Interactions of Lipoprotein(a) with the Plasminogen System: Mechanisms and Pathophysiological Consequences

Elevated plasma concentrations of lipoprotein(a) (Lp(a)) are associated with increased risk of atherothrombotic disease. Lp(a) is a unique lipoprotein consisting of a low density lipoprotein-like moiety covalently linked to apolipoprotein(a) (apo(a)), a homologue of the fibrinolytic proenzyme plasminogen. Apo(a) is extremely heterogeneous in size with small isoforms being independently associated with increased cardiovascular risk. Several in vitro and in vivo studies have shown that Lp(a)/apo(a) can inhibit tissue-type plasminogen activator (tPA)-mediated plasminogen activation on fibrin surfaces, although the mechanism of inhibition by apo(a) remains controversial. Essential to fibrin clot lysis are a number of plasmin-dependent positive feedback reactions that enhance the efficiency of plasminogen activation, including the plasmin-mediated conversion of Glu1-plasminogen to Lys78-plasminogen. Additionally, abnormal fibrin clot structures have been associated with both an increased risk of cardiovascular disease and elevated Lp(a) levels. Similarly, oxidized phospholipids have been implicated in the development of cardiovascular disease, and are not only preferentially carried by Lp(a) in the plasma but have also been shown to covalently-modify both apo(a) and plasminogen. In this thesis, we built upon the understanding of the role of apo(a) in plasminogen activation on the fibrin/degraded fibrin surface by determining that: (i) apo(a) inhibits plasmin-mediated Glu1-plasminogen to Lys78-plasminogen conversion and identifying the critical domains in apo(a) responsible for this effect, (ii) apo(a) isoform size does not affect either the inhibition of tPA-mediated plasminogen activation or the inhibition of plasmin-mediated Glu1-plasminogen to Lys78-plasminogen conversion, (iii) apo(a) modifies fibrin clot structure to form more dense clots with thinner fibers and reduced permeability, modifications that enhance the ability of apo(a) to inhibit tPA-mediated plasminogen activation and (iv) the phosphorus content of apo(a) affects its ability to inhibit tPA-mediated plasminogen activation and the phosphorus content of plasminogen affects its ability to be activated by tPA. By understanding these individual reactions, each of which has the potential to affect the broader fibrin clot lysis process, we have expanded our understanding of the overall effect of Lp(a)/apo(a) in the inhibition of plasminogen activation on the fibrin/degraded fibrin surface and thus broadened our understanding of how Lp(a)/apo(a) may mediate the inhibition of thrombolysis in vivo.

A Quantitative Investigation of Selected Reactions in the Fibrinolytic Cascade

Previous work has shown that thrombin activatable fibrinolysis inhibitor (TAFI) was unable to prolong lysis of purified clots in the presence of Lys-plasminogen (Lys-Pg), indicating a possible mechanism for fibrinolysis to circumvent prolongation mediated by activated TAFI (TAFIa). Therefore, the effects of TAFIa on Lys-Pg activation and Lys-plasmin (Lys-Pn) inhibition by antiplasmin (AP) were quantitatively investigated using a fluorescently labeled recombinant Pg mutant which does not produce active Pn. High molecular weight fibrin degradation products (HMW-FDPs), a soluble fibrin surrogate that models Pn modified fibrin, treated with TAFIa decreased the catalytic efficiency (kcat/Km) of 5IAF-Glu-Pg cleavage by 417-fold and of 5IAF-Lys-Pg cleavage by 55-fold. A previously devised intact clot system was used to measure the apparent second order rate constant (k2) for Pn inhibition by AP over time. While TAFIa was able to abolish the protection associated with Pn modified fibrin in clots formed with Glu-Pg, it was not able to abolish the protection in clots formed with Lys-Pg. However, TAFIa was still able to prolong the lysis of clots formed with Lys-Pg. TAFIa prolongs clot lysis by removing the positive feedback loop for Pn generation. The effect of TAFIa modification of the HMW-FDPs on the rate of tissue type plasminogen activator (tPA) inhibition by plasminogen activator inhibitor type 1 (PAI-1) was investigated using a previously devised end point assay. HMW-FDPs decreased the k2 for tPA inhibition rate by 3-fold. Thus, HMW-FDPs protect tPA from PAI-1. TAFIa treatment of the HMW-FDPs resulted in no change in protection. Vitronectin also did not appreciably affect tPA inhibition by PAI-1. Pg, in conjunction with HMW-FDPs, decreased the k2 for tPA inhibition by 30-fold. Hence, Pg, when bound to HMW-FDPs, protects tPA by an additional 10-fold. TAFIa treatment of the HMW-FDPs completely removed this additional protection provided by Pg. In conclusion, an additional mechanism was identified whereby TAFIa can prolong clot lysis by increasing the rate of tPA inhibition by PAI-1 by eliminating the protective effects of Pn-modified fibrin and Pg. Because TAFIa can suppress Lys-Pg activation but cannot attenuate Lys-Pn inhibition by AP, the Glu- to Lys-Pg/Pn conversion is able to act as a fibrinolytic switch to ultimately lyse the clot.

