The Fibrinolytic System
The human fibrinolytic system plays a pivotal role in hemostasis and thrombosis related
to multiple organ systems. At the heart of the proteolytic cascade lie serine proteases—tissue-type
plasminogen activator (tPA) and urokinase-type plasminogen activator (uPA)—responsible
for catalyzing the conversion of the zymogen, plasminogen, into the active form, plasmin,
through the cleavage of a serine–arginine to valine peptide bond.[1 ] Plasmin, a key player in the fibrinolytic system, directly catalyzes the degradation
of the fibrin protein meshwork that results in the dissolution of thrombi. While tPA
primarily facilitates intravascular fibrinolysis through localization of plasminogen
and fibrin deposits on the endothelium, uPA is observed predominantly in the pericellular
environment, regulating localized proteolysis critical for cell invasion, wound healing,
and tumor metastasis.[2 ]
[3 ] Plasminogen activator inhibitor type 1 (PAI-1), a potent serine protease inhibitor
(serpine), plays a significant regulatory role for uPA/tPA to establish an equilibrium
between plasminogen and plasmin as well as maintaining a stable pericellular environment
for optimal cellular function. Thus, dysregulation of PAI-1 expression leads to multiple
downstream pathologies that contribute to coagulopathy and other systemic disease
states such as cardiovascular disease, metabolic syndrome, and cancer.[4 ] Numerous other disease processes include inflammation and infection, diabetes, and
neurodegeneration, with PAI-1 associated with aging, skeletal muscle repair, and cellular
senescence. PAI-1's role in the fibrinolytic system has clinical relevance, with the
excess or deficiency that results in bleeding disorders. Specifically, hyperfibrinolytic
disorders can present as bleeding of unknown origin but are largely reported in patients
with a PAI-1 deficiency, suggesting a greater need for specific diagnostic testing
and timeliness for treatment.[5 ]
To better recognize the role of PAI-1 as a regulatory checkpoint, it is critical to
understand the mechanism of its substrate. uPA is synthesized as an inactive, single-chain
zymogen (pro-uPA) composed of 411 amino acids distributed across distinct functional
domains: A growth factor domain (GFD; residues 1–49), a Kringle domain (residues 50–131),
an interdomain connecting peptide (residues 132–158), and a serine protease domain
(residues 150–411).[6 ]
[7 ] Activation occurs through a single cleavage event at Lys158–Ile159, resulting in
a two-chain, active enzyme stabilized by a disulfide bond between Cys148 and Cys279.[8 ]
[9 ] Subsequent proteolysis can further process uPA, separating the amino-terminal fragment
(residues 1–135)—responsible for receptor binding and signaling—from the enzymatic
serine protease domain.[8 ]
[9 ]
The effects of uPA are spatially restricted and amplified through its interactions
with its high-affinity receptor, urokinase plasminogen activator receptor (uPAR),
or CD87.[10 ] Belonging to the Ly-6/uPAR superfamily, mature uPAR consists of 283 residues derived
from a 335-residue precursor after subsequent signal peptide removal and the addition
of a glycosylphosphatidylinositol moiety.[10 ]
[11 ] Structurally, uPAR is organized into three homologous Ly-6/uPAR/α-neurotoxin-like
domains—D1, D2, and D3—which form a central hydrophobic cavity that interacts with
the GFD of uPA, inducing conformational changes at the D1–D2 interface that primes
the receptor's affinity for secondary ligands, notably vitronectin (VN), a highly
abundant protein found in the subendothelial matrix.[11 ]
[12 ]
[13 ]
[14 ] Crystal structural studies reveal that while uPA occupies the hydrophobic core of
uPAR, VN associates externally through its somatomedin domain (SBD; residues 1–44),
enabling simultaneous ligand engagement.[12 ] Binding of uPA shifts uPAR into a closed, high-affinity conformation for VN and
initiates intracellular signaling cascades—such as focal adhesion kinase (FAK), the
Janus (JAK) and Src family kinases, and the phosphatidylinositol 3-kinase/protein
kinase-B (PI3K/Akt) axis—that ultimately culminates in enhanced cytoskeletal remodeling,
increased motility, or greater survival signaling.[11 ]
[15 ]
Given the central role of uPA/uPAR interactions in coordinating cellular processes
implicated in the pathogenesis of disease progression, the need for precise regulation
is critical, and thus, has garnered the attention of many researchers and clinicians
as a potential target to inhibit. Consequently, much of the research has been focused
on PAI-1, a key endogenous regulator of uPA that is a part of the serine protease
inhibitor superfamily known as serpins. Interestingly, the most peculiar is the bevy
of recent literature that has implicated PAI-1 in exerting both promigratory and antimigratory
effects independent of its conventionally associated functions within proteolysis.
This paradox highlights the complex interplay among these tightly regulated interactions,
underscoring perhaps a delicate balance within the cellular microenvironment and the
physiological or pathological processes involved. This review aims to summarize our
current knowledge on the diverse cellular functions of PAI-1 that have broad implications
for emerging roles in cancer biology and therapeutics ([Fig. 1 ]).
