The TIMP protein family: diverse roles in pathophysiology

Abstract

The tissue inhibitors of matrix metalloproteinases (TIMPs) are a family of four matrisome proteins classically defined by their roles as the primary endogenous inhibitors of metalloproteinases (MPs). Their functions however are not limited to MP inhibition, with each family member harboring numerous MP-independent biological functions that play key roles in processes such as inflammation and apoptosis. Because of these multifaceted functions, TIMPs have been cited in diverse pathophysiological contexts. Herein, we provide a comprehensive overview of the MP-dependent and -independent roles of TIMPs across a range of pathological conditions. The potential therapeutic and biomarker applications of TIMPs in these disease contexts are also considered, highlighting the biomedical promise of this complex and often misunderstood protein family.

INTRODUCTION

The extracellular matrix (ECM) is a dynamic network of secreted macromolecules that underpin all biological phenotypes in multicellular organisms. The influence of ECM structure and composition on cell phenotype, and vice versa, is referred to as dynamic reciprocity, characterized by reciprocal cross talk between the cellular and extracellular compartments (1). This primordial attribute relies on the continuous remodeling and reshaping of the ECM, a process that is driven by extracellular proteinases. The Matrixin and Adamalysin protease families are the major mediators of ECM degradation, which are members of the Metzincin superfamily of zinc-dependent proteases (2). The Matrixins, also known as matrix metalloproteinases (MMPs), are a family of ∼23 mammalian proteinases that have wide-ranging functions within the ECM and beyond. Similar to the MMPs, the Adamalysins are a large family of important proteinases that are highly active within the ECM, in particular the two subclasses of ADAMs (A Disintegrin and Metalloproteinase) and ADAMTS (ADAM with Thrombospondin Type I repeats). In line with all biological processes, homeostatic mechanisms exist to regulate metalloproteinase activity within tissues, including through degradative processes and direct inhibition. The major mediators of metalloproteinase inhibition are the tissue inhibitors of metalloproteinases (TIMPs). Four paralogous genes encoding TIMP 1 through 4 are conserved throughout mammalian genomes (3). Each contains six disulfide bridges forming six loop regions that produce similar tertiary structures, with sequence similarity between the four proteins ranging from 37 to 51% (4). Despite their sequence and structural similarity, each TIMP possesses unique characteristics and attributes. TIMP1 is an N-linked glycoprotein containing two glycosylation sites at N30 and N78 (5). Similarly, TIMP3 is an N-glycosylated (N184) protein that displays a unique affinity with the ECM through attachments with specific glycosaminoglycans, rendering this TIMP family member anchored with the ECM (6). Furthermore, TIMP3 displays the widest range of inhibitory targets, inhibiting all members of the MMP family, in addition to a range of ADAMs and ADAMTS (7). TIMP2 displays the broadest expression profile of the TIMP proteins and is ubiquitously expressed in normal human tissues (8). TIMP4, however, displays a limited expression profile, with expression detected primarily in the brain (cerebellum), heart, and adipose tissue (4, 9).

Although TIMPs have historically been defined by metalloproteinase inhibitory functions, they contain a wide range of MP inhibition-independent functions (8, 10, 11). TIMPs have a unique relationship with the gelatinases (MMP2 and MMP9), whereby TIMP1/3 can bind to proMMP9 in a noninhibitory complex, and TIMP2/3/4 can similarly bind to proMMP2. These relationships are well-characterized and have direct consequences on the activation of MMP2/9, reviewed extensively elsewhere (11). Furthermore, individual TIMP proteins have been shown to possess specific membrane receptors, including CD63 for TIMP1 and integrin α3β1 for TIMP2 (12–14). These interactions mediate observable signaling pathways that are involved in, but not limited to, granulopoiesis, migration, and growth inhibition (8, 15). The idea that TIMPs are not just tissue inhibitors of metalloproteinases is maturing, and it is anticipated that the catalog of MP-independent functions of TIMPs will continue to expand. Recent work investigating novel functions of TIMP2 has revealed a role in modulating myeloid cell populations (16), expanding on earlier work illustrating how TIMP2 could target myeloid-derived suppressor cells in cancer models (17). The mechanistic details underlying these discoveries are supported by further recent findings describing the expanded proximal interactome for TIMP2 (18).

Most, if not all, diseases in multicellular organisms exhibit signs of ECM dysfunction. In inflammatory conditions, this often manifests as degradation of local ECM architecture and subsequent remodeling to a disease-associated ECM that often amplifies disease phenotypes (19). The Matrixin and Adamalysin proteases play critical roles in these pathological processes, and in multiple instances, they are viable targets for therapeutic intervention (20–23). Consequently, imbalances between metalloproteinases and their inhibitors are often associated with disease etiology, indicative of a breakdown in the homeostatic mechanisms that regulate ECM degradation and remodeling (23, 24). This balance may prove critical, as it was recently described how a molar excess of active MMP9 could cleave and regulate TIMP function, an occurrence that could represent a tipping point for local ECM degradation during development and disease (25). Furthermore, other proteases can cleave TIMP proteins, such as human neutrophil elastase (HNE), which cleaves and inactivates at a site within TIMP1 that is conserved in TIMP2 and substituted in TIMP3/4 (valine to leucine, similar aliphatic hydrophobic amino acids) (26, 27). In addition, a recent report also reveals how ADAMTS7 could cleave TIMP1, modulating atherosclerotic plaque formation (28). Further comprehensive investigation is required to fully map the dynamic relationships between TIMPs and metalloproteinases.

