MMPs are zinc-dependent enzymes involved in ECM remodelling, which play a central role in tissue homeostasis. There are 23 MMPs in humans[
105] and at least ten of them are expressed in the kidney (MMP-1, -2, -3, -9, -13, -14, -24, -25, -27 and -28)[
106]. MMP-12 was thought not to be expressed in the kidney even though some experimental results in an animal model suggest the opposite[
107]. MMPs were hypothesized to be anti-fibrotic due to their function as ECM degradation enzymes. Increasing evidence suggests that MMPs have a more complex role in renal fibrosis[
4,
108,
109]. For example, MMP-9-mediated degradation of collagens creates collagen fragments, which possess chemotactic properties for neutrophils and are able to stimulate MMP-9 production. Apart from their action on ECM components, MMPs are also known to modulate growth factors and their receptors (TGF-β, FGF-R1), adhesion molecules (integrins and cadherins)[
109], cytokines and chemokines. Consequently, MMPs are involved in several processes aside from ECM remodelling, such as destruction of the basement membrane, angiogenesis, cell migration and cell apoptosis, some being pro- and some anti-fibrotic depending on the context[
4,
109,
110]. MMPs are inhibited permanently by degradation or temporarily by tissue inhibitors of metalloproteinases (TIMPs). A balance between MMP and TIMP activity is essential for ECM homeostasis[
109]. Among the four TIMPs that have been identified in vertebrates, TIMP-1, -2 and -3 are expressed in the kidney[
106]. Increased mRNA and protein levels of TIMP-1 were reported in several human and rodent models of different renal diseases, suggesting that TIMP-1 might be involved in the early events during the progression of renal diseases[
5,
73,
109]. TIMP-2 has also been shown to be elevated in various rat models of renal disease[
73]. The exact localization and temporal expression of MMPs in the human kidney is still not completely understood[
108]. Most of the data on MMP expression derive from animal models of kidney diseases (Additional file
1: Table S1). MMP-2 and MMP-9 are known to be involved in the proteolysis of collagen type IV, which accumulates in the basement membranes, for example in early stages of DN. MMP-2 and MMP-9 expression and activity were up-regulated in different animal models of renal fibrosis[
69,
91,
109,
111,
112], but were decreased in cases of DN in both humans and rats[
69,
113]. Changes in MMP-2 and MMP-9 activity might therefore influence the ECM composition causing renal damage at early stages of DN[
106]. However, another study showed that urinary levels of MMP-9, together with collagen type IV, were elevated in type 2 DN patients with macroalbuminuria[
114]. MMP-3 expression and activity during DN was decreased in both humans and rats[
69]. MMP-7 is not expressed in healthy human kidneys but was found in epithelial cells and atrophic tubules in patients with ADPKD and in a mouse model of acute renal tubule injury and chronic progressive renal fibrosis[
115]. Some of the contrasting results, particularly in regards to MMP-2 and MMP-9, can be explained by the impossibility to distinguish between the active and the inactive form of the protease with the commercially available assays. In many cases the findings are based on up- or down-regulated expression of MMP genes, which do not necessarily translate into an increased presence of active proteases. This is the main limitation in the use of MMPs and TIMPs as markers of renal fibrosis. Given the functional complexity of the MMPs, it is likely that they themselves might not be suitable biomarkers of renal fibrosis.