Plasminogen Activator Inhibitor The Expression Biological

Document Sample
Plasminogen Activator Inhibitor The Expression Biological Powered By Docstoc
					                                                                                                                   Review Article

Plasminogen Activator Inhibitor-1: The Expression,
Biological Functions, and Effects on Tumorigenesis
and Tumor Cell Adhesion and Migration
Chun-Chung Lee and Tze-Sing Huang1
National Cancer Research Center, National Health Research Institutes, Taipei, Taiwan

    Plasminogen activator inhibitor-1 (PAI-1) is a primary regulator of urokinase-type plas-
    minogen activator (uPA). It inhibits uPA by forming a covalent complex, thus blocking
    uPA’s interaction with substrates. The half-life of active PAI-1 is < 1 h and it is easily
    transformed to a more stable inactive latent form. By binding with vitronectin, PAI-1
    can be stabilized and remained its activity. PAI-1 expression is regulated by many in-
    trinsic factors (e.g. cytokines, growth factors, hormones, and lipids) and extrinsic fac-
    tors (e.g. physical injury and DNA-damaging agents). PAI-1 is an essential regulator in
    physiological thrombotic/fibrinolytic process in vessel. It is also in the extracellular
    matrix (ECM) where it controls local proteolysis via inhibiting uPA. PAI-1 can regulate
    cell adhesion via either inhibition of uPA or interference with the binding between cellu-
    lar integrins or uPA receptor (uPAR) and vitronectin. In addition, PAI-1 can regulate cell
    migration by inducing clearance receptor-mediated cycled attachment-detachment-                                   Keywords:
    reattachment of integrins. PAI-1 may have a role in regulating tumor invasion, angio-
    genesis and metastasis. Elevated levels of both uPA and PAI-1 are associated with a                                   PAI-1
    poor prognosis in many cancers. However, most in vivo experiments revealed: (1) High                                  uPA
    level (pharmacological level) of PAI-1 prevented angiogenesis and tumorigenesis; (2)
    Low level (physiological level) of PAI-1 conversely facilitated tumor growth and angio-                               uPAR
    genesis, and (3) While in PAI-1-deficient host, tumor growth and angiogenesis could not                              integrin
    progress. The conclusions of PAI-1 in cancer development are still controversial. Be-
    cause PAI-1 may play a role in tumor cell migration and invasion through anti-                                    cell adhesion
    proteinase activity and interference with cell attachment to ECM, some strategies can                             cell migration
    be considered to apply PAI-1 derivatives to cancer therapy.

    Journal of Cancer Molecules 1(1): 25-36, 2005.

Introduction                                                                molecular weight of 45 kDa. It is expressed in many cell
                                                                            types, such as fibroblasts, smooth muscle cells, endothelial
  Plasminogen activator inhibitors (PAIs ) are classified as a              cells, hepatocytes, inflammatory cells, let alone the platelets,
subgroup of the serine protease inhibitor (serpin) super-                   the main source of the circulating PAI-1 [1]. PAI-2 was origi-
family with a common characteristic of possessing an argin-                 nally described in the extracts prepared from placenta and
ine in the reactive center [1]. The members, including PAI-1,               accordingly was called placental PAI [1]. Macrophage line-
PAI-2, PAI-3 (protein C inactivator), and protease nexin 1                  age cells has also been found to express PAI-2 [1]. PAI-3
(PN-1), act as the inhibitors of tissue-type plasminogen acti-              was isolated from human urine and this molecule is identical
vator (tPA), urokinase-type plasminogen activator (uPA) and                 with protein C inactivator found in plasma [4]. PN-1 was
thrombin [1-4]. PAI-1 is a single chain glycoprotein with the               isolated from cultured fibroblasts, but is also produced by
                                                                            other cultured cell types [1-2]. Generally, PAIs are specific
Received 8/10/05; Revised 9/14/05; Accepted 9/20/05.                        and fast-acting inactivators with distinct biological character-
  Correspondence: Dr. Tze-Sing Huang, National Cancer Research              istics and are present in most body fluids, tissues and cell
Center, National Health Research Institutes, 7F, No. 161, Min-Chuan         lines. They can act as a pseudo-substrate to stably couple
East Road Sec. 6, Taipei 114, Taiwan, ROC. Phone: 886-2-26534401-           with target proteases by ester bond formation, and thus con-
25138; Fax: 886-2-27929654. E-mail:
  Abbreviations: PAI-1, plasminogen activator inhibitor-1; serpin, serine   vert the target enzymes into an inactive conformation [5].
protease inhibitor; PN-1, protease nexin 1; tPA, tissue-type plasminogen    PAI-1 is the most extensively studied serpin member as we
activator; uPA, urokinase-type plasminogen activator; uPAR, uPA             know today. In this article, we will review the biological func-
receptor; ECM, extracellular matrix; MMP, matrix metalloproteinase;         tion of PAI-1 and the factors that regulate PAI-1 expression.
TGF-β1, transforming growth factor-β1; bFGF, basic fibroblast growth        Additionally, we will also focus on the effects of PAI-1 on
factor; VEGF, vascular endothelial growth factor; RCL, reactive center      cancer cell growth, adhesion, migration and invasion. The
loop; LRP, lipoprotein receptor-related protein; SMB, somatomedin B;
                                                                            possible role of PAI-1 for cancer therapeutics will be dis-
LPS, lipopolysaccharide; TNF-α, tumor necrosis factor-α; PMA, phor-
bol 12-myristate 13-acetate; HRE, hypoxia-responsive element.               cussed.

 2005 MedUnion Press                                                                                                                  25
Lee et al. J. Cancer Mol. 1(1): 25-36, 2005

Figure1: Plasminogen activator inhibitor-
1 is a regulator of plasminogen activators.
There are two types of plasminogen
activator, i.e. tissue-type plasminogen
activator (tPA) and urokinase-type plasmi-
nogen activator (uPA). In the vessel, tPA
binds to fibrin and converts plasminogen
into plasmin for proteolytic degradation of
the clot. uPA is the major plasminogen
activator for migrating cells, and it is
activated by binding with uPA receptor
(uPAR) and subsequently initiates a
proteinase cascade. Plasmin is a pro-
teinase with a wide range of substrates. It
is able to degrade fibrin or other extracel-
lular matrix (ECM) components, to cleave
and activate other proteinases sush as
matrix metalloproteinases, and to activate
latent transforming growth factor-β1 (TGF-
β1), basic fibroblast growth factor (bFGF)
and vascular endothelial growth factor
(VEGF). Plasminogen activator inhibitor-1
(PAI-1) is a primary inhibitor to inhibit the
activation of both tPA and uPA.

PAI-1 is a regulator of PA system                                   designated as P1-P1’ residues of the RCL (Figure 2). The
                                                                    amino acids located N-terminal from this site are referred to
                                                                    as P-residues P16-P1 and the C-terminal residues are P’-
   PAI-1 is an important regulatory protein involved in the
                                                                    residues P1’-P13’ (Figure 2A). The strength of binding be-
proteolytic [6,7] and fibrinolytic [8,9] pathways of plasmino-
                                                                    tween the P1 residue and target protease determines the
gen activator. It is the major inhibitor of two types of plasmi-
nogen activator, i.e. tPA and uPA. tPA binds to fibrin and          inhibitory activity of the serpin family protein [23]. PAI-1
convert plasminogen to plasmin in the clot and is known as          inhibits PA by the formation of a covalent complex, thus
the primary fibrinolytic activator [10](Figure 1). Distinctly,      blocking PA for further interaction with its substrate [24].
uPA has very low affinity for fibrin, but it is the major PA        The inhibitory process involves the formation of a noncova-
form expressed by the migrating cells and its activity is me-       lent reversible Michaelis-like complex, followed by an acyl
diated by uPA receptor (uPAR). uPA binds to uPAR on cell            intermediate and finally the formation of an ester bond be-
surface to initiate a proteinase activity, which in turn leads to   tween the carboxyl group of the P1 residue and the hydroxyl
the activation of plasmin [11-13]. Plasmin is a proteinase          group of the serine residue of protease [25](Figure 2A upper
exhibiting a wide range of substrate spectrum. It is able to        panel). Once the initial complex is formed, the P1-P1’ bond
degrade many glycoproteins and proteoglycans of extracel-           in the RCL is cleaved, and immediately followed by an inser-
lular matrix (ECM) such as laminin, fibronectin, vitronectin,       tion of the P1-PA complex from the initial interaction site into
fibrin, etc. [14]. It also can activate other proteinases such as   the opposite side of the PAI-1. This causes a distortion of
matrix metalloproteinase (MMP)-1, -3 and -9 [15], and activate      the protease and inhibits its catalytic activity (Figure 2B). In
or release growth factors from ECM including latent trans-          this covalent complex, the PAI-1 is cleaved at the P1-P1’ site
forming growth factor (TGF-β) [16], basic fibroblast growth         implicating that the inhibitory activity of PAI-1 is limited to a
factor (bFGF) [14], and vascular endothelial growth factor          single encounter with its target protease [26-28]. Besides the
(VEGF) [17] (Figure 1). Cellular releasing of plasmin can           active and cleaved forms, there exists a latent form of PAI-1
therefore promote ECM degradation and cell migra-                   that can be transited from both active and cleaved forms.
tion/invasion [18]. By preventing the activation of plasmin,        The active form spontaneously converts to the latent form
the biological function of PAI-1 is not restricted in the vascu-    with a half-life of < 1 h [29]. In this structure, the entire amino
lar fibrinolytic system, PAI-1 can play an important role in        terminal side of the RCL is inserted as the central strand of β-
many other physiological and pathological processes of              sheet A (Figure 2C). This increases the stability of PAI-1 but
tissue remodeling including wound healing, embryogenesis,           lacking of its inhibitory activity. The latent form can be reac-
and tumor cell migration and invasion [19].                         tivated partially by denaturing agents [30].

Molecular structure of PAI-1 and its interaction with               Interaction of PAI-1 with proteins other than PAs
                                                                       In addition to the ability to bind and inactivate PAs, PAI-1
   PAI-1 is a single-chain glycoprotein belonging to the ser-       also has other ligands that might mediate functions of PAI-1.
pin family. All members of this family consist of about 400         These ligands include glucosaminoglycans (eg. heparin)[31],
amino acid residues and have apparent molecular weights             ECM components (eg. vitronectin)[32], and scavenger recep-
from 38 to 70 kDa, depending on their degree of glycosyla-          tors such as members of the low-density lipoprotein recep-
tion [20]. From the nucleotide sequence analysis, PAI-1             tor (LDLR) superfamily, specifically the lipoprotein receptor-
cDNA encodes a protein containing 402 amino acids with a            related protein (LRP)[33-37].
predicted nonglycosylated molecular mass of 45 kDa [21].
The mature secreted form of PAI-1 consists of 379 amino             Vitronectin
acids and contains ~13% carbohydrate, which increases the              Vitronectin is a 459-aa glycoprotein. It is divided into a
molecular weight to ~50 kDa. The tertiary structure of the          number of specific domains, including the 44-residue soma-
serpin family protein exhibits three β-sheets (A, B and C),         tomedin B (SMB) domain at the amino terminus, the connect-
nine α-helices, and an exposed reactive center loop (RCL)           ing region that is immediately adjacent to the SMB domain
that contains the reactive site Arg346-Met347 for the target ser-   and contains a single RGD sequence for binding with in-
ine protease [22]. The reactive residues Arg346-Met347 are          tegrin, and two hemopexin-like repeats. Vitronectin is rela-

26                                                                                                              Print ISSN 1816-0735
                                                                    PAI-1’s Expression, Biological Function, and Effects on Cancer Development

