Biochemical Effect of Fibrinogen
Chapter One
Research Objectives
- To investigate the biochemical properties and structure-function relationships of fibrinogen.
- To elucidate the role of fibrinogen in physiological processes such as blood coagulation, wound healing, and inflammation.
- To examine the implications of fibrinogen dysregulation in pathological conditions such as cardiovascular disease, thrombosis, inflammation, and cancer.
CHAPTER TWO
LITERATURE REVIEW
Cell Culture–These studies employ the HepG2 human hepatocellular carcinoma cell line as an in vitro model system. HepG2 cells, like native hepatocytes, produce the majority of the circulating coagulation factors. Most importantly for our work, HepG2 cells have been shown to produce fibrinogen (40). Moreover, HepG2 cells have been shown to produce both g chain isoforms of fibrinogen, the gA and g’ forms (41). HepG2 cells have been utilized to examine the hepatic acute phase response in many studies and provide a useful tool to examine the acute phase response of the two fibrinogen g chain isoforms in the studies presented here.
HepG2 cells were grown to 90% confluence in 24-well plates in minimal essential medium (MEM) containing 10% fetal calf serum. Cells were then serum-starved for 24 hours and treated with recombinant cytokines (Chapters 3 and 5) or fibrin degradation products (Chapter 4) for 24 hours. In inhibitor experiments (Chapter 3), cells were pretreated with 5 mM epigallocatechin gallate (EGCG), a specific STAT1 inhibitor, for 30 minutes then treated with recombinant IFN-g for 24 hours. Conditioned media were harvested from the wells, and cells were washed with PBS and lysed in RIPA buffer (150mM NaCl/20mM Tris, pH7.4/0.2% SDS/5mM EDTA/1% NP-40/5mM EACA/1mM leupeptin/1mM pepstatin/0.1mM N-ethylmaleimide/0.1mM phenylmethylsulfonyl fluoride (PMSF)). Total cellular protein concentration was measured by bicinchonic assay (BCA) (Pierce) and was used to normalize ELISA results.
Total Fibrinogen ELISAs–To measure total fibrinogen produced by HepG2 cells, 96 well plates were coated with 1.5 mg/ml AXL203 rabbit anti-human polyclonal antibody (Accurate Chemical) and incubated overnight at 4ºC. Wells were washed three times with PBS/0.1% Triton and blocked in PBS containing 1% BSA/0.1% Triton at 37ºC for 2 hours. Wells were washed and 50 ml of conditioned medium was added to each well and incubated at 37ºC for one hour. Wells were washed again with PBS/0.1% Triton and a 1:2500 dilution of sheep anti-human fibrinogen-horseradish peroxidase (HRP) conjugate was added to the wells and incubated for 1 hour at 37ºC. Wells were washed and incubated with 3,3′, 5,5″-tetramethylbenzidine (TMB) substrate for 30 minutes at room temperature. Stop solution was added to each reaction and absorbance was measured at 450 nm.g’ Fibrinogen ELISAs- Previous studies evaluating g’ fibrinogen in patient samples lacked a standardized and well-characterized assay. ELISAs for g’ fibrinogen were developed by our laboratory so as to be able to easily determine g’ fibrinogen concentrations in plasma samples (42). This assay was modified slightly for measuring g’ levels in cell culture supernatant. Briefly, 96 well plates were coated overnight at 4oC with the monoclonal 2.G2.H9 antibody which is directed against the unique C-terminus of the g’ chain and does not bind to the gA chain (42). Wells were washed three times with PBS/0.1% Triton and blocked in PBS containing 1% BSA/0.1% Triton at 37ºC for 2 hours. Wells were washed and 50 ml of conditioned medium was added to each well and incubated at 37ºC for one hour. Wells were washed again with PBS/0.1% Triton and a 1:2500 dilution of sheep anti-human fibrinogen-HRP conjugate was added to the wells and incubated for 1 hour at 37ºC. Wells were washed and incubated with TMB substrate for 30 minutes at room temperature. Stop solution was added to each reaction and absorbance was measured at 450 nm.
