The coordinated regulation of fibrinogen gene transcription by hepatocyte-stimulating factor and dexamethasone.

Glucocorticoids and hepatocyte-stimulating factor (HSF; a monocyte/macrophage-derived polypeptide) are potent regulators of fibrinogen biosynthesis. Using primary rat hepatocytes and a rat hepatoma cell line (FAZA) we have determined, more precisely, the interaction between these two molecules in the control of fibrinogen production. When dexamethasone (DEX) or HSF is added to the cells, there is a substantial increase in fibrinogen production (1.5-3-fold). However, if both agents are administered simultaneously the response is much greater with a 15-20-fold rise in synthesis. Quantitative RNA analysis demonstrates that when the factors are present individually only HSF elevates fibrinogen mRNA levels, but the effect is much enhanced in the presence of DEX. This pattern is also seen in the results of the in vitro transcription assays which allow quantitation of mRNA synthesis in isolated nuclei. Cycloheximide does not significantly interfere with the increased transcription brought about by HSF in either cell type. However, the DEX enhancement is blocked by cycloheximide in FAZA cells, thus indicating that in the transformed cell protein synthesis is required for maximal transcription to occur. Data presented here demonstrates the requirement for two types of regulator molecules in the control of fibrinogen gene expression; a polypeptide hormone (HSF) that increases transcription and a steroid (DEX) that enhances the action of the polypeptide.

Gene transfer and expression of human phenylalanine hydroxylase.

Phenylketonuria (PKU) is caused by a genetic deficiency of the enzyme phenylalanine hydroxylase (PAH). A full-length complementary DNA clone of human PAH was inserted into a eukaryotic expression vector and transferred into mouse NIH3T3 cells which do not normally express PAH. The transformed mouse cells expressed PAH messenger RNA, immunoreactive protein, and enzymatic activity that are characteristic of the normal human liver products, demonstrating that a single gene contains all of the necessary genetic information to code for functional PAH. These results support the use of the human PAH probe in prenatal diagnosis and detection of carriers, to provide new opportunities for the biochemical characterization of normal and mutant enzymes, and in the investigation of alternative genetic therapies for PKU.

The role of carbohydrate in the function of human plasminogen: comparison of the protein obtained from molecular cloning and expression in Escherichia coli and COS cells.

A cDNA library was constructed in the phage lambda gt11 from human liver mRNA enriched for plasminogen mRNA by chromatography on Sepharose 4B. A full-length cDNA clone of human plasminogen was isolated. The 2.7 kb cDNA encoded the entire plasminogen molecule, a signal peptide sequence and two start codons with a 5'-untranslated region of about 80 base pairs. In the 3'-non coding region of 280 base pairs a consensus signal AATAAA was found at a distance of 46 base pairs upstream of the poly(A) tail. The plasminogen cDNA was subcloned in the eukaryotic expression vector p91023 (B), and human plasminogen was expressed in monkey kidney (COS m6) cells and in Escherichia coli. The recombinant molecule obtained from COS cells has physicochemical and biological properties similar to native human plasminogen I, indicating that it has folded in a manner similar to plasminogen synthesized by liver. By contrast, plasminogen expressed in E. coli could not be activated and showed biological properties which are very different from glycosylated forms of plasminogen. However, the non-glycosylated plasminogen was bound by lysine-Sepharose and reacted with a conformation dependent monoclonal antibody to kringles 1 to 3. These data suggest that the protein has properly folded kringle domains. Our studies suggest that the carbohydrate domains may play an important role in the function of the plasminogen molecule.

Immunochemical properties of abnormal hemoglobins C-Harlem (beta 6 Glu replaced by Val, beta 73 Asp replaced by Asn), S (beta 6 Glu replaced by Val), Korle Bu (beta 73 Asp replaced by Asn), Vancouver (beta 73 Asp replaced by Tyr), and Mobile (beta 73 Asp replaced by Val).