Figure 2: Molecular structure of PAI-1 and
its interaction with uPA. (A) PAI-1 contains
an exposed reactive center loop (RCL)(red).
The active site Arg346 (red ball)-Met347 (green
ball), designated as P1-P1’ residues, is
responsible for contacting with its target
protease (blue ball). The inhibitory process
involves the formation of a reversible acyl
intermediate, cleavage at the P1-P1’ bond,
and the final formation of an irreversible
ester bond between the carboxyl group of
the P1 residue and the hydroxyl group of the
serine residue of the protease. (B) Once the
initial acyl intermediate is formed, the P1-P1’
bond in the RCL is cleaved, and immediately
followed by a translocation of the P1-PA
complex from the initial interaction site to
the opposite side of the PAI-1. This causes
a distortion of the PA and thus inhibition of
the proteinase activity. (C) Besides the
active and cleaved forms, there exists a
latent form of PAI-1. In this structure, the
entire amino-terminal part of RCL (i.e. P1-
P16) is inserted as the central strand of β-
sheet A. This increases PAI-1’s stability but
decreases its inhibitory activity.

tively unique among adhesive proteins not only because               tors such as lipoprotein receptor-related protein (LRP)[37].
PAI-1 binds to it with high affinity but also cells can attach to    After PAI-1 is complexed with proteases, the heparin-binding
it through integrins, uPAR, or both [38,39](Figure 3A). Vi-          domain of PAI-1, which contains a cryptic receptor-binding
tronectin-bound PAI-1 inhibits the proteolytic degradation of        site (Lys69) with the high affinity for LRP, becomes exposed
ECM by inhibiting uPA/tPA, which in turn inhibits cell migra-        and accessible for LRP binding. The complex is endocyto-
tion and invasion [40]. The binding site of vitronectin for          sed by binding to clathrin-coated pit-localized endocytosis
PAI-1 is localized to the SMB domain [41]. All of the active         receptors (Figure 3C and Figure 4). The uPA and PAI-1 then
PAI-1 in plasma circulates in complex with vitronectin [32],         undergo lysosomal degradation, whereas uPAR is recycled
and the binding with vitronectin increases the stability of          and reappears on the cell surface [52] (Figure 4). However,
PAI-1 [42]. In tissue, vitronectin also binds to uPAR [43] and       another report demonstrate that GPI-anchored uPAR is also
the binding site for uPAR is located in the SMB domain as            endocytosed by piggybacking on LRP and that direct bind-
well [44]. The binding sites for PAI-1 and uPAR are partially        ing of occupied uPAR to LRP is essential for internalization
overlapping but not identical. The affinity of PAI-1 for the         of occupied uPAR [53]. Especially, direct binding of domain
SMB domain is much higher than that of uPAR [45] (Figure             3 (D3) of uPAR to LRP is required for clearance of uPA-PAI-
3B).    Thus, PAI-1 can competitively inhibit the uPAR-              1-occupied uPAR [53]. PAI-1 also contains a cryptic recep-
dependent       attachment      of    cells    to     vitronectin    tor-binding site that is exposed upon complex formation with
[38,39,45](Figure 3B). Binding of PAI-1 to SMB also inhibits         tPA, and plays a role in the clearance of tPA [54]. These
integrin-mediated cell adhesion by sterically blocking the           studies suggest that PAI-1, by interacting with uPA/uPAR
adjacent RGD site [36] (Figure 3B). The stability of PAI-1 in        and LRP, can be a potent chemoattractant molecule, which
plasma is increased by 2-fold when binding with vitronectin          induces cell migration via regulating morphological attach-
[46]. The crystal structure of the PAI-1–SMB complex sug-            ment/dettachment changes (Figure 4). Recently, PAI-1 was
gests a simple mechanism for the stabilization of PAI-1 activ-       reported to induce cell migration with corresponding cy-
ity: SMB slows the transition of PAI-1 to the latent form by         toskeleton reorganization and phosphotyrosine redistribu-
blocking the associated sliding movement of strands 1A and           tion [55]. Using several PAI-1 mutants and specific inhibi-
2A of the main β-sheet into gap between helices E and F [47]         tors, the activity of PAI-1 was shown not to depend on its
(Figure 2B). By binding to vitronectin, the target specificity       interactions with uPA, tPA or vitronectin, but rather with LRP.
of PAI-1 is altered, because interaction with vitronectin en-        This interaction leads to the activation of the Jak/Stat signal-
ables PAI-1 to inhibit another serine protease, thrombin [48].       ing pathway and the induction of cell migration [55].

Heparin                                                              Biochemical function of glycosylation of PAI-1
    It has been demonstrated that the PAI-1-binding site on vi-
tronectin is adjacent to a heparin-binding site [49]. In addi-         PAI-1 has 3 potential sites for N-linked glycosylation. A
tion, a heparin binding site can be predicted from the amino         heterogeneous glycosylation pattern was noted to occur at
acid sequence of PAI-1, which is conserved in antithrombin           the Asn209 and Asn265 sites of PAI-1, while the Asn329 site is
III and heparin cofactor II. Consequently, It have been inves-       not utilized for glycosylation yet in human cell lines [56].
tigated that PAI-1 indeed interacts with heparin. In the pres-       The latent transition of non-glycosylated PAI-1 was more
ence of heparin, the reactivity of PAI-1 toward thrombin is          easily enhanced by a non-ionic detergent if comparison with
substantially increased. In contrast to vitronectin, heparin         glycosylated PAI-1. The glycosylation at Asn265 seems to
does not stabilize the active conformation of PAI-1 [50,51].         determine the conformation of PAI-1. Also, the PAI-1 binding
                                                                     protein vitronectin may reverse the conformational changes
Lipoprotein receptor-related protein (LRP)                           induced by the lack of glycosylation at Asn265 [56].
  Once PAI-1 forms the complex with uPA that is specifi-
cally bound to uPAR, it loses its high affinity for vitronectin
but instead increases its affinity toward the clearance recep-

 2005 MedUnion Press                                                                                                                    27
Lee et al. J. Cancer Mol. 1(1): 25-36, 2005

Figure 3: PAI-1 involved in cell adhesion.
(A) Vitronectin contains the somatomedin B
(SMB) domain at its amino terminus, and
immediately adjacent to the SMB domain is
the single RGD sequence that is a binding
site for integrins. Vitronectin is relatively
unique among adhesive proteins not only
because PAI-1 binds to it but also because
cells can attach to it through integrins,
uPAR, or both. The binding sites of vi-
tronectin for uPAR and PAI-1 are both
located in the SMB domain. (B) The binding
sites of vitronectin for PAI-1 and uPAR are
partially overlapping but not identical. The
affinity of PAI-1 for the SMB domain is much
higher than the affinity of uPAR for this
domain.      Thus, PAI-1 can competitively
inhibit the uPAR-dependent attachment of
cells to vitronectin. Binding of PAI-1 to SMB
also inhibits integrin-mediated cell adhesion
via a steric hindrance to the adjacent RGD
site. (C) PAI-1 can bind with uPA/uPAR
complex.      This association renders the
cryptic receptor-binding site (CRB) of PAI-1
exposed out and increases PAI-1’s affinity
for the clearance receptor lipoprotein receptor-related protein (LRP). By binding to clathrin-coated pit-localized endocytosis receptors, the complex
of PAI-1/uPA/uPAR can be endocytosed. The PAI-1 and uPA then undergoes lysosomal degradation.

Many factors regulate PAI-1 expression in vitro and in                        Factors regulate PAI-1 expression
vivo                                                                             Expression of PAI-1 can be regulated at the transcriptional
                                                                              level by many factors including: growth factors and cyto-
                                                                              kines (e.g. TGF-β1, interleukin-1, FGF, and VEGF), hormones
PAI-1 promoter                                                                (e.g. glucocorticoids and insulin), inflammatory factors (e.g.
   The human PAI-1 gene is located on the long arm of                         tumor necrosis factor-α and lipopolysaccharide), glucose or
chromosome 7. The genomic DNA of human PAI-1 is ~12.2                         lipid metabolites (e.g. glucose, free fatty acid, triglycerol, and
kb long, comprising nine exons and eight introns [21,57].                     very-low-density lipoprotein), vascular tone regulating fac-
There are two distinct transcripts (3 and 2 kb, respectively)                 tors (e.g. angiogentensin II); chemicals (e.g. phorbol ester),
of PAI-1 mRNA with different lengths of polyadenylation in                    and other environmental or physical factors (e.g. reactive
the 3’ untranslated region (3’UTR)[21,58]. The sequence of                    oxygen species, hypoxia, stress, wound, adhesion to matrix).
5'-flanking DNA contains the essential cis-acting elements                    Some examples are described as below:
CCAAT and TATAA for RNA polymerase binding [59]. Sev-                            TGF-β1: TGF-β1 is one of the earliest cytokines that was
eral known consensus sequences of regulatory elements are                     described to regulate PAI-1 expression at the transcriptional
located in the region of ~800 bp in front of the PAI-1 tran-
                                                                              level. TGF-β1 is abundant in the α-granules of platelets.
scription initiation site. These cis-acting elements and the
                                                                              After vascular injury, it can be released from platelets and
trans-acting regulatory factors involved in the regulation of
                                                                              stimulate local endothelial PAI-1 expression. This will slow
PAI-1gene expression include: SP-1 (-76 to -71 and -44 to -39)
                                                                              down the lytic process of thrombin, and result in increase of
[60,61], CTF/NF-1 (-119 to -105, -477 to -445 and -542 to -519),
                                                                              fibrin deposit and progression of atherosclerosis. The TGF-
AP-1 (-65 to -50), AP-2 and SP-1-like (-82 to -65)[62], and two
                                                                              β1-responsive CAGA boxes in the PAI-1 gene promoter are
repressor elements (-764 to -628 and -266 to -188)[57]. Inter-
                                                                              the binding sites for SMA/Mothers against decapentaplegic
estingly, wild-type p53 acts as a trans-activator and mutant
                                                                              (Smad) proteins, which play the key role in TGF-β1-mediated
p53 (His273) acts as a repressor to the binding region at -160
                                                                              activation of PAI-1 transcription [64].
to -139 of PAI-1 promoter [63]. There are three CAGA con-
                                                                                 Inflammatory factors: Lipopolysaccharide (LPS) induces
served elements (-280, -580 and -70) that are specific for
                                                                              PAI-1 levels in the plasma of patients with Gram-negative
TGF-β1-activated Smad 3 and Smad 4 binding in the PAI-1
                                                                              septicemia and in the cultured human and bovine endothelial
promoter [64]. One hypoxia-responsive element at the hu-
                                                                              cells [71]. LPS stimulates PAI-1 synthesis but decreases tPA
man PAI-I promoter -194 to -187 is identified, and is neces-
                                                                              production in endothelial cells, resulting in an antifibrinolytic
sary and sufficient for hypoxia-mediated response
                                                                              outcome. The stimulatory effect of LPS on PAI-1 expression
[65](Figure 5).
                                                                              is mediated in part by tumor necrosis factor-α (TNF-α). The
                                                                              5' distal TNF-α-responsive element of PAI-1 gene located 15
PAI-1 mRNA stability                                                          kb upstream of the transcription start site that is a conserved
   The level of PAI-1 gene expression is determined not only                  NF-κB-binding site and mediates PAI-1’s response to TNF-α
by the rate of gene transcription but also by the rate at which               [72].
the mRNA is degraded. The AU-rich sequence in the 3’UTR                          Lipids: Free fatty acids induces PAI-1 expression by acti-
of PAI-1 transcript determines its mRNA stability [66,67].                    vating a transcription factor to bind with the sequence 5’-
Moreover, PAI-1 mRNA is especially degraded easily when                       TG(G/C)1-2CTG-3’ that is repeated four times in the PAI-1
cells have been treated with 8-bromo-cAMP [68]. Using se-                     promoter (-528 and –599)[73]. A very-low-density lipoprotein-
rial deletion and insertion of PAI-1 3’UTR sequence, the re-                  responsive element that locates at the PAI-1 promoter region
gion from 2926 to 3054 of PAI-1 mRNA is defined as PAI-1                      from -672 to -657 is responsible for the effect of plasma
cAMP-responsive element (PAI-1 CRE) [69]. Recently, a PAI-                    triglycerides on PAI-1 expression [74]. Notably, the 4G/5G
1 RNA binding protein 1 (PAI-RBP1) that specifically inter-                   polymorphism site locates at -675. Higher level of PAI-1 syn-
acts with the PAI-1 CRE has been identified to involve in                     thesis is associated with the 4G/4G genotype. The 4G/4G
cAMP-mediated PAI-1 mRNA degradation [70].                                    allele is preferentially bound with an enhancer protein,
                                                                              whereas the 5G/5G allele is bound with an enhancer plus a