Fig. 1 Overview of the cell-based model of thrombosis and central role for PAI-1 in regulation
of fibrinolysis. PAI-1, plasminogen activator inhibitor type 1.
Plasminogen Activator Inhibitor Type 1
Structure, Stability, and Regulation
Glycoprotein PAI-1 is a 45-kDa, 379- to 381-amino acid protein expressed within several
tissue beds with a primary role in regulating the fibrinolytic pathway. Being in the
serpin superfamily, PAI-1 exhibits a highly conserved structure seen in other serine
proteases. Specifically, PAI-1 contains the characteristic three β-sheets (A, B, and
C) as well as nine α-helices (hA–hI) that are typical of this family of proteases.
Critical for its function, a reactive center loop (RCL) composed of 26 residues (331-SGTVASSSTAVIVSARMAPEEIIMDR-356)
is critical to its function as a plasminogen inhibitor ([Fig. 2 ]).[16 ]
[17 ]
[18 ] This sequence is particularly important at Arg346 and Met347 (P1-P1′) due to its
role as a substrate mimic toward tPA and uPA. The remainder of the RCL sequence is
thus referred to as P16-P10′ regarding its position around these key residues. Importantly,
the P16-P10′ region confers specificity to the individual proteins within the serpine
family.
Fig. 2 Protein structure of plasminogen activator inhibitor type 1 with reactive center
loop highlighted and labeled residues. Molecular graphics and analyses performed with
University of California, San Francisco (UCSF) ChimeraX, developed by the Resource
for Biocomputing, Visualization, and Informatics at the University of California,
San Francisco, with support from the National Institutes of Health R01-GM129325 and
the Office of Cyber Infrastructure and Computational Biology, National Institute of
Allergy and Infectious Diseases.[18 ]
The remainder of the amino acids outside this region are highly conserved among this
superfamily of proteins, and mutagenic studies have confirmed that alteration of the
RCL changes its function. Specifically, studies indicate that mutations in the P1-P1′
region can alter the structure to act as antithrombin III depending on the alterations
made to the base sequence.[19 ]
[20 ]
[21 ] As a target-specific region, PAI-1 selectively inhibits plasminogen activators (PAs)
to perform its role as an antifibrinolytic agent.
PAI-1 is highly expressed in many tissues, such as the liver, respiratory system,
kidneys, placenta, vascular smooth muscle cells, endothelial cells, and adipose tissue.
Once secreted into the circulation, it exists in an active conformation for only a
few minutes (plasma t
1/2 6 minutes) before being rapidly cleared by the liver.[22 ] Structurally, PAI-1 exists in two forms, active and latent, which modulate its activity
toward PA in a conformationally dependent manner. Once PAI-1 is secreted by endothelial
cells, it exists in two main reservoirs intravascularly—free plasma, and retained
in circulating platelets. The free plasma concentration is predominantly in an active
conformation but only accounts for as much as 5 to 50 ng/mL in serum.[23 ] Conversely, platelet retained PAI-1 accounts for much higher levels up to 300 ng/mL
but exhibits mostly latent conformation.[23 ] Interestingly, research has indicated that platelet activation induced by endothelial
damage shows a significant increase in the active form of PAI-1. This presents a benefit
to reinforcing clot formation at the onset of damage to ensure proper hemodynamic
stability during times of injury.
PAI-1 exerts its effects in a tissue-specific manner due to glycosylation at Asn209
and Asn265 and its interaction with local mediators like PA.[23 ] Inhibitory members of the serpin family, such as PAI-1, interact through the formation
of a covalent 1:1 stoichiometric complex with their target proteinases.[24 ] Formation of an ester bond between the carboxyl group of the P1 residue of the serpin
and the hydroxyl group of the active serine residue of the target protease inhibits
the activity of the target protease.[24 ] The crystal structure of latent PAI-1 was originally elucidated by Mottonen et al
in 1992, followed by determination of the crystal structure of active PAI-1 by Nar
et al.[25 ]
[26 ] Structure–function experimentation has revealed that the major, biologically relevant
difference between latent and active forms of PAI-1 is that active PAI-1 has an exposed
bait region (P1-P1′; [Fig. 2 ]) known as an RCL (R346-M347), as well as accessible secondary binding sites that
enable inhibitory complex formation with a target protease. In latent PAI-1, the bait
region and secondary structures are inaccessible and cannot bind to the active site
of serine proteases.[23 ] The active conformation of PAI-1 in human tissue is maintained by complexing with
VN, a highly abundant protein found in the subendothelial matrix. It is likely that
VN-bound PAI-1 ([Fig. 3 ]) is the physiologically relevant mechanism for maintaining the inhibitory function
of PAI-1 in human tissue. Importantly, the structural characteristics of PAI-1 present
key interactions that can be leveraged during the development of effective therapeutics
discussed later in this review.
Fig. 3 Protein structure of PAI-1 and somatomedin B domain of vitronectin with resultant
depiction of protein–protein binding of PAI-1/vitronectin complex. Coloring of the
protein surface indicates polarity of acidic (red), alkaline (blue), and uncharged
(white) side chains. Vitronectin is composed of relatively acidic R groups that interact
strongly with PAI-1's alkaline-binding pocket.[18 ] PAI-1, plasminogen activator inhibitor type 1.
