University of Southern California Dissertations and Theses

Astrocyte regulation of endothelial haemostasis function via transforming growth factor-beta

(USC Thesis Other)

Astrocyte regulation of endothelial haemostasis function via transforming growth factor-beta

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS This manuscript has been reproduced from the microfihn master. UM I films the text directly fijom the original or copy submitted. Thus, som e thesis and dissertation copies are in typewriter free, while others may be from any type o f computer printer. The quality o f th is reproduction is dependent upon th e quality o f th e copy subm itted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from 1 ^ to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back o f the book. Photographs included in the original manuscript have been reproduced xerographicaUy in this copy. Higher quality 6” x 9” black and w hite photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. UMI A Bell & Howell Information Company 300 North Zed) Road, Ann Arbor MI 48106-1346 USA 313/761-4700 800/521-0600 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ASTROCYTE REGULATION OF ENDOTHELIAL HAEMOSTASIS FUNCTION VIA TRANSFORMING GROWTH FACTOR-P Nam Duy Tran A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment o f the Requirements for the Degree DOCTOR OF PHILOSOPHY (Neuroscience) August 1998 © 1998 Nam Duy Tran Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number: 9919113 UMI Microform 9919113 Copyright 1999, by UMI Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. UMI 300 North Zeeb Road Ann Arbor, MI 48103 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UNIVERSITY OF SOUTHERN CALIFORNIA THE GRADUATE SCHOOL UNIVERSITY PARK LOS ANGELES. CALIFORNIA 90007 This dissertation w ritten by under the direction of h is. Dissertation Committeer and approved by all its members, has been presented to and accepted by The Graduate School in partial fulfillm ent of rg- cfuirements fo r the degree of DOCTOR OF PHILOSOPHY Dean of Craduau Studies Date .97.LW.9.P.. DISSERTA' chairp e rson Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table of Contents List o f figures iv List o f tables v Abstract vi C hapter 1 Introduction 1 References 12 C hapter 2 Regnlation o f B rain C apillary Endothelial Throm bom odulin mRNA Expression 21 Abstract 21 Introduction 23 Materials & Methods 26 Results 35 Discussion 45 References 48 C hapter 3 A strocyte Regulation of B rain Endothelial Tissue Plasminogen A ctivator in a Blood-Brain B arrier Model 52 Abstract 52 Introduction 54 Materials & Methods 56 Results 65 Discussion 74 References 79 C hapter 4 M easurem ent o f Throm bom odulin mRNA in B rain C apillaries by Polym erase C hain Reaction 84 Abstract 84 Introduction 85 Materials & Methods 87 Results 91 Discussion 96 References 98 u Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapters Traasforming Growth Factor-P Mediates Astrocyte-Specific Rcgnlatioa o f brain Capfllaiy Endothelial Aotkoagolaflit Factors 100 Abstract 100 Introductxm 102 Materials & Methods 105 Results 113 Discussion 120 References 124 Chapter 6 Discussion 125 References 136 lU Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. List of Figures Chapter 2 2.1. Confluent monolayer o f brain capillary endothelial cells 37 2.2. Quantitative-competitive PCR analysis o f TM mRNA 38 2.3. PCR analysis o f TM expression by astrocytes and endothelial cells 39 2.4. In situ hybridization demonstrating endothelial TM mRNA expression 40 2.5. Astrocyte glial fibrillary acidic protein expression 41 2.6. y-glutamyl transpeptidase activity by mono- and co-culture preparations 42 2.7. Thrombomodulin and P-actin mRNA expression in mono-and co-culture preparations 43 Chapter 3 3.1. Phase-contrast photomicrograph o f capillary-like structures 67 3.2. Glial fibrillary acidic protein expression by astrocytes 68 3.3. PCR analysis o f tPA and PAI-1 mRNA expression by astrocytes and endothelial cells 69 3.4. Densitrometric analyses o f tPA PCR products 70 3.5. tPA mRNA expression in endothelial monolayer preparations 71 3.6. tPA activity in culture media from mono- and co-culture preparations 72 3.7. PAI-1 mRNA expression in mono- and co-culture preparations 73 Chapter 4 4.1. Kinetics o f PCR amplification 92 4.2. Quantitative-competitive PCR analysis o f TM mRNA 93 4.3. Thrombomodulin mRNA expression by pontine, cerdwllar, and cortical capillaries 94 4.4. RT-PCR analysis o f thrombomodulin mRNA expression by pontine, cerebellar, and cortical capillaries 95 Chapter S 5.1. PCR analysis o f tPA and TM mRNA expression 5.2. tPA mRNA concentration 5.3. TM mRNA concentration 5.4. tPA activity 5.5. TM activity 115 116 117 118 119 IV Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. List of Tables Chapter 1 1.1 Astnx^e-endothelial co-culture and endothelial mono culture preparations 44 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Abstract Endothelial cells r^ u la te hemostasis in part via expression o f thrombomodulin (TM), tissue plasminogen activator (tPA) and plasminogen activator inhibitor (PAI-1). The present study used a blood-brain barrier (BBB) model to analyze regulation o f brain c^U lary endothelial TM, tPA, and PAI-1. This model consists o f co-culture o f astrocytes with brain c^illary endothelial cells. Astrocytes became progressively associated with endothelial capillary-like structures (CS); after seven days o f co-culture, nearly ail astrocytes are associated with CS and the BBB-associated enzyme gamma- glutamyl transpeptidase is expressed by these co-culture preparations. A fter one day o f co-culture, polymerase chain reaction (PCR) assay demonstrated that TM, tPA and PAI-1 mRNA levels were comparable for mono-culture vs. co-culture preparations. However, after seven days (ie, when elements o f the BBB are present), astrocyte-endothelial co cultures (compared to endothelial mono-cultures) showed dramatic reductions in TM and tPA mRNA, and an increase in PAI-1 mRNA. Moreover, seven day co-cultures demonstrated reduced tPA activity compared to endothelial mono-cultures. These data suggest that astrocytes modulate brain capillary expression o f TM, tPA and PAI-1 mRNA in this model. In order to study the mechanism o f this process, we examined the hypothesis that astrocyte regulation o f endothelial tPA and TM is mediated by transforming growth fector-P (TGF-P). Astrocyte-endothelial co-cultures and media conditioned by astrocytes (ACM) exhibited significantly higher levels o f active TGF-P compared to a) brain capillary mono-cultures and b) endothelial cells grown in non- vi Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. conditioned media, respectively. Brain c^U lary endothelial ceils incubated with ACM demonstrated reduced TM and tPA mRNA and activity, compared to cells grown in non conditioned media. Treatment with exogoious TŒF-P produced dose-dependent reductions in TM and tPA that were comparable in magnitude to that o f ACM The effects o f ACM on both TM and tPA were blocked by TGF-P neutralizing antibody. These data indicate that TGf-P mediates astrcxqte r^ u latio n of brain capillary endothelial expression o f TM and tPA. vu Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 1 Introduction Stroke, the third leading cause o f death in North America, has an incidence o f 400,000 including approximately 175,000 annual deaths and stroke survivors numbering nearly 2 million (W olf et al., 1983). Risk Actors for stroke include older age, high blood pressure, diabetes