28                                                                                                                           Print ISSN 1816-0735
                                                                  PAI-1’s Expression, Biological Function, and Effects on Cancer Development

Figure 4: PAI-1 involved in cell migration.
PAI-1 binds uPA/uPAR complex. This leads
to a large shift in PAI-1’s relative affinity
from vitronectin to the clearance receptor
LRP. Subsequently, PAI-1, LRP, uPA/uPAR,
and associated integrins will be endocyto-
sed. Cell will detach not only from the
vitronectin but also the general matrixes. In
the endosome, PAI-1/uPA complex sepa-
rates from the receptors and is targeted to
the lysosome for degradation. The LRP,
uPAR and integrins are then recycled to
another edge of the cell surface for the next
attachment to ECM. This model provides
for the PAI-1-mediated cycled attachment-
detachment-reattachment of integrins that is
necessary for cell migration.

suppressor, resulting in lower levels of PAI-1 transcription.      PMA-treated cells. The PMA response elements (i.e. AP-1
This polymorphism site responds to triglyceride. Increased         site) in the PAI-1 gene promoter are also identified [62].
levels of PAI-1 will occur in the individuals with the 4G/4G          Hypoxia: Oxygen deprivation, as occurred during tissue
genotype and elevated triglyceride levels in plasma [75,76].       ischemia, usually results in coagulation. When mice are
Despite the plasma levels of PAI-1 seem to be the result of        placed in a hypoxic environment, the plasma levels of PAI-1
both environmental and genetic influences, many studies            increase in a time-dependent manner accompanied by an
have reported that higher PAI-1 plasma levels occur in indi-       increase in plasma PAI-1 activity [90]. Northern blot analysis
viduals with 4G/4G genotype [77-80]. However, the data on          of the hypoxic murine lung tissues have demonstrated an
the clinical relevance of these polymorphisms are contro-          increase in PAI-1 mRNA compared with normoxic controls
versial [81]. The transcription factor PPARγ may also par-         [90]. In contrast, the levels of transcripts for both tPA and
ticipate in PAI-1 regulation, and PPARγ might involve in vas-      uPA are decreased under hypoxic conditions. Several puta-
cular diseases [82].                                               tive hypoxia-responsive elements (HRE-1, at -158 to –151;
   Glucose and Insulin: In pig aortic endothelial cells, the       HRE-2, at -194 to -187; HRE-3, at -453 to -446; HRE-4, at -566
increase of PAI-1 secretion is significantly associated with       to -559; HRE-5, at -681 to -674) were identified in the human
extracellular glucose concentration [83]. It has also been         PAI-I gene promoter [65]. Reporter gene assays show that
found that insulin stimulates synthesis of PAI-1 in human          the HRE-2 site is necessary and sufficient for the hypoxia-
hepatic cells [84]. Glucose-induced PAI-1 expression could         induced PAI-1 expression [65].
interpret the frequent occurrence of attenuated fibrinolytic          Cell adhesion: In anchorage-dependent cells, the level of
activity of plasma and premature cardiovascular disease in         PAI-1 mRNA expression is increased when cells begins to
the patients with type II diabetes mellitus. Glucose regulates     attach onto the substrate and subsequently turns to be down
PAI-1 gene expression through two adjacent SP-1 sites lo-          regulated when cells have attached already [91]. The PAI-1
cated between -85 and -42 of the PAI-1 5'-flanking region [61].    gene expression is induced only in adherent but not in non-
Release of a transcriptional repressor from the SP-1 com-          adhered cells. Regulation of PAI-1 gene expression by cell
plexes may explain the induction of PAI-1 gene under high          adhesion can be initiated through the interaction between
glucose conditions in vascular smooth muscle cells [61].           vitronectin, fibronectin or collagen and integrins, and is me-
Hyperglycemia stimulates Rho-kinase activity via PKC and           diated by PI-3 kinase/Akt and MEK/ERK pathways [92-94].
oxidative stress-dependent pathways, leading to increased
PAI-1 gene transcription [85].
                                                                   The role of PAI-1 in regulation of cell adhesion, migra-
   p53 and the cell cycle: A p53 binding site in the region -
                                                                   tion and invasion
160 to -139 of the human PAI-1 gene promoter has been
identified that is responsible for stimulation of the PAI-1
transcription by p53 [63]. Conversely, a p53 mutant (His273)       PAI-1 in cell adhesion
inhibits PAI-I promoter activity. This result suggests that an       PAI-1 has dual roles in regulation of cell adhesion. uPA
oncogenic p53 form may reduce PAI-1 expression and thus            functions as an anti-adhesion molecule by proteolytic deg-
alter the plasminogen/plasmin system during tumor pro-             radation of ECM components. As an inhibitor of uPA, PAI-1
gression. In addition, PAI-1 is expressed by a cell cycle-         would be expected to promote adhesion. Indeed, PAI-1 has
dependent manner with increasing expression during growth          the ability to selectively protect vitronectin from proteolysis
activation (G0-to-G1 transition), and PAI-1 could therefore be     by inhibiting local plasminogen activator and thus stabilizes
an indirect cell proliferation indicator [86,87].                  the vitronectin-dependent cell adhesion [95](Figure 3A).
   Phorbol Ester: Phorbol 12-myristate 13-acetate (PMA) in-        However, PAI-1 also exhibits an anti-adhesive function. Its
duces the PAI-1 levels in human rhabdomyosarcoma and               association with vitronectin blocks the binding of both uPAR
hepatocellular carcinoma cell lines [88,89]. There is also an      and integrin to vitronectin [39,96-99](Figure 3B). In addition,
increase in the PAI-1 mRNA and gene transcription levels in        an anti-adhesive effect of PAI-1 can be through the down-

 2005 MedUnion Press                                                                                                                  29
Lee et al. J. Cancer Mol. 1(1): 25-36, 2005

Figure 5:     The transcription regulatory
sites in the PAI-1 promoter. PAI-1 gene
expression is regulated by many intrinsic
factors (e.g. cytokines, growth factors,
hormones, and lipids) and extrinsic factors
(e.g. physical injury and DNA-damaging
agents). The sequence of 5'-flanking DNA
contains the essential cis-acting elements
CCAAT and TATAA for RNA polymerase
binding. Several known consensus se-
quences of regulatory elements are located
in the region of ~800 bp in front of the PAI-
1 transcription initiation site. These cis-
acting elements and the trans-acting
regulatory factors include: SP-1 (-76 to -71
and -44 to -39), CTF/NF-1 (-119 to -105, -477
to -445 and -542 to -519), AP-1 (-65 to -50),
AP-2 and SP-1-like (-82 to -65), and two
repressor elements (-764 to -628 and -266
to -188). Interestingly, wild-type p53 acts
as a trans-activator and mutant p53 (His273)
acts as a repressor to the binding region
at -160 to -139. There are
three CAGA conserved elements (-280, -580 and -70) that are specific for TGF-β1-activated Smad 3 and Smad 4. One hypoxia-responsive element at
the human PAI-I promoter -194 to -187 is identified, and is necessary and sufficient for hypoxia-mediated response.

regulation of the integrin levels on cell surface by the                   ability to inhibit the proteinase activity (110-112), but co-
mechanism that PAI-1 binds and induces LRP-mediated en-                    expression of uPA, uPAR and PAI-1 is required for optimal
docytosis of PAI-1/uPA/uPAR/integrin complex [100](Figure                  invasiveness of human lung cancer cells [113]. In addition,
3C).                                                                       transfection of PAI-1 cDNA into PC-3 prostate carcinoma
                                                                           cells has been reported not to alter cell invasiveness [114].
PAI-1 in cell migration
   Cell migration is the locomotion of a cell over an ECM
                                                                           PAI-1 in tumor growth, angiogenesis and metastasis
substratum. uPA has been recognized as to stimulate cell
migration by catalyzing plasminogen activation for proteoly-
sis of substratum and thus releasing cells from the substra-                  Association of uPA with cancer development has been no-
tum [101]. PAI-1 would be expected to inhibit plasminogen                  ticed for decades. A simple conception is that ECM and
activation-dependent cell migration. However, it has been                  basement membrane are degraded through the proteinases
shown that PAI-1 can directly block cell attachment and mi-                that are activated by uPA during cancer cell invasion. High
gration by a mechanism independent of its anti-proteolytic                 levels of uPA in extracts of primary breast carcinoma can be
activity [102,103]. PAI-1 can play a role in cell migration                used to predict an early relapse [115,116]. Several studies
owing to its dual roles in regulating the cell adhesion. It can            with specific disruptions of the genes for plasminogen and
be either pro-migratory or anti-migratory, depending on                    uPA also support the hypothesis that uPA activation is a
whether PAI-1 locates at the leading or trailing edge of the               rate-limiting step for tumor growth, local invasion and dis-
cell and on the PAI-1 concentration [97,102-105]. Once PAI-1               tant metastasis [117-121]. However, as an inhibitor of uPA
is binding with the uPA/uPAR complex, PAI-1 decreases its                  and originally regarded as an anti-cancer molecule, PAI-1 is
affinity for vitronectin but increases its affinity for the clear-         surprised to be a poor prognostic marker in breast and other
ance receptors LRP and very-low-density lipoprotein recep-                 types of cancer [101,122-126]. What on earth is the role of
tor [37,51]. Subsequently, cell will detach not only from the              PAI-1 during tumor cell progress toward malignancy? It is
vitronectin but also the general matrixes by the clearance                 an important stimulant for cancer scientists to investigate
receptor-mediated endocytosis mechanism [98], because the                  the biological function(s) of PAI-1 in tumor. PAI-1 has been
endocytosis will recruit the PAI-1/uPA/uPAR complex and                    extensively investigated in tumor growth, invasion, and me-
associated integrins [106-108]. In the endosome, PAI-1/uPA                 tastasis by experimental animal models (some important
complex separates from the receptors and is targeted to the                examples summarized in Table 1). The results obtained are
lysosome for degradation. The clearance receptors, uPAR                    quite controversial as different experimental systems are
and integrins are then recycled to another edge of the cell                used. For example, high levels of PAI-1 expression in hu-
surface. Thus this model provides for the cycled attach-                   man or murine cancer cells were associated with the retarda-
ment-detachment-reattachment of integrins that is necessary                tion of tumor growth, invasion, and metastasis in immunode-
for cell migration [109]. Hence, the cell migration is not only            ficient mice [112,114,127]. Application of high level of PAI-1
as simple as through the proteolysis of ECM or blocking the                protein to the immunodeficient mice bearing transplanted
binding between vitronectin and integrin, but also through                 human tumors caused tumor growth inhibition [128,129], but
the induction of endocytosis by PAI-1-conjugated uPA/uPAR                  low level of PAI-1 enhanced tumor growth [129]. Deficiency
complex (Figure 4).                                                        in PAI-1 or uPA expression in host mice prevented murine
                                                                           T241 fibrosarcoma growth compared with control wild-type
PAI-1 in cell invasion                                                     mice [130]. Also, PAI-1 deficiency in the host mice pre-
   When cells progress from migration to invasion, not only                vented local invasion and tumor vascularization of trans-
locomotion but also penetration through ECM of the cells                   planted malignant keratinocytes [131]. While this prevention
can be observed. In the cell invasion models, the proteoly-                in PAI-1-deficient mice was aborted by the exogenous ade-
sis of ECM by uPA is well associated with the cellular pene-               noviral vector expressing PAI-1, the tumor invasion and an-
tration in ECM. However, the observatory results of PAI-1                  giogenesis were restored [131]. This PAI-1-recovered tumor
action influencing cell invasion are not consistent. Some                  vascularization is through its proteinase inhibition activity
reports suggest that PAI-1 inhibits cell invasion through its              but not through its interference with vitronectin-integrin

30                                                                                                                     Print ISSN 1816-0735
                                                                               PAI-1’s Expression, Biological Function, and Effects on Cancer Development

Table 1: Some important examples studying the roles of PAI-1 in cancer development by experimental animal models
                      Model System                                                   PAI-1 Effect                                    Reference
Overexpression of PAI-1 in PC-3 human prostate carcinoma Overexpression of PAI-1 reduced the aggressive phenotype Soff et al. 1995 [114]
in athymic mice                                          of PC-3 human prostate carcinoma.