Recognizing the importance of the RCL to PAI-1's function, it is critical to elucidate
the mechanisms surrounding this inhibitory process of PAs. Specifically, two major
pathways exist once RCL interacts with the PA active site: The irreversible inhibitory
pathway and the substrate pathway. Prior to entering a specific sequence, the RCL
chain first binds to PA, forming a weak, non-covalent Michaelis complex.[17 ]
[23 ]
[27 ]
[28 ] This step is followed by attack of the PA active site serine (tPA-Ser478 or uPA195)
onto the RCL to produce an unstable tetrahedral intermediate with PAI-1.[17 ]
[29 ] The tetrahedral intermediate undergoes subsequent collapse that forms a more stable
acyl–enzyme intermediate with resultant covalently bound PAI-1/PA complex.[30 ]
[31 ] From the acyl–enzyme complex, PAI-1 activity induces either the irreversible inhibitory
or substrate pathway. Importantly, the inhibitory sequence accounts for most PAI-1
interactions and follows a complex rearrangement of three-dimensional structure. Crystal
structure analysis indicates that the high-energy state of acyl-intermediate formation
encourages protein rearrangement to a lower-energy conformation that results in shifting
of the P16-P1 portion (N-terminal side) into the core β-sheet of PAI-1.[29 ] The thermodynamically driven rearrangement ultimately fuels the compression of PA
against the core of PAI-1, leading to effective blocking of its active site.[32 ] Resultant deformation of PA thus prevents hydrolysis of the PAI-1/PA complex and
yields an irreversible inhibition of function.[30 ]
[31 ] Conversely, reports have identified an alternative substrate pathway that effectively
regenerates PA after the formation of the acyl–enzyme intermediate. It is proposed
that hydrolysis occurs prior to PA deformation in the core of PAI-1, which releases
PA before the P16-P1 region and can undergo conformational change.[17 ]
[23 ]
[33 ]
[34 ] It is thought that a distinct subset of PAI-1 in this alternative substrate pathway
exists or can be induced through local intermediaries or kinetic properties, which
could precipitate more active PA.[35 ]
[36 ]
[37 ]
[38 ]
Interestingly, PAI-1 regulation centers on deactivating the RCL through conformational
change from the active-to-latent transition. Studies have shown that the in vitro
half-life of the active conformer of PAI-1 is around 2 hours and occurs through a
slow, self-insertion of the N-terminal region of the RCL into the core protein.[17 ]
[26 ] This insertion functionally blocks the P1-P1′ region from interacting with PAs such
as tPA and uPA. It is hypothesized that this form of autoregulation presents an ancestral
benefit in reducing the risk of thrombosis due to otherwise extended antifibrinolytic
activity of PAI-1.[16 ]
[17 ] Additionally, in vivo analysis of PAI-1 stability has been extensively studied and
relies upon interaction with the ubiquitous plasma glycoprotein, VN. VN, as discussed
throughout this review, serves as a cofactor for PAI-1 and stabilizes its active form,
approximately twofold, through stabilization of the lower half of its protein structure
without affecting enzymatic activity to regulate PA.[39 ]
[40 ]
[41 ]
[42 ] This stabilization ultimately slows the conversion to latent-type PAI-1 and serves
as an important barrier to keep in mind during the development of therapeutic targets.
Genetic Architecture
PAI-1 is encoded by the SERPINE1 gene, located on chromosome 7q21.3-q22 (NCBI accession number M16006). This gene
spans approximately 12.3 kb and consists of nine exons and eight introns, as shown
([Fig. 4 ]).[22 ]
[43 ]
[44 ]
[45 ]
[46 ]
[47 ]
[48 ]
[49 ]
[50 ]
[51 ] Two distinct mRNA transcripts, 2.3 and 3.2 kb in length, are produced from this
gene due to alternative polyadenylation, with the longer transcript potentially exhibiting
greater stability due to an AT-rich sequence.[52 ]
Fig. 4 Schematic representation of the SERPINE1 gene structure. The SERPINE1 gene, encoding plasminogen activator inhibitor type 1, is located on chromosome 7q21.3-q22
and spans approximately 12.3 kb. The gene consists of nine exons (blue boxes) and
eight introns (connecting lines). The 4G/5G polymorphism is shown at −675 bases upstream
of the transcriptional start site.
The 5′ promoter region of the PAI-1 gene is a complex landscape of regulatory elements with significant implications
in cardiovascular disease ([Fig. 5 ]).[43 ]
[44 ]
[52 ] In the presence of wound healing, p38 mitogen-activated protein kinase (MAPK) is
rapidly activated in the MAP kinase pathways in cells adjacent to the injury, leading
to phosphorylation of the transcription factor upstream stimulatory factor-1 (USF-1),
a member of the helix–loop–helix family. This phosphorylated USF-1 complex then translocates
to the nucleus, where it binds to the E-box motif (CACGTG sequence located at nucleotides
−160 to −165) in the proximal promoter region of the PAI-1 gene; this binding is essential for upregulation of PAI-1 transcription. Notably,
activator protein-1 (AP-1)-binding sites do not play a significant role in the PAI-1
regulation model, with the MAPK–USF-1 axis the primary pathway linking mechanical
injury to PAI-1 gene activation.[53 ] This pathway is not only active in wound healing, but is a critical player in vascular
injury response and cardiovascular disease.[47 ]
[53 ] During vascular injury due to shearing, inflammation, and diabetes, the MAPK–USF-1
pathway is activated as described above, promoting the upregulation of PAI-1 transcription.
