The plasminogen coding sequence is interrupted by 18 introns, as shown schematically in Fig. 178-3. The plasminogen gene (PLG) spans a total of 52.5 kb on human chromosome 6q26-27.7– 9 Plasminogen is a member of a gene family of closely related serine proteases that also includes the plasminogen activators tPA and uPA (see below10). This gene family also includes the highly homologous nonprotease, apolipoprotein (a) (Lp(a)), which may contribute to thrombosis and atherosclerosis by interfering with plasminogen function (Chaps. 116 and 169).
An alignment of the intron/exon structures for plasminogen, tPA, and uPA is shown. Intron positions are indicated by triangles, with the preceding codon number given above the triangle. Labeled protein domains are the signal peptide (SP), preactivation peptide (PAP), “kringle” domains (K1-5), fibronectin-like “finger” domain (F), epidermal growth factor-like domain (E), and the protease domain. The catalytic triad amino acid residues in the latter domain are also indicated (H, D, and S). (Adapted from Hajjar.45 Used with permission.)
The gene for Lp(a) is located ≈50 kb from the PLG gene on 6q. Plasminogen is an abundant plasma protein and a number of variant alleles have been identified based on pI. The most common variants are PLGA and PLGB, with rarer variants designated as PLGM1, M2, and so forth.11 Plasminogen is first synthesized as a single-chain proenzyme with molecular mass of 92 kDa. It circulates in plasma at a concentration of 1.5 μM and is thought to be derived primarily from synthesis within the liver.12 The 791-amino-acid protein contains 24 disulfide bonds (Fig. 178-4). Most of these disulfides (16) are contained within the five repeated homologous loop structures referred to as “kringles,” because of a fanciful resemblance to a Danish pastry of that name.9 Each kringle domain contains approximately 80 amino acids, with a molecular mass of about 10,000 daltons and is held in its unique loop conformation by 3 disulfide bonds present in homologous locations within the structure (Fig. 178-5). Plasminogen interacts with free lysines on fibrin and α2AP. A high-affinity lysine-binding site appears to be localized to the first kringle domain and a lower affinity site to the fourth kringle.13 Binding to surface receptors for plasminogen is also dependent on lysine interactions. All of these lysine-binding interactions can be inhibited by lysine analogues such as ε-aminocaproic acid and tranexamic acid, both of which are used effectively as pharmacologic inhibitors of fibrinolysis.14
The primary structure of plasminogen is depicted in single letter amino acid code. “Pli” indicates the site of cleavage to generate Lys-plasminogen from Glu-plasminogen. UK indicates the Arg560-Val561 bond that is cleaved by both uPA and tPA to activate plasminogen to plasmin. Black bars indicate disulfide bonds. Kringles 1 to 5 are indicated by the numbered brackets. (Adapted from Collen and Lijnen.149 Used with permission.)
The domain structures of plasminogen, tPA, and uPA. P = protease domain and GF = epidermal growth factor-like domain; other domain abbreviations are as in Fig. 178-3. Arrows indicate the cleavage sites that yield the two-chain activated form of each protease. The active sites are depicted by dots. (Adapted from Wiman and Hamsten.150 Used with permission.)
Two circulating isoforms of plasminogen can be biochemically identified containing different amino termini.15,16 Glu-plasminogen contains the full-length molecule formed following cleavage of the signal peptide. Limited proteolysis of the peptide bond between lysines 77 and lysine 78 generates Lys-plasminogen (Fig. 178-4), which is missing the first 77 amino acids of Glu-plasminogen. Limited proteolysis is also seen after Arg68 and Lys78. All three of these variant molecules are referred to together as Lys-plasminogen. All of these forms of Lys-plasminogen show enhanced binding to both fibrin and cellular receptors and can be activated to plasmin by plasminogen activators at a 10 to 20 times more rapid rate than Glu-plasminogen.17 Small amounts of Lys-plasminogen can be detected in plasma. Both Lys- and Glu-plasminogen are activated to plasmin by specific cleavage of the peptide bond between Arg560 and Val561, yielding the two-chain active serine protease.18 Most of the plasminogen activated at the cell surface is of the Lys-form whereas the majority of circulating plasminogen is in the Glu-form.19 Plasminogen contains a typical serine protease active site triad formed by His602, Asp645, and Ser740.