B16 melanomas implanted into PAI-1-overexpression or         No difference in tumor growth, metastasis, and survival was Eitzman et al. 1996 [137]
PAI-1 deficient mice                                         observed between control and PAI-1-deficient or -
                                                             overexpressed mice.

Intraocular melanomas by adenovirus-mediated gene trans- Metastatic tumor burden was reduced and host survival was Ma et al. 1997 [127]
fer of PAI-1 in an athymic mouse model                   significantly prolonged.

LNCaP prostate carcinoma in SCID mice treated with mu-       PAI-1 inhibited tumor growth through its inhibition of PA      Jankun et al. 1997 [128]
tated form of PAI-1 to confirm a correlation between the     activity.
inactivation of uPA and tumor size

Malignant keratinocytes implanted into PAI-1-deficient       Deficient PAI-1 expression in host mice prevented invasion Bajou et al. 1998 [131]
mice                                                         and tumor vascularization of transplanted malignant kerati-
                                                             nocytes. But it was restored by intravenous injection of a
                                                             adenoviral vector expressing human PAI-1.

T241 fibrosarcomas implanted into uPA or PAI-1 deficient     uPA or PAI-1 deficiency disturbed tumor growth. Only PAI- Gutierrez et al. 2000 [130]
mice                                                         1 deficiency prevented angiogenesis.

Chicken chorioallantoic membrane treated with pharmacol- PAI-1 inhibited angiogenesis either through inhibiting pro-        Stefansson et al. 2001 [133]
ogical levels of PAI-1 mutants                           teinase activity or blocking cellular access to vitronectin.

Malignant keratinocytes implanted into PAI-1 deficient       PAI-1 restored tumor vascularization through inhibition of     Bajou et al. 2001 [132]
mice and treated with adenoviral PAI-1 mutants               proteinase activity but not by interacting with vitronectin.

M21 human melanomas in nude mice treated with different PAI-1 stimulated tumor growth and angiogenesis at low      McMahon et al. 2001 [129]
concentrations of PAI-1: matrigel implants in wild type amounts while it inhibited tumor growth at higher amounts.
mice treated with increasing concentrations of PAI-1

Overexpression of PAI-1, PAI-1 (VN-), PAI-1 (INH-) PAI- Only overexpression of wild-type PAI-1 reduced the burden Praus et al. 2002 [112]
1 (VN-, INH-) and tumor cell migration in vitro and metas- of metastasis by 68% compared with the control.
tasis in mice in vivo

Aortic ring explant in PAI-1 deficient mice treated with   PAI-1 stimulated angiogenesis at low doses while it inhib-       Devy et al. 2002 [134]
increasing concentration of PAI-1 or adenovirus expression ited angiogenesis at higher doses.
of PAI-1 mutants.

T241 fibrosarcomas implanted into PAI-1 deficient mice       PAI-1 deficiency had no effect on tumor growth and angio- Curino et al. 2002 [138]

MMTVPym-T-induced breast tumor that mimic human              No difference of breast tumor development and vascular         Almholt et al. 2003 [140]
mammary adenocarcinoma in PAI-1 deficient or normal          density was observed between PAI-1 deficient and normal
mice                                                         mice.

Doses of recombinant PAI-1 protein intraperitonealy in-     PAI-1 exhibited both pro- and anti-angiogenic effects de-       Lambert et al. 2003 [135]
jected in PAI-1 deficient animals to see the angiogenic     pending on the dose and merely through its anti-proteolytic
potential with a model of laser-induced choroidal neovascu- activity.
larization. (CNV)

PAI-1-overexpressed PVDA keratinocytes transplanted into Overexpression of PAI-1 in host or in tumor cells reduced          Bajou et al. 2004 [136]
PAI-1-overexpressed, -deficient or wild type mice        tumor development in vivo. High level of PAI-1 in cancer
                                                         cells did not overcome the PAI-1 deficiency in the host for
                                                         tumor growth.

Low grade (HaCaT II-4) and high grade (HaCaT A5-RT3)         Tumor incidence was reduced for HaCaT II-4 (low grade)         Maillard et al. 2005 [141]
of malignant skin keratinocytes subcutaneously injected in   cells in PAI-1 deficient mice, but not reduced for HaCaT
PAI-1-deficient and wild type mice                           A5-RT3 (high grade) cells.

binding [132]. Many studies have suggested that PAI-1 at                        cells were transplanted into the PAI-1-deficient mice, the
high level prevents tumor growth through inhibiting angio-                      PAI-1-overexpressed cells did not compensate the PAI-1
genesis, while enhances tumor growth by promoting angio-                        deficiency in the host to recover the angiogenesis and tu-
genesis at low level [129,133,134]. Recently, many investiga-                   morigenesis [136]. However, some reports exhibit no sig-
tions more support that null expression of PAI-1 in the host                    nificant difference in the tumor growth and angiogenesis
and supraphysiologic level of PAI-1 expression either by                        between PAI-1-deficient and wild-type mice. For example,
host or tumor cells exhibit preventive effect on tumor growth                   when intravenous inoculation of murine melanoma cells into
and angiogenesis, but the physiological concentration of                        PAI-1-overexpressed or PAI-1-deficient mice in comparison
PAI-1 promotes in vivo tumor invasion and angiogenesis                          with wild-type control, no differences were observed in the
[135,136]. Interstingly, when the PAI-1-overexpressed tumor                     extent of pulmonary metastasis, tumor size and overall sur-

 2005 MedUnion Press                                                                                                                                    31
Lee et al. J. Cancer Mol. 1(1): 25-36, 2005

vival in these mice [137]. PAI-1 deficiency has also no effect     Is PAI-1 a target or drug for cancer therapy
on the tumor growth and angiogenesis of the transplantation
of T241 fibrosarcoma cells [138]. One report illustrates that
                                                                      Since the early observations of elevated PAI-1 expression
using MMTVPym-T-induced breast cancer to mimic sponta-
                                                                   in several tumor types, PAI-1 had been expected to be a po-
neously human mammary adenocarcinoma [139], there are
                                                                   tential target for cancer therapy. But actually, it still remains
no differences in the primary tumor growth and vascular
                                                                   a controversy of the role of PAI-1 in tumorgenesis. From
density between PAI-1-deficient and wild-type mice [140]. In
                                                                   many experimental results, low level (physiological level) of
addition, PAI-1 deficiency is not sufficient to prevent neo-
                                                                   PAI-1 in the host may facilitate angiogenesis and tumor
plastic growth of the aggressive tumors in human skin, it
                                                                   growth [129,131,133-136]. In PAI-1-deficient background,
still exerts its tumor-promoting effect in a tumor stage–
                                                                   angiogenesis and tumor will not progress [130,131]. Con-
dependent manner [141].
                                                                   versely, many other studies indicate that administration of
   The proangiogenic effect of PAI-1 at suitable concentra-
                                                                   high level (pharmacological level) of PAI-1 seems to prevent
tions may be through the inhibitory activity of PAI-1 to limit
                                                                   the    angiogenesis,     tumor    growth     and     metastasis
uPA-dependent plasminogen activation, and thus stabilize
                                                                   [112,114,127-129]. It is notable that all these results are de-
the basement membranes around newly formed vessels and
                                                                   rived from the model of tumor transplanted, and many ex-
also stabilize the matrix scaffold required for endothelial cell
                                                                   periments with different cell types or different protocols
migration and the assembly of endothelial cells into capillar-
                                                                   exhibit the different conclusions. The only one study that
ies. A critical balance between proteases and their inhibitors
                                                                   used spontaneously induced tumors in the PAI-1-deficient
is thought to be essential for optimal tumor cell invasion and
                                                                   host also shows no correlation of PAI-1 with tumorigenesis
angiogenesis [113,142]. Elevated expression of PAI-1 in
tumor cells tends to inhibit tumor growth, invasion, and me-
                                                                      However, administration of pharmacological levels of PAI-
tastasis. This result is opposite to the clinical observation
                                                                   1 to inhibit tumor growth and angiogenesis has been dem-
that PAI-1 overexpression correlates with the malignancy in
                                                                   onstrated by many reports [112,114,127-129,133,134,136].
breast or colon cancer. By immunohistochemistry assay,
                                                                   Because PAI-1 is multifunctional and exhibits short half-life
the PAI-1 is expressed in the leading edge of myofibroblasts
                                                                   in vivo, more studies needed to be performed to render PAI-1
in breast cancer tissues, which is responsible for the corre-
                                                                   really usable in the treatment of cancer. It should be con-
lation between the PAI-1 level in breast tumor extracts and
                                                                   cerned that if treatment of high level of the PAI-1, the inhibi-
poor prognosis [143]. Similar observation is also obtained in
                                                                   tory activity of PAI-1 may increase the rate of the process of
colon cancer [144]. These results are consistent with other
                                                                   atherosclerosis and thrombosis. The strategy of using the
results that PAI-1 is mostly expressed by stromal cells but
                                                                   specific mutant is intriguing to scientists. For example, the
not by tumor cells [131,140].        Moreover, whatever the
                                                                   mutant of PAI-1 that interferes with the binding of integrin
amounts of PAI-1 are produced by tumor cells, blood vessels
                                                                   with vitronectin but without inhibiting PA activity is found to
are limited under the collagen gel and tumor cells fail to in-
                                                                   inhibit angiogenesis in vivo, but not found to promote
vade in the PAI-1-deficient host tissue [136]. These findings
                                                                   thrombosis because it does not inhibit fibrinonlysis [133].
suggest that the site of PAI-1 production (stromal cells
                                                                   Up to date, more than 600 PAI-1 mutants have been prepared
rather than tumor cells) is a critical factor in tumor angio-
                                                                   and all these mutants have been utilized for studying the
genesis rather than the total amount of PAI-1 in tumor.
                                                                   structure-and-function of PAI-1 [148]. The availability of
However, some reports exhibit no differences in the tumor
                                                                   these mutants and the information of the PAI-1’s function
growth and angiogenesis between PAI-1-overexpressed or -
                                                                   may provide us some directions to develop useful agents to
deficient and wild type mice. These controversial results
                                                                   inhibit tumor angiogenesis and growth. As demonstrated by
may implicate that the influence of diverse factors exist on
                                                                   many studies, PAI-1 has been administrated in its protein
animal tumor models, such as cell numbers implant, cell
                                                                   form. Delivery of PAI-1 gene into the target tissue can be
types (stages of malignancy), expression of some other
                                                                   another considerable way. Adenovirus-mediated PAI-1 gene
types of PAI (PAI-2 or PAI-3), and the activation of PA system.
                                                                   expression is found to reduce the tumor growth, migration
Nevertheless, in despite of many reasons for the incompati-
                                                                   and metastasis in vivo [127,149,150]. Some factors, such as
ble results, most of these studies result from the transplan-
                                                                   the efficiency of the gene delivery system (adenovirus, ret-
tation of tumor models that may not be reflected onto the
                                                                   rovirus or liposome), the design for PAI-1 expression con-
real pathogenesis of cancer. Till now, only one report illus-
                                                                   struction (driven by which promoter and what the mutant
trates the spontaneous development of primary breast tu-
                                                                   design), and the possibility of accompanying side effects
mors, but shows no significant difference in the rates of
                                                                   during the delivery process, are all needed to concern. An-
tumor formation, growth, angiogenesis, and metastasis.
                                                                   other intriguing strategy is to develop antibodies that par-
From these results, how does PAI-1 involve in the real proc-
                                                                   tially modulate PAI-1’s function. It has been demonstrated
ess of tumorigenesis is still not certified. It still needs to
                                                                   that antibodies to PAI-1 can suppress the metastasis of hu-
accumulate more investigations using such spontaneous
                                                                   man tumor cells in mouse xenograft models [151-153]. In
carcinogenesis models.
                                                                   addition, the scientists have screened some small molecules
   It has been suggested recently that the prognostic impact
                                                                   bearing the specific inhibitory activity on PAI-1 [154-156],
of PAI-1 is not based on its involvement in angiogenesis
                                                                   although the detailed mechanisms of these small molecules
alone [145]. Another important efferent is that high levels of
                                                                   are unclear yet.
PAI-1 contribute to tumor growth by inhibiting apoptosis of
tumor cells. Addition of PAI-1 to culture medium can inhibit
tumor cell apoptosis [146]. The spontaneous fibrosarcoma           Conclusion
with wild-type PAI-1 shows less sensitivity to chemotherapy-
elicited apoptosis [147]. This phenomenon was not noted in           Further studies should be done to fully realize PAI-1’s bio-
PAI-1-deficient fibrosarcoma. However either in wild type or       logical functions in vivo. For understanding of PAI-1’s role
PAI-1 deficiency, the mice display similar sensitivity to the      in cancer development, accumulation of more data from
treatment of etoposide [147]. Hence, PAI-1 might regulate          consistent animal models and protocols is necessary.
apoptosis of transformed cells but not normal cells, suggest-      Moreover, spontaneous tumorigenesis models should be
ing a differential effect of PAI-1 between cancer cells and        adopted to investigate the actual role of PAI-1 in tumori-
normal cells.                                                      genesis. The developments of several PAI-1 mutants, gene
                                                                   delivery systems, and some antibodies or small molecules
                                                                   for antagonizing the partial or complete PAI-1’s function may
32                                                                                                            Print ISSN 1816-0735
                                                                           PAI-1’s Expression, Biological Function, and Effects on Cancer Development