Additionally, the presence of a glucocorticoid response element site enables upregulation
of PAI-1 expression by glucocorticoids like dexamethasone and mineralocorticoids like
aldosterone.[43 ]
[54 ]
[55 ] This is particularly relevant in hypertension, where angiotensin II stimulates PAI-1
expression and release, which can be further amplified by glucocorticoids and aldosterone.[43 ]
[54 ]
[56 ] A VLDL response element situated within the PAI-1 promoter region at positions −672
to −657 facilitates interaction with very low density lipoprotein (VLDL), a lipoprotein
often elevated in diabetes and metabolic syndrome, suggesting potential clinical implications
for these conditions.[43 ]
[57 ]
[58 ] Sp1 sites in the promoter region are positively regulated by glucose and glucosamine,
linking elevated glucose levels to increased PAI-1 levels, and potentially contributing
to the development of diabetes.[43 ]
[59 ]
[60 ] Additionally, the presence of Smad-binding element and p53-responsive element in
the PAI-1 promoter suggests potential involvement in various cancers.[43 ]
[61 ]
[62 ]
Fig. 5 Schematic representation of the 5′ promoter region of the PAI-1 gene and its regulatory elements. The PAI-1 promoter contains multiple transcriptional
regulatory elements influencing gene expression. The −675 4G/5G polymorphism (yellow)
is associated with higher expression. A VLDL response element (−672 to −657, black)
links PAI-1 expression to lipid metabolism. Smad-binding elements (red) at −730, −580,
and −280 indicate TGF-β/Smad pathway regulation. A p53-responsive element (−160 to
−139, green) suggests a role in cellular stress responses and tumorigenesis. Sp1-binding
sites (−118, −73, −42, light blue) mediate glucose-induced upregulation. A glucocorticoid
response element (GRE; −73, dark blue) suggests modulation by glucocorticoids. The
TATA box (−28, gray) is essential for transcription initiation. PAI-1, plasminogen
activator inhibitor type 1.
Another important aspect of the PAI-1 promoter is the −675 4G/5G polymorphism. This
insertion/deletion polymorphism affects PAI-1 levels, with the 4G allele associated
with higher levels than the 5G allele.[43 ]
[63 ]
[64 ]
[65 ] Although the exact clinical consequences of this polymorphism are under debate,
its association with various diseases, including stroke and venous thromboembolism,
emphasizes the significance of PAI-1 in a wide range of clinical conditions.[43 ]
[66 ] Collectively, these conserved regulatory elements provide insights into the mechanisms
that modulate PAI-1 levels and function in vivo and influence various clinical conditions
ranging from cardiovascular diseases and metabolic disorders to cancer. Understanding
these regulatory mechanisms enhances our grasp of PAI-1's role as both a biomarker
and a potential therapeutic target across different disease states characterized by
thrombotic risk and altered fibrinolysis.
Intracellular and Extracellular Functions
It is extensively characterized that the uPA/uPAR system plays a central role in tightly
regulating extracellular matrix (ECM) remodeling essential for supporting cell migration.
Disruption of this balance can have pathological consequences.[11 ]
[67 ]
[68 ]
[69 ] Excessive ECM degradation leads to compromised structural integrity and enhanced
cell invasion, while insufficient matrix turnover results in abnormal ECM accumulation
and impaired cellular dynamics.[9 ]
[68 ] Early studies substantiate that PAI-1 critically modulates ECM dynamics by interacting
with multiple extracellular components, including PAs, VN, and low-density lipoprotein
receptor-related protein 1 (LRP1).[70 ]
[71 ]
[72 ]
[73 ] Through these interactions, PAI-1 has been shown to both inhibit and promote cell
migration depending on the contextual molecular environment.[68 ]
[71 ]
[74 ]
Regulation of Plasmin Generation
One of the major functions of PAI-1 is its ability to modulate coagulation through
its interaction with tPA and uPA. By blocking tPA/uPA agonist activity on the conversion
of plasminogen to plasmin, PAI-1 effectively inhibits fibrinolysis through a deficiency
of plasmin levels.[75 ] In the absence of PAI-1, uPA bound to uPAR and tPA/fibrin activates plasminogen
through cleavage of Arg561–Val562 residues of each proenzyme.[75 ] This cleavage results in the generation of plasmin, which binds to fibrin to produce
fibrin degradation products. Mechanistically, PAI-1's binding to uPA/tPA (as described
above) leads to the collapse of the tPA and uPA active site for plasminogen. Therefore,
plasminogen is unable to effectively bind and undergo cleavage into plasmin products.