CANCER AND THE TUMOR MICROENVIRONMENT

Excess metalloproteinase (MP) expression and activity have been explicitly linked with cancer for decades. Early findings culminated in the development of a series of broad-spectrum MP inhibitors that failed clinical trials due to a lack of efficacy and severe side-effects (29). Despite these trial failures, inhibition of MPs remains a viable therapeutic approach in oncology. The TIMP family has a complex relationship with cancer, which can be both tumor-suppressive and tumor-promoting. MPs can break down virtually every component of the ECM and their activity is often used by cancer cells to facilitate tumor invasion (23). The activity of different MPs within the tumor microenvironment (TME) affects many aspects of tumor progression. For example, MMP2, 9, and 14 can release the latent form of TGFβ from the ECM, later activating it, resulting in a decreased T-lymphocyte reaction, enhanced expression of several other MMPs, induction of epithelial-to-mesenchymal transition (EMT), and tumor cell migration (30). In this regard, the MP-inhibitory functions of TIMPs are often described to be tumor-suppressive (31). However, it is important to note that not all metalloproteinase activity within the tumor microenvironment is tumor-promoting, and there are various examples of tumor-suppressive metalloproteinase functions that may be lost with excess TIMP activity (19, 32). Notably, MMP8, which works to tighten the adhesions between tumor cells and the ECM, directly suppresses metastasis (32).

Of the TIMP family members, TIMP1 has the most consistent findings regarding its role in tumorigenesis. It is traditionally described as a driver of tumor progression through its distinct MP-independent functions (33). The range of functions TIMP1 exhibits is unique to its two distinct domains. The N-terminal domain is responsible for MP-inhibition, in addition to CD74 and CD82 binding (12, 15). The C-terminal domain of TIMP1 mediates binding to proMMP9 and CD63, the latter of which has been cited as a primary driver of TIMP1’s MP-independent functions and classification as a proinflammatory cytokine (15, 34). Indeed, spatial transcriptomics reveals that TIMP1 expression is strongly associated with regions of pulmonary inflammation, adding credence to this notion (9). A positive correlation between TIMP1 overexpression and disease progression/prognosis has been observed in multiple malignancies, in both solid and hematological tumors (35). Furthermore, the consequences of high TIMP1 levels can be influenced by the glycan status of TIMP1, as aberrant glycosylation of TIMP1 has been shown to decrease its ability to bind to and inhibit MPs (11, 33). Independent of MPs, TIMP1 can stimulate cell growth and displays antiapoptotic functions, contributing to tumor proliferation and migration. Cancer-associated fibroblasts (CAFs) have well-established protumorigenic roles in the TME, and TIMP1 has been reported to mediate fibroblast activation and survival in models of fibrosis through CD63 signaling (36–39). High TIMP1 expression within the tumor-associated stroma has been associated with the accumulation of CAFs within the TME in a manner that likely occurs through CD63 (40). Furthermore, multiple studies have shown that knockdown of TIMP1 results in an inhibition of CAF proliferation in addition to an overall decrease in cancer cell proliferation and metastasis (40–42).

TIMP2 has been the subject of conflicting reports regarding its role in tumor progression. Early studies revealed how TIMP2 could stimulate the proliferation of fibrosarcoma and normal fibroblasts in a dose-dependent manner (43). Later studies using the same fibrosarcoma cell line (HT1080) revealed that TIMP2 could promote an invasive phenotype through interaction with the membrane-type MMP, MMP14 (MT1-MMP) (44). In addition, the TIMP2-MMP14 interaction has been described to protect tumor cells from apoptosis through the Akt pathway (45). Despite these limited reports, numerous studies have described TIMP2 as a tumor suppressor. TIMP2 can indirectly inhibit receptor tyrosine kinase signaling through interaction with integrin α3β1, a mechanism that is mediated through the activity of the tyrosine phosphatase SHP1 (13, 46). This pathway leads to inhibition of growth factor-induced angiogenesis and proliferation in an MP-independent manner. Later studies revealed that TIMP2 could inhibit tumor growth, invasion, and metastasis in various solid cancer models (17, 47–50). In one of these studies that used an orthotopic model of triple-negative breast cancer, the authors found that daily treatment with recombinant TIMP2 decreased vascular permeability, suppressed tumor growth by up to 50%, and strongly inhibited pulmonary metastasis (48). Furthermore, TIMP2 has been shown to display immune regulatory functions, mediated through targeting of myeloid-derived suppressor cells and through inhibition of TGFβ-induced formation of decidual-like natural killer cells, which are classically protumorigenic. In addition, IGF1R is a key regulator of cell proliferation and angiogenesis, and therapeutic targeting of IGF1R has demonstrated promise for the treatment of solid tumors (51, 52). It has been shown that the loop 6 region of TIMP2 can bind to IGF1R in an antagonistic manner to inhibit angiogenesis in endothelial cells (53). Although TIMP2 expression is itself generally quite stable across the spectrum of human cancers, many of its molecular targets and interacting partners are not and show changes that likely result in altered biological activities for TIMP2 in the TME (24).

