plasmin system in transgenic mice

Evaluation of the PlasminogenIplasmin

P. Carmeliet,

System in Transgenic Mice

D. Collen

SUMMARY: Indirect evidence suggests a crucial role for the fibrinolytic system, and its physiological triggers tissue-type plasminogen activator and urokinase-type plasminogen activator in many proteolytic processes, including blood clot dissolution (thrombolysis), thrombosis, hemostasis, atherosclerosis, neointima formation/restenosis, reproduction, embryo implantation, embryogenesis, wound healing, malignancy and brain function. The implied role of the fibrinolytic system in viva is, however, deduced from correlations between fibrinolytic activity and (patho)physiological phenomena, which does not allow to definitively establish a causal role of this sytem in these processes. Recently, several transgenic mice, over- or under-expressing fibrinolytic system components, have been generated. This article reviews briefly the physiological consequences of gain or loss of function of these ftbrinolytic system components on thrombolysis, thrombosis, hemostasis, neointima formation, cell invasion, brain function and the associated effects on reproduction, development, health and survival.

prekallikrein and possibly u-PA)~ as well as plasminogen-independent mechanisms5x6 have been proposed to contribute to clot lysis, but their role in vivo remains to be established. Deficient fibrinolytic activity, e.g. resulting from increased plasma PAI- levels or reduced plasma t-PA or plasminogen levels might participate in the development of thrombotic events. Elevated plasma PAI- 1 levels have indeed been correlated with a higher risk of deep venous thrombosis’ and of thrombosis during sepsis,s surgery and trauma.” PAI- 1 plasma levels were also elevated in patients with coronary artery disease,‘a angina pectoris” and recurrent myocardial infarction.12 Although a possible causal relationship between high PAI- levels and the occurrence of thrombosis may be inferred from observations that PAI- specific antibodies enhance endogenous thrombolysis and reduce thrombus extension in rabbits,‘” the acute phase reactant behaviour of PAI- I4 does not always allow to deduce whether increased PAI- 1 levels are cause or consequence of thrombosis. To date, genetic deficiencies of t-PA or u-PA have not been reported in man but quantitative and qualitative deficiencies of plasminogen have been associated with increased thrombotic tendency.‘”

THROMBOSIVTHROMBOLYSIS Mammalian blood contains an enzymatic system, called the fibrinolytic or the plasminogen/plasmin system, which comprises an inactive proenzyme, plasminogen, that is activated to the proteolytic enzyme plasmin by two physiological plasminogen activators, tissue-type plasminogen activator (t-PA) and urokinase-type plasminogen activator (u-PA). ‘,2 Inhibition of the fibrinolytic system may occur at the level of plasmin, mainly by a,-antiplasmin or at the level of the plasminogen activators by specific plasminogen activator inhibitors (PAI’s) of which PAI- appears to be the principcal inhibitor.3 t-PA is believed to be primarily responsible for removal of fibrin from the vascular tree; it has a specific affinity for fibrin and produces clot-restricted plasminogen activation.2 The role of u-PA in thrombolysis is less well defined. It lacks affinity for fibrin and requires conversion from a single-chain precursor to a catalytically active two-chain derivative.2 Whether, to what extent and under which conditions u-PA might participate in fibrin clot dissolution in viva remained to be identified. Other plasminogen activation pathways, such as the ‘intrinsic’ pathway (involving blood coagulation factor XII, high molecular weight kininogen,

Spontaneous P. Carmeliet, D. Collen, Center for Molecular and Vascular Biology, Campus Gasthuisberg, Herestraat 49, University of Leuven, Leuven, B-3000, Belgium. Tel: 32-16-34.57.72. Fax: 32-16-34.59.90. Correspondence to D. Collen.

Thrombosis

Mice with single deficiency of t-PA and of u-PA were generated via homologous recombination in embryonic stem cells.“,” Targeting of the t-PA genes was accom269

270

Evaluation of the Plasminogerdplasmin

System in Transgenic Mice

plished by deleting the genomic sequences encoding most of the kringle 2 domain and part of the proteinase domain, including the catalytically essential HisX2” residue. The u-PA genes were targeted by deleting all but 23 amino acids of the coding sequences. Microscopical analysis of tissues from u-PA deficient mice revealed occasionally minor fibrin deposits in liver and intestines. No spontaneous fibrin deposits were observed in t-PA deficient mice. Microscopical analysis of mice with a combined deficiency of t-PA and u-PA, which were generated by interbreeding mice with single deficiencies, revealed extensive fibrin deposits in liver, intestines, lung, gonads, kidney and intestinal adhesions. In addition, ischemic necrosis of uterus and intestines was observed, possibly resulting from thrombotic occlusions. Overexpression of the principal plasminogen activator inhibitor, plasminogen activator inhbitor- 1 (PAI- l), in transgenic mice resulted in swollen hindlegs and necrotic amputation of the tail shortly after birth.‘s Microscopical analysis revealed cellular, fibrin-platelet rich venous occlusions in tail and hindlegs. Thrombosis