Interestingly, research on patients with homozygous PAI-1 deficiency exhibited increased
peak plasmin levels and resulted in hyperfibrinolytic bleeding.[76 ] Due to its effective inhibition of tPA/uPA as well as its rapid onset of action
when released, PAI-1 remains a critical component of the balance between coagulation
and fibrinolysis.
Plasminogen Activator Inhibitor Type 1 and Vitronectin: A Central Player in Cell Migration
through Urokinase Plasminogen Activator Receptor
VN is established to play a pivotal role in this process. The somatomedin B domain
of VN binds to PAI-1, stabilizing its active conformation and enhancing its inhibitory
effect on uPA-dependent proteolysis, thereby reducing ECM degradation and limiting
the cellular ability to migrate and invade.[77 ]
[78 ] Furthermore, PAI-1 exhibits a stronger binding affinity to the SMB domain of VN
compared with uPAR.[71 ] During binding of the somatomedin B domain of VN, PAI-1 structurally distorts the
arginine-glycine-aspartate (RGD) sequence of VN, the primary binding site for αvβ3
integrins, affecting the formation and turnover of focal adhesions, essential for
cell migration.[24 ]
[69 ] This competitive binding is a key mechanism through which PAI-1 indirectly modulates
integrin function without directly binding to the integrins themselves.[72 ] As a result, elevated levels of PAI-1 can lead to detachment of cells from VN-rich
matrices by blocking integrin-mediated adhesion, promoting cell motility or dysmotility
in certain contexts.[72 ]
PAI-1 also contributes to cell signaling through its effects on uPAR–integrin associations.
The binding of uPA to uPAR promotes the assembly of uPAR–integrin complexes, which
in turn facilitates the activation of FAK and Src family kinases through caveolin-mediated
clustering at the cell surface.[9 ]
[15 ] These complexes are necessary for initiating adhesion-dependent signal transduction.[9 ]
[15 ] Disruption of uPAR–integrin interactions by PAI-1, through VN competition, impairs
this signaling cascade.[79 ]
[80 ] Notably, excessive uPAR activity has also been shown to drive extracellular-signal
related kinase (ERK) activation and exit from tumor dormancy by promoting α5β1-dependent
FAK/Src signaling.[81 ] Thus, PAI-1-mediated interference with uPAR–integrin clustering may serve as a regulatory
checkpoint in these pathways.
Furthermore, PAI-1 indirectly influences G-protein-coupled receptor signaling via
its control over uPAR cleavage. When uPA binds to uPAR or when the D1 domain is proteolytically
removed, a chemotactic epitope (Ser88–Tyr92) becomes exposed, allowing the receptor
to engage G-protein coupled receptors (GPCRs) such as formyl peptide receptor-like
1.[82 ]
[83 ]
[84 ]
[85 ]
[86 ]
[87 ] These interactions promote cytoskeletal rearrangement and cell movement in a pertussis
toxin-sensitive manner.[88 ] By modulating uPAR availability and surface retention, PAI-1 can indirectly regulate
the activation of these chemotactic pathways.
Experimental studies have further confirmed the relevance of PAI-1–integrin dynamics
to cell movement. Treatment of cells with recombinant PAI-1 mimics the effects of
αvβ5 integrin blockade, resulting in cell detachment from VN and enhanced migration
toward other ECM proteins such as fibronectin or collagen type IV.[72 ]
[79 ]
[89 ] These findings highlight PAI-1's ability to remodel adhesion patterns by interfering
with integrin–ECM engagement, independent of its classical role as a serine protease
inhibitor. Further investigations by Sathyamoorthy et al[90 ] employed RNA interference to inhibit PAI-1 expression in HL-60 human monocytes.
Transfection of small interfering RNAs targeting PAI-1 mRNA achieved an 85% reduction
in PAI-1 mRNA and complete inhibition of glycosylated PAI-1 protein expression after
72 hours.[90 ] This posttranscriptional silencing revealed a significant increase in monocyte adhesion
attributable to enhanced interactions between VN and its receptors. Further investigations
by Sathyamoorthy et al[91 ] demonstrated the role of PAI-1 in endothelial progenitor cell (EPC) regulation.
Using bone marrow-derived EPCs from wild-type and PAI-1-deficient mice, they found
that PAI-1 deficiency led to a 75% increase in early outgrowth EPCs and enhanced proliferative
capacity.[91 ] Similarly, in a model of chronic vascular injury using N()- nitro-L-arginine methyl
ester (L-NAME), PAI-1-deficient mice showed a 65% increase in circulating EPCs in
comparison to c57BL6 wild-type controls.[91 ] These results suggest that PAI-1 deficiency can enhance EPC proliferation and mobilization,
potentially offering protective effects against vascular injury.
Conversely, PAI-1 has also been reported to promote cell migration by protecting the
ECM against excessive plasmin-mediated degradation, providing a provisional matrix
that tumor and endothelial cells use as a scaffold for migration or capillary formation.[68 ]
[92 ] In fact, a recent study by McCann et al demonstrated that a decrease in TGF-β signaling
caused a decrease in miR-30 expression, which subsequently resulted in the upregulation
of PAI-1, decreased uPA activity, and fibrin-mediated angiogenesis.[93 ] Restoration experiments in PAI-1-deficient mice confirmed that reintroducing recombinant
PAI-1—particularly functional at the VN-binding site—could restore tumor angiogenesis,
further highlighting the dual requirement for matrix preservation and controlled proteolysis
during new vessel formation.[94 ]
[95 ] Collectively, these studies underscore the context-dependent nature of PAI-1 in
cell migration, suggesting that the relative stoichiometry of PAI-1, VN, and uPA,
as well as the other factors in the cell microenvironment, significantly influences
these processes.