TIMP3 possesses antiproliferative/angiogenic/metastatic and proapoptotic functions that align with antitumor capabilities (11, 54). This is in line with multiple reports that describe a silencing and downregulation of TIMP3 expression across a range of cancer types (11, 24, 55–57). Overexpression of TIMP3 inhibits tumor invasion and proliferation in vitro and in vivo across many cancer models, including prostate, thyroid, ocular, brain, and colorectal cancers (58–64). There are multiple mechanisms by which TIMP3 imparts these beneficial effects that are independent of MP inhibition. Its activity has been shown to stabilize important death receptor proteins thereby increasing the sensitivity of cancer cells to apoptosis, inhibit angiogenesis through competitive inhibition of VEGF at VEGFR2, and protect the ECM through the inhibition of a range of MMPs, ADAMs, and ADAMTSs (6, 65, 66). TIMP3 can inhibit the widest range of MPs, an attribute that is supported by its C-domain glycosylation site. Mutations at this glycosylation site result in a significant reduction of TIMP3’s MP-binding affinities, suggesting that TIMP3’s unique glycosylation properties regulate its target-binding affinity (11, 67). Interestingly, recent reports describe how active MMP9 can cleave TIMP3 at its C-terminal tail, removing this glycosylation site and potentially producing a truncated TIMP3 with reduced metalloproteinase inhibitory capabilities (25). The effects exerted by TIMP3 on tumor progression may vary between early- and late-stage cancers. In a rare report describing a potential protumor role for TIMP3, it was found that the loss of TIMP3 resulted in delayed tumor initiation and tumor suppression during early-stage mammary tumorigenesis. This was postulated to involve an increase in the shedding of TNF receptor 1, directly as a result of loss of MP inhibition around the mammary epithelium. The same study also highlighted how tumor growth was accelerated in late-stage mammary tumors of TIMP3-deficient mice, highlighting the complex intricacies within the outcomes of TIMP expression at different stages of tumor progression (68).

The mechanisms by which TIMP4 is involved in cancer progression are not well studied, likely due to its highly restricted expression. Within the limited reports available, it has been described that TIMP4 regulates cancer by stabilizing the tumor progenitor cell population. Consequently, upregulation of TIMP4 was found to correspond with rapid tumor growth in nude mouse models (69). Another study found that primary breast tumors with increased levels of TIMP4 were associated with a decreased likelihood of long-term disease-free survival, and an average survival of less than 3 years indicating that TIMP4 may serve as a useful biomarker for aggressive breast carcinomas. The study also showed that patients with early-stage infiltrating ductal carcinoma expressing TIMP4 had on average shorter cancer-free periods, compared with their TIMP4-negative counterparts (70). Like TIMP1, TIMP4 can bind to CD63, although the effects of this interaction remain unstudied (71). The roles of each of the TIMP proteins in cancer are outlined in Fig. 1.

Figure 1.

Summary of the MMP-independent functions of TIMPs 1–3 in the tumor microenvironment. Created with BioRender.com. MMP, matrix metalloproteinase; TIMPs, tissue inhibitors of matrix metalloproteinases.

DEMENTIA AND OTHER NEUROLOGICAL DISORDERS

Metalloproteinases play a multifaceted role in both the developing and adult brains and consequently are involved in the pathophysiology of ischemia, epilepsy, and dementias (72, 73). Dementia is characterized by cognitive decline that is severe enough to interfere with daily functioning. Alzheimer’s disease (AD) is the most common cause of dementia and occurs due to the accumulation of abnormal proteins within the brain (74, 75). Most research on AD focuses on 1) the accumulation of amyloid beta (Aβ) peptides that represent the predominant components of the β-amyloid plaques, 2) hyperphosphorylation of Tau protein, which in excess results in the formation of neurofibrillary tangles (NFTs), and 3) sustained inflammatory responses within AD brains (76, 77). The roles of TIMPs and MPs in AD progression are poorly defined, despite the established roles that MPs play in the processing of amyloid precursor protein (APP) and Aβ. Nonamyloidogenic (nontoxic) processing of APP is mainly performed by ADAM10 and ADAM17, which are inhibited by TIMP1/3 and TIMP3, respectively (78). Amyloidogenic processing of APP through sequential cleavage by β-secretases and γ-secretase produces Aβ, whose levels are tightly controlled due to their toxicity when in high abundance. MMPs such as MMP2/9 are Aβ-degrading enzymes that help to control the levels of Aβ in healthy tissues (79). As such, dysregulation within the TIMP-MP system may play an important role in the early stages of AD and related dementias. In line with all chronic diseases, inflammation (neuroinflammation) is a major component of the pathogenesis of AD for which MPs (and by association, their inhibitors) play a key role (80, 81).

TIMPs can impact AD progression via both MP-dependent and MP-independent mechanisms. TIMP1 displays enriched expression in the cerebral cortex and the thalamus, and its activity is frequently described as protective in models of AD, often through stimulated cell survival and protection of the blood-brain barrier (BBB), a mechanism that likely relies on interaction with CD63 (82–87). Interestingly, TIMP1 has been described as a ligand for APP in monocytes, suggesting a potential role for TIMP1 in the biology and/or pathogenesis of APP in the brain (88). TIMP2 is ubiquitously expressed throughout the brain and, like TIMP1, it has reported beneficial effects on BBB function that could prove beneficial in AD (85, 89). Several reports have described how TIMP2 can regulate neuronal function, while others have described how circulating TIMP2 can support profound beneficial effects on aged brains, the predominant risk factor for AD (90–94). Specifically in AD, a unique population of AD-limiting glial cells referred to as disease-associated microglia are neuroprotective, and are defined by a series of markers that include TIMP2 (95). Whether TIMP2 plays a role in these microglial-dependent neuroprotective functions remains to be investigated. TIMP3 displays low regional specificity in the brain, with the choroid plexus being the sole region with a moderate expression level (85). Consequently, TIMP3’s functions are poorly defined in AD, despite its inhibitory actions against APP-targeting ADAM10 and ADAM17. The sparse reports describing TIMP3 prevalence in AD are limited to a few inconclusive studies, with an individual report describing how TIMP3 supports AD pathology through inhibition of ADAM10/17 (96). On the contrary, TIMP3 levels in cerebrospinal fluid have been reported to negatively correlate with cognitive decline in AD cases (97). Similar across the TIMPs, TIMP3 has reported beneficial effects supporting BBB function, thus TIMP3 may display conflicting effects upon disease etiology (98, 99). TIMP4 exhibits minimal expression within the central nervous system (CNS), with transcript detection limited to the cerebellum (85). This restricted expression has produced few studies investigating potential roles for TIMP4 AD, with individual reports purporting TIMP4 as a serum biomarker in AD (97, 100–102).

