Chronic kidney disease (CKD) is a pathological condition that occurs when the structure and function of the kidney changes, which in turn can lead to complications, such as hypertension, edema, and oliguria, as well as high levels of serum creatinine or blood urea nitrogen [1]. Kidney fibrosis (KF), the most prevalent histological hallmark of CKD, is a condition marked by an increased deposition of extracellular matrix (ECM) in the interstitium. KF disrupts kidney architecture and exacerbates kidney dysfunction [2]. Nowadays, it is widely accepted that KF involves an epithelial–mesenchymal transition (EMT) of renal tubular epithelial cells prompted by abnormal activation of myofibroblasts and subsequent ECM deposition. Transforming growth factor $\beta$ (TGF-$\beta$) regulates the most important processes of this complex mechanism. The first phase of EMT is characterized by both the loss of stability of the various cell–cell contacts and the disruption of apical–basal polarity. This destabilization is caused by stimulating tubular epithelial cells through cytokines, such as the epidermal growth factor (EGF), basic fibroblast growth factor (FGF-2), and TGF-$\beta$1. These cytokines are secreted mainly by infiltrating mononuclear cells, partially reducing epithelial membrane junctions and tight junction proteins, such as E-cadherin and zonula occludens. In the intermediate phase, there is a shift towards the expression of new mesenchymal proteins, including $\alpha$-smooth muscle actin ($\alpha$-SMA) and fibroblast-specific protein 1 (FSP-1) synthesis. In this stage, there is a protein reorganization of the cytoskeleton and an upregulation of both matrix metalloproteinases (MMP-2 and MMP-9) and enlarged spindle-shaped myofibroblasts, which show increased capacity for cell migration and invasion into the tubular interstitium. The observation that epithelial markers are expressed in interstitial cells in biopsies from End Stage Renal Disease (ESRD) patients suggests that EMT is involved in KF. The detection of FSP-1 in the tubular epithelium is another sign of EMT [3]. In human biopsy samples, there is a correlation between tubular cells presenting EMT markers, the appearance of interstitial fibrosis, and functional involvement [4]. However, there is a lack of support from in vivo cell-fate tracking experiments in showing the direct involvement of EMT in contributing to the myofibroblast population [5]. Recent studies have shown that a partial EMT (pEMT—wherein tubular epithelial cells acquire one or two phenotypic markers without leaving the local microenvironment) was induced by Snail1 or Twist-related protein 1 (TWIST1) in tubular epithelial cells, leading to the onset of fibrosis. Consistent with this, pEMT has been observed in clinical samples and was found to correlate with the progression of fibrosis in renal allografts [5]. Inflammatory processes, beginning with the infiltration and activation of macrophages, are considered to be major factors leading to fibrotic diseases [6]. The recruitment of macrophages in the glomerular or tubular interstitium is necessary to promote innate immune responses and important defensive and disruptive reactions in renal disease. The subsequent macrophage infiltration leads to kidney structure disruption and tissue fibrosis development. Glomerular and interstitial macrophages, although believed to be associated with the development of interstitial fibrosis, play an important role in stromal remodeling during tissue repair [7, 8]. In addition to releasing inflammatory cytokines, activated macrophages also secrete matrix metalloproteinases (MMPs), including high levels of MMP-1, -3, -7, -9, -10, -12, -14, and -25, along with lower levels of MMP-2, -3, -8, -10, -11, and -12 [9]. An inflammatory injury in the kidney is also the result of these MMPs contributing to the degradation of the extracellular matrix [10]. The profibrotic changes in tubular epithelial cells [11] and the proteolytic activation of osteopontin in kidney fibrosis led to macrophage recruitment, indicating that MMP-9 causes fibrosis [12].

Matrix metalloproteinases (MMPs) are zinc-dependent endopeptidase enzymes. They function to remodel the ECM and are fundamental to tissue development and homeostasis. Previously, MMPs have been considered to have a negative impact on human health; today, their role in physiology and general health is being progressively clarified [11, 12, 13]. In fact, MMPs regulate many physiological processes. These include angiogenesis, embryogenesis, morphogenesis, and wound healing, although they may also induce various pathological conditions, similar to myocardial infarction, fibrotic diseases, osteoarthritis, and cancer.