Plasminogen Activator Inhibitor Type 1 and Low-Density Lipoprotein Receptor-Related
Protein 1-Mediated Endocytosis and Integrin Regulation
Beyond its interaction with VN, PAI-1 can influence cell migration through its interaction
with LRP1, a member of the LDL receptor family.[96 ] The low-density lipoprotein receptor-related protein family is involved in numerous
physiological processes such as regulating cholesterol homeostasis and lipid transport,
supporting brain development and function, facilitating nutrient and vitamin transport,
aiding the uptake of essential molecules, and participating in signal transduction.[97 ]
Following uPA/PAI's detachment from VN, the cryptic receptor-binding site of PAI-1
becomes exposed, allowing the molecule to bind to LRP1.[59 ]
[60 ]
[61 ] This interaction initiates endocytosis of the uPA/uPAR/PAI-1 complex along with
associated integrins. Inside the cell, PAI-1 and uPA are targeted for lysosomal degradation,
while uPAR and LRP1 are recycled back to the plasma membrane, particularly at the
leading edge of migrating cells.[9 ]
[98 ]
[99 ]
[100 ] This recycling process enables cells to establish new adhesion points in different
ECM regions, contributing to persistent migration across tissue substrates. Notably,
while LRP1-mediated integrin endocytosis can occur, it is not strictly required for
PAI-1-initiated cell detachment, suggesting that PAI-1 can promote cell detachment
through multiple mechanisms, not solely dependent on LRP1-mediated integrin internalization.[98 ] Despite losing their antiprotease activity, latent or cleaved forms of PAI-1 retain
the capacity to bind to LRP1 independently and promote cell migration.[96 ]
[101 ]
[102 ] These forms can, therefore, remain embedded in the ECM, potentially serving as a
reservoir for active PAI-1 to maintain cell movement.[103 ]
LRP1 plays a dual role in endocytosis and signal transduction. PAI-1 facilitates the
internalization of uPA/uPAR complexes by binding LRP1 via its cryptic receptor-binding
domain exposed after detachment from VN.[68 ]
[104 ] In doing so, PAI-1 regulates receptor trafficking while also modulating intracellular
signaling. Studies have shown that LRP1 suppresses Rac1 and ERK activation in an uPAR-dependent
manner.[73 ]
[96 ]
[105 ] In LRP1-deficient mouse embryonic fibroblasts or in cells treated with receptor-associated
protein to block LRP1, Rac1 activity, and cell motility increase—an effect absent
from uPAR-knockout cells, confirming the requirement of uPAR for this pathway.[106 ] By controlling LRP1-mediated trafficking, PAI-1 thus influences Rho GTPase-driven
cytoskeletal dynamics and downstream MAPK activation, which in turn leads to cytoskeletal
rearrangements, including the formation of lamellipodia and filopodia, which are essential
for cell motility.
PAI-1 also plays a role in modulating intracellular signaling through the JAK/STAT
pathway, albeit indirectly, by regulating the availability and clustering of the uPA/uPAR
complex at the cell surface. When uPA binds uPAR, receptor clustering can recruit
JAK1 and activate STAT1, leading to STAT1 dimerization, nuclear translocation, and
transcription of target genes via interferon-γ activation site and interferon-stimulated
response elements.[74 ] This signaling is facilitated by co-localization with gp130.[74 ]
[107 ] By inhibiting uPA binding and promoting internalization of uPA/uPAR complexes through
LRP1, PAI-1 reduces surface receptor density and attenuates JAK/STAT activation. These
effects have been observed in epithelial and vascular smooth muscle cells where PAI-1
influences migration and gene expression through upstream control of this pathway.[108 ]
[109 ]
The intricate interplay between the components of the uPA system, particularly the
multifaceted role of PAI-1, emphasizes the complexity of cell migration regulation.
PAI-1's ability to influence cell behavior through interactions with uPA, its receptor
uPAR, LRP1, integrins, and VN highlights its significance as a potential therapeutic
target in diseases characterized by dysregulated cell migration and ECM remodelling.