Multiple sclerosis (MS) is an autoimmune disorder caused by the immune cell targeting of the myelin sheath surrounding neurons. In MS, the myelin sheath is degraded leading to the formation of scar tissues called sclerosis, causing nerve damage manifesting as neurological disabilities, decreased motor capabilities, muscle abnormalities, vision problems, and fatigue (103). Studies have described how TIMP1 and TIMP2 display anti-inflammatory properties within the CNS, at least partially through the maintenance of the BBB (87, 89, 104–106). Dysregulation within the BBB and immune cell infiltration is described as a hallmark of MS (107). In addition, TIMP1 has been shown to directly regulate CD4+ T-cell migration across the CNS parenchyma to modulate the neuroinflammatory response (108). TIMPs also exhibit a neuroprotective effect by promoting the survival of both oligodendrocytes and neurons through neurotrophic factor signaling, which is critical since axonal degeneration and demyelination are characteristic features of MS (99, 109, 110). Furthermore, research points toward an influential role for TIMP1 in oligodendrocyte differentiation and remyelination, both of which are crucial for repairing damaged nerves (111). It was described that TIMP1 knockout mice exhibit not only an impaired ability to remyelinate their nerves following injury but also delayed developmental myelination of the CNS (109, 111). These results indicate that TIMP1 overexpression might hold potential as a therapeutic factor for remyelination. Further studies revealed that astrocyte-derived TIMP1 plays a central role in the differentiation of oligodendrocytes in vitro, and mice that were TIMP1-deficient could not spontaneously remyelinate their axons (112). This further bolsters the idea that TIMP1 might have the therapeutic ability to potentiate recovery in demyelinating disorders such as MS. Other neurological sclerosing diseases such as amyotrophic lateral sclerosis have poorly defined associations with TIMP activity, possibly due to poor disease models and a reliance on postmortem tissue. The roles that MPs, and by association TIMPs, play in neurodegenerative diseases are reviewed extensively elsewhere (73). In addition, a recent report highlights important mechanistic details describing TIMP2’s protective effects over BBB function in a model of traumatic brain injury (113). Mediated through α3β1 integrin binding, TIMP2 could limit BBB disruption through inhibition of Src-dependent VE-Cadherin internalization in a manner that is MP-independent, reiterating the therapeutic potential of TIMP2 in neurological disorders. The signaling roles for each of these TIMP proteins in neurological disorders are depicted in Fig. 2.

Figure 2.

TIMP proteins in neurological disease. Diverse TIMP signaling and inhibitory pathways underlying roles in Alzheimer’s disease, neuronal health and inflammation, and blood-brain barrier integrity. Created with BioRender.com. APP, amyloid precursor protein; CNS, central nervous system; MMP, matrix metalloproteinase; TIMP, tissue inhibitors of matrix metalloproteinase.

ARTHRITIC DISEASES

Arthritis is a hypernym describing a variety of inflammatory joint diseases, among them rheumatoid arthritis (RA) and osteoarthritis (OA). While both are common conditions, they differ in their etiology. RA is an autoimmune disease that produces chronic inflammation, often affecting many joints (114, 115). OA is a degenerative joint disease caused by erosion of the joint architecture. Both of these joint disorders have been associated with elevated expression and activity of extracellular proteinases that can degrade ECM components of the afflicted joint (116). Within the synovial tissue of RA-afflicted joints, neutrophils represent the predominant cell type, and their activity has been implicated in the pathogenesis of the disease (117, 118). Neutrophils are a major source of TIMP-free MMP8 and MMP9, collagenolytic MPs whose activity is involved in both RA and OA disease progression (116, 119–121). With regards to the roles that MP activity assumes in disease etiology, it is logical to assume a protective role for TIMP proteins in arthropathies. Despite this assumption, there is little concrete evidence describing the direct role of TIMPs in the pathophysiology or resolution of arthritic conditions. Punctate reports describing changes in TIMP expression in the synovial fluid and serum of individuals with arthritis are conflicting or poorly conclusive (122–126). Despite the lack of definitive studies, there are several reports describing the potential therapeutic effect of TIMP proteins present in arthritis. TIMP2-deficient mice exhibit accelerated OA through a mechanism that involves enhanced angiogenesis, a pathway that likely relies on the MP-independent function of TIMP2. TIMP3 knockout mice also display mild cartilage degradation similar to that observed in patients with OA through enhanced aggrecanase activity, which are MPs of the ADAMTS family. The aggrecanase inhibitory activity of TIMP3, which is unique among the TIMP family members, has led to the description of TIMP3 as a chondroprotective factor that may be clinically used in cases of OA (127–129). In a comparable study, TIMP1 was described as therapeutically ineffective in a murine model of collagen-induced arthritis (130). In line with TIMP4’s highly restricted expression, there have been no reports linking TIMP4 with the pathophysiology of arthritis, and only individual instances describing an increase in the expression of TIMP4 in OA (131, 132). Figure 3 summarizes the functional roles of each of the TIMP proteins in arthritic diseases.

Figure 3.

TIMP proteins in rheumatoid arthritis and osteoarthritis. Created with BioRender.com. ECM, extracellular matrix; MMP, matrix metalloproteinase; TIMP, tissue inhibitors of matrix metalloproteinase.