Elucidation of the interactions between the transcription factor E2A and the transcriptional co-activator CBP/p300

E2A is a transcription factor that plays a particularly critical role in lymphopoiesis. The chromosomal translocation 1;19, disrupts the E2A gene and results in the expression of the fusion oncoprotein E2A-PBX1, which is implicated in acute lymphoblastic leukemia. Both E2A and E2A-PBX1 contain two activation domains, AD1 and AD2, which comprise conserved ΦxxΦΦ motifs where Φ denotes a hydrophobic amino acid. These domains function to recruit transcriptional co-activators and repressors, including the histone acetyl transferase CREB binding protein (CBP) and its paralog p300. The PCET motif within E2A AD1 interacts with the KIX domain of CBP/p300, the disruption of which abrogates the transcriptional activation by E2A and the transformative properties of E2A-PBX1. The generation of a peptide-based inhibitor targeting the PCET:KIX interaction would serve useful in further assessing the role of E2A and E2A-PBX1 in lymphopoiesis and leukemogenesis. An interaction between E2A AD2 and the KIX domain has also been recently identified, and the TAZ domains of CBP/p300 have been shown to interact with several transcription factors that contain ΦxxΦΦ motifs. Thus the design of an inhibitor of the E2A:CBP/p300 interaction requires the full complement of interactions between E2A and the various domains of CBP/p300 to be elucidated. Here, we have used nuclear magnetic resonance (NMR) spectroscopy to determine that AD2 interacts with KIX at the same site as PCET, which indicates that the E2A:KIX interaction can be disrupted by targeting a single binding site. Using an iterative synthetic peptide microarray approach, a peptide with the sequence DKELQDLLDFSLQY was derived from PCET to interact with KIX with higher affinity than the wild type sequence. This peptide now serves as a lead molecule for further development as an inhibitor of the E2A:CBP/p300 interaction. Fluorescence anisotropy, peptide microarray technology, and isothermal titration calorimetry were employed to characterize interactions between both TAZ domains of CBP/p300 and the PCET motif and AD2 of E2A. Alanine substitution of residues within PCET demonstrated that the ΦxxΦΦ motif is a key mediator of these interactions, analogous to the PCET:KIX interaction. These findings now inform future work to establish possible physiological roles for the E2A:TAZ1 and E2A:TAZ2 interactions.