be the advantageous strategies for developing novel cancer                  27.   Lindahl TL, Ohlsson PI, Wiman B. The mechanism of the reac-
therapeutic agents.                                                               tion between human plasminogen-activator inhibitor 1 and tis-
                                                                                  sue plasminogen activator. Biochem J 265: 109-113, 1990.
                                                                            28.   Wilczynska M, Fa M, Ohlsson PI, Ny T. The inhibition mecha-
References                                                                        nism of serpins. Evidence that the mobile reactive center loop is
                                                                                  cleaved in the native protease-inhibitor complex. J Biol Chem
1.    Saksela O, Rifkin DB. Cell-associated plasminogen activation:               270: 29652-29655, 1995.
      regulation and physiological functions. Annu Rev Cell Biol 4:         29.   Hekman CM, Loskutoff DJ. Kinetic analysis of the interactions
      93-126, 1988.                                                               between plasminogen activator inhibitor 1 and both urokinase
2.    Loskutoff DJ, Edgington TE. Synthesis of a fibrinolytic activator           and tissue plasminogen activator. Arch Biochem Biophys 262:
      and inhibitor by endothelial cells. Proc Natl Acad Sci USA 74:              199-210, 1988.
      3903–3907, 1977.                                                      30.   Hekman CM, Loskutoff. Endothelial cells produce a latent inhibi-
3.    Baker JB, Low DA, Simmer RL, Cunningham DD. Protease-nexin:                 tor of plasminogen activators that can be activated by denatur-
      a cellular component that links thrombin and plasminogen acti-              ants. J Biol Chem 260: 11581-11587, 1985.
      vator and mediates their binding to cells. Cell 21: 37–45,1980.       31.   Lawrene D, Strandberg, L, Grundstrom T, Ny T. Purification of
4.    Heeb MJ, Espana F, Geiger M, Collen D, Stump DC, Griffin JH.                active human plasminogen activator inhibitor 1 from Es-
      Immunological identity of heparindependent plasma and urinary               cherichia coli. Comparison with natural and recombinant forms
      protein C inhibitor and plasminogen activator inhibitor-3. J Biol           purified from eucaryotic cells. Eur J Biochem 186: 523-533, 1989.
      Chem 262: 15813–15816,1987.                                           32.   Declerck PJ, DeMol M, Alessi MC, Baudner S, Paques EP, Pre-
5.    Potempa J, Korzus E, Travis J. The serpin superfamily of pro-               issner KT, Muller-Berghaus G, Collen D. Purification and charac-
      teinase inhibitors: structure, function, and regulation. J Biol             terization of a plasminogen activator inhibitor 1 binding protein
      Chem 269: 15957–15960, 1994.                                                from human plasma. Identification as a multimeric form of S
6.    Chapman HA, Vavrin Z, Hibbs JB. Macrophage fibrinolytic activ-              protein (vitronectin). J Biol Chem 263: 15454-15461, 1988.
      ity: identification of two pathways of plasmin formation by intact    33.   Nykjaer A, Petersen CM, Moller B, Jensen PH, Moestrup SK,
      cells and of a plasminogen activator inhibitor. Cell 28: 653-662,           Holtet TL, Etzerodt M, Thogersen HC, Munch M, Andreasen PA,
      1982.                                                                       Gliemann J. Purified alpha 2-macroglobulin receptor/LDL recep-
7.    Chapman HA. Plasminogen activators, integrins, and the coor-                tor-related protein binds urokinase.plasminogen activator in-
      dinated regulation of cell adhesion and migration. Curr Opin                hibitor type-1 complex. Evidence that the alpha 2-macroglobulin
      Cell Biol 9: 714-724, 1997.                                                 receptor mediates cellular degradation of urokinase receptor-
8.    Loskutoff DJ, Curriden SA. The fibrinolytic system of the vessel            bound complexes. J Biol Chem 267: 14543-14546, 1992.
      wall and its role in the control of thrombosis. Ann NY Acad Sci       34.   Stefansson S, Kounnas MZ, Henkin J, Mallampalli RK, Chappell
      598: 238-247, 1990.                                                         DA, Strickland DK, Argraves WS. gp330 on type II pneumocytes
9.    Collen D. The plasminogen (fibrinolytic) system Thromb                      mediates endocytosis leading to degradation of pro-urokinase,
      Haemost 82: 258-270, 1999.                                                  plasminogen activator inhibitor-1 and urokinase-plasminogen
10.   Strickland S. Tissue plasminogen activator in nervous system                activator inhibitor-1 complex. J Cell Sci 108: 2361-2368, 1995.
      function and dysfunction. Thromb Haemost 86: 138–143, 2001.           35.   Heegaard, CW, Simonsen AC, Oka K, Kjoller L, Christensen A,
11.   Mondino A, Resnati M, Blasi F. Structure and function of the                Madsen B, Ellgaard L, Chan L, Andreasen PA. Very low density
      urokinase receptor. Thromb Haemost 82 (Suppl 1): 19–22, 1999.               lipoprotein receptor binds and mediates endocytosis of
12.   Ploug M. Structure-function relationships in the interaction                urokinase-type plasminogen activator-type-1 plasminogen acti-
      between the urokinase-type plasminogen activator and its re-                vator inhibitor complex. J Biol Chem 270: 20855-20861, 1995.
      ceptor. Curr Pharmaceut Design 9: 1499–1528, 2003.                    36.   Stefansson S. Lawrence DA, Argraves WS. Plasminogen activa-
13.   Behrendt N, Ronne E, Dano K. The structure and function of the              tor inhibitor-1 and vitronectin promote the cellular clearance of
      urokinase receptor, a membrane protein governing plasmino-                  thrombin by low density lipoprotein receptor-related proteins 1
      gen activation on the cell surface. Biol Chem Hoppe Seyler 376:             and 2. J Biol Chem 271: 8215-8220, 1996.
      269-279, 1995.                                                        37.   Stefansson S. Muhammad S, Cheng XF, Battey FD, Strickland
14.   Sahni A, Francis CW. Plasmic degradation modulates activity of              DK, Lawrence DA. Plasminogen activator inhibitor-1 contains a
      fibrinogen-bound fibroblast growth factor-2. J Thromb Haemost               cryptic high affinity binding site for the low density lipoprotein
      1: 1271-1277, 2003.                                                         receptor-related protein. J Biol Chem 273: 6358-6366, 1998.
15.   Murphy G, Stanton H, Cowell S, Butler G, Knauper V, Atkinson S,       38.   Waltz DA, Natkin LR, Fujita RM, Wei Y, Chapman HA. Plasmin
      Gavrilovic J. Mechanisms for pro-matrix metalloproteinase acti-             and plasminogen activator inhibitor type 1 promote cellular mo-
      vation. APMIS 107: 38-44, 1999.                                             tility by regulating the interaction between the urokinase recep-
16.   Khalil N, Corne S, Whitman C, Yacyshyu H. Plasmin regulates                 tor and vitronectin. J Clin Invest 100: 58-67, 1997.
      the activation of cell-associated latent TGF-beta 1 secreted by       39.   Deng G, CurridGen SA, Hu G, Czekay RP, Loskutoff DJ. Plasmi-
      rat alveolar macrophages after in vivo bleomycin injury. Am J               nogen activator inhibitor-1 regulates cell adhesion by binding to
      Respir Cell Mol Biol 15: 252-259, 1996.                                     the somatomedin B domain of vitronectin. J Cell Physiol 189:
17.   Rifkin DB, Mazzieri R, Munger JS, Noguera I, Sung J. Proteolytic            23-33, 2001.
      control of growth factor availability. APMIS 107: 80-85, 1999.        40.   Blasi F. Proteolysis, cell adhesion, chemotaxis, and invasive-
18.   Sidenius N, Blasi F. The urokinase plasminogen activator sys-               ness are regulated by the u-PA-u-PAR-PAI-1 system. Thromb
      tem in cancer: recent advances and implication for prognosis                Haemost 82: 298-304, 1999.
      and therapy. Cancer Metastasis Rev 22: 205–222, 2003.                 41.   Seiffert D, Ciambrone G, Wagner NV, Binder BR, Loskutoff DJ.
19.   Lijnen HR. Pleiotropic functions of plasminogen activator inhibi-           The somatomedin B domain of vitronectin. Structural require-
      tor-1. J Thromb Haemost 3: 35-45, 2005.                                     ments for the binding and stabilization of active type 1 plasmi-
20.   Travis J, Salvesen GS. Human plasma proteinase inhibitors.                  nogen activator inhibitor. J Biol Chem 269: 2659–2666, 1994.
      Annu Rev Biochem 52: 655-709, 1983.                                   42.   Mimuro J, Loskutoff DJ. Binding of type 1 plasminogen activa-
21.   Ginsburg D, Zeheb R, Yang AY, Rafferty UM, Andreasen PA,                    tor inhibitor to the extracellular matrix of cultured bovine endo-
      Nielsen L, Dano K, Lebo RV, Gelehrter TD. cDNA cloning of hu-               thelial cells. J Biol Chem 264: 5058–5063, 1989.
      man plasminogen activator-inhibitor from endothelial cells. J         43.   Wei Y, Waltz DA, Rao N, Drummond RJ, Rosenberg S, Chapman
      Clin Invest 78: 1673-1680, 1986.                                            HA. Identification of the urokinase receptor as an adhesion re-
22.   Huber R, Carrell RW. Implications of the three-dimensional                  ceptor for vitronectin. J Biol Chem 269: 32380–32388, 1994.
      structure of alpha 1-antitrypsin for structure and function of        44.   Okumura Y, Kamikubo Y, Curriden SA, Wang J, Kiwada T, Futaki
      serpins. Biochemistry 28: 8951-8966, 1989.                                  S, Kitagawa K, Loskutoff DJ. Kinetic analysis of the interaction
23.   Gils A, Declerck PJ. Structure-function relationships in serpins:           between vitronectin and the urokinase receptor. J Biol Chem
      current concepts and controversies Thromb Haemost 80: 531-                  277: 9395–9404, 2002
      541, 1998.                                                            45.   Deng G, Curriden SA, Wang S, Rosenberg S, Loskutoff DJ. Is
24.   Laskowski M Jr, Kato I. Protein inhibitors of proteinases. Annu             plasminogen activator inhibitor-1 the molecular switch that
      Rev Biochem 49: 593-626, 1980.                                              governs urokinase receptor-mediated cell adhesion and release?
25.   Egelund R, Rodenburg KW, Andreasen PA, Rasmussen MS,                        J Cell Biol 134: 1563-1571, 1996.
      Guldberg, RE, Petersen TE. An ester bond linking a fragment of        46.   Lawrence DA, Berkenpas MB, Palaniappan S, Ginsburg D. Lo-
      a serine proteinase to its serpin inhibitor. Biochemistry 37:               calization of vitronectin binding domain in plasminogen activa-
      6375-6379, 1998.                                                            tor inhibitor-1. J Biol Chem 269: 15223-15228, 1994.
26.   Lawrence DA, Ginsburg D, Day DE, Berkenpas MB, Verhamme               47.   Zhou A. How vitronectin binds PAI-1 to modulate fibrinolysis
      IM, Kvassman JO, Shore JD. Serpin-protease complexes are                    and cell migration. Nat Struct Biol 10: 541-544, 2003.
      trapped as stable acyl-enzyme intermediates. J Biol Chem 270:         48.   Ehrlich HJ, Gebbink RK, Keijer J, Linders M, Preissner KT, Pan-
      25309- 25312, 1995.                                                         nekoek H. Alteration of serpin specificity by a protein cofactor.
                                                                                  Vitronectin endows plasminogen activator inhibitor 1 with