Some studies demonstrated that specific increases in MMP could play a role in arterial remodeling, aneurysm conformation, venous dilatation, and pulmonary embolism diseases [14, 15]. Many MMPs are secreted as pro-MMPs, requiring activation through prodomain cleavage, typically by plasmin or other MMPs. Indeed, 24 MMP genes have been identified in the human genome. Notably, MMP-23A is a duplicated version of MMP-23B and is likely a pseudogene [16, 17]. Each MMP contains the following components: (1) an N-terminal signal peptide that directs the enzyme to the secretory pathway and (2) a prodomain with a conserved PRCGXPD sequence responsible for the latency of the enzyme bound to a conserved HEXGHXXGXXH motif [18]. Four twisted $\beta$-sheets are present in the hemopexin-like domain’s ‘propeller’, allowing substrate recognition [14] Together, the hinge region and the hemopexin-like domain break down substrate proteins for further degradation. The C-terminal domain interacts with Tissue inhibitors of metalloproteinases (TIMPs) and participates in substrate recognition [19]. MMPs are also able to cleave a range of non-ECM substrates, such as cell adhesion molecules (cadherins and integrins), as well as growth factors and their receptors (TGF-$\beta$), fibroblast growth factor receptor 1 (FGF-R1) [20]. Tissue inhibitors of metalloproteinases or TIMPs, which have been identified in vertebrates, control the activities of MMPs. There are four TIMP members, TIMP-1, -2, -3, and -4, that inhibit all MMPs, except for TIMP-1, which does not inhibit MMP-14, -16, or -24 [21]. TIMPs can induce intracellular signaling such as proliferation, angiogenesis, and apoptosis [22]. Based on substrate and sequence homology, MMPs are classified into six distinct groups: collagenases (MMP-1, -8, -13, -18), gelatinases (MMP-2, -9), stromelysins (MMP-3, -10, -11), matrilysins (MMP-7, -26), membrane-bound matrix MMPs (MMP-14, -15, -16, -17, -24, -25), and the “other MMPs” (MMP-12, -19, -21, -23a, -23b, -28) [21]. TIMPs are also classified numerically from TIMP-1 to TIMP-4 [23]. Gelatinases are responsible for cleaving and further degrading triple-helical fibrillar collagens. They also degrade basement membrane constituents; for example, MMP-2 (but not MMP-9) can degrade laminin and fibronectin [24]. Stromelysins degrade extracellular matrix proteins but not triple-helical fibrillar collagens. Membrane-bound MMPs activate many proteases and growth factors. All TIMP molecules have 190 amino acids, which are folded into two distinct domains. These domains comprise a larger N-terminal domain (approximately 125 amino acid residues) and a smaller C-terminal domain. Three conserved disulfide bonds stabilize each domain. The N-terminal domain alone can fold individually and is fully functional for MMP inhibition. While the function of the C-terminal domain is not fully understood, it has been shown to tightly bind to the hemopexin domain of latent MMPs [20]. Regulation of MMP activity may occur through several mechanisms: (1) gene expression, by both transcriptional and posttranscriptional modifications; (2) compartmentalization that occurs by incorporating cell-specific and tissue-specific release of extracellular MMPs; or (3) through the zymogen activation of the pro-enzyme forms of MMPs achieved by the removal of the 80–84 residue terminal domain; (4) inhibition by TIMPs. While TIMPs 1–4 all share some sequence similarity, TIMPs 2, 3, and 4 are more identical than TIMP-1. Furthermore, TIMP-2 is unique in that it can both inhibit and activate MMPs [25]. All TIMPs inhibit the pro- and active forms of the MMPs, although with relatively low selectivity, and with which they form tight non-covalent complexes. TIMP-1 has shown low inhibitory activity against MMP-19 and membrane-bound MMP-14, -16, and -24 while displaying greater potency against MMP-3 and MMP-7 than TIMP-2 and TIMP-3 [26]. TIMP-2 has an effective inhibitory activity against all MMPs. Similarly, TIMP-3 inhibits all MMP members while extending its inhibitory profile to include members of the ADAMs (a disintegrin and metalloprotease) family of metalloproteinases. MMP-1, -2, -3, -9, -11, -13, -14, -24, -25, -27, -28 and TIMP-1, -2, -3, -4 are expressed in the kidney [27]. The distribution of MMPs and TIMPs in human kidney tissue and their localization at the subcellular level and immunohistochemical study are available at https://www.proteinatlas.org/[28].