mellitus, cardiovascular diseases, cholesterol, alcohol, nutrition factors, and smoking (Klag & Whelton, 1989; W olf et al., 1991; Jacobs et al., 1992; Neaton et al., 1992). The mechanisms relating these factors to stroke has been undefined. Nevertheless, these risk factors indicate the influence o f a genetic and/or environmental component that affects the hemostatic pathways; thus, increasing the propensity for stroke. Stroke, or cerebral in&rction, is known to be associated with hemostatic alterations. Alterations o f systemic hemostatic variables have been extensively described in association with acute stroke (Fisher & Francis, 1990). A number o f these abnormalities have also been described beyond the acute phase o f stroke (Fisher & Francis, 1990). The occurrence of cerebral infarctions correlates strongly with circadian cycles o f low endogenous thrombolytic activity (Marsh et al., 1990). Lastly, the incidence o f cerebral infarction can be significantly reduced by pharmacological Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. treatment (e g. aspirin and warfarin) designed to interfere with components o f the hemostasis pathway (Albers et al., 1991). Understanding the mechanisms o f hemostasis r%ulation ultimately requires the understanding o f the biology of the endothelial cells (EC) that line blood vessels. Blood microvessels were first observed by Malpigi in 1661 using a fiea-lens microscope. The simplicity o f M algigi’s microscope is reflective of the simplistic view o f the cellular nature o f blood vessels- an inert barrier between the tissue and circulating blood. This idea dominated our view o f endothelial biology until nearly three decades ago, whereby advances in cellular and molecular biology allowed the culture o f EC in vitro and their examination in detail. These studies helped shed light on the complexity o f functions in the endothelial repertoire. EC function as a semi-permeable membrane between the blood and tissue that permit exchange o f oxygen, carbon dioxide, hormones, nutrients, and metabolic waste (Pardridge, 1991). In addition to these roles, the endothelium plays a key role in hemostasis by maintaining blood in a fluid state under physiological conditions. Under pathological conditions, as in the case o f vascular injury, this system must respond by sealing the defect in the vessel wall in order to minimize blood loss. This response must not only be rapid, but must also be localized to the site o f injury. A thrombotic event may ensue if the coagulation occurs unregulated. The vascular endothelium maintains blood fluidity by inhibiting blood coagulation and promoting fibrinolysis. EC modulate these functions via expression and/or secretion o f proteins that interact with blood constituents. Among the most important proteins are thrombomodulin (TM), a key factor in the protein C anticoagulant 2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. pathway, and tissue plasminogen activator (tPA), a component o f the jQbrinolytic pathway. TM is an int%ral membrane protein that resembles the LDL receptor in structure (Jackman et al., 1986; Wen et al., 1987). TM, like IFN-ot, IFN-P, ihodopsin, and angiogenin, is transoibed from a gene that lacks introns (Jackman et al., 1987; Dittman & Majerus, 1990). The implications o f this finding is unclear; however, it may suggest that TM protein expression is more dq>endent on transcriptional, as opposed to translational, processes. The mature peptide contains 5 structural domains. The amino terminal domain is followed by a domain that contains 6 tandem EGF-like repeats (Doolittle et al., 1984). This domain is the site o f thrombin and protein C binding (Kurosawa et al., 1988; Steams et al., 1989; Suzuki et al., 1989; Zushi et al., 1989). Additionally, the EGF-like repeats may help to preserve ability o f TM to release ligand, avoid degradation, and recycle to the cell surface, as has been suggested for the LDL receptor (Goldstein et al., 1979; Brown & Goldstein, 1986; Davis et al., 1987). The next domain in rich in serine and threonine residues and is likely to be the sight o f O-linked glycosyiation (Russell et al., 1984). This is also a r% ion o f least homology between the human, bovine and mouse proteins (Jackman et al., 1986; Wen et al., 1987; Dittman et al, 1988). The fourth domain is the transmembrane region containing 23 hydrophobic amino acids and is the most highly conserved among species (Jackman et al., 1986; Wen et al., 1987; Dittman et al, 1988). The carboxyl terminal cytoplasmic terminal contains potential phosphorylation sites fisr cellular kinases (Dittman et al., 1988). This region also contains sites that could interact with other intracellular proteins. Unlike the LDL Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. recqïtor, the cytoplasmic domain o f TM doesn’t contain the putative consensus signal for internalization by coated pit mediated cytolysis (Goldstein et al., 1985). TM binds to the normally pro-coagulant thrombin with 1 ; 1 stoichiometry (Sadler et al., 1993). Thrombin bound to TM stimulates m docytosis o f the complex (Maruyama & Majerus, 1985). Thrombin is then d^raded in lysosomes, while TM is recycled back to the membrane surface. Endocytosis o f the TMrthrombin complex can be inhibited by protein C (Maruyama & Majerus, 1987). The TMrthrombin complex enhances the activation o f the circulating zymogen protein C nearly 20,000 fold over that by thrombin alone (Esmon & Owen, 1981). Thus, TM plays an important role in the hemostasis r^^latio n and affects the coagulant pathway in 2 different ways: 1) TM inhibits the procoagulant activity o f thrombin, and 2) TM serves as a coÊictor for the activation o f protein C by thrombiiL Activated protein C (APC) functions as a critical circulating anticoagulant by inactivating clotting factors Va and VUIa (Walker et al., 1979; Fulcher et ai., 1984) and by inhibiting PAI-1, an inhibitor o f tPA (Fouw et al., 1988). TM-dependent protein C activation has a major role in maintaining normal blood fluidity and preventing intravascular thrombosis. A high incidence o f neonatal purpura fulminans occurs in infants with severe protein C deficiency (Griffin et al., 1981; Seligsohn et al., 1984). Resistance to the effects o f APC correlates strongly with venous thrombosis in humans (Svensson et al., 1994), while decreased levels o f protein C are associated with patients with disseminated intravascular coagulation (Griffin et al., 1982). Experimental models further substantiate the antithrombotic role o f APC. APC provides substantial protection against procoagulant and lethal effects in a sepsis model (Taylor et al., 1987). 