 2005 MedUnion Press                                                                                                                           33
Lee et al. J. Cancer Mol. 1(1): 25-36, 2005

      thrombin inhibitory properties. J Biol Chem 265: 13029–13035,              tor and tissue-type plasminogen activator gene expression in
      1990.                                                                      HTC rat hepatoma cells by glucocorticoids and cyclic nucleo-
49.   Preissner KT, Koyama T, Muller D, Tschopp J, Muller-Berghaus               tides. Mol Endocrinol 6: 53-60, 1992.
      G. Domain structure of the endothelial cell receptor thrombo-        69.   Heaton JH, Tillmann-Bogush M, Leff NS, Gelehrter TD. Cyclic
      modulin as deduced from modulation of its anticoagulant func-              nucleotide regulation of type-1 plasminogen activator-inhibitor
      tions. Evidence for a glycosaminoglycan-dependent secondary                mRNA stability in rat hepatoma cells. Identification of cis-acting
      binding site for thrombin. J Biol Chem 265: 4915-4922, 1990.               sequences. J Biol Chem 273: 14261-14268, 1998.
50.   Ehrlich HJ, Keijer J, Preissner KT, Gebbink RK, Pannekoek H.         70.   Heaton JH, Dlakic WM, Dlakic M, Gelehrter TD. Identification and
      Functional interaction of plasminogen activator inhibitor type 1           cDNA cloning of a novel RNA-binding protein that interacts with
      (PAI-1) and heparin. Biochemistry 30:1021-1028, 1991.                      the cyclic nucleotide-responsive sequence in the type-1 plasmi-
51.   Lawrence DA, Palaniappan S, Stefansson S, Olson ST, Francis-               nogen activator inhibitor mRNA. J Biol Chem 276: 3341-3347,
      Chmura AM, Shore JD, Ginsburg D. Characterization of the bind-             2001.
      ing of different conformational forms of plasminogen activator       71.   Zhang WJ, Wojta J, Binder BR. Notoginsenoside R1 counteracts
      inhibitor-1 to vitronectin. Implications for the regulation of             endotoxin-induced activation of endothelial cells in vitro and
      pericellular proteolysis. J Biol Chem 272: 7676-7680, 1997.                endotoxin-induced lethality in mice in vivo. Arterioscler Thromb
52.   Nykjaer A, Conese M, Christensen EI, Olson D, Cremona O,                   Vasc Biol 17: 465-474, 1997.
      Gliemann J, Blasi F. Recycling of the urokinase receptor upon        72.   Hou B, Eren M, Painter CA, Covington, JW, Dixon JD, Schoen-
      internalization of the uPA:serpin complexes. EMBO J 16: 2610-              hard JA, Vaughan DE. Tumor necrosis factor alpha activates the
      2620, 1997.                                                                human plasminogen activator inhibitor-1 gene through a distal
53.   Czekay RP, Kuemmel TA, Orlando RA, Farquhar MG. Direct                     nuclear factor kappaB site. J Biol Chem 279: 18127-18136, 2004.
      binding of occupied urokinase receptor (uPAR) to LDL receptor-       73.   Chen Y, Billadello JJ, Schneider DJ. Identification and localiza-
      related protein is required for endocytosis of uPAR and regula-            tion of a fatty acid response region in the human plasminogen
      tion of cell surface urokinase activity. Mol Biol Cell 12: 1467-           activator inhibitor-1 gene. Arterioscler Thromb Vasc Biol 20:
      1479, 2001.                                                                2696-2701, 2000.
54.   Horn IR, van den Berg BM, Moestrup SK, Pannekoek H, van              74.   Eriksson P, Nilsson L, Karpe F, Hamsten A. Very-low-density
      Zonneveld AJ. Plasminogen activator inhibitor 1 contains a                 lipoprotein response element in the promoter region of the hu-
      cryptic high affinity receptor binding site that is exposed upon           man plasminogen activator inhibitor-1 gene implicated in the
      complex formation with tissue-type plasminogen activator.                  impaired fibrinolysis of hypertriglyceridemia. Arterioscler
      Thromb Haemost 80: 822-828, 1998.                                          Thromb Vasc Biol 18: 20-26, 1998.
55.   Degryse B, Neels JG, Czekay RP, Aertgeerts K, Kamikubo Y,            75.   Panahloo A, Mohamed-Ali V, Lane A, Green F, Humphries SE,
      Loskutoff DJ. The low density lipoprotein receptor-related pro-            Yudkin JS. Determinants of plasminogen activator inhibitor 1
      tein is a motogenic receptor for plasminogen activator inhibitor-          activity in treated NIDDM and its relation to a polymorphism in
      1. J Biol Chem 279: 22595-22604, 2004.                                     the plasminogen activator inhibitor 1 gene. Diabetes 44: 37–42,
56.   Gill A, Pedersen KE, Skottrup P, Christensen A, Naessens D,                1995.
      Deinum J, Enghild JJ, Declerck PJ, Andreasen PA. Biochemical         76.   Mansfield MW, Stickland MH, Grant PJ. Plasminogen activator
      importance of glycosylation of plasminogen activator inhibitor-1.          inhibitor-1 (PAI-1) promoter polymorphism and coronary artery
      Thromb Haemost 90: 206-217, 2003.                                          disease in non-insulin-dependent diabetes. Thromb Haemost 74:
57.   Loskutoff DJ, Linders M, Keijer J, Veeman H, vanHeerikhuizen H,            1032–1034, 1995.
      Pannekoek H. Structure of the human plasminogen activator in-        77.   Ye S, Green FR, Scarabin PY, Nicaud V, Bara L, Dawson, SJ,
      hibitor 1 gene: nonrandom distribution of introns. Biochemistry            Humphries SE, Evans A, Luc G, Gambou JP. The 4G/5G genetic
      26: 3763-3768, 1987.                                                       polymorphism in the promoter of the plasminogen activator in-
58.   Cicila GT, O’Connell TM, Hahn WC, Rheinwald JG. Cloned cDNA                hibitor-1 (PAI-1) gene is associated with differences in plasma
      sequence for the human mesothelial protein 'mesosecrin' dis-               PAI-1 activity but not with risk of myocardial infarction in the
      closes its identity as a plasminogen activator inhibitor (PAI-1)           ECTIM study. Etude CasTemoins de I'nfarctus du Mycocarde.
      and a recent evolutionary change in transcript processing. J               Thromb Haemost 74: 837–841, 1995.
      Cell Sci 94 (Pt 1): 1-10, 1989.                                      78.   Eriksson P, Kallin B, van’t Hooft FM, Bavenholm P, Hamsten A.
59.   Bosma PJ, van den Berg EA, Kooistra T, Siemieniak DR, Sligh-               Allele-specific increase in basal transcription of the plasmino-
      tom JL. Human plasminogen activator inhibitor-1 gene. Pro-                 gen-activator inhibitor 1 gene is associated with myocardial in-
      moter and structural gene nucleotide sequences. J Biol Chem                farction. Proc Natl Acad Sci USA 92: 1851–1855, 1995.
      263: 9129-9141, 1988.                                                79.   Burzotta F, Di Castelnuovo A, Amore C, D’Orazio A, Di Bitondo
60.   Johnson MR, Bruzdzinski CJ, Winograd SS, Gelehrter TD. Regu-               R, Donati MB, Iacoviello L. 4G/5G promoter PAI-1 gene polymor-
      latory sequences and protein-binding sites involved in the ex-             phism is associated with plasmatic PAI-1 activity in Italians: a
      pression of the rat plasminogen activator inhibitor-1 gene. J Biol         model of gene-environment interaction. Thromb Haemost 79:
      Chem 267: 12202-12210, 1992.                                               354–358, 1998.
61.   Chen YQ, Su M, Walia RR, Hao Q, Covington JW, Vaughan DE.            80.   Margaglione M, Grandone E, Vecchione G, Cappucci G, Giuliani
      Sp1 sites mediate activation of the plasminogen activator inhibi-          N, Colaizzo D, Celentano E, Panico S, Di Minno G. Plasminogen
      tor-1 promoter by glucose in vascular smooth muscle cells. J               activator inhibitor-1 (PAI-1) antigen plasma levels in subjects at-
      Biol Chem 273: 8225–8231, 1998.                                            tending a metabolic ward: relation to polymorphisms of PAI-1
62.   Descheemaeker KA, Wyn S, Nelles L, Auwerx J, Ny T, Collen D.               and angiontensin converting enzyme (ACE) genes. Arterioscler
      Interaction of AP-1-, AP-2-, and Sp1-like proteins with two dis-           Thromb Vasc Biol 17: 2082–2087, 1997.
      tinct sites in the upstream regulatory region of the plasminogen     81.   Grubic N, Stegnar M, Peternel P, Kaider A, Binder BR. A novel
      activator inhibitor-1 gene mediates the phorbol 12-myristate 13-           G/A and the 4G/5G polymorphism within the promoter of the
      acetate response. J Biol Chem 267: 15086-15091, 1992.                      plasminogen activator inhibitor-1 gene in patients with deep
63.   Kunz C, Pebler S, Otte J, von der Ahe D. Differential regulation           vein thrombosis Thromb Res 84: 431-443, 1996.
      of plasminogen activator and inhibitor gene transcription by the     82.   Marx N, Bourcier T, Sukhova GK, Libby P, Plutzky J. PPARγ
      tumor suppressor p53. Nucleic Acids Res 23: 3710-3717, 1995.               activation in human endothelial cells increases plasminogen ac-
64.   Dennler S, Itoh S, Vivien D, ten Dijke P, Huet S, Gauthier JM.             tivator inhibitor type-1 expression: PPARgamma as a potential
      Direct binding of Smad3 and Smad4 to critical TGF beta-                    mediator in vascular disease. Arterioscler Thromb Vasc Biol 19:
      inducible elements in the promoter of human plasminogen acti-              546-551, 1999.
      vator inhibitor-type 1 gene. EMBO J 17: 3091–3100, 1998.             83.   Nordt TK, Klassen KJ, Schneider DJ, Sobel BE. Augmentation of
65.   Fink T, Kazlauskas A, Poellinger L, Ebbesen P, Zachar V. Identi-           synthesis of plasminogen activator inhibitor type-1 in arterial
      fication of a tightly regulated hypoxia-response element in the            endothelial cells by glucose and its implications for local fibri-
      promoter of human plasminogen activator inhibitor-1. Blood 99:             nolysis. Arterioscler Thromb 13: 1822-1828, 1993.
      2077-2083, 2002.                                                     84.   Kooistra T, Bosma PJ, Tons HA, van den Berg AP, Meyer P,
66.   Chen CY, Xu N, Shyu AB. mRNA decay mediated by two distinct                Princen HM. Plasminogen activator inhibitor 1: biosynthesis and
      AU-rich elements from c-fos and granulocyte-macrophage col-                mRNA level are increased by insulin in cultured human hepato-
      ony-stimulating factor transcripts: different deadenylation kinet-         cytes. Thromb Haemost 62:723-728, 1989.
      ics and uncoupling from translation. Mol Cell Biol 15: 5777-5788,    85.   Rikitake Y, Liao JK. Rho-kinase mediates hyperglycemia-
      1995.                                                                      induced plasminogen activator inhibitor-1 expression in vascu-
67.   van den Berg EA, Sprengers ED, Jaye M, Burgess W, Maciag T,                lar endothelial cells. Circulation 111: 3261-3268, 2005.
      van Hinsbergh VW. Regulation of plasminogen activator inhibi-        86.   Ryan MP, Higgins PJ. Growth state-regulated expression of p52
      tor-1 mRNA in human endothelial cells. Thromb Haemost 60: 63-              (PAI-1) in normal rat kidney cells. J Cell Physiol 155: 376–384,
      67, 1988.                                                                  1993.
68.   Heaton JH, Kathju S, Gelehrter TD. Transcriptional and post-         87.   Mu XC, Staiano-Coico L, Higgins PJ. Increased transcription
      transcriptional regulation of type 1 plasminogen activator inhibi-         and modified growth state-dependent expression of the plasmi-