Antibodies were made to three mutant hemoglobins, each containing different single amino acid substitutions at beta 73:Hb Korle Bu (Asp replaced by Asn), Hb Mobile (Asp leads to Val), Hb Vancouver (Asp replaced by Tyr); and to one mutant hemoglobin, Hb C-Harlem, containing two substitutions in the beta chain (beta 6 Glu replaced by Val, as in Hb S, and beta 73 Asp replaced by Asn, as in Hb Korle Bu). The antiserum to Hb C-Harlem contained two antibody populations, each specific for one mutant amino acid residue. The antiserum to Hb Vancouver was completely specific for this mutant and did not cross-react with Hb Mobile and Hb Korle Bu; however, antiserum to Hb Korle Bu partially cross-reacted with Hb Mobile and to a smaller degree with Hb Vancouver. Antiserum to Hb Mobile exhibited even less cross-reactivity with Hb Korle Bu and C-Harlem and none with Hb Vancouver. These and previous studies indicate the involvement of at least three independent areas in the beta chain as antigenic determinant sites. It appears that the three mutants at beta 73 elicit the formation of antibodies which have a gradation in their specificity due to the nature of the amino acid sidechain.

Identification of a human cDNA with high homology to yeast omnipotent suppressor 45.

Omnipotent suppression is a well-established phenomenon in yeast and bacteria in which nonsense mutations are misread. Wild-type (wt) suppressors are presumed to be involved in ensuring the fidelity of translation. We report a human homolog to wt yeast omnipotent suppressor 45 which shares 63% identity at the nucleotide level in the area of open reading frame (ORF) and 73% similarity at the amino acid (aa) level. The aa sequence of the human protein was deduced from a 2.3-kb cDNA (TB3-1) isolated from an adenocarcinoma T84 cell line cDNA library. The cDNA contains an ORF of 1284 bp which encodes a 47.8-kDa protein. Two transcripts for the clone were identified (2.6 and 4.0 kb) in a variety of human cell types. The strong structural similarity to yeast omnipotent suppressor 45, and its widespread expression suggest that this cDNA may play a role in the accurate recognition of nonsense codons in mammalian cells.

Retroviral mediated transfer and expression of human alpha 1-antitrypsin in cultured cells.

Genetic deficiency of alpha 1-antitrypsin in man is a predisposing factor to emphysema and a disorder potentially correctable by somatic gene therapy. A full-length human alpha 1-antitrypsin cDNA was cloned into a retroviral vector and introduced into cells which package the recombinant gene in a retroviral capsule. Cells infected with the recombinant retrovirus express human alpha 1-antitrypsin mRNA and protein. The recombinant protein is glycosylated, secreted and exhibits anti-protease activity against human neutrophil elastase.

Isolation and characterization of biologically active murine interleukin-6 produced in Escherichia coli.

Interleukin-6 (IL-6) is a multi-functional cytokine produced and secreted by several different cell types, including those of the immune system. A cDNA coding for the mature murine IL-6 (mIL-6), which extends from amino acid (aa) 25 through 211, was cloned into a prokaryotic vector and then expressed in Escherichia coli. The recombinant mIL-6 (remIL-6) was isolated from bacterial inclusion bodies by solubilization in 4 M guanidine hydrochloride followed by gel-filtration chromatography. The protein was refolded to an active conformation by dialysis against 25 mM Na. acetate pH 5.5. A final step of purification and concentration on a cation exchange resin yielded pure and biologically active remIL-6. The purified preparation had the expected aa composition, as confirmed by aa analysis and pI of 7.0-7.1. The biological activity of the recombinant protein was measured in two systems; a proliferation assay employing 7TD1 cells, and a fibrinogen biosynthesis assay employing primary rat hepatocytes. Both assay systems demonstrated that the remIL-6 was active in the range of 10(8) units/mg, which is similar to that estimated for native cytokine. Antibodies raised in rabbits against remIL-6 neutralized the biological activity of both recombinant and native IL-6.

Identification and characterization of a C-terminally extended form of recombinant murine IL-6.

Murine interleukin-6 (mIL-6) was expressed in Escherichia coli in the insoluble fraction of cell lysates. Approximately equal amounts of two polypeptide species, reactive with anti-IL-6 antibodies, were produced. The two forms of mIL-6 were isolated and found to have identical N-terminal sequences initiated by Met-Phe-Pro-Thr-Ser-Gln-. Peptide mapping after endoproteinase glu-C digestion led to isolation and characterization of the C-terminal peptides from each of the two forms and allowed the source of the heterogeneity to be identified as a C-terminal addition of three amino acids, Gln-Lys-Leu, to authentic mIL-6. Inspection of the nucleotide sequence of the plasmid containing the mIL-6 gene and expression of the plasmid in other strains suggested that the addition of three amino acids was caused by a readthrough of the termination codon arising from an unexpected suppressor mutation in the original host strain. Although the C-terminus of IL-6 is critical for the activity of this cytokine, the IL-6 variant with extended C-terminus was fully active in two separate bioassays. This suggests that the additional amino acids do not disrupt the structure or function of this important region of the molecule.