Recently, an internal fragment of plasminogen containing the first four kringle domains was shown to be a potent inhibitor of angiogenesis termed angiostatin.20,21 Angiostatin is currently under study as a potential therapeutic agent for the treatment of human malignancies.
Tissue Plasminogen Activator (TPA)
tPA appears to be the primary physiological activator of plasminogen within the vascular space. tPA is a serine protease with significant homology to plasminogen (Figs. 178-3 and 178-5). The human tPA gene (PLAT) is localized to chromosome 822 and is composed of 14 exons spanning approximately 36 kb.23,24 The mature protein contains 527 amino acids with a molecular mass of 72 kDa. tPA was originally assumed to begin with a serine designated as amino acid 1.25 Although the N-terminus was subsequently shown to include three additional proximal amino acids,26 the original numbering scheme is still generally used. As in plasminogen, the serine protease domain is at the C-terminus and is preceded by two kringle domains (compared to the five in plasminogen, Fig. 178-5). At the immediate amino terminus following the signal peptide is a 47-residue domain referred to as the finger domain, which is homologous to a similar segment in fibronectin. The next 38 amino acids following the finger domain are homologous to epidermal growth factor and are referred to as the EGF or growth factor domain (Fig. 178-5). The finger and kringle 2 domains appear to be most important for fibrin binding.27 Note that the finger, EGF, and kringle domains are all encoded by single exons or two adjacent exons (Fig. 178-3). A comparison of the domain and intron/exon structures of plasminogen, tPA, and uPA, along with the homologous serine proteases of the coagulation cascade, suggests that these genes arose by assembly of common modules, and is one of the original observations supporting the notion of exon shuffling.10
tPA is cleaved by plasmin at the bond between Arg275 and Ile276 to generate the two-chain, disulfide-linked form of tPA.28 Although single-chain tPA has less activity than the two-chain form in solution, both molecules show similar activity toward plasminogen when bound to fibrin. Fibrin binding increases the activity of both forms of tPA toward plasminogen by at least two orders of magnitude, primarily by lowering the Km, 29 presumably through the colocalization of both molecules on the fibrin surface. The significant functional activity of the single-chain, uncleaved form of tPA is unique among serine proteases, with other members of this family exhibiting more typical zymogen behavior.30,31 tPA exists in at least two variants, referred to as type 1 and type 2, defined by differences in glycosylation. The carbohydrate structures of tPA are thought to play an important role in clearance and may also modulate binding to cell-surface receptors.32
tPA is thought to be derived primarily from synthesis in the endothelial cell. Its secretion from the endothelial cell is regulated by a variety of agonists including thrombin, histamine, epinephrine, and DDAVP and is also induced by exercise or venous occlusion, as well as shear stress. All of these responses may be important in maintaining the integrity of the vascular space under a variety of physiological and pathologic conditions.33– 35
Urokinase-Type Plasminogen Activator (UPA)
The mature uPA molecule contains 411 amino acids with a molecular mass of approximately 54 kDa.36,37 Similar to tPA, uPA exists in a single-chain, uncleaved form referred to as prourokinase or single-chain uPA (abbreviated scu-PA). Though prourokinase has been reported to exhibit a small amount of intrinsic PA activity,38– 40 the significance of this observation is controversial.41 uPA is cleaved by plasmin or kallikrein at the peptide bond between Lys158 and Ile159 to generate the two-chain form.42 The uPA domain structure (Fig. 178-5) consists of an N-terminal EGF domain, followed by a single kringle, and, finally, by the C-terminal protease domain.
The human uPA gene (PLAU) contains 11 exons and spans 6.4 kb on chromosome 10,22,37,43 again, with considerable homology to the gene structures of other members of the serine protease gene family (Fig. 178-3). Two-chain uPA can be cleaved between Lys135 and Lys136 by plasmin to remove the N-terminal 135 residues, generating a low-molecular-weight form (33 kDa).44 Both the high- and low-molecular-weight forms of uPA activate plasminogen with similar efficiency, although only the high-molecular-weight form of uPA can bind to its specific cellular receptor, the uPA receptor (uPAR see below). In contrast to tPA, uPA has a low affinity for fibrin and appears to function primarily in the extravascular space.