34                                                                                                                         Print ISSN 1816-0735
                                                                             PAI-1’s Expression, Biological Function, and Effects on Cancer Development

       nogen activator inhibitor type-1 gene characterize the senescent       111. Kobayashi H, Moniwa N, Gotoh J, Sugimura M, Terao T. Role of
       phenotype. J Cell Physiol 174: 90–98, 1998.                                 activated protein C in facilitating basement membrane invasion
88.    Mayer MJ, Lund LR, Riccio A, Skouv J, Nielsen LS, Stacey SN,                by tumor cells. Cancer Res 54: 261-267, 1994.
       Dano K, Andreasen PA. Plasminogen activator inhibitor type-1           112. Praus M, Collen D, Gerard RD. Both u-PA inhibition and vi-
       protein, mRNA and gene transcription are increased by phorbol               tronectin binding by plasminogen activator inhibitor 1 regulate
       esters in human rhabdomyosarcoma cells. J Biol Chem 263:                    HT1080 fibrosarcoma cell metastasis. Int J Cancer 102: 584-591,
       15688-15693, 1988.                                                          2002.
89.    Bosma PJ, Kooistra T. Different induction of two plasminogen           113. Liu G, Shuman MA, Cohen RL. Co-expression of urokinase,
       activator inhibitor 1 mRNA species by phorbol ester in human                urokinase receptor and PAI-1 is necessary for optimum inva-
       hepatoma cells. J Biol Chem 266: 17845-17849, 1991.                         siveness of cultured lung cancer cells. Int J Cancer 60: 501-506,
90.    Pinsky DJ, Liao H, Lawson CA, Yan SF, Chen J, Carmeliet P,                  1995.
       Loskutoff DJ, Stern DM. Coordinated induction of plasminogen           114. Soff GA, Sanderowitz J, Gately S, Verrusio E, Weiss I, Berm S,
       activator inhibitor-1 (PAI-1) and inhibition of plasminogen acti-           Kwaan HC. Expression of plasminogen activator inhibitor type 1
       vator gene expression by hypoxia promotes pulmonary vascu-                  by human prostate carcinoma cells inhibits primary tumor
       lar fibrin deposition. J Clin Invest 102: 919-928, 1998.                    growth, tumor-associated angiogenesis, and metastasis to lung
91.    Lee CC, Shyu KG, Lin S, Wang BW, Liu YC, Chang H. Cell adhe-                and liver in an athymic mouse model. J Clin Inves 96: 2593-2600,
       sion regulates the plasminogen activator inhibitor-1 gene ex-               1995.
       pression in anchorage-dependent cells. Biochem Biophys Res             115. Duffy MJ, Reilly D, O’Sullivan C, O’Higgins N, Fennelly JJ, An-
       Commun 291: 185-190, 2002.                                                  dreasen P. Urokinase-plasminogen activator, a new and inde-
92.    Chang H, Shyu KG, Lin S, Wang BW, Liu YC, Lee CC. Cell adhe-                pendent prognostic marker in breast cancer. Cancer Res 50:
       sion induces the plasminogen activator inhibitor-1 gene expres-             6827-6829, 1990.
       sion through phosphatidylinositol 3-kinase/Akt activation in an-       116. Janicke F, Schmitt M, Ulm K, Gossner W, Graeff H. Urokinase-
       chorage dependent cells. Cell Commun Adhes 9: 239-247, 2002.                type plasminogen activator antigen and early relapse in breast
93.    Chang H, Shyu KG, Lin S, Tsai SC, Wang BW, Liu YC, Sung YL,                 cancer. Lancet 2: 1049. 1989.
       Lee CC. The plasminogen activator inhibitor-1 gene is induced          117. Shapiro RL, Duquette JG, Roses DF, Nunes I, Harris MN,
       by cell adhesion through the MEK/ERK pathway. J Biomed Sci                  Kamino H, Wilson EL, Rifkin DB. Induction of primary cutaneous
       10: 738-745, 2003.                                                          melanocytic neoplasms in urokinase-type plasminogen activa-
94.    Hapke S, Kessler H, Arroyo dePrada N, Benge A, Schmitt M,                   tor (uPA)-deficient and wild-type mice: cellular blue nevi invade
       Lengyel E, Reuning U. Integrin alpha(v)beta(3)/vitronectin inter-           but do not progress to malignant melanoma in uPA-deficient
       action affects expression of the urokinase system in human                  animals. Cancer Res 56: 3597-3604, 1996.
       ovarian cancer cells. J Biol Chem 276: 26340-26348, 2001.              118. Sabapathy KT, Pepper MS, Kiefer F, Mohle-Steinlein U, Tacchini-
95.    Ciambrone GJ, McKeown-Longo PJ. Plasminogen activator                       Cottier F, Fetka I, Breier G, Risau W, Carmeliet P, Montesano R,
       inhibitor type I stabilizes vitronectin-dependent adhesions in              Wagner EF. Polyoma middle T-induced vascular tumor forma-
       HT-1080 cells. J Cell Biol 111: 2183-2195, 1990.                            tion: the role of the plasminogen activator/plasmin system. J
96.    Palmieri D, Lee JW, Juliano RL, Church FC. Plasminogen activa-              Cell Biol 137: 953-963, 1997.
       tor inhibitor-1 and -3 increase cell adhesion and motility of MDA-     119. Bugge TH, Kombrinck KW, Xiao Q, Holmback K, Daugherty CC,
       MB-435 breast cancer cells. J Biol Chem 277: 40950-40957, 2002.             Witte DP, Degen JL. Growth and dissemination of Lewis lung
97.    Isogai C, Laug WE, Shimada H, Declerck PJ, Stins MF, Duredn                 carcinoma in plasminogen-deficient mice. Blood 90: 4522-4531,
       DL, Erdreich-Epstein A, DecClerck YA. Plasminogen activator                 1997.
       inhibitor-1 promotes angiogenesis by stimulating endothelial           120. Bugge TH, Lund LR, Kombrinck KK, Nielsen BS, Holmback K,
       cell migration toward fibronectin. Cancer Res 61: 5587-5594,                Drew AF, Flick MJ, Witte DP, Dano K, Degen JL. Reduced metas-
       2001.                                                                       tasis of Polyoma virus middle T antigen-induced mammary can-
98.    Stahl A, Mueller BM. Melanoma cell migration on vitronectin:                cer in plasminogen-deficient mice. Oncogene 16: 3097-3104,
       regulation by components of the plasminogen activation system.              1998.
       Int J Cancer 71: 116-122, 1997.                                        121. Almholt K, Lund LR, Rygaard J, Nielsen BS, Dano K, Romer J,
99.    Loskutoff DJ, Curriden SA, Hu G, Deng G. Regulation of cell                 Johnsen M. Reduced metastasis of transgenic mammary cancer
       adhesion by PAI-1. APMIS 107: 54-61, 1999.                                  in urokinase-deficient mice. Int J Cancer 113: 525-532, 2005
100.   Czekay RP, Aertgeerts K, Curriden SA, Loskutoff DJ. Plasmino-          122. Andreasen PA, Egelund R, Petersen HH. The plasminogen acti-
       gen activator inhibitor-1 detaches cells from extracellular matri-          vation system in tumor growth, invasion, and metastasis. Cell
       ces by inactivating integrins. J Cell Biol 160: 781-791, 2003.              Mol Life Sci 57: 25-40, 2000.
101.   Andreasen PA, Kjoller L, Christensen L, Duffy ML. The                  123. Knoop A, Andreasen PA, Andersen JA, Hansen S, Laenkholm
       urokinase-type plasminogen activator system in cancer metas-                AV, Simonsen AC, Andersen J, Overgaard J, Rose C. Prognostic
       tasis: a review. Int J Cancer 72: 1-22, 1997.                               significance of urokinase-type plasminogen activator and plas-
102.   Stefansson S, Lawrence DA. The serpin PAI-1 inhibits cell mi-               minogen activator inhibitor-1 in primary breast cancer. Br J
       gration by blocking integrin alpha V beta 3 binding to vitronectin.         Cancer 77: 932-40, 1998.
       Nature 383: 441-443, 1996.                                             124. Harbeck N, Kates RE, Gauger K, Willems A, Kiechle M, Magdo-
103.   Kjoller L, Kanse SM, Kirkegaard T, Rodenburg KW, Ronne E,                   len V, Schmitt M. Urokinase-type plasminogen activator (uPA)
       Goodman SL, Preissner KT, Ossowski L, Andreasen PA. Plas-                   and its inhibitor PAI-I: novel tumor-derived factors with a high
       minogen activator inhibitor-1 represses integrin- and vi-                   prognostic and predictive impact in breast cancer. Thromb
       tronectin-mediated cell migration independently of its function             Haemost 91: 450-456, 2004.
       as an inhibitor of plasminogen activation. Exp Cell Res 232: 420-      125. Duffy MJ. Urokinase plasminogen activator and its inhibitor,
       429, 1997.                                                                  PAI-1, as prognostic markers in breast cancer: from pilot to
104.   Sugiura Y, Ma L, Sun B, Shimada H, Laug WE, Seeger RC De-                   level 1 evidence studies. Clin Chem 48: 1194-1197, 2002.
       Clerck YA. The plasminogen-plasminogen activator (PA) system           126. Duffy MJ. Proteases as prognostic markers in cancer. Clin Can-
       in neuroblastoma: role of PA inhibitor-1 in metastasis. Cancer              cer Res 2: 613-618, 1996.
       Res 59: 1327-1336, 1999.                                               127. Ma D, Gerard RD, Li XY, Alizadeh H, Niederkorn JY. Inhibition of
105.   Tanaka S, Koyama H, Ichii T, Shioi A, Hosoi M, Raines EW, Ni-               metastasis of intraocular melanomas by adenovirus-mediated
       shizawa Y. Fibrillar collagen regulation of plasminogen activator           gene transfer of plasminogen activator inhibitor type 1 (PAI-1) in
       inhibitor-1 is involved in altered smooth muscle cell migration.            an athymic mouse model. Blood 90: 2738-2746, 1997.
       Arterioscler Thromb Vasc Biol 22: 1573-1578, 2002.                     128. Jankun J, Keck RW, Skrzypczak-Jankun E, Swiercz R. Inhibitors
106.   Chapman HA, Wei Y. Protease crosstalk with integrins: the                   of urokinase reduce size of prostate cancer xenografts in severe
       urokinase receptor paradigm. Thromb Haemost 86: 124-129,                    combined immunodeficient mice. Cancer Res 57: 559-563, 1997
       2001                                                                   129. McMahon GA, Petitclerc E, Stefansson S, Smith E, Wong MK,
107.   Tarui TJ, Mazar AP, Cines DB, Takada Y. Urokinase-type plas-                Westrick RJ, Ginsburg D, Brooks PC, Lawrence DA. Plasmino-
       minogen activator receptor (CD87) is a ligand for integrins and             gen activator inhibitor-1 regulates tumor growth and angio-
       mediates cell-cell interaction. J Biol Chem 276: 3983-3990, 2001.           genesis. J Biol Chem 276: 33964-33968, 2001.
108.   Kjoller L. The urokinase plasminogen activator receptor in the         130. Gutierrez LS, Schulman A, Brito-Robinson T, Noria F, Ploplis VA,
       regulation of the actin cytoskeleton and cell motility. J Biol              Castellino FJ. Tumor development is retarded in mice lacking
       Chem 383: 5-19, 2002                                                        the gene for urokinase-type plasminogen activator or its inhibi-
109.   Stefansson S, Lawrence DA. Old dogs and new tricks: prote-                  tor, plasminogen activator inhibitor-1. Cancer Res 60: 5839-5847,
       ases, inhibitors, and cell migration. Sci STKE. 2003: pe24, 2003.           2000.
110.   Bruckner A, Filderman AE, Kirchheimer JC, Binder BR, Remold            131. Bajou K, Noel A, Gerard RD, Masson V, Brunner N, Holst-
       HG. Endogenous receptor-bound urokinase mediates tissue in-                 Hansen C, Skobe M, Fusenig NE, Carmeliet P, Collen D, Foidart
       vasion of the human lung carcinoma cell lines A549 and Calu-1.              JM. Absence of host plasminogen activator inhibitor 1 prevents
       Cancer Res 52: 3043-3047, 1992.                                             cancer invasion and vascularization. Nat Med 4: 923-928, 1998