CARDIOVASCULAR DISEASES

Altered expression patterns of MPs and TIMPs, such as decreased expression of TIMP4, are known to occur following myocardial infarction (MI) (133). These manifest as region-specific imbalances of specific MPs and TIMPs throughout the myocardium (134). Such modifications of protein expression take place in stages, with some occurring as late as 16 wk post-MI (135). In a mouse model, broad-spectrum MP inhibition has been found to reduce post-MI left ventricular dilation, indicating a role of MPs and TIMPs in left ventricular remodeling after infarction (136, 137). In a comparable study in pigs, broad-spectrum inhibition of MPs using PD166793 helped to attenuate infarct expansion and also produced an early-phase increase in TIMP1 levels, with chronic treatment (8 wk post-MI) also increasing the region-specific levels of TIMP4 (137). Unlike the other TIMPs, TIMP4 is largely restricted to structures of the cardiovascular system, where it is upregulated in inflammatory conditions such as atherosclerosis and giant cell arteritis (138). The role of TIMP4 in the heart is not thoroughly understood, although it seems to contribute to cardiac contractility and cardiac cell differentiation, becoming epigenetically silenced in heart failure (139, 140). Spatial transcriptomics in the murine heart reveals that TIMP4 expression is most prevalent in the microvasculature of the cardiac parenchyma, an observation that may impart key context regarding the role of TIMP4 in healthy and infarcted hearts (9). In any event, the impact of TIMP4 on cardiovascular pathology is multifaceted. TIMP4 may play a role in the prevention of atrial fibrosis, which appears to involve an interaction between TIMP4 and TGFβ1 (141, 142). Polymorphisms in TIMP4 are associated with changes in angiographic coronary plaque progression (143). Furthermore, TIMP4 knockout mice show abnormally decreased cardiac function with age, as well as increased mortality following MI. This occurs primarily through left ventricle (LV) rupture, although this effect is reduced by the administration of a broad-spectrum MMP inhibitor or genetic deletion of MMP2 (144). These observations illuminate the importance of TIMP4-mediated MP inhibition in maintaining the structural integrity of the heart. Beyond cardiac protection, TIMP4 may also be important for cardiac recovery, having been found to play an essential role in the recovery of normal cardiac function following myocardial ischemia-reperfusion, in large part due to its ability to inhibit MT1-MMP (MMP14) (145). An important discriminating factor between TIMP2 and TIMP4’s affinity for MMP14 is the fact that the TIMP4-MMP14 interaction is insufficient for the activation of MMP2 (146). On the contrary, trimolecular complex formation between MMP14-TIMP2-MMP2 is a well-established mechanism of proMMP2 activation (147, 148).

Though TIMP4 may be the most prominently studied TIMP in the context of cardiovascular disease, the importance of other TIMPs in this area is not to be left unnoticed. For example, increased TIMP1 levels post-MI appear to be important in the regulation of MP-mediated left ventricular remodeling, as TIMP1 deletion induces adverse remodeling that is rescued by broad-spectrum MP inhibition (149). The important yet complex functions of TIMP1 in cardiac health are seen in other findings showing that TIMP1 knockout mice displayed markedly decreased cardiac fibrosis but increased coronary atherosclerosis (150). Correspondingly, overexpression of TIMP1 reduces atherosclerosis, at least in apolipoprotein E-deficient mice (151). Plasma TIMP1 is associated with risk factors and indices of cardiac disease, although no causal connection has been determined, it likely represents a collateral observation that is linked to TIMP1’s long-standing association with inflammation, a definitive risk factor in cardiomyopathy (152, 153).

The contributions of TIMP2 and TIMP3 to cardiovascular disease are equally intriguing. As with TIMP1 and TIMP4, deficiency of TIMP2 increases post-MI disease severity, largely due to decreased inhibition of MMP14 (154). As mentioned, TIMP1 was previously found to protect against atherosclerosis, but other work comparing TIMP1 and TIMP2 suggests TIMP2 is a more effective cardioprotective factor (155). In accord with all the other TIMPs, TIMP3 knockout mice exhibited worsened pathology and survival following MI, but marked improvement was observed when the knockout mice were treated with a broad-spectrum MP inhibitor quickly after MI (156). TIMP3 may also play a protective role against the development and rupture of atherosclerotic plaques (157, 158). A causal association between elevated serum TIMP3 and lower risks of coronary artery disease and myocardial infarction has been found (159). Interestingly, although both TIMP2 and TIMP3 seem to play protective roles in the context of cardiovascular disease, the pathways underlying these roles may be very different. TIMP2-deficient mice infused with angiotensin II displayed aggravated myocardial hypertrophy and decreased cardiac fibrosis, whereas TIMP3-deficient mice receiving the same infusion displayed severe fibrosis without hypertrophy (160). Although not a focus of the present review, it is also worth noting that associations have been reported between TIMPs and other cardiovascular conditions, such as TIMP3 and hypertension and TIMP2 and atrial fibrillation (161, 162).

Many of the aforementioned roles for TIMPs in cardiovascular diseases lean on the MP-inhibitory functions of TIMP proteins within cardiac tissue, yet the MP-independent functions of TIMPs likely distinguish their beneficial effects. This can be exemplified by an intriguing study that overexpressed individual TIMP family members in cardiac fibroblasts, concluding that each TIMP had heterogenous MP-independent effects on cardiac fibroblast proliferation and apoptosis, despite each instance producing a phenotype reminiscent of myofibroblasts (163). The latter point may be crucial in the early stages of MI, where the time-dependent evolution of activated fibroblast phenotypes is crucial for tissue repair (164). Furthermore, TIMPs have well-established regulatory functions on endothelial cells and cardiomyocytes, reviewed extensively elsewhere (165).