Western Blotting–HepG2 cells were prepared and treated as described above. Following 24 hours of cytokine treatment, cells were washed with PBS and lysed in RIPA buffer. Total cellular protein concentration was determined using a BCA assay and equivalent amounts of protein from the cell lysates were separated by 10% resolving/4% stacking sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Protein was transferred to a nitrocellulose membrane and the membrane was blocked overnight in Western A buffer containing 50mM Tris, pH 7.4/5 mM EDTA/0.05% NP-40/150 mM NaCl/0.25% gelatin . The membrane was probed using rabbit anti-human STAT1 or STAT3 and GAPDH antibodies (Cell Signaling) in Chapter 3, or polyclonal rabbit anti- fibrinogen antibody in Chapter 4. The membrane was stripped and reprobed for phospho- STAT1 or phospho-STAT3 using a rabbit anti-human pSTAT1 or rabbit anti-human pSTAT3 antibody (Cell Signaling) in Chapter 3. Goat anti-rabbit Alexafluor 680 secondary antibodies were used for detection and membranes were imaged using the LiCor Odyssey Imaging System.
CHAPTER THREE
MATERIALS AND METHODS
Reagents–HepG2 cells were purchased from ATCC and grown in MEM containing 0.2 mM L-glutamine/100 units/ml penicillin/100 mg/ml streptomycin/10% fetal calf serum (Invitrogen). Recombinant human IFN-g and IL-6 were purchased from Leinco Technologies. Rabbit anti-human pSTAT1 (Y701), rabbit anti-human STAT1, rabbit anti-human STAT3, rabbit anti-human pSTAT3 (Y705) and rabbit anti-human GAPDH polyclonal antibodies were purchased from Cell Signaling Technologies. IR-700 labeled oligonucleotides were purchased from IDT Technologies. AXL203 rabbit anti-fibrinogen antibody and HRP-labeled sheep anti-human fibrinogen antibody were purchased from Accurate Chemical Corp. TMB HRP substrate and 450nm Stop Solution were purchased from BioFX. Epigallocatechin gallate (EGCG), a specific STAT1 inhibitor from green tea extract, was purchased from Sigma.
CHAPTER FOUR
RESULTS
Characterization of the Effects of IFN-g on Basal Fibrinogen Production in HepG2 Cells–Previous studies have shown that the presence of IFN-g leads to a decrease in fibrin deposition during Toxoplasma gondii infection in mice (49) and delayed wound healing (47). Additionally, the association between a locus located near the gene encoding IRF-1 and circulating fibrinogen levels (50) led us to investigate the role of IFN-g in regulating fibrinogen production. As shown by ELISA, treatment of HepG2 cells with increasing concentrations of IFN-g within the physiological range caused a dose dependent, although not statistically significant, decrease in the amount of total fibrinogen secreted into the media after 24 hours (Figure 3.2A). At the highest concentration, production of fibrinogen from HepG2 cells was decreased by approximately 25%. The production of total fibrinogen mRNA was also decreased following treatment of HepG2 cells with IFN-
g, as shown by quantitative real-time PCR (Figure 3.2B) but again, this data was not statistically significant. These data therefore show that at physiological concentrations, IFN-g may be able to inhibit production of fibrinogen at the protein and mRNA level in HepG2 cells, although the data are inconclusive.
CHAPTER FIVE
SUMMARY AND CONCLUSIONS
Total fibrinogen and g’ fibrinogen are two independent risk factors that have been shown to increase the risk for myocardial infarction, coronary artery disease and stroke.
Determining novel mechanisms that regulate the production of both g chain isoforms of fibrinogen is an important step in understanding of the etiology of cardiovascular disease.
In addition to being an important hemostatic protein, fibrinogen is also a known acute phase reactant. Fibrinogen levels increase in response to certain inflammatory- and infection-mediated stimuli due to increased hepatic synthesis. Most importantly for this work, IL-6 is known to increase the production of fibrinogen up to 6-fold in HepG2 cells (53). Previously, little was known about the ability of IFN-g to regulate fibrinogen levels.
