplant rejection [267]. Among potential MMP-7 inhibitors, doxycycline seems to be effective in decreasing the protein loss in urine of patients with glomerulonephritis [268] and diabetic nephropathy [269].
MMP-7 expression has been associated with enhanced renal fibrotic changes [116]. Active MMP-7 was not detected in the rodent aging renal tissue, but the pro-form of the enzyme, pro-MMP-7, has been linked to the increase in collagen expression promoting kidney fibrosis. The enhanced expression of MMP-7 activates the Col1a2 and Col3a1 genes through PIK3, p38, ERK, src, and PKA signaling, resulting in the accumulation of collagen in the renal tissue [254]. There is no information on the ability of MMP-7 to cleave Col1a2 and Col3a1. The only collagens that are known to be degradable by MMP-7 are collagen IV [270] and collagen type XVIII [271]. However, MMP-7 activates the gelatinases MMP-2 and MMP-9 [272] and the collagenases MMP-1 and MMP-8, which in turn degrade collagen. Interestingly, the total collagenase activity is decreased and the gelatinase activity is increased in ageing rat kidneys [270]. Furthermore, MMP-7 gene expression was found to be upregulated in aged kidneys and was associated with the typical histopathological fibrotic changes of old kidneys such as glomerulosclerosis, interstitial fibrosis, and tubular atrophy [273]. In a cohort of 102 CKD patients, high urinary MMP-7 levels were observed [274]. The close association of MMP-7 with kidney fibrosis score supports the potential use of MMP-7 as a biochemical marker of renal scarring [274]. Altogether, these results point out the importance of MMP-7 changes in ECM formation and the breakdown accompanied with either acute or chronic loss of kidney tissue [1]. A recent review evaluated the pivotal effect of MMP-7 in kidney fibrosis on the interaction of three pathways, including epithelial-mesenchymal transition, TGF-β signaling activation, and ECM accumulation, and suggested that MMP-7 not only acts as a target in kidney scarring treatment, but also as a marker of renal fibrosis [275].
The prevalent cause of kidney allograft dysfunction is the fibrosis of interstitium and tubular atrophy. Chronic allograft dysfunction results from both immunological and nonimmunological factors causing slow allograft dysfunction, usually with proteinuria. A key structure that determines the outcome of the transplant is vessels. Nonexistent collateral vasculature accompanying progressive vascular narrowing and occlusion leads to downstream hypoxia and ischemia predominantly in the interstitium [276]. The imbalance in MMP/TIMPs activity plays a leading role in ECM remodeling. Although serum levels of MMPs and TIMPs remained unchanged during subclinical tubulitis linked to existing or newly developing interstitial fibrosis and tubular atrophy, the urinary levels of MMP-1, MMP-7, TIMP-1 were several times higher, and were not found to be elevated in quiescent interstitial fibrosis and tubular atrophy [277].
Apparently, MMP-7 participates in the development and evolution of renal carcinoma [278]. Renal cell carcinomas are adenocarcinomas arising from tubular epithelium. MMPs synthesized by stromal or tumor cell lines are included in tumor advancement. The mechanisms of MMP-7 involvement in tumor progression include modulation of invasion and angiogenesis in renal cell carcinoma, and in multivariate analysis MMP-7 acted as a predictor of poor prognosis in renal carcinoma [279]. In addition, it has been shown that the expression of MMP-7 in clear cell renal cell carcinoma was significantly elevated when compared with controls, and it even correlated with the degree of malignancy and the predicted poor prognosis [280]. On the other hand, the TIMP-2 expression level in the same study in clear cell renal cell carcinoma was clearly lower than in the control group [280]. Therefore, MMP-7 and TIMP-2 can potentially be used as novel biochemical markers for the prognosis evaluation in clear cell renal cell carcinoma patients.