Therapeutic applications of TIMPs in cardiovascular disease are promising but remain limited to preclinical testing. The approaches often involve targeted inhibition of elevated MPs that contribute to adverse cardiac remodeling. TIMP3’s unique matrix-binding characteristic makes it a predominant interest regarding therapeutic intervention in MI. This can be exemplified by a study that used hydrogel-embedded recombinant TIMP3 implanted post-MI that was successful in reducing inflammation, MI expansion, and LV dilatation (166). Despite the apparent success, implementation of this approach requires an open chest procedure. Intracoronary delivery of recombinant TIMP3 yielded similarly promising results, and the pharmacokinetics of this treatment may be improved through the engineering of TIMP3 with additional glycosylation or fusion with other proteins (167, 168). Other approaches to increase TIMP3 levels post-MI have included adenovirus-based or cell-based (cell transplantation) gene therapy (169–171). Indeed, TIMP3 adenovirus-coated stents were demonstrated in pigs to significantly decrease in-stent restenosis (172). Each of these TIMP3-based cardiovascular therapeutics is reviewed extensively by Fan and Kassiri (6). Applications of TIMP3 to non-MI cardiovascular pathologies are far more limited, although overexpression of TIMP3 by macrophages was found to reduce atherosclerotic plaque size, inflammation, and oxidative stress signaling in mice (173). Along the same lines, adenovirus-based TIMP3 expression following vein grafting in pigs increases neointimal apoptosis, reducing neointima formation considerably, which holds clinical potential in the context of procedures such as coronary artery bypass (174). Exogenous TIMP3 also holds promise for the treatment of abdominal aortic aneurysms, although experiments with live animal models remain to be performed (175). Importantly, a consideration for any exogenous TIMP3 treatment is the potential for TIMP3 to alter the cellular secretome through its interactions with ADAM10 and LRP1 (176). Notably, some therapeutics commonly administered post-MI, such as angiotensin-converting enzyme (ACE) inhibitors, angiotensin II type I receptor blockers, and β-adrenergic receptor blockers, may reduce MP expression and/or activity, although the specific effects vary and more work is needed to understand the relevant pharmacological mechanisms (177). The TIMP functional roles outlined in this section are summarized in Fig. 4.

Figure 4.

Cardioprotective effects of TIMP proteins through MP-dependent and MP-independent mechanisms. Created with BioRender.com. MP, metalloproteinase; TIMP, tissue inhibitors of matrix metalloproteinase.

NONNEOPLASTIC PULMONARY DISEASES

Much of our present understanding of TIMPs in pulmonary disease focuses on the balance between TIMPs and MPs in conditions such as pulmonary fibrosis and emphysema. ECM degradation and remodeling are hallmarks of both pulmonary fibrosis and emphysema, and each harbors phenotypes consistent with dysregulation in MP expression and activity (178–180). Fibrosis has been associated with an imbalance between TIMP and MP expression that favors the expression of TIMPs and suggests a nondegradative environment (181). However, gene-targeted studies reveal that most MPs promote pulmonary fibrosis through diverse mechanisms (179). These conflictions, in addition to the many causal factors that induce pulmonary diseases, create a disjointed understanding of the roles that TIMPs play in pathogenesis and their potential role in disease resolution. TIMP2 and TIMP3 are strongly expressed throughout all compartments of the lung, with TIMP1 displaying a low, but highly inducible, expression profile (9). In line with the inducible nature of TIMP1, increased TIMP1 expression is associated with increased fibrosis in mouse models of silica exposure (182, 183) One study suggests that TIMP1 regulates the fibrotic response by stimulating fibroblast activation and proliferation through a pathway involving the formation of a cell surface TIMP1/CD63/integrin β1 complex that activates ERK signaling (184). Indeed, in its early stages, pulmonary fibrosis may be suppressed using antisense TIMP1 retroviral vectors (185). Somewhat paradoxically, TIMP3 knockout mice displayed both increased total MP activity in the lungs and increased fibrosis following lung injury, and it was determined that this effect was mediated through persistent inflammation and neutrophil recruitment at the injury site, leading to delayed wound resolution (186). This knockout model could be rescued with an MP inhibitor, suggesting that metalloproteinase activity was the causative factor.

In contrast to the high TIMP/MP ratio that may be found in fibrosis, a high MP/TIMP ratio is thought to cause a degradative environment that can destroy normal tissue structure and aggravate injury, which may be the case in acute lung injury and radiation-induced lung injury (187, 188). MP-favored imbalances may occur following acute lung injury that augments disease severity, as has been suggested previously in the case of ventilator-induced lung injury (189). MP and TIMP expression patterns vary significantly between different benign pulmonary diseases, contributing to the incohesive understanding of TIMP biology in the lungs. The MMP9/TIMP1 imbalance is of special interest not only in pulmonary fibrosis but also in other diseases, such as chronic obstructive pulmonary disease (COPD), bronchiectasis, and asthma. However, the magnitude of the MMP9/TIMP1 ratio may vary between these diseases, and the details are complicated by other TIMPs (190, 191). In asthma, overexpression of not only TIMP1 but also TIMP3 has been reported, along with abnormally high or low expression of various ADAM and ADAMTS proteinases (192). MMP9:TIMP1 balance has been found to differ between fibrosis and other pulmonary conditions like COPD, even when inflammatory profiles are similar; for instance, COPD is associated with increased expression of both TIMP1 and MMP9, along with MMP2 (193, 194). Sputum analysis suggests that the increase in MMP9 activity appears to be disproportionately large relative to the increase in TIMP1 expression (195). Furthermore, the increase in TIMP1 may be macrophage-independent, as alveolar macrophages from patients with COPD release less TIMP1 than those from individuals with normal lung function (196). Complicating analysis is the fact that pulmonary patients may possess comorbidities that impact systemic MP-TIMP balance. Osteoporosis, which is common in patients with COPD and is associated with increased MMP9 activity, is often seen in patients with COPD (197). Interestingly, blood from patients with COPD revealed a similar pattern of increased MMP9 expression with no significant difference in TIMP1 or TIMP2 expression (197). This effect was specifically attributed to the patients’ osteoporosis as opposed to their pulmonary disease (197). Other studies of circulating MMP9, TIMP1, and TIMP2, however, have yielded very different results and interpretations, with one set of findings showing increased expression of all three proteins in the plasma of patients with COPD, perhaps suggesting roles for these circulating proteins in COPD-related airway remodeling (198).