Following a recent study identifying a member of the interferon signaling pathway as a mediator of circulating fibrinogen levels (50), we examined the role of IFN-g in fibrinogen regulation. As discussed in Chapter 3, we have shown that IFN-g is a novel regulator of basal and IL-6 mediated fibrinogen production and that this regulation occurs at the promoter level. We saw an approximate 25% decrease in the production of fibrinogen from HepG2 cells treated with IFN-g for 24 hours. We also saw a concomitant decrease in the amount of fibrinogen mRNA produced. During this time period, the downstream mediator of IFN-g, STAT1, was phosphorylated and activated. We also identified a novel GAS within the promoter of the fibrinogen gamma gene, which we have shown is functional and binds to activated STAT1 within the nucleus. We have also shown that pSTAT1 is able to bind to an identified IL-6 response element, blocking the ability of STAT3 to bind to this element downstream of IL-6 signaling. This led to a decrease in the induction of fibrinogen synthesis by IL-6 when co-treated with IFN-g. These findings support observations by others that the loss of IFN-g leads to increased fibrin deposition during T. gondii infection (48). While some of these effects have been attributed to an increase in fibrinolysis following IFN-g administration (49), our results suggest that a decrease in the production of fibrinogen may also play a role. While our results point to IFN-g as a regulator of fibrinogen synthesis in vitro, the role of IFN-g must also be assessed in vivo. In the future, administration of IFN-g to normal and hyperfibrinogenemic mice, in the presence and absence of infection, will allow us to more definitively assign a role for IFN-g in fibrinogen synthesis. There are also two additional putative GAS sequences that we have identified, one each within the a and b fibrinogen gene promoters. While our preliminary results suggest that pSTAT1 is not able to bind to the b promoter GAS, the a GAS is able to bind to pSTAT1. Future studies will allow us to determine whether this GAS site is functional and whether binding of STAT1 to this GAS positively or negatively influences the production of fibrinogen.
Because little is known about the mechanisms regulating g’ fibrinogen synthesis, identifying potential modulators of g’ fibrinogen levels will allow us to expand our understanding of its physiological and pathophysiological functions. At this point in time, our understanding of how the g’ isoform of fibrinogen is produced is, for the most part, an unproven theory that is nonetheless accepted by the field. This theory states that g’ fibrinogen mRNA is produced when intron 9 of the fibrinogen gene is polyadenylated and cleaved before it can be removed (13,14). This suggests that spatiotemporal regulation of both the spliceosome and enzymes involved in polyadenylation may play a significant role in generating more of one isoform over the other. Moreover, any modifications to these enzymes may alter their affinity for the target pre-mRNA, in turn altering the splicing or polyadenylation of intron 9. In Chapter 4, we discuss the novel role of fibrin degradation products, namely D-dimer, in the regulation of g’ fibrinogen synthesis. Upon treatment with D-dimer, HepG2 cells produce less g’ fibrinogen protein and mRNA while the amount of total fibrinogen protein and mRNA produced remains stable. This suggests that D-dimer is somehow altering the splicing or polyadenylation of the fibrinogen g chain. This data partially agrees with previous work that has shown the addition of D-dimer and D-fragment to cultured cells does not lead to a change in the amount of total fibrinogen produced (75). Previous reports have shown that D-dimers are able to affect the transcription of target genes, such as PAI-1, by activating the transcription factor AP-1 components c-fos/junD (69). While we see an decrease in the amount of g’ fibrinogen produced by HepG2 cells treated with D-dimer, we do not see a concomitant decrease in the amount of total fibrinogen that would suggest a promoter level effect of D-dimers, similar to the effect seen on PAI-1. A limitation of the work described here is that we are using an immortalized cell line, which has been altered in order to survive multiple passages. In order to truly determine the role of D-dimer signaling as it occurs naturally, future studies include the administration of murine D- dimer to mice in order to determine its effects on circulating plasma fibrinogen levels.