Analysis of Lipoprotein(a) Catabolism

Elevated plasma concentrations of lipoprotein(a) [Lp(a)] have been identified as an independent risk factor for vascular diseases including coronary heart disease and stroke. In the current study, we have examined the binding and degradation of recombinant forms of apolipoprotein(a) [r-apo(a)], the unique kringle-containing moiety of Lp(a), using a cultured cell model. We found that the incubation of human hepatoma (HepG2) cells with an iodinated 17 kringle-containing (17K) recombinant form of apo(a) resulted in a two-component binding system characterized by a high affinity (Kd = 12 nM), low capacity binding site, and a low affinity (Kd = 249 nM), high capacity binding site. We subsequently determined that the high affinity binding site on HepG2 cells corresponds to the LDL receptor. In the HepG2 cell model, association of apo(a) with the LDL receptor was shown to be dependent on the formation of Lp(a) particles from endogenous LDL. Using an apo(a) mutant incapable of binding to the high affinity site through its inability to form Lp(a) particles (17KΔLBS7,8), we further demonstrated that the LDL receptor does not participate in Lp(a) catabolism. The low affinity binding component observed on HepG2 cells, familial hypercholesterolemia (FH) fibroblasts and human embryonic kidney (HEK) 293 cells may correspond to a member(s) of the plasminogen receptor family, as binding to this site(s) was decreased by the addition of the lysine analogue epsilon-aminocaproic acid. The lysine-dependent nature of the low affinity binding site was further confirmed in HepG2 binding studies utilizing r-apo(a) species with impaired lysine binding ability. We observed a reduction maximum binding capacity for 17K r-apo(a) variants lacking the strong lysine binding site (LBS) in kringle IV type 10 (17KΔAsp) and the very weak LBS in kringle V (17KΔV). Degradation of Lp(a)/apo(a) was found to be mediated exclusively by the low affinity component on both HepG2 cells and FH fibroblasts. Fluorescence confocal microscopy, using the 17K r-apo(a) variant fused to green fluorescent protein, further confirmed that degradation by the low affinity component on HepG2 cells does not proceed by the activity of cellular lysosomes. Taken together, these data suggest a potentially significant route for Lp(a)/apo(a) clearance in vivo.

A Quantitative and Mechanistic Assessment of Activated Thrombin-Activatable Fibrinolysis Inhibitor and its Role in Pathological Bleeding and Thrombosis

The coagulation and fibrinolytic systems are linked by the thrombin-thrombomodulin complex which regulates each system through activation of protein C and TAFI, respectively. We have used novel assays and techniques to study the enzymology and biochemistry of TAFI and TAFIa, to measure TAFI activation in hemophilia A and protein C deficiency and to determine if enhancing TAFI activation can improve hemostasis in hemophilic plasma and whole blood. We show that TAFIa not TAFI attenuates fibrinolysis in vitro and this is supported by a relatively high catalytic efficiency (16.41μM-1s-1) of plasminogen binding site removal from fibrin degradation products (FDPs) by TAFIa. Since the catalytic efficiency of TAFIa in removing these sites is ~60-fold higher than that for inflammatory mediators such as bradykinin it is likely that FDPs are a physiological substrate of TAFIa. The high catalytic efficiency is primarily a result of a low Km which can be explained by a novel mechanism where TAFIa forms a binary complex with plasminogen and is recruited to the surface of FDPs. The low Km also suggests that TAFIa would effectively cleave lysines from FDPs during the early stages of fibrinolysis (i.e. at low concentrations of FDPs). Since individuals with hemophilia suffer from premature fibrinolysis as a result of insufficient TAFI activation we quantified TAFI activation in whole blood from hemophilic subjects. Both the rate of activation and the area under the TAFI activation time course (termed TAFIa potential) was determined to be reduced in hemophilia A and the TAFIa potential was significantly and inversely correlated with the clinical bleeding iii phenotype. Using a novel therapeutic strategy, we used soluble thrombomodulin to increase TAFI activation which improved the clot lysis time in factor VIII deficient human plasma and hemophilic dog plasma as well as hemophilic dog blood. Finally, we briefly show in a biochemical case study that TAFI activation is enhanced in protein C deficiency and when afflicted individuals are placed on Warfarin anticoagulant therapy, TAFI activation is reduced. Since TAFIa stabilizes blood clots, this suggests that reducing TAFI activation or inhibiting TAFIa may help restore blood flow in vessels with pathological thrombosis.