While PAI-1 inhibition of uPA-mediated proteolysis can suppress cell migration, its
interactions with other molecules like VN and LRP1 can promote cell detachment and
signaling, leading to increased motility. This dual nature of PAI-1 explains the paradoxical
observations in clinical settings, which have been observed to be associated with
high levels of PAI-1 with poor prognosis in specific cancer patient populations.[110 ]
[111 ]
[112 ]
Sustaining Proliferative Signaling
Building on its established role in ECM regulation and modulation of uPA/uPAR signaling,
PAI-1 also contributes to proliferative signaling in cancer by shaping the extracellular
and pericellular environments that control receptor activation. While some studies
have shown PAI-1 to inhibit proliferation in certain cancer types, like hepatocellular
carcinoma,[92 ] ovarian,[113 ] and prostate cancer,[92 ] most research indicates that PAI-1 promotes cancer cell proliferation through various
mechanisms.[114 ]
[115 ] Giacoia et al demonstrated that PAI-1 downregulation inhibited cell proliferation
by inducing G0–G1 cell cycle arrest, whereas PAI-1 overexpression enhanced S-phase
entry and increased tumor size in xenograft models.[116 ] This is consistent with findings by Mashiko et al, who demonstrated that PAI-1 knockdown
or inhibition blocked ovarian cancer cell proliferation through G2–M cell cycle arrest
and intrinsic apoptosis.[117 ] Similarly, Li et al found that PAI-1 acts as a tumor enhancer in aggressive fibromatosis
but reduces tumor cell proliferation and motility when deficient.[118 ]
These proliferative effects are largely mediated through PAI-1's indirect influence
on surface receptor availability and proteolytic microenvironment. By inhibiting uPA-
and tPA-mediated plasminogen activation, PAI-1 limits plasmin generation and maintains
thrombin activity in the tumor microenvironment. Elevated thrombin levels can activate
protease-activated receptors (PARs), particularly PAR-1 and PAR-2, which are G-protein-coupled
receptors implicated in tumor cell proliferation, migration, and angiogenesis.[119 ]
[120 ]
[121 ] Thrombin-induced PAR-1 activation triggers intracellular signaling via MAPK/ERK
and PI3K/Akt pathways, linking PAI-1-mediated fibrinolytic suppression to enhanced
mitogenic responses. Furthermore, PAI-1 fibronectin-dependent cell growth by regulating
cell detachment from VN and adhesion to fibronectin.[120 ]
[121 ] PAI-1 also modulates growth factor receptor signaling by affecting uPA/uPAR complex
dynamics. Stabilization or clearance of these complexes can influence epidermal growth
factor receptor (EGFR) activation, as well as ERK and AKT signaling, as observed in
lung fibroblasts.[122 ] In other contexts, such as neuronal cells, PAI-1 enhances nerve growth factor-mediated
responses by maintaining the proteolytic balance necessary for receptor engagement.[123 ]
Despite these growth-promoting roles, PAI-1 can also induce senescence or dormancy
under stress or in the presence of p53 activation.[124 ]
[125 ]
[126 ] PAI-1 is a well-established transcriptional target of p53, and its upregulation
is essential for enforcing p53-mediated replicative senescence. Kortlever et al demonstrated
that PAI-1 expression alone was sufficient to induce a senescent phenotype in fibroblast
cells even in the absence of functional p53, permitting cells to bypass senescence.[124 ] Mechanistically, PAI-1 has been shown to inhibit proteasome activity, resulting
in stabilization of p53 and sustained activation of the p53–p21–pRb axis, thereby
enforcing irreversible cell cycle arrest.[124 ] Aguirre-Ghiso et al showed that PAI-1 inhibition of uPA/uPAR activity can induce
tumor cell dormancy by altering the ERK/p38 activity ratio.[127 ] Kubala and DeClerck[92 ] highlighted this ongoing controversy, stating that while some studies[116 ] show that PAI-1 stimulates cell cycle progression through cyclin D3/CDK4/6 upregulation,
other studies[128 ] indicate PAI-1 has an inhibitory function on proliferation in breast cancer cell
lines. Thus, without a precise molecular mechanism identified, the direct impact of
PAI-1 on tumor cell cycles remains a subject of debate.
Resisting Cell Death
PAI-1 contributes to tumor cell survival by modulating apoptosis through both indirect
protease inhibition and direct intracellular signaling mechanisms. PAI-1 has been
shown to suppress both extrinsic and intrinsic apoptotic pathways.[129 ] Extrinsically, it protects cells from Fas-mediated apoptosis by regulating pericellular
plasmin activity, thereby preventing the cleavage and shedding of Fas ligand (FasL)
in various human cancer cell lines.[129 ]
[130 ]
[131 ] This function is primarily mediated through PAI-1's ability to inhibit uPA and maintain
low plasmin levels, thereby preventing proteolytic activation of proapoptotic pathways.