Induction and regulation of interleukin-6 gene expression in rat astrocytes.

Cells that produce interleukin-6 (IL-6) require the presence of signaling molecules since this cytokine is not normally constitutively expressed. It is now established that astrocytes produce IL-6; however, the precise inducing molecules and the kinetics of their action have not yet been clearly identified. In the current study, we show that either interleukin-1 beta (IL-1 beta) or tumor necrosis factor-alpha (TNF-alpha) exert a strong inducing signal for IL-6 in primary rat astrocytes. When the two cytokines are added together the response is synergistic, suggesting that each cytokine may induce IL-6 gene expression by different pathways. Interferon-gamma (IFN-gamma) does not affect IL-6 expression although if it is added in conjunction with IL-1 beta, an augmented induction of IL-6 occurs. In addition to the cytokines, bacterial lipopolysaccharide (LPS) and the calcium ionophore, A23187, induce IL-6 expression. IL-6 expression can be blocked by the glucocorticoid analogue, dexamethasone. IL-6 induction by LPS/Ca2+ ionophore is more sensitive to the suppressive effects of dexamethasone than is IL-6 induction by TNF-alpha/IL-1 beta. Cycloheximide (CHX), an inhibitor of protein synthesis, markedly increased levels of IL-6 mRNA in both unstimulated and stimulated astrocytes, indicating that ongoing protein synthesis is not required for astrocyte IL-6 gene expression. We propose that astrocyte-produced IL-6 may have a role in augmenting intracerebral immune responses in neurological diseases such as multiple sclerosis (MS), AIDS dementia complex (ADC), and viral infections. These diseases are characterized by infiltration of lymphoid and mononuclear cells into the central nervous system (CNS), and intrathecal production of immunoglobulins. IL-6 may act to promote terminal differentiation of B cells in the CNS, leading to immunoglobulin synthesis.

Expression of plasminogen activator inhibitor type I in genotyped human endothelial cell cultures: genotype-specific regulation by insulin.

Patients with non-insulin-dependent diabetes mellitus frequently have been associated with elevation in plasma levels of PAI-1. Part of the variations in individual plasma PAI-1 levels have been attributed to variations in the PAI-1 gene. In order to determine whether insulin regulates PAI-1 expression in a genotype-specific manner, individual human umbilical vein ECs (HUVECs) were genotyped using a Hind III RFLP and incubated in the absence/presence of insulin. Treatment of 1/1 PAI-1 genotype HUVECs with insulin increased secretion of PAI-1 antigen approximately 1.7 to 2.2-fold and mRNA levels were increased approximately 1.8 to 2.8-fold. Treatment of HUVECs with actinomycin D or puromycin completely abolished the induction of PAI-1 by insulin. The nuclear run-on assays indicated approximately 3-4 fold increase in PAI-1 transcription rates. These in vitro studies with the 1/1 PAI-1 genotyped cultured HUVECs, suggests that hyperinsulinemia may be expected to increase EC PAI-1 synthesis in those patients with the responsive 1/1 genotype.

Ethanol downregulates transcription of the PAI-1 gene in cultured human endothelial cells.