Associated

with Inflammation

u-PA deficient mice and combined t-PA:u-PA deficient mice developed rectal prolapse of non-infectious origin and chronic non-healing ulcerations at the ears, skin and mouth, possibly resulting from minor scratch and feeding trauma. Microscopical analysis revealed excessive fibrin deposits in these inflamed lesions.‘h,‘7 Although mice with a single deficiency of t-PA or uPA developed no or only a subtle spontaneous thrombotic phenotype, respectively, they were significantly more susceptible to development of venous thrombosis following local injection of proinflammatory endotoxin in the footpad. Indeed, a half-maximal dose of endotoxin induced more widespread venous thrombosis in tPA deficient and in u-PA deficient mice than in their respective wild type littermates.16 On the contrary, mice with a deficiency of PAI-1, which were generated by deleting all coding sequences of the PAZ-1 genes via homologous recombination in embryonic stem cells, were virtually protected against development of venous thrombosis following a similar dose of endotoxin.‘9*20 Plasma Clot Lysis The role of t-PA and u-PA in spontaneous clot lysis was tested by determining the ex viva lysis of ‘251-fibrin labelled plasma clots, injected via the jugular vein and embolized into the pulmonary arteries.16 No difference in the rate of spontaneous plasma clot lysis was observed between wild-type and u-PA deficient mice. In contrast, t-PA deficient mice lysed pulmonary plasma clots at a significantly reduced rate but pulmonary clot lysis was most prominently reduced in combined tPA:u-PA deficient mice. Pulmonary plasma clot lysis remained significantly lower in combined t-PA:u-PA deficient mice than in t-PA deficient mice for up to 48

hours, but was not significantly different after 72 hours. On the other hand, PAI- deficient mice lysed plasma clots at a significantly higher rate than wild-type mice. This difference became even more significant after intraperitoneal injection of endotoxin, which elevated plasma PAI- 1 levels.20 Macrophage-associated

Fibrin Degradation

u-PA binds to a specific cell-surface receptor2’ and is believed to be primarily involved in cell-mediated proteolysis either via direct proteolytic activity of plasmin or via activation of latent matrix-degrading proteinases.22 Since mono- and polymorphonuclear blood cells might aid in dissolving blood clots via such mechanism, thioglycollate-activated macrophages (which are known to express 10 to lOO-fold higher levels of membrane associated u-PA)~~ were tested for their ability to lyse a ‘251-labelled fibrin matrix. Macrophages from u-PA deficient mice but not from t-PA deficient or PAI- 1 deficient mice indeed lacked plasminogen-dependent matrix breakdown,‘6*‘7 possibly explaining the mild spontaneous and significant inflammation-induced thrombotic phenotype observed in u-PA deficient mice. Further support for a possible role of cell-associated clot lysis in vivo was suggested by the observations that transgenic mice, overexpressing the hematopoietic granulocyte-macrophage colony-stimulating factor (GM-CSF) frequently suffered spontaneous peritoneal bleeding, possibly due to increased expression of u-PA by accumulated macrophages.24 Collectively, analysis of transgenic mice with over- or underexpression of fibrinolytic enzymes has provided the following insights in their role in thrombosis and thrombolysis. (1) The impaired thrombolytic potential of t-PA deficient mice, in combination with their increased susceptibility to endotoxin-induced thrombosis, confirmed and established its important role in maintaining vascular patency. Surprisingly, however, tPA deficient mice did not develop spontaneous thrombosis, suggesting that other plasminogen dependent or independent proteinases might compensate for the loss of t-PA mediated thrombolysis. (2) The apparently normal ability of u-PA deficient mice to lyse ‘251-labelled pulmonary plasma clots might seem to contradict their minor spontaneous and signficant endotoxin-induced thrombotic phenotype. The absent plasminogen-dependent ‘251-labelled fibrin breakdown by activated macrophages suggested, however, that u-PA dependent cellular fibrinolysis might play a significant role in maintaining clot surveillance in basal and especially in inflamed conditions. (3) Despite circumstantial evidence that increased PAI- 1 levels contribute to thrombosis, it remained to be identified whether acute phase reactant PAI- 1 was causally involved. The observations that PAI- 1 overexpressing animals developed spontaneous thrombosis and that PAI- 1 deficient animals were resistant to endotoxin-induced thrombosis, in combination with an enhanced rate of thrombolysis, indicated that PAI- 1 plays a causative role in venous thrombosis.

Fibrinolysis

(4) The significantly more severe thrombotic phenotype in the combined t-PA:u-PA deficient mice than in the single mutants indicated that both plasminogen activators cooperate in a variety of biological phenomena.5 The lack of an even more severe thrombotic phenotype in the combined t-PA:u-PA deficient mice suggested, however, that yet other plasminogen activator-dependent or -independent mechanisms might be involved in normal fibrin clot surveillance. Analysis of recently generated plasminogen deficient mice (unpublished data) might aid in resolving this issue.