Other Plasminogen Activators
In addition to the specific plasminogen activators tPA and uPA, proteases of the intrinsic coagulation system, including kallikrein, factor XIa, and factor XIIa, can also activate plasminogen under some conditions, although this is thought to account for <15 percent of the total plasmin-generating activity of plasma.45 The observation that knockout mice deficient for both tPA and uPA46 have a phenotype quite similar to plasminogen knockout mice47,48 suggests that the physiologic contribution of these additional pathways for plasminogen activation is minimal.
A number of pathogenic bacteria produce their own plasminogen activators, which are sometimes referred to as exogenous plasminogen activators, in contrast to the endogenous PAs discussed above. In several examples, these exogenous activators have been shown to play an important role in microbial pathogenesis.49,50 Streptokinase is a 414-amino-acid polypeptide produced by streptococci that does not exhibit direct protease activity. However, when bound to plasminogen, the complex induces a conformational change, exposing the active site of the enzyme, which is then capable of activating a second plasminogen molecule to plasmin.51– 53 Plasminogen in complex with streptokinase is also converted to plasmin and the plasmin/streptokinase complex is resistant to inhibition by α2AP, in contrast to free plasmin, which is rapidly inhibited by this serpin (see below). A 36-amino-acid protein produced by S. aureus, a staphylokinase-like streptokinase, forms a 1:1 complex with plasminogen that is also capable of activating additional plasminogen molecules.54 However, the staphylokinase/plasmin complex is inhibited by α2AP. Both staphylokinase and streptokinase are used as pharmacologic agents to induce clot lysis by activating plasminogen.
α2AP has a molecular mass of 70,000 and contains 452 amino acids.55 α2AP is a member of the serpin gene family, with its P1 and P1′ residues corresponding to Arg364 and Met365. α2AP circulates in plasma at concentrations of approximately 1 μm and has a half-life of approximately 2.5 days. The human α2AP gene (PLI) is located on chromosome 17 and contains 10 exons, spanning approximately 16 kb.56 The primary site of synthesis of α2AP is in the liver, although the protein is also found in the platelet α-granule. Free plasmin in the circulation is promptly neutralized through the formation of a 1:1 complex with α2AP via the standard serpin inhibitory mechanism and the α2AP/plasmin complex is rapidly cleared from plasma.57 A modified form of α2AP lacking the 26 carboxy-terminal residues constitutes a small portion of circulating α2AP, but is inactive. α2AP is efficiently cross-linked to the fibrin α-chain in the blood clot by the action of factor XIII. This localization serves to stabilize the fibrin clot against plasmin degradation.58
α2-Macroglobulin, a large 725-kDa dimeric protein, can also inhibit plasmin, although only with ≈10 percent of the efficiency of α2AP. Unlike other protease inhibitors of the fibrinolytic system, α2-macroglobulin is not a serpin and its complex with plasmin is noncovalent. α2-Macroglobulin is synthesized in endothelial cells and macrophages, and is also found in the platelet α-granule.45
Plasminogen Activator Inhibitors
There are two major plasminogen activator inhibitors (PAIs), PAI-1 and PAI-2. PAI-1 is the major inhibitor found in normal plasma. It is a single-chain, 52-kDa molecule that is devoid of cysteines, and thus of disulfide bonds.59– 61 PAI-1 is synthesized in many cell types in tissue culture, although in vivo its primary sites of synthesis appear to be megakaryocytes (leading to storage in the platelet α-granule), smooth muscle cells, and adipocytes.62– 64 In response to inflammation, PAI-1 expression is markedly induced in endothelial cells and several other cell types, including hepatocytes. The synthesis of PAI-1 is regulated by a broad range of inflammatory mediators and cytokines, and its transcriptional regulation has been extensively studied.65,66
The PAI-1 gene spans approximately 12 kb on human chromosome 7q and consists of 9 exons.67,68 PAI-1 is unique among serpins in its propensity to decay to an inactive, latent form as a result of a conformational change with a half-life of approximately 2 h.69– 71 PAI-1 forms a complex with the adhesive glycoprotein vitronectin that serves to stabilize PAI-1 in the active form and also to localize it to specific regions within the extracellular matrix.72 PAI-1 is an efficient inhibitor of both tPA and uPA and appears to be the major physiological inhibitor of these proteases.73,74