Intrinsically, PAI-1 has been shown to interact with and inhibit caspase-3, a key
executioner of apoptosis, as demonstrated in vascular smooth muscle cells and endothelial
models.[129 ]
[132 ]
[133 ] This inhibition may occur through a cross-class serpin mechanism involving the RCL
of PAI-1, although this interaction remains incompletely defined. Furthermore, PAI-1
has been shown to upregulate antiapoptotic proteins, such as B-cell lymphoma 2 (Bcl-2)
and B-cell lymphoma-extra large (Bcl-xL), while downregulating proapoptotic proteins,
including Bcl-2-associated X protein (Bax) and B-cell lymphoma-extra small (Bcl-xS).[129 ]
[132 ]
[133 ]
These effects are mediated primarily through activation of the PI3K/Akt and ERK signaling
pathways. By promoting the internalization of uPA/uPAR complexes via LRP1, PAI-1 activates
PI3K, which phosphorylates and activates Akt.[9 ]
[134 ] Activated Akt enhances p53 transcriptional activity, suppressing transcription of
proapoptotic genes such as Bax while also enhancing the stability and expression of antiapoptotic proteins, Bcl-2
and Bcl-xL. Akt also phosphorylates and inactivates Bad, preventing it from displacing
Bcl-2 from its inhibitory complexes. Simultaneously, ERK activation downstream of
LRP1 signaling may contribute to transcription of Bcl-2 family genes, supporting cell
survival in stress conditions. These pathways have been observed in endothelial and
vascular smooth muscle cells in which PAI-1 expression correlates with resistance
to serum starvation or detachment-induced apoptosis.[135 ]
[136 ] Notably, the antiapoptotic activity of PAI-1 can be attenuated using neutralizing
antibodies, as demonstrated by Balsara et al, who accelerate its conformational inactivation,
suggesting that only the active form of PAI-1 possesses this prosurvival function.[137 ]
Despite this, several studies indicate that PAI-1's role in apoptosis is not uniformly
protective. Depending on the tumor type and microenvironment, PAI-1 overexpression
has been associated with increased apoptosis and reduced tumor growth.[138 ]
[139 ]
[140 ] As seen in certain breast and colon cancer models, PAI-1 overexpression or transfection
enhanced apoptotic sensitivity.[130 ]
[141 ] This functional duality underscores the complexity of PAI-1's role in cancer biology.
These multifaceted roles make PAI-1 an important determinant of tumor cell fate and
a compelling candidate for targeted therapy in apoptosis-resistant cancers.
Angiogenesis
PAI-1 plays a complex, dose-dependent role in tumor angiogenesis by regulating endothelial
cell migration, matrix remodeling, and protease activity through both direct and indirect
mechanisms.[95 ]
[140 ]
[142 ] At physiological concentrations, PAI-1 exhibits proangiogenic activity through multiple
mechanisms. It inhibits FasL cleavage by plasmin, preventing Fas-mediated apoptosis
in endothelial cells.[143 ] One mechanism involves inhibition of plasmin-mediated cleavage of FasL, thereby
protecting endothelial cells from Fas-dependent apoptosis and allowing sustained angiogenic
activity.[94 ]
[130 ] In parallel, PAI-1 facilitates endothelial migration by disrupting adhesion to VN—through
competitive binding at the somatomedin B domain—and promoting movement toward fibronectin-rich
areas typical of tumor stroma.[130 ]
[144 ]
By inhibiting plasminogen activation, PAI-1 contributes to fibrin accumulation, which
can enhance endothelial organization and the release of proangiogenic cytokines such
as interleukin-8.[145 ]
[146 ] This matrix stabilization not only provides a scaffold for new vessel formation
but also creates a localized environment favorable to cell guidance and alignment
during neovascular sprouting.[144 ] Moreover, PAI-1 promotes interleukin-8-mediated neutrophil transendothelial migration
by preventing the proteolytic shedding of IL-8/heparan sulfate/syndecan-1 complexes.
By stabilizing IL-8 on the endothelial surface, PAI-1 enhances IL-8 bioavailability
and augments its proangiogenic signaling effects.[147 ] These functions underscore PAI-1's essential role in coordinating immune cell-driven
and matrix-dependent processes that support angiogenic remodeling.
Importantly, VEGF-induced activation of uPA on endothelial cells leads to the formation
of uPA/PAI-1 complexes on the surface, which are then internalized via LRP1. This
internalization is required for targeted recycling of uPAR to the leading edge, a
process essential for directional migration during sprouting angiogenesis.[74 ] The recycling of uPAR was significantly impaired in PAI-1-deficient cells, further
underscoring PAI-1's non-proteolytic role in orchestrating the migratory behavior
of endothelial cells via receptor trafficking.[74 ]
Conversely, at supraphysiological concentrations, PAI-1 inhibits angiogenesis by suppressing
uPA and plasmin activity, both of which are required for matrix degradation and endothelial
invasion. Bajou et al showed in PAI-1 knockout models that normal physiological levels
of host-derived PAI-1 are essential for effective tumor vascularization, whereas excessive
PAI-1 impairs this process by overly restricting proteolysis.[140 ] This biphasic behavior explains the paradox wherein PAI-1 is associated with aggressive
tumor progression despite its protease-inhibiting function, reflecting its role in
maintaining the proteolytic balance needed for controlled angiogenesis. Notably, the
cellular source of PAI-1 is a critical determinant of its angiogenic function. Host-derived
PAI-1 at physiological concentrations promotes tumor invasion and angiogenesis; whereas,
PAI-1 produced by tumor cells, even at high concentrations, cannot compensate for
its absence in the host.[140 ] Transgenic mice overexpressing PAI-1 in tumor cells or tumor cell lines transfected
with PAI-1 cDNA exhibited impaired vascularization, suggesting that excessive or localized
PAI-1 disrupts the spatial and temporal proteolysis needed for neovessel formation.[140 ]
[148 ] This suggests that PAI-1's angiogenic function depends not only on its concentration
but also on its cellular origin and its ability to regulate endothelial receptor localization,
inflammatory signaling, and matrix composition.