HEMOSTASIS Hemostasis involves platelet deposition and coagulation factor-mediated fibrin deposition to stabilize the clot. Failure to stabilize the clot, e.g. as a result of hyperfibrinolytic activity might result in delayed rebleeding.15 A hemorrhagic tendency has indeed been observed in patients with absent or reduced plasma PAI- 1 or alpha,antiplasmin activity levels and increased plasma t-PA levels.3,‘5.25 Delayed rebleeding might also explain the hemorrhagic tendency in several transgenic mice. Overexpression of u-PA in the liver of transgenic mice resulted in significantly elevated plasma u-PA activity levels, hypofibrinogenemia and severe bleeding in the gastro-intestinal tract during the first 4 days after birth.26 Interestingly, the timing of the bleeding (not immediately associated with the birth trauma but shortly after birth) suggested delayed rebleeding due to premature dissolution of the fibrin clot as a result of hyperfibrinolytic activity. Hyperfibrinolytic activity was, however, not the sole determinant of bleeding in these transgenic mice since high levels of u-PA activity in the plasma of both hemorrhaging and nonhemorrhaging transgenic mice were observed. GM-CSF overexpressing transgenic mice developed intraperitoneal bleeding, possibly resulting from locally increased u-PA levels on accumulated macrophages in the peritoneal cavity.24 Furthermore, inactivation of the genes encoding the low density lipoprotein receptor-related protein (LRP) resulted in embryonic death with associated gastrointestinal bleeding. *’ Whether increased proteolytic activity resulting from deficient clearing of t-PA and u-PA via the LRP might explain this bleeding phenotype remains, however, to be further defined. Contrary to patients with low or absent plasma PAI- 1 levels, PAI- deficient mice did not reveal spontaneous or delayed rebleeding, even when a segment of the tail or a partial amputation of the caecum was performed. This suggested that proteinase inhibitor control of fibrinolysis is different between mice and man (as suggested by the lower plasma PAI- levels in mice) and might explain the mild hyperfibrinolytic phenotype in PAIdeficient mice (as evidenced by the minor increases in spontaneous pulmonary plasma clot lysis in vivo and in lysis of diluted whole blood clots in vitro).*” In aggregate, excessive systemic or local production of u-PA dependent fibrinolytic activity may lead to

27

I

bleeding and hypofibrinogenemia in mice, similar to what has been documented in man following thrombolytic therapy with u-PA. The lack of rebleeding in PAI- deficient mice is suggestive of a different proteinase inhibitor control of hemostasis in mice and in man.

NEOINTIMA FORMATION AND ATHEROSCLEROSIS Vascular reconstructions including coronary angioplasty, endarterectomy, bypass surgery, vascular stents and heart transplantation have become succesful and widely used treatments for patients with atherothrombotic disease. However, chronic restenosis in 30 to 50% of patients remains one of the major limitations of these procedures. *’ Besides elastic recoil, thrombosis and a neoadventitial reaction, smooth muscle cell accumulation and extracellular matrix deposition in the neointima with subsequent narrowing of the lumen and tissue ischemia, frequently characterize vascular wound healing following such traumatic procedures.29 Contrary to an almost quiescent state in a resting (uninjured) vessel, the fibrinolytic system becomes significantly activated following vascular trauma (Figure 1) as evidenced by expression of t-PA, u-PA and PAI- 1 in in,jured or activated smooth muscle cells, endothelial cells and platelets (Fig. 1).‘@s4 Such local changes in fibrinolytic activity could affect mural thrombosis (a possible modulator of restenosis by itself) or participate in passivation of denuded vessels. Expression of uPA/u-PAR in smooth muscle and endothelial cells following injury in vivo,“i’32 or following treatment with mitogens including bFGF, PDGF and thrombin in vitro,33.‘4 and analysis of the effect of fibrinolytic enzymes on cultured cells in vitro suggests, however, that the plasminogen/plasmin system may also participate in migration and proliferation of endothelial and smooth muscle cells. In effect, it has been proposed that t-PA might mediate, at least in part, the effect of PDGF on migration of smooth muscle cells.‘* Besides possible non-proteolytic effects of plasminogen activators on cell proliferation,36 proteolytic activation of latent TGFI3 and liberation of bFGF from the extracellular matrix, or activation of latent matrix-degrading proteinases by plasminogen activators might affect cell migration, proliferation and extracellular matrix remodeling (reviewed in 22). Support for the involvement in matrix remodeling is provided by observations that degradation of ‘H-proline labelled subendothelial matrix by activated macrophages (primarily reflecting collagen breakdown under the experimental conditions used) was indeed specifically impaired in u-PA deficient mice.‘6,‘7 The role of the fibrinolytic system in restenosis in vivo remains, however, undefined since administration of u-PA had no effect on restenosis” whereas the anti-fibrinolytic drug tranexamic acid reduced migration but enhances proliferation of smooth muscle cells in rat carotid artery.j2