Human endothelial cells are a major site of synthesis for plasminogen activator inhibitor type-1. Elevated plasminogen activator inhibitor type-1 levels in young survivors of myocardial infarction [1] suggest that plasminogen activator inhibitor type-1 may have an important pathologic role in the development of coronary artery disease. Epidemiological studies indicate that moderate alcohol consumption (1-2 drinks/day) reduces the risk for cardiovascular mortality. This cardioprotective benefit has been attributed in part to an increase in fibrinolysis, which decreases fibrin-based thrombosis. The studies described herein were performed to determine whether moderate levels of ethanol affect plasminogen activator inhibitor type-1 gene expression. Cultured human endothelial cells were exposed to 0.1% v/v ethanol for 1 hour. Following incubation in the absence of ethanol plasminogen activator inhibitor type-1, mRNA levels were decreased in a time- and dose-dependent manner, reaching a maximum decrease of 3- to 4-fold at 2 to 4 hours following ethanol challenge. This decline in mRNA occurs at the transcription level; therefore, nuclear transcription run-on assays were performed. A 2.5- to 5-fold decrease in the rate of plasminogen activator inhibitor type-1 gene transcription was measured at 2 and 4 hours following ethanol challenge. Next, a 3.4- and a 1.1-kb fragment from the plasminogen activator inhibitor type-1 promoter region were linked to a luciferase reporter gene, and these constructs were transfected into human endothelial cells. Treatment of these transiently transfected human endothelial cells with ethanol showed a 2- to 3.5-fold decrease in promoter activity, respectively. These results indicate that low doses of ethanol downregulate transcription of the plasminogen activator inhibitor type-1 gene in cultured human endothelial cells. However, the mechanism(s) for this transcriptional decrease is currently unknown.

Expression of PAI-1, t-PA and u-PA in cultured human umbilical vein endothelial cells derived from racial groups.

To determine whether inherent fibrinolytic differences may exist in racial groups (black americans, BA vs. white americans, WA), 55 different individual racially-derived human umbilical vein endothelial cell (HUVEC) cultures (35 BA and 20 WA) were analyzed in terms of their fibrinolytic protein (t-PA, u-PA and PAI-1) antigen and mRNA levels. Values (mean +/- SD) for measured fibrinolytic component levels include: cell-associated t-PA antigen (ELISA), 1.14 +/- 0.82 ng/ml/8.6 x 10(5) cells/24 hr in BA and 0.70 +/- 0.85 ng/ml in WA (p = 0.0624); secreted t-PA antigen, 18.65 +/- 17.06 ng/ml in BA and 10.37 +/- 6.38 ng/ml in WA (p = 0.0422); t-PA/cyclophilin mRNA ratios (Northern blot analysis), 1.90 +/- 1.34 in BA and 1.32 +/- 0.70 in WA (p = 0.0776); cell-associated PAI-1 antigen, 71.10 +/- 30.16 ng/ml/8.6 x 10(5) cells/24 hr in BA and 108.85 +/- 56.89 ng/ml in WA (p = 0.0022); secreted PAI-1 antigen, 1,582.13 +/- 612.67 ng/ml in BA and 1,992.17 +/- 711.50 ng/ml in WA (p = 0.0285); 2.4 kb PAI-1/cyclophilin mRNA ratios, 0.59 +/- 0.39 in BA and 0.79 +/- 0.31 in WA (p = 0.1085); 3.4 kb PAI-1/cyclophilin mRNA ratios, 0.70 +/- 0.47 in BA and 0.77 +/- 0.54 in WA (p = 0.6322). These combined data suggest that cultured HUVECs from BA express significantly higher levels of t-PA, lower levels of PAI-1 and approximately 1.72-fold lower molar ratio of PAI-1/t-PA antigen (183.99 +/- 168.81 vs. 315.92 +/- 164.99) (p < 0.05) than cultured HUVECs from WA, presumably reflecting an apparent inherent increased fibrinolytic potential in cultured HUVEC derived from BA.

Identification of the Hind III polymorphic site in the PAI-1 gene: analysis of the PAI-1 Hind III polymorphism by PCR.

The identification of the Hind III polymorphic site in the 3' end of the plasminogen activator inhibitor 1 (PAI-1) gene and a simple method to identify the Hind III polymorphism rapidly in the PAI-1 gene using PCR is described. The Hind III restriction site was identified by restriction site mapping and sequence analysis from a cosmid DNA clone. Genomic DNA was isolated from individual human umbilical cords and a 754-bp fragment of the human PAI-1 gene was amplified by PCR. Aliquots of the PCR products were digested with Hind III and analyzed by agarose gel electrophoresis. The presence of two fragments, 754 and 567 bp, was identified, and they were designated as 1/1 (750-bp band), 1/2 (754- and 567-bp bands), and 2/2 (567-bp band). The PCR method is considerably less time consuming than the conventional DNA genotyping using Southern blot analysis. To ensure that this new method identified the same PAI-1 genotypes as previously identified by Hind III restriction fragment length polymorphism (RFLP), samples were simultaneously genotyped by PCR and Southern blot analysis. Both methods identified the same Hind III genotypes in all the samples, confirming the reliability of this new PCR method for the rapid identification of the Hind III polymorphism in the human PAI-1 gene.