Gene polymorphisms can influence both the expression and activity of MMP-9. Two functional polymorphisms affecting the transcriptional activity of the MMP-9 gene were identified in the promoter region, namely the C-1562T (rs3918242) and microsatellite (CA)${}_{n13-25}$ (rs3222264) polymorphisms [44]. In particular, the microsatellite (CA)n region located near the -90 positions functions as a binding site for a specific DNA regulatory protein and also promotes the unwinding of the DNA helix and the transcription process by allowing the DNA to adopt a Z structure. The longest repeat alleles of the microsatellite were correlated with the highest activity, as demonstrated in vitro. The single nucleotide polymorphism C-1562T prevents the nuclear repressor protein from binding to the MMP-9 gene promoter region, resulting in increased MMP-9 expression, as observed in some in vitro studies [45, 46]. A large Japanese cohort study has highlighted that the MMP-9-1562T/279R/668Q haplotype reduced the risk of CKD compared with the MMP-9-1562C/279R/668R. Furthermore, the MMP-9-1562TT/279RR/668QQ genotype combination has been linked to a decreased CKD risk compared with the major combination of homozygous allele MMP-9 -1562CC/279RR/668RR. The association of these genotypes supports the idea that they play a protective role in kidney disease and could cause an increase in MMP-9 expression [47]. Earlier studies have also shown that the CC genotype for the -1562 C/T (rs3918242) polymorphism in the promoter region causes a significant increase in serum MMP-9 in hemodialysis (HD) patients compared with the CT genotype [48]. Conversely, in nephrolithiasis, the -1562 TT (rs3918242) genotype, but not the Q279R (rs17576) single nucleotide polymorphism (SNP) of the MMP-9 gene, was associated with a higher risk of lithiasis [49]. It has been shown, in a 3-year study of HD patients, that the 2G/2G homozygous genotype for MMP-1 was associated with increased mortality in contrast to the 6A/6A homozygous genotype for MMP-3 in the same population [48]. These findings suggest that investigating the genetic polymorphisms in MMP genes might help identify high-risk populations that can benefit from a targeted treatment approach.

Identifying the molecular components involved in the signal transduction pathways activated by MMPs may allow interventions to be developed by modifying their activity. However, there have only been a few studies investigating the signaling pathways involved in the cellular actions of MMP-9, with several molecules identified as playing a role in MMP-9 signaling, as discussed in the following subsections.

4.3 Notch Signaling

Zhao et al. [56] reported the induction of Endo-MT in human kidney glomerular endothelial cells (HKGECs) by recombinant MMP-9 (rMMP-9). It was observed that rMMP-9 caused a decrease in Notch1 receptor levels while simultaneously increasing the levels of the Notch intracellular domain (NICD) [56]. The Notch signaling pathway consists of the Notch receptors together with several ligands, although in the canonical signaling pathway, Notch is cleaved to form the NCID, which is then transported to the nucleus to regulate transcription of its target genes [57]. Zhao et al. [58] also demonstrated that using a Notch pathway inhibitor, a $\gamma$-secretase inhibitor (GSI), inhibited the rMMP-9-induced morphological changes and decreased NICD levels in HKGECs. In a following study performed by the same group, Zhao et al. [58] used primary mouse renal peritubular endothelial cells (MRPECs) to observe Endo-MT following induction by MMP-9. Once again, rMMP-9 caused an increase in NICD, whose levels were decreased by GSI. Furthermore, in MRPECs isolated from MMP-9 knockout (KO) mice and subjected to TGF-$\beta$1, higher levels of Notch1 and lower levels of NICD were observed. Finally, Zhao et al. [58] performed unilateral ureteral obstruction (UUO) of the kidneys in MMP-9 KO and wild-type mice to induce fibrosis. Zhao et al. [58] noted that the MMP-9 deficient UUO kidneys had less fibrosis, and levels of hairy/enhancer-of-split related with YRPW motif 1 (Hey-1), a downstream target gene in the Notch signaling pathway, decreased, implicating Notch signaling in the MMP-9 mechanism of action. Interestingly, it is worth mentioning that a complex interaction between Notch and NF-$\kappa$B exists, which could activate the NF-$\kappa$B pathway (Fig. 1) [59].

Fig. 1.