272

Evaluation

of the Plasminogen/plasmin

System in Transgenic

EEL

IEL

+

t-PA

INJURY

1 EEL

EEL

IEL

/CELL PROLIFERATION NEOINTIMA -CELL MIGRATION 'MATRIX REMODELING

Figure Schematic model of the involvement of the fibrinolytic or plasminogen/plasmin system in a resting (uninjured) vessel (upper panel), its induction after injury (middle panel) and subsequent formation of a neointima (lower panel). In a resting, uninjured vessel, the fibrinolytic system is quiescent. Following injury, production of t-PA by smooth muscle cells is proposed to mediate their migration across the internal elastic membrane (3 I ,32,35); expression of u-PA and u-PA receptor (uPAR) by smooth muscle cells and endothelial cells might participate in their proliferation and migration, respectively (22.3 I ,34). PAI- 1, produced by migrating injured or TGF-I.3 primed smooth muscle and endothelial cells (22) or released from platelets in mural thrombi, might control excessive proteolysis. Analysis of neointima formation following mechanical or electrical injury in plasminogen activator deficient mice revealed reduced, normal and increased rate of neointima formation and neointimal cell accumulation in u-PA deficient, t-PA deficient and PAI- I deficient mice, respectively (unpublished observations), suggesting that u-PArather than t-PA-mediated plasmin proteolysis promotes vascular wound healing and development of restenosis. It remains to be defined whether cell proliferation, migration or extracellular matrix remodeling are affected. SMC: smooth muscle cell; EC: endothelial cell; EEL: external elastica lamina; IEL: internal elastica lamina.

Mice

Although several methods are available to inflict vascular trauma and induce neointima formation in larger animals, the small size of the mouse presents a serious technical challenge for such analysis. We have developed a vascular injury model based on the use of a guidewire (mechanical trauma) or electrical current to induce neointima formation in mice (unpublished data). Preliminary analysis of neointima formation after vascular trauma in mice with deficiencies of t-PA, u-PA or PAI- suggests that deficiency of t-PA does not affect the degree or rate of neointima formation nor neointimal cell accumulation, whereas deficiency of u-PA significantly reduces and deficiency of PAI- significantly increases the rate of neointimal formation and neointima1 cell accumulations.‘* Interestingly, an inverse relationship between neointima and residual thrombosis was observed in these knock-out mice, suggesting that the role of the fibrinolytic system in neointima formation is probably more related to its effects on cell proliferation or migration than on clot lysis. These data suggest that u-PA- rather than t-PA-mediated plasmin proteolysis promotes vascular wound healing and development of restenosis and warrant further investigation of the involvement of the fibrinolytic system and its underlying mechanisms in neointima formation. Considering the relative refractoriness of restenosis to current therapies, balanced control of local antifibrinolytic activity (by a.o. gene transfer, antisense oligonucleotides or recombinant polypeptides) might constitute a possible alternative therapeutical avenue. The plasminogen/plasmin system might also be involved in the development and/or progression of atherosclerosis. Known risk factors for atherosclerosis including obesity, noninsulin-dependent diabetes, hyperinsulinemia and hypertriglycedemia have been correlated with increased plasma PAI- 1 levels (reviewed in 39). Furthermore, elevated PAI- levels constitute a risk factor for coronary atherosclerosis in survivors of myocardial infarction with glucose intolerance.40 At the molecular level, in situ identification of PAI- in plaque lesions suggests a role of the fibrinolytic system in this process.4’,42 Preliminary analysis suggests that t-PA deficient mice develop fatty streak lesions to the same extent as their wild type littermates (unpublished data). However, more complete evaluation whether, to what extent and at what stage the fibrinolytic system might affect atherosclerotic lesions, requires further study. This could be achieved by cross breeding of t-PA, u-PA and PAI- deficient mice with other atherosclerosis-prone transgenic mice such as the apo-lipoprotein E deficient mice. An interesting but unresolved issue is whether the atherothrombotic activity of the plasminogen homologue lipoprotein(a) might, at least in part, be attributed to an inhibitory role on plasmin formation. 43 Transgenic mice overexpressing only the apo-lipoprotein(a) component develop fatty streak lesions upon feeding of a cholesterol-rich diet.44 Interestingly, a significant correlation between high levels of ape(a) and reduced in situ plasmin activity and active TGF-l3 levels were observed in atherosclerotic

Fibrinolysis

vessels of these transgenic mice.45 Although TGF-B has a concentration-dependent bimodal effect on smooth muscle cell proliferation in vitro,46 the authors proposed that reduced TGF-l3 activation (resulting from reduced plasmin activity) might constitute a growth stimulus for vascular smooth muscle cells in vivo as well as in vitr~.~’ Similar analyses in t-PA, u-PA and PAI- deficient mice and in recently obtained u-PA receptor deficient or plasminogen deficient mice (unpublished data) will allow to further elucidate the interaction between the fibrinolytic system, TGF-B and vascular smooth muscle growth.