Quercetin induced tissue-type plasminogen activator expression is mediated through Sp1 and p38 mitogen-activated protein kinase in human endothelial cells.

BACKGROUND: Wine polyphenol quercetin upregulates tissue-type plasminogen activator (t-PA) transcription in cultured human umbilical cord vein endothelial cells (HUVECs). However, the regulatory elements and signaling pathways involved in this regulation are unknown. OBJECTIVES: We aimed to localize quercetin-responsive t-PA promoter elements, identify the proteins that bind these elements, and decipher signaling pathways involved in the regulation of t-PA. METHODS: To localize quercetin-responsive elements, HUVECs were transiently transfected with various t-PA promoter-reporter constructs. Element functionality was evaluated by mutational analysis. Nuclear protein-t-PA element interactions were evaluated by electrophoretic mobility shift assays (EMSAs) and chromatin immunoprecipitation (ChIP) analysis. Mitogen-activated protein kinase (MAPK) inhibitors were used to determine the signaling pathways involved in t-PA regulation. MAPK inhibition effects were evaluated by real-time PCR, immunoblotting analysis, and transfections. Coimmunoprecipitation was used to evaluate MAPK and transcription factor interaction. RESULTS: Deletion of the t-PA promoter region - 288 to - 250 resulted in loss of quercetin responsiveness. This region contains putative Sp1-binding elements, which we termed Sp1a and Sp1b. Sp1b mutation abolished the quercetin-inducible response, whereas Sp1a mutation had no effect. EMSA and ChIP analysis demonstrated quercetin-enhanced Sp1 binding to Sp1b. Inhibition of p38 MAPK abrogated basal and quercetin-induced t-PA expression and promoter activity, as well as quercetin-induced Sp1 binding to Sp1b. Quercetin enhanced p38 MAPK and Sp1 physical association, which was similarly diminished by p38 MAPK inhibition. CONCLUSIONS: We showed, for the first time, the presence of a functional Sp1-binding element in the t-PA promoter controlling quercetin induction via the p38 MAPK pathway. Understanding these mechanisms may provide new insights into polyphenol cardioprotective effects.

Identification and quantitation of sickle cell hemoglobin with an enzyme-linked immunosorbent assay (ELISA).

The experimental details of ELISA for the identification and quantitation of Hb S are presented; the assay is based upon the passive adsorption of Hb S top a solid phase (polystyrene tubes) and the addition of monospecific rabbit antibodies capable of recognizing the (beta 6 Glu leads to Val) substitution in Hb S. After the addition of alkaline phosphatase-conjugated goat antibody to rabbit IgG and substrate, the yellow color produced by hydrolysis of substrate is measured spectrophotometrically. For the identification and quantitation of Hb S in unknown samples, the hemolysate is added to the Hb S-coated tubes before the addition of antibody to Hb S, thus causing an inhibition of the antigen-antibody reaction as evidenced by an absence or reduction of color formation. With this procedure, there is no cross-reactivity with normal hemoglobins, and the immunoassay has a sensitivity in detecting 50 ng quantities of the abnormal hemoglobin in a 5 microgram hemolysate. The assay can be performed on multiple samples in 1 day and offers many advantages over other techniques currently used for the identification and quantitation of Hb S and other abnormal hemoglobins in the clinical laboratory.

Biochemical characterization of recombinant human phenylalanine hydroxylase produced in Escherichia coli.

A full-length human phenylalanine hydroxylase cDNA has been recombined with a prokaryotic expression vector and introduced into Escherichia coli. Transformed bacteria express phenylalanine hydroxylase immunoreactive protein and pterin-dependent conversion of phenylalanine to tyrosine. Recombinant human phenylalanine hydroxylase produced in E. coli has been partially purified, and biochemical studies have been performed comparing the activity and kinetics of the recombinant enzyme with native phenylalanine hydroxylase from human liver. The optimal reaction conditions, kinetic constants, and sensitivity to inhibition by aromatic amino acids are the same for recombinant phenylalanine hydroxylase and native phenylalanine hydroxylase. These data indicate that the recombinant human phenylalanine hydroxylase is an authentic and complete phenylalanine hydroxylase enzyme and that the characteristic aspects of phenylalanine hydroxylase enzymatic activity are determined by a single gene product and can be constituted in the absence of any specific accessory functions of the eukaryotic cell. The availability of recombinant human phenylalanine hydroxylase produced in E. coli will expedite physical and chemical characterization of human phenylalanine hydroxylase which has been hindered in the past by inavailability of the native enzyme for study.