4 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Additionally, animals pretreated with TM are protected against thromboembolism (Kumada et al., 1987; Gomi et al., 1990). Assuming that TM is uniformly distributed among EC throughout a vascular bed, the major site o f action o f TM is in the miCTO-drculation ^ e r e capillary EC constitute greater than 99 % o f the EC surface area (Dittman & Majerus, 1990). As the vessel size increases, the ratio o f EC surâce area per vessel volume deo-eases, and results in decreased blood-endothelial contact. This difference in blood-endothelial interaction would produce an effective capillary TM concentration nearly 3 orders o f magnitude greater than that in a large vessel. Therefore, minute changes in capillary TM expression could potentially yield dramatic changes in the protein C anticoagulant system. TM has been detected in endothelium from nearly all vascular beds with an estimated 30-50,000 TM molecules per EC (Esmon et al., 1982; Salem et al., 1984; Kurosawa & Aoki, 1985; Maruyama & Majerus, 1985; Jakubowski & Owen, 1986; Kumada et al, 1987). A number o f investigators have examined TM expression in the CNS. It was initially suggested that TM was absent in brain capillaries (Ishii et al., 1986). Later studies demonstrated TM expression in the brain (DeBauh et al., 1986; Wong et al., 1991; Isaka et al., 1994; Maruno et al., 1994) and th ro u ^o u t the CNS (DeBault et aL, 1986; Boffa et al., 1991). Nevertheless, brain TM expression was limited (Wong et al., 1991; Isaka et aL, 1994; Maruno et al., 1994) and found to be substantially reduced in subcortical r iio n s where inforction is common (Wong et al., 1991). Tissue plasminogen activator (tPA) is synthesized and secreted as a single polypeptide chain (Pennica et aL, 1983). Following proteolytic cleavage, the peptide is converted to a two-chain molecule bridged by a disulfide bond. tPA is composed o f five 5 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. functional domains: the finger, EGF, kringie-1 and -2, and proteolytic domain. The finger and kringle-2 domains o f tPA are involved in interactions with fibrin (van Zonneveld et al., 1986; Veiheijen et al., 1986), while the finger and EGF domains are involved in the binding o f tPA to cells (Beebe et al., 1989). The proteolytic domain in the C-terminal part o f the molecule is required for interaction with plasminogen. tPA activation is not dependent on its proteolytic conversion to the two-chain polypeptide isoform, as in the case with urokinase plasminogen activator (uPA). Rather, tPA activity is dependent on its interaction with fibrin (Ranby 1982; Rijken et al., 1982). Bound fibrin confers PA activity to both single- and two-chain tPA isoforms. It is unclear how these two isoforms differ in bioactivity. Active tPA proteolytically converts plasminogen to plasmin, a protease involved in the dissolution of the fibrin clot (Collen & Lijnen, 1995). tPA can be rapidly inhibited in the presence o f PAI-1 (Wiman, 1995). PAI-1 complexes irreversibly with tPA with 1:1 stoichiometry. The fibrinolytic capacity o f blood is largely determined by the plasma tPA concentration (Wun & Capuano, 1985). Impairment o f the fibrinolytic capacity may be due to a decrease in tPA synthesis and secretion and/or to an increase in PAI-1 expression. Decreased fibrinolytic activity is associated with deep vein thrombosis and increased risk o f thrombotic recurrency (Collen & Lijnen, 1986). Reduced levels o f brain tPA are associated with increased infarct size following transient middle cerebral artery occlusion in diabetic and nicotine stroke models (Kittaka et al., 1996; Wang et al., 1996). In contrast, administration o f fibrinolytic drugs reduced infarct size in both thromboembolic and non-thromboembolic stroke models (Zivin et al., 1985; del Zoppo et al., 1986). The importance o f the fibrinolytic system has been further emphasized by the 6 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. demonstration o f improved outcome in acute ischemic stroke foUowing treatment with ocogenous tPA (MINDS it-PA Stroke Study Group, 1995). EC are the primary source o f circulating plasma tPA (Erickson et al., 1985; van Hinsbergh, 1988). Limited expression o f tPA has been demonstrated in the brain microvasculature (Levin & del Zoppo, 1994; Zlokovic et al., 1995) and in cultured brain microvascular endothelium (Saksela et al., 1990; KoUros et al., 1994; Shatos et al., 1995). In primate models, brain cq)illaries exhibit only 3 % o f the total brain tPA expression (Levin & del Zoppo, 1994). The presence o f tPA suggests a role for brain endothelium in the r%ulation o f fibrinolysis. The preponderance of evidence indicate limited TM and tPA expression by brain EC compared to peripheral vascular EC; however, little is known o f the regulatory mechanisms that contribute to this limited expression. EC lining capillaries o f the brain are unique in several ways: 1) they comprise a central component o f the BBB, a physiological stm cture separating the vascular bed fi'om the CNS, and 2) they are intimately associated with astrocytes. Studies have definitively shown that brain EC make up the anatomical barrier between the blood and brain. Brain EC have tight intercellular junctions, or zonal occludens, to prevent interendothelial passage (Reese & Kamovsky, 1967; Brightman & Reese, 1969). The sparsity o f pinocytic vesicles, pores and fenestrations further contribute to the limited permeability o f the barrier. Studies using markers, such as Evans Blue dye or horseradish peroxidase, showed that these markers could not diffuse fi'om the blood into the brain tissue because of the EC barrier (Reese & Kamovsky, 1967; Brightman & Reese, 1969; Pardridge et al., 1986). 7 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The primary role o f the BBB is to sdectiveiy control transport o f essential nutrients, hormones and protans to and from the CNS (Goldstein & Betz, 1983; Pardridge et aL, 1986; Betz & Goldstein, 1986; loo, 1987; Pardridge, 1991). Substances entering or exiting the brain must pass th ro u ^ the EC and both luminal and abluminal membranes. Small polar substances usually pass through the EC unimpeded. Non-polar or large molecules, however, are dependent on other biochemical mechanisms for their passage. Brain EC have developed highly specialized transport mechanisms for delivery o f substances into the CNS, including y-glutamyl transpeptidase (GGTP) (DeBault & (Dancilla, 1980), GLUT-1 ^ c o s e transporter (Pardridge et al., 1990), and Na-K-Cl cotransporter (O'Donnell et al., 1995; Sun et al., 1995). Further, brain EC also exhibit other unique characteristics, such as high mitochondria density, high electrical resistance, lack o f pinocytic vesicles, and expression o f alkaline phosphatase (ALP) that contribute to its specialized function (Betz & Goldstein, 1980; Tio et al., 1990; Pardridge, 1991). In a hallmark experiment using chick-quail chimeras, it was demonstrated that endothelial BBB phenotype is dependent on the tissue environment (Stewart & Wiley, 1981). Chick and quail cells in these chimeras can be distinguished by nuclear morphology. Chicks have a large central nucleolus and their chromatin are concentrated more centrally with sparse peripheral distribution. Quails have chromatin that are more dCTsely distributed along the periphery (LeDouarin, 1973). Avascularized embryonic quail brain tissues were transplanted into chide coelemic cavities, where they were vascularized by non-CNS vessels, and avascularized embryonic quail somites were transplanted into chick brains, \^iere they were vascularized by CNS vessels. Non-CNS capillaries that vascularized brain tissue exhibited high mitochondrial density, lack of 8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. pinocytic vesicles and tight junctions, indicating the presence o f a functional BBB. Conversely, CNS capillaries that vascularized somites lacked BBB properties, as demonstrated by low mitochondrial density and large numbers o f pinocytic vesicles. Thus, charactoistics o f the BBB are determined in large by the cellular environment and not by the origin o f the oidothelium. In the CNS, astrocytic processes ensheathe capillaries and form the glia limitans (Goldstein, 1988). Because the formation o f the glia limitans coincides with the development o f the BBB, perivascular astrocytes are thought to be responsible