 2005 MedUnion Press                                                                                                                             35
Lee et al. J. Cancer Mol. 1(1): 25-36, 2005

132. Bajou K, Masson V, Gerard RD, Schmitt PM, Albert V, Praus M,                     broblasts in human colon cancer express plasminogen activator
     Lund LR, Frandsen TL, Brunner N, Dano K, Fusenig NE, Weidle                      inhibitor-1. Am J Clin Pathol 122: 256-265, 2004.
     U, Carmeliet G, Loskutoff D, Collen D, Carmeliet P, Foidart JM.           145.   Hansen S, Overgaard J, Rose C, Knoop A, Laenkholm AV, An-
     The plasminogen activator inhibitor PAI-1 controls in vivo tumor                 dersen J, Sorensen FB, Andreasen PA. Independent prognostic
     vascularization by interaction with proteases, not vitronectin.                  value of angiogenesis and the level of plasminogen activator
     Implications for antiangiogenic strategies. J Cell Biol 152: 777-                inhibitor type 1 in breast cancer patients. Br J Cancer 88: 102-
     784, 2001.                                                                       108, 2003.
133. Stefansson S, Petitclerc E, Wong MK, McMahon GA, Brooks PC,               146.   Kwaan HC, Wang J, Svoboda K, Declerck PJ. Plasminogen
     Lawrence DA. Inhibition of angiogenesis in vivo by plasminogen                   activator inhibitor 1 may promote tumour growth through inhibi-
     activator inhibitor-1. J Biol Chem 276: 8135-8141, 2001.                         tion of apoptosis. Br J Cancer 82: 1702-1708, 2000.
134. Devy L, Blacher S, Grignet-Debrus C, Bajou K, Masson V,                   147.   Lademann U, Romer MU, Jensen PB, Hofland KF, Larsen L,
     Gerard RD, Gils A, Carmeliet G, Carmeliet P, Declerck PJ, Noel A,                Christensen IJ, Brunner N. Malignant transformation of wild-
     Foidart JM. The pro- or antiangiogenic effect of plasminogen ac-                 type but not plasminogen activator inhibitor-1 gene-deficient fi-
     tivator inhibitor 1 is dose dependent. FASEB J 16: 147-154, 2002                 broblasts decreases cellular sensitivity to chemotherapy-
135. Lambert V, Munaut C, Carmeliet P, Gerard RD, Declerck PJ, Gils                   mediated apoptosis. Eur J Cancer 41: 1095-100, 2005.
     A, Claes C, Foidart JM, Noel A, Rakic JM. Dose-dependent                  148.   De Taeye B, Gils A, Declerck PJ. The story of the serpin plasmi-
     modulation of choroidal neovascularization by plasminogen ac-                    nogen activator inhibitor 1: is there any need for another mutant?
     tivator inhibitor type I: implications for clinical trials. Invest Oph-          Thromb Haemost 92: 898-924, 2004.
     thalmol Vis Sci 44: 2791-2797, 2003.                                      149.   Hjortland GO, Bjornland K, Pettersen S, Garman-Vik SS, Emil-
136. Bajou K, Maillard C, Jost M, Lijnen RH, Gils A, Declerck P, Car-                 sen E, Nesland JM, Fodstad O, Engebraaten O. Modulation of
     meliet P, Foidart JM, Noel A. Host-derived plasminogen activa-                   glioma cell invasion and motility by adenoviral gene transfer of
     tor inhibitor-1 (PAI-1) concentration is critical for in vivo tumoral            PAI-1. Clin Exp Metastasis 20: 301-309, 2003.
     angiogenesis and growth. Oncogene 23: 6986-6990, 2004.                    150.   Praus M, Wauterickx K, Collen D, Gerard RD. Reduction of tu-
137. Eitzman DT, Krauss JC, Shen T, Cui J, Ginsburg D. Lack of                        mor cell migration and metastasis by adenoviral gene transfer
     plasminogen activator inhibitor-1 effect in a transgenic mouse                   of plasminogen activator inhibitors. Gene Ther 6: 227-236, 1999.
     model of metastatic melanoma. Blood 87: 4718-4722, 1996.                  151.   Tsuchiya H, Katsuo S, Matsuda E, Sunayama C, Tomita K, Ueda
138. Curino A, Mitola DJ, Aaronson H, McMahon GA, Raja K, Keegan                      Y, Binder BR. The antibody to plasminogen activator inhibitor-1
     AD, Lawrence DA, Bugge TH. Plasminogen promotes sarcoma                          suppresses pulmonary metastases of human fibrosarcoma in
     growth and suppresses the accumulation of tumor-infiltrating                     athymic mice. Gen Diagn Pathol 141: 41–48, 1995.
     macrophages. Oncogene 21: 8830-8842, 2002                                 152.   Brooks TD, Slomp J, Quax PH, De Bart AC, Spencer MT, Verhei-
139. Guy CT, Cardiff RD, Muller WJ. Induction of mammary tumors                       jen JH, Charlton PA. Antibodies to PAI-1 alter the invasive and
     by expression of polyomavirus middle T oncogene: a transgenic                    migratory properties of human tumour cells in vitro. Clin Exp
     mouse model for metastatic disease. Mol Cell Biol 12: 954-961,                   Metastasis 18: 445-453, 2000.
     1992.                                                                     153.   Ngo TH, Hoylaerts MF, Knockaert I, Brouwers E, Declerck PJ.
140. Almholt K, Nielsen BS, Frandsen TL, Brunner N, Dano K, John-                     Identification of a target site in plasminogen activator inhibitor-1
     sen M. Metastasis of transgenic breast cancer in plasminogen                     that allows neutralization of its inhibitor properties concomitant
     activator inhibitor-1 gene-deficient mice. Oncogene 22: 4389-                    with an allosteric up-regulation of its antiadhesive properties. J
     4397, 2003.                                                                      Biol Chem 276: 26243-26248, 2001.
141. Maillard C, Jost M, Romer MU, Brunner N, Houard X, Lejeune A,             154.   Liang A, Wu F, Tran K, Jones SW, Deng G, Ye B, Zhao Z, Snider
     Munaut C, Bajou K, Melen L, Dano K, Carmeliet P, Fusenig NE,                     RM, Dole WP, Morser J, Wu Q. Characterization of a small mole-
     Foidart JM, Noel A. Host plasminogen activator inhibitor-1 pro-                  cule PAI-1 inhibitor, ZK4044. Thromb Res 115: 341-350, 2005.
     motes human skin carcinoma progression in a stage-dependent               155.   Crandall DL, Elokdah H, Di L, Hennan JK, Gorlatova NV, Law-
     manner. Neoplasia 7: 57-66, 2005.                                                rence DA. Characterization and comparative evaluation of a
142. Pepper MS, Belin D, Montesano R, Orci L, Vassalli JD. Trans-                     structurally unique PAI-1 inhibitor exhibiting oral in-vivo effi-
     forming growth factor-beta 1 modulates basic fibroblast growth                   cacy. J Thromb Haemost 2: 1422-1428, 2004.
     factor-induced proteolytic and angiogenic properties of endo-             156.   Folkes A, Brown SD, Canne LE, Chan J, Engelhardt E, Epshteyn
     thelial cells in vitro. J Cell Biol 111: 743-755, 1990.                          S, Faint R, Golec J, Hanel A, Kearney P, Leahy JW, Mac M, Mat-
143. Offersen BV, Nielsen BS, Hoyer-Hansen G, Rank F, Hamilton-                       thews D, Prisbylla MP, Sanderson J, Simon RJ, Tesfai Z, Vicker
     Dutoit S, Overgaard J, Andreasen PA. The myofibroblast is the                    N, Wang S, Webb RR, Charlton P. Design, synthesis and in vitro
     predominant plasminogen activator inhibitor-1-expressing cell                    evaluation of potent, novel, small molecule inhibitors of plasmi-
     type in human breast carcinomas. Am J Pathol 163: 1887-1899,                     nogen activator inhibitor-1. Bioorg Med Chem Lett 12: 1063-1066,
     2003.                                                                            2002.
144. Illemann M, Hansen U, Nielsen HJ, Andreasen PA, Hoyer-
     Hansen G, Lund LR, Dano K, Nielsen BS. Leading-edge myofi-

36                                                                                                                              Print ISSN 1816-0735