Compared with TIMP1, there are relatively few TIMP2-specific findings regarding benign pulmonary disease, but some fundamental questions have been raised. Within the lungs, TIMP2 has been found to colocalize with Ki67 in fibroblast foci (181), an observation in line with previous reports describing proliferative effects of TIMP2 on cardiac fibroblasts and fibrosarcoma cells (43, 163). TIMP2 expressed by myofibroblasts in idiopathic pulmonary fibrosis is thought to contribute to the stable deposition of ECM that is associated with the condition (199). Single nucleotide polymorphisms in TIMP2 have been associated with the development of different lung conditions, including paraseptal emphysema and additional forms of COPD (200–202). Yet, increased TIMP2 expression is associated with improved pulmonary function in COPD and chronic asthma (203). These contradictions may result due to the varying roles of TIMP2 as both an inhibitor of MPs and a signaling molecule in inflammation-related pathways. Interestingly, TIMP2 expression in the lungs seems to increase with age, as does the expression of TGFβ1 and MMP2 (204). The distinct roles that TIMP2 plays in the lungs are specifically reviewed elsewhere (205).

Shifting focus to TIMP3, again the TIMP-MP balance is often described as contributing to pulmonary pathophysiology. TIMP3 knockout mouse studies have revealed critical anti-inflammatory roles for TIMP3 in the lungs including regulation of MP activity, neutrophil influx, and macrophage differentiation (186, 206). MP-TIMP balance is regulated at least in part by TGFβ signaling, which becomes dysregulated in patients with COPD, and the regulation of TIMP3 by TGFβ appears especially important in the pathogenesis of both COPD and idiopathic pulmonary fibrosis (IPF) (207, 208). Indeed, TIMP3 knockout mice are known to exhibit emphysema-like enlargement of the airway with decreased tissue elasticity, and in humans, mutation of TIMP3 may result in the autosomal dominant condition Sorsby fundus dystrophy, which is associated with pulmonary in addition to ocular disease (209, 210). Although TIMP3 expression appears to have a protective effect against pulmonary fibrosis, an excess of TIMP3 may be problematic, with proteomic analysis of human lung tissue indicating increased TIMP3 expression in COPD (211, 212).

Pulmonary hypertension (PH) is another disease of interest regarding TIMP biology. PH induced by hypoxia or monocrotaline is associated with increased abundance of MPs and TIMPs (213, 214). TIMP4 levels may contribute to PH, including in cases of systemic sclerosis (215). This is notable considering the relatively limited expression patterns of TIMP4 in the lungs (9). In a surgical mouse model of pulmonary hypertension, folic acid supplementation was found to raise right ventricular TIMP4 abundance that had been decreased after the surgery and to decrease TIMP1 levels that had been increased after the surgery, improving MP/TIMP balance (216). TIMP4 may also play roles in other pulmonary conditions, notably COPD. TIMP4, along with TIMP1, is increased in COPD following exposure to cigarette smoke (217, 218). Similarly, patients with COPD with mild-to-moderate acute exacerbation displayed increased pulmonary levels of TIMP4 compared with healthy subjects (219).

Several TIMP-related therapeutic approaches have been studied for the treatment of pulmonary disease, though clinical trials with such biologics remain to be completed. By and large, these approaches involve a restoration of normal pulmonary MP/TIMP balance to prevent and possibly reverse pathological airway remodeling. For example, doxycycline administration to inhibit MP activity and thus restore the MMP9/TIMP1 ratio represents a promising treatment modality for COPD (220). Similarly, treating emphysematous mice using glycyl-l-histidyl-l-lysine peptide was found to attenuate the emphysema and partially restore normal MMP9/TIMP1 balance in the lungs (221). This action appears to involve multiple mechanisms, including NF-κB and oxidative stress pathways (221). Mice exposed to cigarette smoke and then treated with rosiglitazone, a drug prescribed for type II diabetes mellitus, also showed reduced MP activity, possibly due to MAPK- or NF-κB-dependent signaling pathways, which helped restore protease/inhibitor balance and ameliorated the development of emphysema-related pathology (222). That work also uncovered a potential role for the peroxisome proliferator-activated receptor γ (PPARγ) in emphysema, with the rosiglitazone treatment raising PPARγ expression levels (222). A subsequent study supports a protective role for PPARγ in maintaining proper MP/TIMP balance, particularly MMP9/TIMP1 balance, in COPD, putting PPARγ forward as a potential therapeutic target in COPD intervention (223). Other promising therapeutics that appear to similarly quell MP activity and reduce MP/TIMP imbalance include the antibiotic erythromycin, thegallic acid (a naturally occurring phenol), and recombinant human keratinocyte growth factor (rHuKGF), also known as palifermin (224–226). An additional interesting player in this therapeutic arena is SIRT1, a protein deacetylase. Reduced SIRT1 is associated with MMP9/TIMP1 imbalance in the lungs of smokers and patients with COPD, which appears to relate to increased TIMP1 acetylation at specific lysine residues and reduced interaction with MMP9 (227, 228). In this regard, it has been suggested that SIRT1 is involved in the expression and deacetylation of TIMP1, facilitating TIMP1-mediated MP inhibition in emphysema and COPD. In any case, redressing the deficiencies in SIRT1 activity in smokers through pharmacological activation of SIRT1 has the potential to restore a normal TIMP1 expression and acetylation patterns, and in turn, reestablish the homeostatic protease/inhibitor balance. The complex roles of TIMP proteins in pulmonary disease are summarized in Fig. 5.