Endothelial cell fibrinolysis: transcriptional regulation of fibrinolytic protein gene expression (t-PA, u-PA, and PAI-1) by low alcohol.

Epidemiological studies have associated moderate alcohol consumption with a reduced risk for coronary artery disease (CAD) and myocardial infarction (MI). This cardioprotection may be attributed to alcohol-induced changes in a variety of cellular functions, including increased fibrinolysis. Fibrinolysis is important in regulating normal hemostasis. Endothelial cells (ECs) synthesize fibrinolytic proteins, t-PA, u-PA, and PAs inhibitor, PAI-1. Systemic factors, i.e., alcohol, that affect one or more of these components, resulting in increased EC fibrinolysis, will reduce the risk for thrombosis, CAD, and MI and afford cardioprotection. These studies will identify/define the effects of low ethanol (< 0.1%, v/v) on the expression of PAs, PAI-1, and surface-localized fibrinolytic activity in cultured ECs. Low ethanol exerted a short-term time- and dose-dependent increase (approximately 5- to 8-fold) in activity at approximately 20 min and 0.05% ethanol, which was sustained for approximately 1 hr. On the other hand, a single brief exposure to low ethanol (< 0.1%, < 120 min), followed by 4-24 hr incubation in the absence of ethanol, showed a time- and dose-dependent increase (approximately 2- to 3-fold) in PAs antigen/mRNA and a concomitant approximately 2- to 3-fold sustained increase (approximately 24 hr) in fibrinolytic activity. Further nuclear transcription run-on assays and transient transfection experiments, using pPAs/luc and pPAI-1/luc promoter constructs, demonstrated that low ethanol transcriptionally upregulates t-PA and u-PA gene expression and downregulates PAI-1 gene expression. These combined studies have described a feasible molecular mechanism by which low ethanol can induce and sustain increased surface-localized EC fibrinolysis that may underlie and contribute, in part, to the cardioprotective benefit associated with moderate alcohol consumption.

Retroviral-mediated gene transfer of human phenylalanine hydroxylase into NIH 3T3 and hepatoma cells.

Phenylketonuria (PKU) is caused by deficiency of the hepatic enzyme phenylalanine hydroxylase (PAH). A full-length human PAH cDNA sequence has been inserted into pzip-neoSV(X), which is a retroviral vector containing the bacterial neo gene. The recombinant has been transfected into psi 2 cells, which provide synthesis of the retroviral capsid. Recombinant virus was detected in the culture medium of the transfected psi 2 cells, which is capable of transmitting the human PAH gene into mouse NIH 3T3 cells by infection leading to stable incorporation of the recombinant provirus. Infected cells express PAH mRNA, immunoreactive PAH protein, and exhibit pterin-dependent phenylalanine hydroxylase activity. The recombinant virus is also capable of infecting a mouse hepatoma cell line that does not normally synthesize PAH. PAH activity is present in the cellular extracts and the entire hydroxylation system is reconstituted in the hepatoma cells infected with the recombinant viruses. Thus, recombinant viruses containing human PAH cDNA provide a means for introducing functional PAH into mammalian cells of hepatic origin and can potentially be introduced into whole animals as a model for somatic gene therapy for PKU.

Mouse phenylalanine hydroxylase. Homology and divergence from human phenylalanine hydroxylase.

The laboratory mouse represents an important model for the study of phenylalanine metabolism and the pathochemistry of phenylketonuria, yet mouse phenylalanine hydroxylase (PAH) has not been extensively studied. We report the cloning and sequencing of a mouse PAH cDNA, the expression of enzymic activity from the mouse PAH cDNA clone and the identification of mouse PAH and human PAH by two-dimensional PAGE of liver samples. These data confirm the expected homology of mouse PAH and human PAH and suggest differences in the primary sequence and the phosphorylation state of the two enzymes.