DEVELOPMENT

AND REPRODUCTION

Much circumstantial evidence based on expression of fibrinolytic system components, implicates the plasminogen/plasmin system in ovulation, sperm migration, fertilization, embryo implantation and embryogenesis, and associated tissue remodeling of the ovary, prostate and mammary gland (reviewed in 48). In addition, serine-proteinase inhibitors and/or plasminogen activatorspecific antisera suppress ovulation and embryo implantation in rodents, supporting the implication of the fibrinolytic system in this phenomenon. Since no genetic deficiencies involving t-PA or u-PA have been described in man, inactivation of these genes in mice might have been anticipated to result in a lethal phenotype. Surprisingly, single- and double-deficient mice appeared normal at birth suggesting that neither t-PA nor u-PA, individually or in combination, are required for normal embryonic development.‘“,” Furthermore, the observations that PAI- 1 deficient breeding pairs produced normal offspring also suggest a non-essential role of PAI- 1 in development.‘” Mice with single deficiency of t-PA, u-PA or PAIare fertile.‘6.‘7~‘y Normal fertility was also observed in a t-PA antisense transgenic mouse strain, expressing less than 50% of wild-type t-PA activity in their oocytes4’ Combined t-PA:u-PA deficient mice were able to reproduce but were significantly less fertile than wild-type mice or mice with a single deficiency of t-PA or u-PA. Although poor general health conditions (low body weight, dyspnea, anemia, rectal prolapse and cachexia) and the presence of large fibrin deposits in gonads” might explain their reduced fertility, some apparently normal and healthy combined t-PA:u-PA deficient mice did not produce any litter at all, possibly suggesting a primary defect of reproduction. Inactivation of the LRP gene, which encodes a functional receptor involved in clearance of o.a. plasminogen activators, resulted in embryonic lethality at mid-gestation secondary to intestinal bleeding. 27 It remains to be determined whether deficient clearance of plasminogen activators might not result in local overexpression and possibly, associated uncontrolled proteolysis. Collectively, these observations suggest that proteinases and proteinase inhibitors other than t-PA, u-PA or PAI- may be more

273

essential in reproduction and embryonic development than previously suspected and that the plasminogen/plasmin system might be rather involved in more subtle modulation of embryonic development, its exact role remaining to be further defined.

HEALTH

AND SURVIVAL

Although certain species, including chicken, only possess u-PA and no t-PA, genetic deficiencies of t-PA or uPA in man or mice have not been reported. Consequently, the role of the fibrinolytic system in general health and survival remained to be determined. No effects on health and survival were observed in t-PA deficient and PAI- deficient mice. A small percentage of u-PA deficient mice developed chronic (non-healing) ulcerations and rectal prolapse but without effect on survival. Although combined t-PA:u-PA deficient mice also developed such chronic ulcerations and rectal prolapse, these mice suffered significant growth retardation, developed a wasting-syndrome with anemia, dyspnea, lethargia and cachexia and had a significantly shorter life span. ‘6,‘7 Generalized thrombosis in the gastro-intestinal tract (with chronic ulcerations and rarely ischemic necrosis possibly causing hypoalimentation), in the lungs (with lung atelectasis contributing to dyspnea) and in other organs (including gonads, liver and kidney) might, at least in part, explain increased morbidity and mortality. It remains, however, to be further defined whether other (yet unidentified) processes, such as the observed chronic anemia, may further contribute to the poor general health and reduced survival of these combined t-PA:u-PA deficient mice.

CELL

INVASION

Invasion by cells through anatomical barriers requires proteolysis. Cell-associated production of plasmin by uPA has been proposed to endow cells with such directional proteolysis. Substantial circumstantial evidence based on expression and neutralization data suggests indeed an important role of u-PA in invasion of trophoblasts during embryogenesis, macrophages during inflammation, endothelial cells during reendothelialization and angiogenesis, tumor cells during local invasion or metastasis, keratinocytes during wound healing, osteoclasts during bone remodeling and neurons during development, nerve regeneration and plasticity (reviewed in 48). Cell invasion has only been examined to a limited extent in transgenic mice. Despite absence of plasminogen-dependent fibrin or subendothelial matrix degradation by u-PA deficient macrophages, accumulation in the peritoneal cavity of u-PA deficient macrophages 3 days after injection of thioglycollate was similar to that of wild-type, t-PA deficient or PAIdeficient

274

Evaluation

of the Plasminogen/plasmin

System in Transgenic

Mice

macrophages. 16*” No evidence was obtained for deficient angiogenesis in u-PA deficient mice as indicated by the absence of any anatomical abnormalities in vascularized tissues.16 Invasion of u-PA deficient trophoblasts also appeared unaffected, based on normal structure of placenta and normal reproductive behaviour of u-PA deficient mice.i6.” Other data suggest, however, involvement of cellular fibrinolysis in cell invasion such as reduced metastasis of Lewis lung carcinoma cells in PAI- overexpressing mice,50 reduced rate of neointima formation and neointimal cell accumulation in u-PA deficient mice (unpublished data). More detailed kinetic analysis and use of other models need, however, to be further performed to elucidate the role of the fibrinolytic system in cellular invasion.