Figure 5.

Complex roles of TIMP proteins in pulmonary disease. TIMP family members are frequently characterized by conflicting functional and associative roles in pulmonary diseases. Created with BioRender.com. MP, metalloproteinase; TIMP, tissue inhibitors of matrix metalloproteinase.

TIMPs AS DISEASE BIOMARKERS

From a clinical standpoint, TIMPs represent an impractical therapeutic due to their short blood half-life of less than 4 h (229), although efforts are described to address this (168, 230). In fact, in considering TIMP biomarker applications, we need not limit the discussion to the diseases discussed above. For example, the NephroCheck test score, which considers urinary TIMP2 and IGFBP7 abundance, is used in the early detection of acute kidney injury (231, 232). Furthermore, TIMPs have shown promise as diagnostic and prognostic biomarkers in diverse cancers, including nonsmall cell lung cancer, prostate cancer, pancreatic cancer, breast cancer, and ovarian cancer (233–237). Serum TIMP1 was found to have upper moderate sensitivity and specificity as a potential diagnostic value for colorectal cancer (238). Though more research is needed, decreases in plasma and cerebrospinal fluid TIMP3 levels may be indicative of Alzheimer’s disease progression (97, 238), and it has been proposed that the measurement of TIMPs and MPs in the cerebrospinal fluid of patients with multiple sclerosis may hold promise for diagnosis and treatment decisions (239).

The relevance of the matrisome to cardiovascular disease is well appreciated. Higher circulating MP levels, especially MMP2 and MMP9, are known to correlate with disease states including both acute coronary disease and post-MI congestive heart failure (240, 241). Also, TIMPs, most notably TIMP1, have been found to possess a significant correlation with vascular events. Associations between increased circulating TIMP1 and MI, heart failure severity, stroke, and cardiovascular death have been reported (242–245). Similarly, in pulmonary hypertension, circulating TIMP1 is elevated, and high TIMP1 appears to reflect more severe disease and predict poorer outcomes (246, 247). In addition to TIMP1, TIMP4 is a very strong biomarker for pulmonary hypertension, with high plasma TIMP4 associated with increased pulmonary arterial pressure, right ventricular hypertrophy severity, and poor survival (248–250). In idiopathic pulmonary hypertension, an increased MMP2/TIMP4 ratio is associated with clinical worsening events and death (251). Measurements of circulating plasma MPs and TIMPs may hold promise for patients with pulmonary hypertension in association with standard hemodynamic indices and various values indicative of pulmonary arterial stiffness (252). Some interest has even been shown in the potential for MPs and TIMPs to be applied as biomarkers for Takayasu’s arteritis, a relatively rare vascular disease (252, 253).

In pulmonary disease, just as the relevant pathways vary within and between conditions, so too do the biomarkers. This translates to a lack of concordance in the utility of TIMPs as biomarkers for pulmonary diseases that often have confounding variables, such as smoking status and pollution exposure. Linked to its emerging role as an inflammatory cytokine, TIMP1 presents with the most consistency as a biomarker in pulmonary diseases such as COPD, as well as respiratory infections involving COVID-19 and respiratory syncytial virus (15, 254, 255). Furthermore, it has been described that plasma MP and TIMP levels have potential use as biomarkers for treatment failure, relapse, and death in patients with pulmonary tuberculosis that may eventually guide clinicians’ judgment regarding treatment regimens (257).

FUTURE PERSPECTIVES

The contribution of TIMPs to the pathophysiology of human disease is becoming increasingly apparent. In large part, these roles are described through their inhibitory action against MPs, whose expression and activity are explicitly intertwined with pathogenesis. However, the everincreasing characterization of MP-independent functions in TIMP proteins across tissues and the spectrum of human disease models adds credibility to their importance in tissue homeostasis and response to pathology. Many investigations probe the ratios between certain MPs and TIMPs, but increasing evidence suggests partial interrogation of an inherently complex system is insufficient, and greater context and understanding will be derived from a holistic approach. This concept is a difficult challenge to engage, considering the complex mechanisms that mediate the levels and activity of MPs and TIMPs. TIMPs and MPs are regulated at multiple levels, through gene expression, microRNA regulation, alternative splicing, zymogen activation, degradation, and posttranslational modifications. The latter point is an important factor, considering that TIMP1 and 3 are glycosylated and TIMP2 can be phosphorylated (3, 258). Furthermore, it is also important to consider the consequences of TIMP binding on the proteinase-independent functions of MPs that have not been fully addressed, whereby TIMP binding can facilitate MP scavenging by the endocytic receptor LRP1 (259, 260). Further work is required to mechanistically map the disease- and tissue-specific functions of each TIMP family member, the dependencies of these functions, and the consequences thereafter. Only then can researchers begin to unlock the therapeutic potential hidden beneath the tissue inhibitors of metalloproteinases appellation.

GRANTS

This research was supported by the Center for Cancer Research, NCI/NIH research grants ZIA BC011204 and ZIA SC 009179 to WGSS (NCI Intramural Research program).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

S.C-P. and D.P. conceived and designed research; S.C-P., J.R., and D.P. prepared figures; S.C-P., J.R., and D.P. drafted manuscript; S.C-P., J.R., and D.P. edited and revised manuscript; S.C-P., J.R., W.G.S-S., and D.P. approved final version of manuscript.