for the induction and maintenance o f the endothelial BBB phenotype. In vitro experiments have implicated astrocytes in the induction of the BBB Direct co-cultures o f astrocytes with endothelial cells resulted in increased GGTP (DeBault & Cancilla, 1980; Bauer et al., 1990; Meyer et al., 1991; Tontsch & Bauer, 1991; Hayashi et al., 1997), GLUT-1 glucose transporter (Hayashi et al., 1997), Na-K-Cl cotransporter (Sun et al., 1995), ALP (Tio et al., 1990; Meyer et al., 1991), Na-K-ATPase (Tontsch & Bauer, 1991), electrical resistance (Dehouck et al., 1990), mitochondria density (Hayashi et al., 1997), and tight junctions (Tao-Cheng et al., 1987; Hayashi et aL, 1997). Further, many o f these phenotypes could be induced by astrocyte-conditioned media (Maxwell et al., 1987; Tao- Cheng et al., 1987; Rubin et al., 1991; Raub et al., 1992; O'Donnell et al., 1995), suggesting that these effects were mediated in part by a &ctor released by astrocytes. Astrocytes have the capacity to secrete transforming growth foctor-P (TGF-P), a factor known to affect endothelial function (Constam et al., 1992). TGF-P belongs to a superfomily o f peptide growth foctors that regulate cellular processes, including cell growth, differentiation, and mctracellular matrix formation (Massague J, 1990). Five Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. distinct isoforms of TGF-3 have been identified, but only three of which (TGF-01, -^2 and -g3) are expressed by mammalian cells (Massague, 1990, Rifidn et al., 1993; Attisano et al., 1994). TGF-3 is normally secreted in a biologically inactive latent form, composed o f the mature TGF-3, a TGF-3 latency-associated peptide (LAP), and a latent TGF-3 binding protein (LTBP) (Massague, 1990, Rifidn et al., 1993; Attisano et al., 1994). Latency is conferred by interactions of the mature TGF-3 with LAP (Massague, 1990, Rifidn et al., 1993). LTBP is not necessary for the maintenance of TGF-3 latency (Nfiyazono et al., 1993). The mechanism of TCT-3 activation in vivo remains unclear; however, TGF-3 can be activated in vitro by treatment with proteases, heat, acidification, or alkalinization (Nfiyazono et al., 1993). Additionally, TGF-3 activation can be demonstrated in heterotypic pericyte-endothelial or smooth muscle cell-endothelial co- cultures (Antonelli-Orlidge et al., 1989; Sato et al., 1989; Sato et al., 1990). Exogenous treatment with TGF-3 has been shown to negatively regulate the protein C anticoagulant system by downregulating TM mRNA, protein, and activity in a dose- and time-dependent manner (Ohji et al., 1995). Downregulation of tPA and enhancement of PAI expression following treatment with TGF-3 has also been reported (Saksela et al, 1987). These data support the role of TGF-3 in the regulation of hemostasis. In chapter 2, an in vitro model of the BBB is used to investigate the hypothesis that astrocytes play a regulatory role on brain capillary endothelial hemostasis function. Astrocytes are able to induce properties of the BBB in this model. A novel quantitative- competitive polymoase chain reaction ((^-PCR) is developed for the sensitive and 10 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. accurate measurement o f TM mRNA in this investigation. This woiic demonstrates for the first tim e that astrocytes downr%ulate oidothelial expression o f TM, an anticoagulant protein. Chapter 3 examines the r% ulatoiy role o f astrocytes on endothelial fibrinolysis. A QC-PGR assay is also developed for the quantification o f tPA mRNA. Further, data derived fi'om QC-PCR is determined to be similar to those derived fi'om standard RT- PCR analysis. This work shows that astrocytes downregulate endothelial fibrinolytic capacity by downr%ulating tPA and upr^ulating PAI-1 expression. Chapter 4 is a comparative study examining the accuracy and reliability o f data derived fi'om QC-PCR and standard RT-PCR for TM. Chapter 5 examines the mechanism o f astrocyte regulation o f endothelial hemostasis function. This study demonstrates that astrocytes mediate downregulation o f endothelial TM and tPA expression through TGF-p. Moreover, these astrocyte-induced effects can be attenuated by TGF-P naitralizing antibody. In conclusion this woik provides evidence that astrocytes negative regulate endothelial anticoagulant and fibrinolytic fonction in part by TGF-(3. 11 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reference 1. Albers G, Sherman D, Gress D, Pauleeth J, Peterson P (1991) Stroke prevention in nonvalvular atrial fibrillation: A review of prospective randomized trials. Arm N eurol 30,511-8 2. Antonelli-Orlidge A, Saunders KB, Smith SR, D'Amore PA (1989) An activated form o f transforming growth A ctor beta is produced by cocultures o f endothelial cells and peri<qftes. /Voc N atl A cad S ci USA 86,4544-8 3. Attisano L, Wrana JL, Lopez-Cassila F, M assague J (1994) TGF-3 receptors and actions. B itxh im B io p h ys A cta 1222,1\- % Q 4. Bauer HC, Tontsch U, Amberger A, Bauer H (1990) Gamma-glutamy 1-transpeptidase (GGTP) and Na-K-ATPase activities in different subpopulations o f cloned cerebral endothelial cells: responses to glial stimulation. Biochem B iophys R es Comm 168, 358-63 5. Beebe DP, M iles LA, Plow EF (1989) A linear amino acid sequence involved in the interaction o f tPA with its endothelial cell receptor. B lood 74, 2034-7 6. Betz AL, Goldstein GW (1986) Specialized properties and solute transport in brain capillaries. Arm Rev P hysiol A%,2A\~5Q 7. Betz AL, Goldstein GW (1980) Polarity o f the blood-brain barrier: distribution of enzyme between luminal and antiluminal membranes of brain capillary endothelial cells. B rain R es 192, 17-28 8. Boffa MC, Jackman RW, Peyri N, Boffa JF, George B (1991) Thrombomodulin in the central nervous system. N oitv Rev Fr H em atol 33, 423-9 9. Brightman MW, Reese TS (1969) Junctions between intimately apposed membranes in the vertdsrate brain. J C ell B iol 40, 648-77 10. Brown MS, Goldstein JL (1986) A receptor-mediated pathway for cholesterol homeostasis. Science 232, 34-47 11. Collen D, Lijnen HR (1986) The fibrinolytic system in man. CRC C ritical Review in H aem atology arui O ncology 4, 249-301 12. Collen D, Lijnen HR (1995) Molecular basis o f fibrinolysis, as relevant for thrombolytic therapy. Thromb H aem ostlA , 167-71 12 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 13. Constam DB, Philipp J, Malipiero UV, Dijke P, Schachner M, Fontana A (1992) Differential expression o f transforming growth fàctor-31, -^2, and -^3 by glioblastoma cells, astrocytes and microglia. J Im m unol 148,1404-10 14. Davis CG, Goldstein JL, Sudhof TC, Anderson RGW, Russell DW, Brown MS (1987) Acid-dependent ligand dissociation and recycling o f LDL receptor mediated by growth factor homology r% ion. Proc Neal A cad Sci USA 84, 368-72 15. DeBault LE, Cancilla PA (1980) Gamma-glutamyl transpeptidase in isolated brain endothelial cells: induction by glial cells in vitro. Senence 207, 653-5 16. DeBault LE; Esmon NL; Olson JR; Esmon CT (1986) Distribution o f the thrombomodulin antigen in the rabbit vasculature. Lab Invest 54:172-8 17. Dehouck B, Dehouck MP, Fruchart JC, Cecchelli R (1994) U pr^ulation o f the low density lipoprotein at the blood-brain barrier: intercommunications between brain capillary endothelial cells and astrocytes. J C ell B io l 126,465-73 18. del Zoppo GJ, Copeland BR, W altz TA, Zyroff J, Plow EF, Harker LA (1986) The beneficial effect o f intracarotid urokinase on acute stroke in a baboon model. Stroke 17, 638-43 19. Dittman WA, Majerus PW (1990) Structure and function o f thrombomodulin : a natural anticoagulant. Blood 75, 329-36 20. Dittman WA, Kumada T, Sadler JE, Majerus PW (1988) The structure and function o f mouse thrombomodulin. J B io lC h e m ld i, 15815-22 21. Doolittle RF, Feng DF, Johnson MS (1984) Computer-based characterization o f epidermal growth 6 cto r precursor. Nature 307, 558-60 22. Erickson LA, Schleef RR, Ny T, Loskutoff DJ (1985) The fibrinolytic system o f the vascular wall. C lin H aem atol 14, 513-29 23. Esmon NL, Owen WG (1981) Identification o f an endothelial cell co6ctor for the thrombin-catalyzed activation o f protein C. Proc Neal A cad Sci USA 78, 2249-52 24. Esmon NL, Owen WG, Esmon CT (1982) Isolation o f a membrane-bound cofector for thrombin-catalyzed activation o f protein C J B io l Chem 257, 859-64 25. Fisher M, Francis R (1990) Altered coagulation in cerebral ischemia: platelet, thrombin and plasmin activity. A rch N eurol AT, 1075-9 13 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 26. Fouw NT, Jong YF, Haverkate F, Bertina RM (1988) Activated protein C increase fibrin lysis by neutralization o f plasminogen activator inhibitor- no evidence for a coActor role o f protein S. Thromb Haemost 60,328-33 27. Fulcher CA, Gardiner IE , Griffin JH, Zimmerman TS (1984) Proteolytic inactivation o f human factor V m procoagulant protein by activated iniman protein C and its analogy with fector V. B /oot/63,486-9 28. Gomi K, Zushi M, Honda G, Kawahara S, Matsuzaki O, Kanabayashi T, Yamamato S, Maruyama I, Suzuki K (1990) Antithrombotic effect o f recombinant human thrombomodulin on thrombin-induced thromboembolism in mice. B lood 75, 1396-9 29. Goldstein GW (1988) Endothelial ceil-astrocyte interactions. A nnals N Y A ca d Sci 529, 31-9 30. Goldstein GW, Betz AL (1983) Recent advances in understanding brain capillary fimction. A nn N eurol 14, 389-95 31. Goldstein JL, Anderson RG, Brown MS (1979) Coated pits, coated vesicles, and receptor-mediated endocytosis. Nature 279, 679-85 32. Goldstein JL, Brown MS, Anderson TG, Russell DW, Schneider WJ (1985) Receptor mediated endocytosis: concepts emerging from the LDL system. Ann R ev C ell B iol 1,1 33. Griffin JH, Evatt B, Zimmerman TS, Kleis AJ, Wildeman C (1981) Deficiency of protein C in congenital thrombotic disease. J C lin Invest 68, 1370-3 34. Griffin JH, Mosher DF, Zimmerman TS, Kleis AJ (1982) Protein C, an anticoagulant protein, is reduced in hospitalized patients with intravascular coagulation. B lo o d 60, 261-4 35. Hayashi Y, Nomura M, Yamagishi S, Harada S, Yamashita J, Yamamoto H (1997) Induction o f various blood-brain barrier properties in non-neural endothelial cells by close apposition to co-cultured astrocytes. G lia 19, 13-26 36. Isaka T, Yoshimine T, Motokiko M, Kurode R, Ishii H, Hayakawa T (1994) Altered expression o f antithrombotic molecules in human glioma vessels. A cta N europaihol 87, 81-5 37. Ishii H, Salem HH, Bell CE, LaporataEA, Majerus PW (1986) Thrombomodulin, an anticoagulant protein, is absent in the human brain. B lood 67, 362-5 14 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 38. Jackman RW, Beeler DL, Fritze L, Sofif G, Rosœberg RD (1987) Human thrombomodulin gene is intron depleted: Nucleic acid sequences o f the cDNA and gene predict proton structure and suggest site o f r% ulatoiy control. Proc N eal A cad Soi USA 84, 6425-9 39. Jackman RW, Beeler DL, VanDeWater L, Rosenberg RD (1986) Characterization o f a thrombomodulin cDNA reveals structural similarity to the low density lipoprotein receptor. Proc N atl A cad Sci USA 83, 8834-8 40. Jacobs D, Blackburn H, ICggins MW et al (1992) Report o f the conference on low cholesterol. Mortality associations. C irculation 86,1046-60 41. Jakubowski HV, Owen WG (1986) The effect o f bovine thrombomodulin on the specificity o f bovine thrombin. J B io l Chem 264, 111 17-21 42. Joo F (1987) Current aspects to the development o f the blood-brain barrier. N it J D iv Neurotics 5, 369-72 43. Kittaka M, Wang L, Sun N, Schreiber SS, Seeds NW, Fisher M, Zlokovic BV (1996) Brain capillary tissue plasminogen activator in a diabetes stroke model. Stroke 27, 712-9 44. Klag MJ, Whelton PK (1989) Stroke and nutrition. Clin N utr 8, 34-41 45. KoUros PR, Knoll BA, Cambrian AP, Henrikson P (1994) Plasminogen activator inhibitor-1 eq)ression by brain microvessel endothelial cells is inhibited by elevated glucose. JN eurochem 63, 903-9 46. Kumada J, Dittman WA, Majerus PW (1987) A role for thrombomodulin in the pathogenesis o f thrombin-induced thromboembolism in mice. B lood 71, 728-33 47. Kurosawa S, Aoki N (1985) Preparation o f thrombomodulin from human placenta. T hrom bR esZ l, 353-64 48. Kurosawa J, Steams DJ, Jackson KW, Esmon CT (1988) A 10-kDa cyanogen bromide fragment from the epidermal growth factor homology domain o f rabbit thrombomodulin contains the primary thrombin binding site. J B io l Chem 263, 5993-6 49. LeDouarin NM (1973) A biological cell labeling technique and its use in eq)erimental embryology. D ev B iol 30,217-22 50. Levin EG, del Zoppo GJ (1994) Localization o f tissue plasminogen activator in endothelium o f a limited number o f vessels. Am J Path 144, 855-61 15 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 52. Maruno M, Yoshimine T, Isaka T, Kurode R, Ishii H, Mayakawa T (1994) Expression o f thrombomodulin in astro<^om as o f various malignancy and in giiotic and normal brain. J N euro-O nœ logy 19, 155-60 53. Maruyama I, Majerus PW (1985) The turnover o f thrombin-thrombomodulin complex in cultured human umbilical vein endothelial cells and A549 lung canco* cells. Endocytosis and degradation of thrombin. J B io l Chem 260, 15432-8 54. Maruyama I, Majerus PW (1987) Protein C inhibits endocytosis o f thrombin- thrombomodulin complexes in A549 lung cancer cells and human umbilical vein endothelial cells. B lood 69, 1481-4 55. Massague J (1990) The transforming growth foctor-P family. Arm Rev C ell B iol 6, 597-601 56. Maxwell K, Berliner JA, Cancilla PA (1987) Induction o f y-glutamyl transpeptidase in cultured cer^ ral endothelial cells by a product released by astrocytes. Brain Res 410,309-14 57. Meyer J, Rauh J, Galla H (1991) The susceptibility of cerebral endothelial cells to astroglial induction o f blood-brain barrier enzymes depends on their proliferative state. JN eurochem 57, 1971-7 58. Miyazono K, Ichijo H., Heldin CH (1993) Transforming growth factor-beta; latent forms, binding proteins and receptors. G rowth Factors 8, 11-22 59 The National Institute o f Neurological Disorders and Stroke rt-PA Stroke Study Group (1995) Tissue plasminogen activator for acute ischemic stroke. N E ngl J M ed 333,1581-7 60. O'Donnell ME, Martinez A, Sun D (1995) Cerebral microvascular endothelial cell Na-K-Cl cotransport: regulation by astrocyte-conditioned media. Am J P hysiol 268, C747-54 61. Ohji T, Urano H, Shirahata A, Yamagishi M, Higashi K, Gotoh S, Karasaki Y (1995) Transforming growth factor betal and beta2 induce down-modulation of thrombomodulin in human umbilical vein endothelial cells. Thromb H aem ost 73, 812-8 62. Pardridge WM (1991) Advances in cell biology o f blood-brain barrier transport Sem 2, 419-26 63. Pardridge WM, Oldendorf WH, Cancilla PA, Frank H3L (1986) Blood-brain barria: interfoce between internal medicine and the brain. Arm Intern M ed 105, 82-95 16 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 64. Pardridge WM, Boado RJ, Farrell CR (1990) Brain-type glucose transporter (GLUT- 1) is selectively localized to the blood-brain barrier. Studies with quantitative