In addition, MMP-7 is able to release the membrane-bound Fas ligand, a transmembrane stimulator of the apoptotic receptor Fas [281]. The released Fas ligand then induces apoptosis of surrounding cells [281] or decreases cancer cell apoptosis [282].
In summary, these discussed studies suggest that both acute and chronic reduction of kidney function may be associated with MMP-7-caused changes in ECM turnover.
Enamelysin (MMP-20)
MMP-20 and Kidney
A recent study has shown that MMP-20, together with dentin sialophosphoprotein, is expressed not only in the ductal epithelial cells of human salivary glands, but it was also found in all sites of the human and monkey nephron exempting the glomeruli, with a high predominance in the proximal and distal tubular cells [283]. The dentin sialoprotein and MMP-20 pairing may actively participate in the restoration and turnover of ECM proteins of the basement membranes in the monkey and human kidney epithelial cells [283].
Membrane-Type MMPs (MMP-14, MMP-15, MMP-16, MMP-17, MMP-24, MMP-25)
Membrane-Type MMPs: Structure and Function
Enzymatic activation of MMPs occurs by the removal of the pro-domain, triggered by other proteases such as plasmin, or by cellular membrane-type MMPs. Four of these belong to type I transmembrane proteins, namely MT1-MMP (MMP-14), MT2-MMP (MMP-15), MT3-MMP (MMP-16), and MT5-MMP (MMP-24), while the remaining two represent glycosylphosphatidylinositol-anchored proteins MT4-MMP (MMP-17) and MT6-MMP (MMP-25). MT-MMPs are a unique class of MMPs bound to the cellular surface by a transmembrane domain. MT-MMPs activities are various, with a leading role in guiding the proteolytic events adjacent to the pericellular microenvironmental processes [258]. All MT-MMPs, excluding MT4-MMP, can activate pro-MMP-2.
MT1-MMP (MMP-14) mainly works as the cell surface activator of MMP-2. MT-1-MMP is capable of cleaving several ECM proteins, namely collagen I, collagen III, fibronectin, and laminin [284, 285]. Additionally, MT1-MMP functions as a “sheddase” for CD44 and syndecan-1 surface proteins [286].
MT-MMPs are capable of activating both pro-MMP-2 and pro-MMP-13 on the cell surface. These enzymes play a leading role in the tissue remodeling which accompanies different physiological and pathological processes. In the presence of MT-1-MMP (MMP-14) situated on the cellular surface, a ternary complex comprised of pro-MMP-2, TIMP-2, and MT1-MMP is formed, resulting in the activation of MMP-2 [1, 287]. The participation of MT1-MMP is essential to this activation.
In general, the aim in the MMP activation or inhibition is the maintenance of matrix homeostasis. For instance, MMP-2 in association with MT1-MMP (MMP-14) is needed for the epithelial mesenchymal transition of NRK-52E cells promoted by TGF-β1 in vitro [117].
Membrane-Type MMPs and Kidney Disease
The role of MT-MMP in the development of kidney injury was investigated in a crescentic glomerulonephritis caused by antiglomerulus basement membrane antibodies in Wistar Kyoto rats. In the invading macrophages in crescent lesions, the activated MT1-MMP (MMP-14) was probably responsible for the pathological breakdown of glomerular ECMs via pro-MMP-2 activation [288].
In the human kidney MMP-24 (MT5-MMP), transcription and production are increased in diabetic kidney disease [162]. Moreover, higher levels of MMP-14 (MT1-MMP) are identified in the urine from patients with diabetic nephropathy [153]. MMP-14 is known to be present as a membrane-bound MMP, or as MMP with various soluble forms. The mechanisms of increased MMP-14 production in urine are not entirely clear, but auto-cleaving is thought to be the most important one [289]. Thus, the increased MMP-14 concentrations in urine could be explained either as a result of the enhanced cleavage of soluble MMP-14 from cells located within the nephron, or as a result of the cellular losses of express-bound MMP-14 cells within the nephron.