Although we do not expect to see a promoter effect following treatment of our cells with D-dimer, EMSAs may be performed in order to assess the binding of transcription factors to the fibrinogen gene promoters following treatment with D-dimers. Chapter 4 also discusses the search for a receptor for D-dimer on the surface of HepG2 cells. Previous work suggested that fibrin degradation products, especially D-dimer, were taken up by murine hepatocytes as a clearance mechanism (70). The macrophage receptor for murine FDPs containing the D domain was later characterized (118), but the identity of a receptor on the surface of liver cells was not examined. We were unable to detect the presence of a receptor for human D-dimer on the surface of our HepG2 cells, as our binding studies did not suggest saturable binding. In fact, the results of our binding studies coupled with the finding of several endocytic plasma membrane proteins eluted from a D-dimer affinity column strongly suggests that D-dimers are taken up by HepG2 cells via endocytosis. However, these studies remain novel in that we have shown that once HepG2 cells have taken up D-dimers, the D-dimers travel to the perinuclear space where they presumably signal to the nucleus to exert their effects on g’ fibrinogen synthesis. Future studies looking at the exact location of internalized D-dimers will be very informative and will help to determine how D-dimers are exerting their effects intracellularly. These studies would involve staining of the individual cellular compartments with several different fluorescent markers and determining by microscopy which of these cellular markers co-localizes with D-dimers.
In addition to determining the effects of D-dimer on the production of the fibrinogen g chain isoforms, this work also examined the effects of several inflammatory cytokines on total and g’ chain production, as described in Chapter 5. We have identified IL-6, TNF-a and IFN-g as molecules that are able to differentially regulate the production of total and g’ fibrinogen. Each of these cytokines seems to alter the ratio of g’/total fibrinogen in a unique way, as IL-6 increases total fibrinogen but not g’ fibrinogen, TNF-a decreases g’ fibrinogen but does not affect total fibrinogen levels, and IFN-g decreases total fibrinogen but does not alter the levels of g’ fibrinogen. This is an important finding in that all three of these mediators are involved in the acute phase response, and are most likely working either synergistically or antagonistically during the acute phase. The identification of known signaling molecules as modulators of total and g’ fibrinogen levels allows us to better understand the origins of cardiovascular disease and may also point to possible future therapeutic targets. While these findings present novel regulators of g’ fibrinogen synthesis, the molecular mechanisms behind these observations must be elucidated in order to fully comprehend the spatial and temporal regulation of fibrinogen g gene splicing and polyadenylation. An in vitro splicing assay coupled with qRT-PCR would be an incredibly useful tool to help determine whether the addition of these cytokines to a splicing reaction truly alters splicing. Should these assays show differential splicing patterns upon the addition of cytokines, western blots and qRT-PCR could be performed in order to detect differences in the expression, localization or phosphorylation of known splicing factors such as the SR family of splicing proteins. This same set of experiments could also be performed using antibodies directed against enzymes known to function in polyadenylation in order to detect differences between treated and untreated cells as well as differences between different cytokine treatments. Finally, the role of two recently identified polymorphisms within intron 9 of the g chain of fibrinogen in alternative splicing should be investigated. Previous studies have proven that IL-6 was independently associated with total fibrinogen levels and, conversely, a single nucleotide polymorphism (SNP) within the fibrinogen gamma gene, FGG 9340 T/C, influenced serum IL-6 levels (119). Furthermore, two additional gamma gene SNPs, FGG 9615 C/T and FGG 10034 C/T have been identified as belonging to a haplotype with an increased g’ fibrinogen level in a case-control study of venous thrombosis (99). The authors speculate that the FGG 10034 C/T SNP, located within a region downstream of the polyadenylation site which binds the Cleavage Stimulatory Factor (CstF), may be associated with increased g’ fibrinogen levels as its presence may increase polyadenylation and cleavage (99). The interaction of this set of SNPs with circulating IL-6 levels as well as SNPs within the IL-6 gene could be examined using a genome wide association study and may provide a mechanistic rationale as to why IL-6 may be differentially regulating g’ fibrinogen levels in our experiments.