Western blotting and in situ hybridization. J B io l Chem 265, 18035-40 65. Pennica D, Holmes WE, Kohr WJ, Haridns RN, Vehar GA, Ward CA, Bennett WF, Yelverton E, Seeburg PH, Heynericer HL, Goeddel DV (1983) Cloning and expression o f human tissue-type plasminogen activator cDNA in K colL Nature 301, 214-21 66. Ranby M (1982) Studies on the kinetics of plasminogen activation by tissue plasminogen activator. Biochàm B iophys Acta 704, 461-9 67. Raub TJ, Kuentzel SL, Sawada GA (1992) Permeability o f bovine brain microvessels endothelial cells in vitro: barrier tightening by a factor released from astroglioma cells. Exp C ell R es 199,330-40 68. Reese TS, Kamovslqr MJ (1967) Fine structural localization of a blood-brain barrier to exogenous protein. J C ell B iol 34,207-17 69. Rifldn DB, Kojima S, Abe M, Harpel JG (1993) TGF-P: Structure, function, and formation. Thromb Haemost 10,111-9 70. Rijken DC, Hoylaerts M, Collen D (1982) Fibrinolytic properties o f one-chain and two-chain human extrinsic (tissue-type) plasminogen activator. J B io l Chem 257, 2920-5 71. Rubin LL, Hall DE, Porter S, Barbu K, Cannon C, Homer HC, Janatpour M, Liaw CW, Manning K, Morales J, Tarmer LI, Tomaselli KJ, Bard F (1991) A cell culture o f the blood-brain barrier. J C ell B io l 115, 1725-35 72. Russell DW, Schneider WJ, Yamamoto T, Luskey KL, Brown MS, Goldstein JL (1984) Domain map o f the LDL receptor: sequence homology with the epidermal growth Actor precursor. C ell 37, 577-85 73. Sadler JE, Lentz SR, Sheehan JP, Tsiang M, Wu Q (1993) Structure-function relationships o f the thrombin-thrombomodulin interaction. H aem ostasis 23, 183-93 74. Saksela O, Moscatelli D, Rifkin DB (1990) The opposing effects o f basic fibroblast growth Actor and transforming growth factor beta on the regulation o f plasminogen activator activity in capillary endothelial cells. J C ell B io l 105, 957-63 75. Salem HH, Maruyama L Ishii H, Majerus PW (1984) Isolation and characterization o f thrombomodulin from human placenta. J B io l Chem 259, 12246-51 17 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 76. Sato Y, Rifldn DB (1989) Inhibition o f endothelial ceils by pericytes and smooth muscle cells: activation o f a latent transforming growth factor-beta 1-like molecule by plasmin during co-culture. J C ell B iol 109, 309-15 77. Sato Y, Tsuboi R, Lyons RM, Moses H, Rifldn DB (1990) Characterization o f the activation o f latent TGF-beta by cocultures o f endothelial cells and pericytes in smooth muscle cells: a self r e la tin g system. J C e ll B iol 111, 757-63 78. Seligsohn U, Berger A, Abend M, Rubin L, Attias D, Zivelin A, Rapaport SI (1984) Homozygous protein C deficiency manifested by massive venous thrombosis in the newborn. N E ngl J M ed 310, 559- 79. Shatos MA, Orfeo T, Doherty JM, Penar PL, Collen D, Mann KG (1995) (a - Thrombin stimulates urokinase production and DNA synthesis in cultured human cerd)ral microvascular endothelial cells. A rterioscler Thromb Vase B io l 15,903-11 80. Steams DJ, Kurosawa S, Esmon CT (1989) Nficrothrombomodulin residues 310-486 fi'om the q}idermal growth factor precursor homology domain o f thrombomodulin will accelerate protein C activation. J B io l Chem 264, 3352-6 81. Stewart PA, Wiley MJ (1981) Developing nervous tissue induces formation o f blood- brain barrier characteristics in invading endothelial cells: a study using quail-chick transplantation chimeras. Dev B iol 183-92 82. Sun D, Lytle C, O'Donnell ME (1995) Astroglial cell-induced expression o f Na-K-Cl cotransporter in brain microvascular endothelial cells. Am J Physiol 269, C l 506-12 83. Suzuki K, Hayashi T, Nishioka J, Kosada Y, Zushi M, Honda G, Yamamoto S (1989) A domain composed of epidermal growth &ctor-like structures of human thrombomodulin is essential for thrombin binding and for protein C activation. J B io l Chem 264, 4872-6 84. Svensson PJ, Dahlback B (1994) Resistance to activated protein C as a basis for venous thrombosis. N E ngl JM ed 330, 517-22 85. Tao-Cheng JH, Nagy Z, Brightman MW (1987) Tight junctions to brain endothelium in vitro are enhanced by astroglia. JN eurosci 7, 3293-9 86. Taylor FB, Chang A, Esmon CT, D*Angelo A, Vigano-D'Angelo S, Blick KE (1987) Protein C prevents coagulopathic and lethal effects o f Escherichia coli injection in the baboon. JC lin Invest 79, 918-25 87. Tio S, Deenen M, Marani E (1990) Astrocyte-mediated induction o f alkaline phosphatase activity in human umbilical cord vein endothelium. E ur J M orphol 28, 289-300 18 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 88. Tontsch U, Bauer H (1991) Glial cells and neurons induce blood-brain barrier related enzymes in cultured c a d r a i endothelial cells. Brcmt R es 539,247-53 89. van Hinsbergh VWM (1988) Regulation o f the synthesis and secretion o f plasminogen activators by endothelial cells. H aem ostasis 18:307-27 90. van Hinsb«igh VWM, Kooistra T, Emeis JJ, Koolwijk P (1991) R ^ulation o f plasminogen activator production by endothelial cells: role in fibrinolysis and local proteolysis. In t JR a d ia t B io l 60,261-72 91. van Zonneveld AJ, Veerman H, Panndcoek H (1986) On the interaction o f the finger and kringle-2 domain o f tissue-type plasminogen activator with fibrin. J B io l Chem 261, 14214-8 92. VeAeijen JH, Casper MPM, Chang GTG, DeMunk GAW, Pouwels PH, Enger-Valk BE (1986) Involvement o f the finger domain and kringle-2 domain o f tissue-type plasminogen activator in fibrin binding and stimulation o f activity by fibrin. Eur M ol OrgJS, 3525-30 93. W alker FJ, Secton PW, Esmon CT (1979) The inhibition o f blood coagulation by activated protein C through the selective inactivation o f activated A ctor V. Biochim B iophys A cta 571, 333-42 94. Wang L, Kittaka M, Sun N, Schreiber SS, Zlokovic BV (1996) Chronic nicotine treatment enhances focal ischemic brain injury and depletes fi'ee pool o f brain microvascular tissue plasminogen activator in rats. J Cereb B lood Flow M etab 17, 136-46 95. Wen D, Dittman WA, Ye RD, Deaven LL, Majerus PW, Sadler JE (1987) Human thrombomodulin: complete cDNA sequence and chromosome localization of the gene. Biochem 26,4350-7 96. Wiman B (1995) Plasminogen activator inhibitor 1 (PAI-1) in plasma: its role in thrombotic disease. Thromb Haemost 74, 71-6 97. W olf PA, Kannel WR, Verter J (1983) Current status o f risk A ctors for stroke. N eurol C lin 1, 317-43 98. W olf PA, D'Agostino RB, Belanger AJ, Kannel WB (1991) Probability o f stroke: a risk profile fium the Framingham Study. Stroke 22, 983-8 99. Wong VLY, Hofinan FM, Ishii H, Fisher M (1991) Regional distribution o f thrombomodulin in the human brain. Brain R es 556, 1-5 19 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 100. Wim TC, Gq)uano A (1985) Spontaneous fibrinolysis in whole human plasma. Identification o f tissue activator-related protein as the major plasminogen activator causing qwntaneous activity in vitro. J B io l Chem 260, 5061-6 101. Zivin AJ, Fisher M, deGiroiami U, Hemenway CC, Stashak JA (1985) Tissue plasminogen activator reduces neurological damage afto ' cer^ ral embolism. Science 230, 1289-93 102. Zlokovic BV, Wang L, Sun N, Hafike S, Verrall S, Seeds NW, Fisher MJ, Schreiber SS (1995) Expression o f tissue plasminogen activator in cerebral capillaries; possible fibrinolytic function o f the blood-brain barrier. N eurosurg 37, 955-61 103. Zushi M, Gomi K, Yamamoto S, Maruyama I, Hayashi T, Suzuki K (1989) The last three consecutive epidermal growth 6ctor-like structures o f human thrombomodulin comprise the minimum functional