Signaling molecules affected by MMP-9. MMP-9, matrix metalloprotease-9; Ac-PGP, N-acetyl–proline–glycine–proline; NICD, Notch intracellular domain; IL-8, interleukin-8; NF-$\kappa$B, nuclear factor-$\kappa$B; ERK1/2, extracellular signal-regulated protein kinase 1/2; p38, p38 mitogen-activated protein kinase.

Several studies have indirectly demonstrated the involvement of MMP-9 in renal fibrosis through the inhibition of t-PA (as t-PA is an inducer of MMP-9) and through knockout mice. To determine the role of MMP-9 in vivo, direct inhibition of MMP-9 activity is the recommended approach [60]. Indeed, a study [60] has shown that inhibiting the activity of the MMPs early, particularly of MMPs-2, 3, and 9, seems to protect against Alport’s disease in mice deficient in the $\alpha$3(IV) chain of type IV collagen. The study showed that late-stage inhibition of MMP activity results in more rapid disease progression, which is associated with interstitial fibrosis and early mortality. Some macrophage phenotypes may contribute to fibrosis, and evidence suggests that neutralizing proinflammatory macrophages may prevent renal fibrosis [61]. In vitro studies have shown that MMP-9 cleaves osteopontin, inducing migration of macrophages, and that MMP-9 from both tubular epithelial cells (TECs) and macrophages induces EMT in tubular cells [12]. In the same study, a unilateral ureteral obstruction murine model of renal fibrosis was used to show that TECs were the main source of MMP-9 in the early stages of UUO, whereas TECs, macrophages, and myofibroblasts were the main sources in the later stages. Using a murine renal tubular epithelial cell line, a study investigated the effects of incubating these cells in an activated macrophage-conditioned medium (AMCM) containing MMP-9 secreted by Lipopolysaccharide (LPS)-stimulated macrophages. They noted that the AMCM could induce EMT in tubular cells, and this effect was inhibited in the presence of MMP-9 inhibitors and by immunoprecipitation of MMP-9 from the AMCM [11]. A recent study also demonstrated how diosmin has a potential multicomponent molecular mechanism of action in treating renal fibrosis. The study indicates that caspase 3, MMP-9, annexin A5, and heat shock protein 90-alpha family class A member 1 could be important direct targets of diosmin and that the MAPK, Ras, phosphatidylinositol-3 kinase (PI3K) Protein kinase B (Akt), FoxO, and hypoxia inducible factor-1 (HIF) signaling pathways could play a major role in its mechanism of action. Further studies on diosmin in treating renal fibrosis should be performed to demonstrate its efficacy [62]. However, as evidence for the role of MMP-9 in renal fibrosis increases, understanding its secretion, regulation, and mechanism of action is necessary to develop therapeutics to regulate its action.

Renal fibrosis is caused by excessive accumulation of the extracellular matrix (ECM), and its pathogenesis is a progressive process that leads to end-stage renal disease. The accumulation of ECM and uncontrolled matrix degradation play an important role, mainly involving MMPs and TIMPs, which are particularly important in the development and progression of renal glomerulosclerosis and tubular interstitial fibrosis. MMPs have traditionally been considered antifibrotic factors due to their proteolytic activity and extracellular matrix degradation. However, a decrease in MMP proteolytic activity or an increase in MMP tissue inhibitors (TIMPs) is thought to be responsible for extracellular matrix accumulation and fibrosis [27]. In particular, in this review, we have highlighted the important role that MMP-9 plays in up-regulating pathways involved in renal fibrosis [58]. Therefore, further studies are needed to demonstrate that inhibition of MMP-9 pathways could represent a therapeutic strategy for treating renal fibrosis in chronic kidney disease. In addition, it would be interesting to understand the intracellular signaling pathways involved in the action of MMPs, particularly the role of MMPs and TIMPs in the kidney, to elucidate the underlying mechanisms and develop biomarkers and therapeutic targets for renal fibrosis.

ALR, RS and MA designed the study. ALR, RS, TF, GCr, AB, GCo, DB, YB, NI, DC, AM and MA performed the search and analyzed the data. ALR, RS, AM and MA wrote the manuscript. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

Not applicable.

Not applicable.

This research received no external funding.

The authors declare no conflict of interest. Raffaele Serra is serving as one of the Editorial Board members of this journal. We declare that Raffaele Serra had no involvement in the peer review of this article and has no access to information regarding its peer review. Full responsibility for the editorial process for this article was delegated to Amancio Carnero Moya.