While regulation of the levels of total and g’ fibrinogen are critical in understanding their role in cardiovascular disease, a dysfunctional fibrinogen presents a different set of physiological challenges. In Chapter 6 we discuss the identification of a fibrinogen g chain mutation, which was previously shown to be predominantly thrombotic (115), in a patient presenting with bleeding. This patient had a diagnosed dysfibrinogenemia that we identified as a mutation in the fibrinogen g gene at residue 275. This mutation caused a C>T transition, leading to a missense mutation in which a free cysteine reside was incorporated at g chain residue 275 in place of an arginine. We show that this free cysteine does not bind to circulating albumin, nor does it drastically change the separation pattern of the g chain. These results suggest that the R275C alone is capable of producing a bleeding phenotype as well as thrombotic and asymptomatic phenotypes.
In conclusion, this work has shown that IFN-g is a novel regulator of total fibrinogen synthesis via a STAT1 dependent mechanism, while D-dimers, IL-6, TNF-a and IFN-g are differential regulators of total and g’ fibrinogen production, both of which have been identified as independent risk factors for cardiovascular disease. These findings may lead to the use of agonists or inhibitors of these cytokines in a therapeutic manner to decrease total or g’ fibrinogen, thus decreasing the risk for cardiovascular disease. Finally, this work has shown that the fibrinogen R275C gamma chain mutation is sufficient to cause a hemorrhagic phenotype. The phenotype of this mutation is known to be predominantly asymptomatic, and symptomatic patients have been overwhelmingly thrombotic.
Therefore the finding of a patient with this mutation presenting with a hemorrhagic phenotype is novel and increases our knowledge of the molecular mechanisms of dysfibrinogenemias.
References
- Collen, The plasminogen (fibrinolytic) system. Thrombosis and Haemostasis. 1999 82:259-70.
- Hoylaerts, M, Rijken, DC, Linjen, HR, Collen, Kinetics of the activation of plasminogen. Journal of Biological Chemistry. 1982 257:2912-9.
- Medved, L and Nieuwenhuizen, Molecular mechanisms of initiation of fibrinolysis by fibrin. Thrombosis and Haemostasis. 2003 89:409-19.
- Kant JA, Fornace AJ, Saxe D, Simon MI, McBride OW, Crabtree GR. Evolution and organization of the fibrinogen locus on chromosome 4: Gene duplication accompanied by transposition and inversion. Proceedings of the National Academy of Sciences of the United States of America. 1985 April;82(8):2344-8.
- Simpson Haidaris Induction of fibrinogen biosynthesis and secretion from cultured pulmonary epithelial cells. Blood. 1997 February 1;89(3):873-82.
- Roy S, Overton O, Redman Overexpression of any fibrinogen chain by hep G2 cells specifically elevates the expression of the other two chains. Journal of Biological Chemistry. 1994 January 7;269(1):691-5.
- Suh TT, Holmbäck K, Jensen NJ, Daugherty CC, Small K, Simon DI, et Resolution of spontaneous bleeding events but failure of pregnancy in fibrinogen-deficient mice. Genes & Development. 1995 August 15;9(16):2020-33.
- Yu S, Sher B, Kudryk B, Redman Intracellular assembly of human fibrinogen. Journal of Biological Chemistry. 1983 November 25;258(22):13407-10.
- Hartwig R, Danishefsky Studies on the assembly and secretion of fibrinogen.Journal of Biological Chemistry. 1991 April 5;266(10):6578-85.
- Roy SN, Mukhopadhyay G, Redman CM. Regulation of fibrinogen assembly. transfection of hep G2 cells with B beta cDNA specifically enhances synthesis of the three component chains of fibrinogen. Journal of Biological Chemistry. 1990 April 15;265(11):6389-93.