domain for protein C activation and anticoagulant activity. JB io l Chem 264, 10351-3 20 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 2 Regulation of Brain Capillary Endothelial Thrombomodulin mRNA Expression Abstract Endothelial ceils regulate hemostasis in part via expression o f thrombomodulin (TM), a potent anticoagulant piotein. The purpose o f this study was to analyze brain capillary endothelial cell expression o f TM mRNA. Bovine brain capillary endothelial cells were grown in a blood-brain barrier model in which endothelial cells form capillary- like structures (CS). In situ hybridization and polymerase chain reaction (PCR) were used to examine TM expression. Endothelial cells were then co-cultured with astrocytes. We examined both mono- and co-culture preparations for y-glutamyl transpeptidase (GGTP), a marker o f the blood-brain barrier. We then used quantitative-competitive PCR to compare TM expression in endothelial mono-cultures and astrocyte-endothelial co-cultures after one day and seven days o f culture. Both in situ hybridization and PCR studies demonstrated TM mRNA expression by endothelial cells. During one week o f astrocyte-endothelial co-culture, there was a) progressive association o f astrocytes with CS and b) expression o f GGTP; endothelial mono-cultures did not express GGTP. There 21 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. was no significant difiference in TM mRNA expression for co-cuhure vs. mono-cultures after 1 day. A fter one week, however, astrocyte-endothelial co-cultures had markedly decreased TM mRNA compared to mono-cuhures (9 ± 2 pg/ml vs. 189 ± 62 pg/ml; p<0.025). This TM mRNA decrease thus occurred when elements o f the blood-brain barrier phenotype were demonstrable, i.e., when astrocyte-CS association was maximal and when GGTP was expressed in co-cultures. These findings indicate astrocyte regulation o f TM mRNA expression in vitro, and suggest an important role for the blood- brain barrier in the regulation o f thrombomodulin. 22 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Introduction Thrombomodulin (TM), an endothelial integral membrane protein, has a crucial hemostasis regulatory role. TM binds the normally procoagulant thrombin, and the TM- thrombin complex activates the circulating zymogen protein C (Esmon & Owen, 1981; Salem et al., 1984). Activated protein C functions as a critical circulating anticoagulant by inactivating clotting Actors Va and V nia (Walker et al., 1979; Fulcher et al., 1984). TM-dependent protan C activation has a major role in maintaining normal blood fluidity and preventing intravascular thrombosis. Animals pretreated with purified or recombinant TM are protected against thromboembolism (Kumada et al., 1987; Gomi et al., 1990). Activated protein C provides protection against procoagulant and lethal effects in a sepsis model (Taylor et al., 1987). Resistance to the effects of activated protein C is closely linked to venous thrombosis in humans (Svensson & Dahlback, 1994). TM thus offers crucial protection against a variety o f thrombotic events. Although the actions of TM are well defined, little is known o f the mechanisms that govern endothelial expression o f TM in the brain. Endothelial cells lining capillaries o f the central nervous system (GNS) are unique in that they comprise a central component o f the blood-brain barrier (BBB), a physiological structure separating the vascular bed from the CNS. A primary role o f the BBB is to control selective transport of essential nutrients, hormones and proteins to the CNS (Pardridge, 1991). Endothelial cells at the BBB are intimately associated with several cell types including astrocytes, whose processes ensheathe capillaries and form the glia limitans (Goldstein, 1988; Risau & Wolburg, 1990). Astrocyte effects on brain 23 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. endothelial cells include tight junction formation ((Stewart & Wiley, 1981; Tao-Cheng et al., 1987), increased electrical resistance (Dehouck et al., 1990; Raub et al., 1992), and induction o f y-glutamyl-glutamyl transpeptidase (GGTP) (DeBault & Cancilla, 1980). Astrocytes also r% ulate expression o f the iow-density lipoprotein receptor by brain capillary endothelial cells (Dehouck et al., 1994). Thrombosis is o f paramount importance in the pathophysiology o f ischemic stroke (Fisher, 1995). Recent work has b%un to define the role o f microcirculatoiy hemostasis factors in large vessel infarction models. Tissue plasminogen activator (tPA) has been identified at the BBB in rats (Zlokovic et al., 1995). In a diabetic model, dow nr^^lation o f brain microvascular tPA mRNA and protein is associated with increased infarct size following reversible middle cerebral artery (MCA) occlusion (Kittaka et al., 1996). Chronic infusion o f nicotine depletes brain microvascular tPA protein, and is also associated with increased infarct size following MCA occlusion (Wang et al, 1996). In a primate MCA occlusion model, microcirculatoiy deposition o f fibrin is partially mediated by tissue factor (Okada et al., 1994). The role o f TM in stroke models has yet to be defined. However, in a coronary artery occlusion model, blockade o f activated protein C produces impaired cardiac outcome (Snow etal, 1991). A number o f investigators have studied TM expression in the brain using immunocytochemistry (Ishii et al., 1986; Wong et al., 1991; Bofia et al., 1991; Isaka et al., 1994; Maruno et al., 1994). These investigations, while yielding somewhat contradictory findings, indicate limited expression o f TM protein in the brain compared to other organs. The restricted expression of brain TM is unexplained, and the relationship between TM expression and the BBB has not been delineated. In the present 24 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. study, we used a BBB model to test the hypothesis that astrocytes regulate brain TM mRNA expression. Tight junction formation and extensive astrocyte-endothelial interactions have been demonstrated in this BBB model (Minakawa et al., 1991). 25 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Materials & Methods Cell Culture Bovine brain capillary endothelial cells were isolated by modification of techniques o f Carson and Haudenschild (1986). Following transportation at 4°C from a local meat processing company, bovine brains were rinsed in a medium containing Dulbecco's Modified Eagle's Medium (DMEM), 1 % bovine albumin serum, 100 U/ml penicillin, 100 mg/ml amphotericin B, and 2 rmnol/L L-glutamine (Irvine Scientific, Santa Ana, CA). Under sterile conditions, the pial membrane was removed, cortical grey matter was aspirated with a pasteur pipette and centrifuged at lOOX g for 10 minutes. Following a rinse with medium, the tissue was homogenized and serially passed through nylon meshes o f 149, 74, and 20 pm. The tissue retained by the 74 and 20 pm meshes was digested at 37®C overnight by 1 mg/mL coUagenase (Sigma, St Louis, MO). Following the overnight digestion, the tissue was incubated with trypsin-EDTA (2.5 and 0.2 mg/ml, respectively) for 30 minutes. The tissue was resuspended in medium containing DMEM, 15 % plasma derived serum (Cocalico Biological, PA), 100 U/ml penicillin, 100 mg/ml streptomycin, 2 mmol/L L-glutamine and plated on dishes coated with 1 % gelatin (Sigma, St Louis, MO). Twenty-four hours following plating, adherent cells were washed and fed fresh medium. Pancreatin (2.5 mg/ml for 3-5 minutes at 25°C) was used to passage subconfluent cells 3-4 days following