Characterization of Glucocorticoid Receptor Promoter Methylation in Breast Cancer

Epidemiological studies have identified psychological stress as a significant risk factor in breast cancer. The stress response is regulated by the HPA axis in the brain and is mediated by glucocorticoid receptor (GR) signalling. It has been found that early life events can affect epigenetic programming of GR, and methylation of the GR promoter has been reported in colorectal tumourigenesis. Decreased GR expression has also been observed in breast cancer. In addition, it has been previously demonstrated that unliganded GR can serve as a direct activator of the BRCA1 promoter in mammary epithelial cells. We propose a model whereby methylation of the GR promoter in the breast significantly lowers GR expression, resulting in insufficient BRCA1 promoter activation and an increased risk of developing cancer. Antibody-based methylated DNA enrichment was followed by qPCR analysis (MeDIP-qPCR) in a novel assay developed to detect locus-specific methylation levels. It was found that 13% of primary breast tumours were hypermethylated at the GR proximal promoter whereas no methylation was detected in normal tissue. RT-PCR and 5’ RACE analysis identified exon 1B as the predominant alternative first exon in the breast. Tumours methylated near exon 1B had decreased GR expression compared to unmethylated samples, suggesting that this region is important for transcriptional regulation of GR. It was also determined that GR and BRCA1 expression was decreased in breast tumour compared to normal tissue. Furthermore, the relative expression of GR and BRCA1 measured by qRT-PCR was correlated in normal tissue but this association was not found in tumour tissue. From this, it appears that lower GR levels with associated decreased BRCA1 expression in tissues may be a predisposing factor for breast cancer. Based on these results we propose a role for GR as a potential tumour suppressor gene in the breast due to its association with BRCA1, also a tumour suppressor gene, as well as its consistently decreased expression in breast tumours and methylation of its proximal promoter in a subset of cancer patients.

Identification and Development of Enzymes Associated With the Sperm Inner Acrosomal Membrane and Their Function during Fertilization

In order for mammalian fertilization to transpire, spermatozoa must transit through the female reproductive tract and penetrate the outer investments of the oocyte: the cumulus oophorus and the zona pellucida. In order to penetrate the oocyte, spermatozoa must undergo the acrosome reaction. The acrosome reaction results in the exposure of the inner acrosomal membrane (IAM) and proteins that coat it to the extracellular environment. After the acrosome reaction, the IAM becomes the leading edge of spermatozoa undergoing progressive movement. Thus the enzymes which effect lysis of the oocyte investments ought to be located on the IAM. An objective of this study was to identify and characterize enzymatic activity detected on the IAM and provide evidence that they play a role in fertilization. This study also describes procedures for fractionating spermatozoa and isolating the IAM and proteins on its intra- and extra-vesicular surfaces, and describes their development during male gametogenesis. Since the IAM is exposed to the extracellular environment and oviductal milieu after the acrosome reaction, this study also sought to characterize interactions and relationships between factors in the oviductal environment and the enzymes identified on the IAM. The data presented provide evidence that MMP2 and acrosin are co-localized on the IAM, originate from the Golgi apparatus in gametogenesis, and suggest they cooperate in their function. Their localization and results of in vitro fertilization suggests they have a function in zona pellucida penetration. The data also provide evidence that plasminogen, originating from the oviductal epithelium and/or cumulus-oocyte complex, is present in the immediate environment of sperm-egg initial contact and penetration. Additionally, plasminogen interacts with MMP2 and enhances its enzymatic action on the IAM. The data also provide evidence that MMP2 has an important function in penetration of the cumulus oophorus. Holistically, this thesis provides evidence that enzymes on the IAM, originating from the Golgi apparatus in development, have an important function in penetration of the outer investments of the oocyte, and are aided in penetration by plasminogen in the female reproductive tract.