BRAIN

FUNCTION

Evidence has been provided that the plasminogen/plasmin system might also be involved in brain function based on expression of fibrinolytic system components in specialized areas of the brain during development5’ or following different forms of brain activity.52,53 In addition, in vitro studies with cultured neurons revealed that neurons are able to produce and respond to plasminogen activators. 54 Restricted and temporal specific expression of t-PA in the nervous system during development has been also observed in transgenic mice, expressing the LacZ marker gene driven by various t-PA promotor constructs. 55*56Ectopic expression of murine u-PA in the brain (e.g. in the hippocampus and limbic system) was associated with impaired learning of tasks in transgenic mice. 57 Although production of u-PA and plasminogen in mouse brain is controversial, the ectopic u-PA expression experiments suggest that abnormal proteolysis in the brain may cause behavioral abnormalities, possibly due to an effect on tissue remodeling. Imbalanced plasmin proteolysis, such as in t-PA deficient mice, also resulted in abnormal brain function. Preliminary results suggested indeed that long term potentiation was specifically impaired in t-PA deficient but not in u-PA deficient mice (unpublished data). No however, were, abnormalities neuroanatomical observed in t-PA deficient mice (unpublished data). Further studies are being performed to examine possible learning deficits in t-PA deficient mice. Thus, although evidence for a physiogical role of plasmin-mediated proteolysis in the brain is accumulating, the exact mechanisms underlying possible alterations in long term potentiation or learning need to be further defined.

CONCLUSIONS

AND PERSPECTIVES

Studies with transgenic mice over- or under-expressing components of the fibrinolytic system, have revealed a

significant role of this system in fibrin clot surveillance, reproduction, (vascular) wound healing, brain function, health and survival. The distinct phenotypes associated with single loss and the more severe phenotype associated with combined loss of plasminogen activator gene function suggest that through evolution, both plasminogen activators have evolved with specific but overlapping biological properties. Interestingly, the role of the fibrinolytic system in thrombosis and vascular wound healing became more apparent after challenging mice with single deficiencies of plasminogen activators with an inflammatory or traumatic challenge, respectively. It therefore seems warranted to examine possible consequences of loss of plasminogen activator gene function in other processes including atherosclerosis, inflammatory lung and kidney disease, malignancy etc. Transgenic mice further promise to be valuable for the study of the fibrinolytic system. Gene-inactivation of other components of the fibrinolytic system have been undertaken (alpha,-antiplamin, PAI-2) or achieved (u-PA receptor, plasminogen). In addition, through targeted genetic manipulation, it is possible not only to inactivate but also to mutate genes or to express them in a conditional, temporal or restricted pattern. Expression of different u-PA receptor splice variants might for example allow to evaluate their differential role in viva. Furthermore, the significantly more severe phenotype of combined t-PA:u-PA deficient mice compared to mice with single deficiencies indicates that intercrossing transgenic mice is a powerful genetic tool. The plasminogen activator knock-out mice with their thrombotic phenotypes are also valuable models to evaluate whether gene-transfer of wild-type or mutant plasminogen activator genes is able to restore normal thrombolytic function or to prevent thrombosis. Preliminary evidence suggests indeed that impaired thrombolysis of t-PA deficient mice can be completely restored using adenoviral mediated gene transfer of rtPA. In addition, analysis of neointima formation in plasminogen activator deficient mice suggests that controlled reduction of fibrinolytic activity in the vessel wall might be beneficial for the prevention or reduction of restenosis. Whether this can be achieved by current gene transfer technology, remains to be defined.

REFERENCES I Astrup T. In: Davidson JF, Rowan RM, Samama MM, Desnoyers PC, eds. Progress in chemical fibrinolysis and thrombolysis. Vol 3. New York: Raven Press, 1978; pp l-57. 2. Cohen D, Lijnen HR. Basic and clinical aspects of fibrinolysis and thrombolysis. Blood 1991; 78: 3 I 14-3 124. 3. Schneiderman J, Loskutoff DJ. Plasminogen activator inhibitors. Trends Cardiovasc Med I99 1; I : 99- 102. 4. Kluft C, Dooijewaard G, Emeis JJ. Role of the contact system in fibrinolysis. Sem Thromb Hemost 1987; 13: 50-68. 5. Plow EF, Edgington TS. An alternative pathway for fibrinolysis. 1. The cleavage of fibrinogen by leukocyte proteases at physiologic pH. J Clin Invest 1975; 56: 3&38. 6. Simon DI, Ezratty AM, Francis SA, Rennke H, Loscalzo J. Fibrincogen) is internalized and degraded by activated human