“Other MMPs” (MMP-12, MMP-19, MMP-21, MMP-23A, MMP-23B, MMP-27, MMP-28)
Overview of “Other MMPs”
Other MMPs, namely MMP-12, MMP-19, MMP-21, MMP-23A, MMP-23B, MMP-27, MMP-28, do not fit easily into the main groups of MMPs. Their expression is typically bound to a single cell or a tissue type, or they are produced in special circumstances only [86, 290].
MMP-12, also known as macrophage metalloelastase, cleaves both soluble and insoluble elastin. MMP-12 may be implicated in aneurysm formation [291] and in the development of emphysema [292]. MMP-12 produced by macrophages infiltrating the glomeruli modulates high-fat diet-induced glomerular fibrogenesis and inflammation in a mouse model of obesity [293]. MMP-19 (MMP RASI-1) expression is linked to the human epidermis and endothelial cells, and MMP-19 participates in such processes as cellular migration, proliferation, angiogenesis, and adhesion. MMP-19 expression has been found in synovial vascularization in patients with rheumatoid arthritis [294].
MMP-21 degrades ECM proteins in such physiological processes as embryogenesis, reproduction and tissue remodeling, as well as in such pathological disease processes as metastasis development and asthma [295]. The MMP23A gene and the almost identical MMP23B gene can be found in the duplicated region of chromosome 1p36.3. It is most probable that MMP23A functions intracellularly. MMP-23B is strongly expressed in the ovary and the heart [296].
In contrast to the majority of MMPs, MMP-27 is not a secretory protease but is efficiently accumulated in the endoplasmic reticulum in three cellular lines of mammals [297]. Despite its sequence homology with MT-MMPs (the unique C-terminal extension which is necessary and sufficient for endoplasmic reticulum), MMP-27 is not provided with a stable membrane anchorage and therefore does not interact permanently with the cellular membrane. There is sparse information about MMP-27 expression during developmental, physiological or pathological processes. In the adult rat, MMP-27 mRNA is abundantly expressed in antiGG/IgM B-stimulated lymphocytes [298] and in the bone and kidney, with much lower levels found in the heart [299].
The expression of MMP-28, also known as epilysin, is bound to testes and keratinocytes where it is seen as a damage response. MMP-28 is involved in such disease processes as asthma and metastases [300].
Conclusion
In this review, we have discussed the significance of MMPs in multiple processes essential for tissue remodeling, with a special focus on the kidney. MMP expression and activation is orchestrated by various mechanisms including cytokines, growth factors, cell-to-cell and cell-matrix interactions and soluble substances. MMPs are vitally important for kidney development and ECM turnover and intervene in their repair and adaptation. The activity of MMPs facilitates the alterations in ECM components, and thus MMPs are implicated in the development of a number of kidney diseases. Profound knowledge of the impact of MMP actions may grant insight into finding new options to effectively cope with kidney disease. New insights, obtained from numerous studies, imply that at early stages MMPs facilitate glomerular basal membrane breakdown (in relation to a specific background and time frame), while at the later stages MMPs assist to digest ECM components accompanying kidney scarring and fibrosis.
Thorough understanding of MMP actions and different facets of renal physiology and disease relies on both clinical and tissue studies. Further progress in mastering control and regulation of MMP actions depends on the development of certain MMP inhibitors, choice of suitable animal models, and finding specific MMP-targeted therapeutic strategies. These future novel medical strategies regulating and controlling MMP activity may help to elucidate the precise ways for conquering kidney diseases, therefore preventing their detrimental clinical outcome.
Acknowledgement
The study was supported by research projects Progres Q25/LF1/2 and MH CZ DRO VFN 64165. The authors are thankful to Assoc. Prof. Věra Čertíková Chábová, MD, PhD, for drawing and preparation of figures 1 and 2.
Disclosure Statement
The authors declare they have no conflicts of interest.
Get Permission
PubMed ID