PAI-2 is a 393-amino-acid member of the serpin family with an 8-exon, approximately 16-kb gene localized to human chromosome 18q. PAI-2 was originally purified from human placenta and is only found in the circulation at significant concentrations during pregnancy. PAI-2 is a poor inhibitor of single-chain tPA, but efficiently inhibits both two-chain tPA and two-chain uPA.75 PAI-2 does not contain a typical signal peptide, similar to other members of the ovalbumin subgroup of serine protease inhibitors. Approximately 70 percent of PAI-2 remains intracellular, where its function is unknown, although it has been proposed to play a role in regulating apoptosis.76 Approximately 20 percent of PAI-2 is secreted via a unique mechanism not involving a signal peptide.77,78 The transcriptional regulation of PAI-2 has been extensively studied and, as for PAI-1, this gene also appears to be regulated by a broad range of inflammatory mediators.75,79
Other Inhibitors of Fibrinolysis
Another serpin, the protein C inhibitor, is also capable of inhibiting uPA and is sometimes referred to as PAI-3.80 Protease nexin, designated as PAI-4, can also inhibit a broad range of proteases including the PAs. Protease nexin is a much more efficient inhibitor of uPA than tPA, and also inhibits plasmin, thrombin, trypsin, and factor Xa.81 Finally, another inhibitor of fibrinolysis termed TAFI (thrombin activated fibrinolysis inhibitor) is produced by the coagulation cascade82 and serves as a connection between these two limbs of hemostasis. TAFI has been shown to be identical to carboxypeptidase B and functions by removing free lysines (the binding sites for plasminogen) from the fibrin clot.83 Carboxypeptidase B is activated by thrombin. Thus, coincident with the formation of the fibrin clot by thrombin, TAFI generation also serves to protect the clot from rapid degradation by plasmin.84
Cellular Receptors for the Fibrinolytic System
Specific cellular receptors have been proposed for plasminogen as well as for the natural PAs, uPA and tPA.85 The best characterized of these is the uPA receptor or uPAR. This protein is widely expressed on monocytes, macrophages, fibroblasts, endothelial cells, platelets, and many tumor cells.85 uPAR has a molecular mass of 55 to 60 kDa and binds single-chain uPA as well as the high-molecular-weight form of two-chain uPA. The uPAR protein contains 313 amino acids with a 21-residue signal peptide.86 uPAR does not have a typical transmembrane domain and belongs to the family of glycosylphosphatidyl-inositol-linked surface proteins.87 uPAR may serve to localize uPA activity to specific sites on the cell surface and has been proposed to play a role in cell migration and possibly also in uPA clearance.88
Annexin II has been proposed as a specific receptor for tPA on the cell surface and may also bind plasminogen, potentially coordinating their interaction.89 α-Enolase has been proposed as a receptor for plasminogen.90 Plasminogen has also been shown to bind to the glycoprotein IIb/IIIa receptor on the platelet surface.91 Finally, complexes formed between the uPA and tPA proteases and PAI-1 or PAI-2 are cleared through the low-density lipoprotein receptor-related protein, LRP.92,93
Fibrin Degradation by Plasmin
Plasmin is capable of degrading both intact fibrinogen in the circulation as well as fibrin deposited in a blood clot. Fibrinogen thus serves as a substrate for both plasmin and thrombin. Thrombin cleaves at the C-termini of fibrinogen α-chains to liberate fibrinopeptide A and expose the Gly-Pro-Arg tripeptide facilitating polymerization to form insoluble fibrin (see Chap. 169). Plasmin first cleaves fibrinogen to release the C-terminal fragments of the α-chain and simultaneously, but more slowly, liberates fibrinopeptide B from the β-chain. Alhough plasmin can cleave fibrinogen in the circulation, its primary physiological role is the degradation of insoluble fibrin. Fibrin degradation by plasmin liberates a number of soluble fragments, including dimers of the central “coiled coil” D domains of fibrin. The latter fragments, termed D-dimers, are measured clinically as a marker of disseminated intravascular coagulation. These fragments can inhibit platelet function and also may contribute to the inflammatory response through effects on blood pressure regulation, chemotaxis, and immune modulation.45
As noted above, both tPA and plasminogen bind to fibrin, resulting in an increase in the rate of plasminogen activation by at least two orders of magnitude in the presence of fibrin.29 Plasmin molecules on the fibrin surface have their lysine binding sites and active sites occupied, interfering with their interaction with α2AP.94 Thus, fibrin both enhances the catalytic efficiency of tPA for plasminogen activation and also protects active plasmin from inhibition by α2AP.