Fibrinolysis

monocytoid cells via Mac- 1 (CD I I b/CD 18): a nonplasmin librinolytic pathway. Blood 1993; 82: 2414-2422. 7. Nilsson IM, Ljungner H, Tengborn L. Two different mechanisms in patients with venous thrombosis and defective fibrinolysis: low concentration of plasminogen activator or increased concentration of plasminogen activator inhibitor. Br Med J Clin Res Ed 1985; 290: 145331456. 8. Colucci M, Paramo JA, Cohen D. Generation in plasma of a fastacting inhibitor of plasminogen activator in response to endotoxin stimuation. J Clin Invest 1985; 75: 818-824. 9. Aillaud MF, Juhan-Vague I, Alessi MC et al. Increased PAinhibitor levels in the postoperative period. No cause- effect relation with increased cortisol. Thromb Haemost 1985: 54: 466468. IO. Paramo JA, Colucci M, Collen D, Van de Werf F. Plasminogen activator inhibitor in the blood of patients with coronary artery disease. Brit Med J Clin Res Ed 1985: 291: 573-574. I I. Aznar J, Estelles A, Tormo G et al. Plasminogen activator inhibitor activity and other librinolytic variables in patients with coronarv arterv disease. Br Heart J 1988; 59: 535-541. 12. Hamsten A, de Faire U, Walldius G et al. Plasminogen activator inhibitor in plasma: risk factor for recurrent myocardial infarction. Lancet 1987; 2: 3-9. 13. Levi M, Biemond BJ, van Zonneveld AJ, ten Cate JW, Pannekoek H. Inhibition ofplasminogen activator inhibitor-l activity results in promotion of endogenous thrombolysis and inhibition of thrombus extension in models of experimental thrombosis. Circulation 1992; 85: 305-3 12. 14. Juhan-Vague I, Aillaud MF, De Cock F et al. The fast-acting inhibitor of tissue-type plasminogen activator is an acute phase protein. Progr Fibrin01 1985; 7: 146-149. 15. Aoki N. Hemostasis associated with abnormalities of fibrinolysis. BloodRev 1989; 3: I l-17. 16. Carmeliet P, Schoonjans L, Kieckens Let al. Physiological consequences of loss of plasminogen activator gene function in mice. Nature 1994; 368: 419424. 17. Carmeliet P, De Clercq C, Janssen S et al. Biological effects of disruption of the tissue-type plasminogen activator, urokinasetype plasminogen activator and plasminogen activator inhibitor- I genes in mice. Ann NY Acad Sci (1994; in press). 18. Erickson LA, Fici GJ, Lund JE, Boyle TP, Politea HG, Marotti KR. Development of venous occlusions in mice transgenic for the plasminogen activator inhibitor- I gene. Nature 1990; 346: 74-76. 19. Carmeliet P, Kieckens L, Schoonjans L et al. Plasminogen activator inhibitor- I gene-deficient mice. 1. Generation by homologous recombination and characterization. J Clin Invest 1993; 92: 274662755. 20. Carmeliet P, Stassen JM, Schoonjans L et al. Plasminogen activator inhibitor-l gene-deficient mice. II. Effects on hemostasis, thrombosis and thrombolysis. J Clin Invest 1993; 92: 2756-2760. 21. Blasi F, Vassalli JD, Dano K. Urokinase-type plasminogen activator: proenzyme, receptor, and inhibitors. J Cell Biol 1987; 104: 80 I-804. 22. Saksela 0, Rifkin DB. Cell-associated plasminogen activation: regulation and physiological functions. Ann Rev Cell Biol 1988; 4: 93-l 26. 23. Unkeless JC, Gordon S. Reich E Secretion of plasminogen activator by stimulated macrophages. J Exp Med 1974; 139: 834-850. 24. Elliott MJ, Faulkner-Jones BE, Stanton H, Hamilton JA, Metcalf D. Plasminogen activator in granulocyte-macrophage-CSF transgenic mice. J lmmunol 1992; 149: 3678-368 I. 25. Fay WP, Shapiro AD, Shih JL, Schleef RR, Ginsburg D. Complete deficiency of plasminogen-activator inhibitor type I due to a frameshift mutation. N Engl J Med 1992; 327: 1729-1733. 26. Heckel JL, Sandgren EP, Degen, JL, Palmiter RD, Brinster RL. Neonatal bleeding in transgenic mice expressing urokinase-type plasminogen activator. Cell 1990; 62: 447456. 27. Herz J, Clouthier DE, Hammer RE. LDL receptor-related protein internalizes and degrades uPA-PAI- I complexes and is essential for embryo implantation. Cell 1992: 7 I : 4 I I42 I. 28. Forrester JS. Fishbein M, Helfant R, Fagin J. A paradigm for restenosis based on cell biology: clues for the development of new preventive therapies. J Am Coll Cardiol 1991; 17: 7.58-769.

275

29. Reidy MA, Fingerle J, Lindner V. Factors controlling the development of arterial lesions after injury. Circulation 1992; 85 (Suppl III): 111~3-111-46. 30. Pepper MS, Sappino AP, Montesano R, Orci L, Vassalli JD. Plasmingen activator- I is induced in migrating endothelial cells. J.CellPhysiol. 1992; 153: 1299139. 3 I. Clowes AW, Clowes MM, An YPT, Reidy MA, Belin D. Smooth muscle cells express urokinase during mitogenesis and tissuetype plasminogen activator during migration in injured rat carotid artery. Circulation Res 1990; 67: 61-67. 32. Jackson CL, Raines EW, Ross R, Reidy MA. Role of endogenous platelet-derived growth factor in arterial smooth muscle cell migration after balloon catheter injury. Arterioscl Thromb 1993; 13:1218-1226. 33. Mignetti P, Mazziesi R, Rifkin DB. Expression of the urokinase receptor in vascular cells is stimulated by basic fibroblast growth factor. J. Cell Biol. 1991; 113: Il93-1201. 34. Reuning U, Bang NU. Regulation of the urokinase-type plasminogen activator receptor on vascular smooth muscle cells is under the control of thrombin and other mitogens. Arterioscl Thrombosis 1992; 12:l 161-l 170. 35. Herbert JM, Lamarche 1, Prabonnaud V, Dol F, Gauthier T. Tissue-type plasminogen activator is a potent mitogen for human aortic smooth muscle cells. J Biol Chem 1994; 269:307&3080. 36. Dumler I, Petri T, Schleuning WD. Induction of C-&OSgene expression by urokinase-type plasminogen activator in human ovarian cancer cells. FEBS Letters 1994; 343: I033 106. 37. Mehan VK, Meier B, Urban P. Influence on early outcome and restenosis of urokinase before elective coronary anpioplasty. Am J Cardiol 1993: 72: 106-108. 38. Carmeliet P, Stassen JM, De Mol M, Declercq C, BouchC A, Collen D. Arterial neointima formation after trauma in mice with inactivation of the t-PA, u-PA or PAI- I genes. Abstract of the XIlth International Congress on Fibrinolysis. Fibrinolysis (Suppl. I) 1994; 8: Abst. 280. 39. Juhan-Vague I, Collen D. On the role of coagulation and fibrinolysis in atherosclerosis. Ann Epidemiol 1992; 2: 427438. 40. Juhan-Vague I, Alessi MC, Vague P. Increased plasma plasminogen activator inhibitor- I levels. A possible link between insulin resistance and atherothrombosis. Diabetologia 1991; 34: 457462. 4 I. Schneiderman J, Sawdey MS, Keeton MR et al. Increased type I plasminogen activator inhibitor gene expression in atherosclerotic human arteries. Proc Nat1 Acad Sci USA 1992; 89: 6998-7002. 42. Lupu F, Bergonzelli GE, Heim DA, Cousin E. Genton CY, Bachmann F, Kruithof EKO. Localization and production of plasminogen activator inhibitor-l in human healthy and atherosclerotic arteries. Arterioscler Thromb 1993; 13: 1090-I 100. 43. Liu AC and Lawn RM. Lipoprotein(a) and atherogenesis. Trends Cardiovasc Medicine 1994; 4:40+4. 44. Lawn RM, Wade DP, Hammer RE, Chiesa G, Verstuyft JG. Rubin EM. Atherogenesis in transgenic mice expressing human apolipoprotein (a). Nature 1992; 360: 670-672. 45. Kemp P, Metcalfe J, Liu A. Lawn R, Grainger D. Activation of TGF-P is depressed in the sera and in the vessels of transgenic apo (a) mice. J Cell Biochem 1994; Suppl I8A: 279 46. Battegay EJ, Raines EW, Seifert RA, Bowen-Pope DF, Ross R. TGF-8 induces bimodal proliferation of connective tissue cells via complex control of an autocrine PDGF loop. Cell 1990; 63: 5 I S-524 47. Grainger DJ, Kirschenlohr HL, Metcalfe JC, Weissberg PL, Wade DP, Lawn RM. Proliferation of human smooth muscle cells promoted by lipoprotein(a). Science 1993; 260: 1655-1658. 48. Vassalli JD, Sappino JD, Belin D. The plasminogen activator/plasmin system. J Clin Invest 1991; 88: 1067-1072. 49. Richards W.G., Carroll P.M., Kinloch R.A., Wassarman P.M., Strickland S. Creating maternal effect mutations in transgenic mice: antisense inhibition of an oocyte gene product. Develop Biol 1993; 160.543-553. 50. Poggi A, Bellelli E, Castelli MP et al. Fibrinolysis and metastasis: the role of plasminogen activator inhibitor- I (PAI- 1) in a model of transgenic mice. Thromb Haemost 1993; 69: 836 (abstract).

276

Evaluation

of the Plasminogenlplasmin

System in Transgenic

Mice

5 I. Menoud PA, Debrot S, Schowing J. Roux’s Arch Devel Biol 1989; 198: 219-226. 52. Qian Z, Gilbert ME, Colicos MA, Kandel ER, Kuhl D. Tissuetype plasminogen activator is induced as an immediate-early gene during seizure, kindling and long-term potentiation. Nature 1993; 361: 453457. 53. Sappino AP, Madani R, Huarte J, Beiin D, Kiss JZ, Wohlwend A, Vassalli JD. Extracellular proteolysis in the adult murine brain. J Clin Invest 1993; 92:679-685. 54. Krystosek A, Seeds NW Plasminogen activator release at the neuronal growth cone. Science 1981; 213: 1532-1534. 55. Carroll PM, Tsirka S, Richards WG, Frohman MA, Strickland S.

Promoter sequences of the tissue-type plasminogen activator gene are able to confer tissue-specific expression of LacZ during mouse development (in press). 56. Theuring F, Kropp C, Aguzzi A, Wohn KD, Schleuning WD. Analysis of t-PA gene promoter activity during embryonic development in transgenic mice. Abstract Book of the Vllth International Symposium on the Biology of Vascular Cells, San Diego, CA, November 1992, p.30. 57. Meiri N, Masos T. Rosenblum K, Miskin R, Dudai Y. Overexpression of urokinase-type plasminogen activator in transgenic mice is correlated with impaired learning. Proc Nat1 Acad Sci USA 1994; 9 1: 3 1963200.