plating. This step is followed by a trypsin-EDTA treatment for 2-3 minutes, resulting in selective release o f endothelial cells. Pancreatin-trypsin-EDTA treatment was repeated for 4-5 passages to obtain a pure bovine brain capillary endothelial cell population. Endothelial cells were maintained on 26 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 % gelatin-coated culture dishes in DMEM supplemented with 2.5 % equine serum (Hyclone Labs* Logan* UT)* 100 U/ml penicillin, 100 mg/ml streptomycin* and 2 mmol/L L-glutamine in a humidified 5 % CO2 95 % air incubator at 37^C. Endothelial cells* passaged twice weekly using trypsin-EDTA at a split ratio o f 1:2, were characterized by cobblestone-like morphology (Fig. 2.1)* uptake o f acetyiated low density lipoprotein labeled with l-r-dioctacecyl-l-l-3-3-3'-3'-tetram ethyl-indocarbocyanine perchlorate (Biomedical Technologies* Stoughton* MA), and immimoreactivity for von Willebrand factor (vWF)* as previously described (Minakawa et al., 1991). Experiments were performed on endothelial cells between passages 15 and 25. Neonatal mouse astrocytes were isolated according to the methods o f McCarthy and deVellis (1980), performed within institutional guidelines. Briefly, cerebral hemispheres were removed from 1-2 day old Swiss-Webster pups, cleaned o f meninges and choroid plexus and serially sieved through meshes o f 230 and 140 ^un. The filtrate was centrifuged at 200X g for 5 minutes at 25^C and resuspended in DMEM supplemented with 10 % fetal bovine serum (Hyclone lab, Santa Ana, CA), 100 U/ml penicillin, 100 mg/ml streptomycin, and 2 mmol/L L-glutamine. Cells were plated at a density o f 30 cm^/brain and maintained on DMEM with 10 % fetal bovine serum in a humidified 5 % CO2 95 % air incubator at 37® C . At confluence, oligodendroglia were removed by orbital shaking at 37®C. Astrocytes were characterized by >99 % immunoreactivity for glial fibrillary acidic protein (GFAP). Astrocytes used for these experiments were taken from primary cultures for establishment o f astrocyte-endothelial co-cultures. Mouse liver cells (CCL 9.1, ATCC, Rockville, MD) were maintained under conditions similar to astrocytes. 27 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. finmunohistochemistry Methanol-fixed frozen slides were rehydrated with phosphate buffered saline (PBS, pH 7.4), incubated with 0.3 % H2O2 for 5 minutes, and washed with PBS. A fter a IS minute incubation with 5 % normal goat serum, the slides were incubated for 1 hour at 25®C with rabbit anti-bovine GFAP antibody (Dako, Carpinteria, CA) or rabbit anti human vWF antibody (Dako) at a dilution o f 1:200 and 1:100, respectively. Control slides were incubated with PBS in place o f the primary antibody. Slides were then incubated with biotinylated goat anti-rabbit immunoglobulin (Vector, Burlingame, CA) at 1:200 dilution for 30 minutes, avidin-biotin peroxidase complex (Vector, Burlingame, CA) for 15 minutes, and amino-ethyl carbazole for 10 minutes. Slides were counterstained in Mayer's hematoxylin and mounted in glycerol. Blood-Brain Barrier Model Capillary-like structures (CS) were prepared according to the method of Minakawa et al (1991) by first coating two 2x2 cm chamber Lab-Tek glass slides with 1 % gelatin (Sigma, St Louis, MO) and then adding 4x10* endothelial cells/chamber in 1.0 ml DMEM with 2.5 % equine serum. After incubation for 24 hours, the cells were washed with cold PBS, then 0.4 ml o f a second collagen solution (pH 7.4) containing 80 % type I collagen (Vitrogen, Celtrix Lab, Palo Alto, CA), 10 % lOX Minimum Essential Medium (Gibco, Gaithersburg, MD), and 10 % 0.1 mol/L NaOH was added to the subconfluent monolayer and excess solution was then aspirated. The slides were incubated for 10 minutes at 37°C and culture medium (2.5 % equine serum-supplemented DMEM) was then added to the slides. Endothelial cells elongated and formed CS within 28 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 24 hours. We established astrocyte-endothelial or liver-endothelial co-cultures 3 days after the addition o f the second collagen layer. Prior to the addition o f astrocytes (4x10^ cells/chamba’ ) or liver cells (4x10^ cells/chamber) to the c ^ illa iy preparations, serum- supplemented media from astro<^e or liver preparations were removed, the cells were treated with trypsin-EDTA, and resuspended in endothelial culture medium. One day and 7 days following the addition o f astrocytes, the cultures were fixed with 80 % ethanol for 10 minutes and stained for GFAP to demonstrate association with CS. Cultures were monitored and photographed (lOX magnification) with an Olympus CK-2 phase-contrast microscope. The extent o f CS formation for both mono- and co-cultures was determined by computer assisted image analysis o f photomicrographs using a Quantimet 970 Image Analysis System and the Quips software package (Cambridge Instruments Limited, U.K.). Photomicrographs from CS preparations were digitized using a video camera and stored as a 512x512 pixel matrix. The image was displayed as a combination o f a gray image and a binary overlay, representing the detected region. An Amend computer algorithm was used to measure CS length. Digitized images of CS photomicrographs were calibrated against a digitized image o f a 10 cm ruler. Quantifications were performed in duplicate. Cells from mono- and co-culture preparations were treated with trypsin and counted. Cytopreps (Shannon Inc., Pittsburgh, PA) were prepared from mono- and co- culture preparations. Slides were fixed in 80 % ethanol and stained for vWF and GFAP. The percentage o f endothelial cells in culture was determined by the percentage o f cells showing immunoreactivity for vWF. 29 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In Sitn Hybridization Frozen slides were fixed in 4 % paraformaldehyde in PBS for 10 minutes, fol lowed by three 10-minute washes in PBS. The cells were permeabilized for S minutes in 0.05 % Triton-X 100 in PBS, then deproteinized in 0.2 mol/L HCl for 10 minutes and 5.0 pg/ml Proteinase K for 10 minutes. The slides were re-fixed in 4 % paraformaldehyde for 5 minutes and stored at 4°C until use. Synthetic sense (5'CTCGGCAACTACACGTGCATCTGCGAG3') and anti-sense (5GCCACCACCAGAGACAGGCTTGCAATGG3') oUgonucleotides (National Biosciences Inc., Plymouth, MN) were chosen fi-om bovine thrombomodulin mRNA coding regions and used as probes for in situ hybridization (Jackman et al., 1986). Probes were labeled with the Genius 3 Oligonucleotide 3'-End Labeling Kit (Boehringer Mannheim Biochemicals, Germany) according to the manufacturer's instructions. Briefly, 5 pmoI/L of the probe was mixed with the reaction buffer (1 mol/L Potassium cacodylate, 125 mmol/L Tris-HCl, 1.25 mg/ml bovine serum albumin; pH 6.6), 5 nunol/L cobalt chloride, 0.2 mmol/L digoxigenin-11-dUTP, and 2.5 units/pl terminal transferase. The reaction was carried out at 37^C for 15 minutes and stopped by adding 1 pi glycogen solution (20 mg/ml) and 1 pi EDTA (200 mmol/L, pH 8.0) at 4°C. The labeled oligonucleotide was precipitated with 0.1 volume lithium chloride (4 mol/L) and 2.5 volume ethanol at -70°C for 30 minutes. The pellet was washed with 80 % ahanol, dried, resuspended in 20 ml Tris-EDTA/sodium dodecyl sulfate buffer and stored at - 20°C. Hybridization was carried out in 4X SSC (3 mol/L sodium chloride, 0.3 mol/L sodium citrate), 50 % formamide, IX Denhardt's solution, 5 % dectran sulfate, 0.5 mg/ml 30 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. salmon sperm DNA, 0.25 mg/'ml yeast tRNA, and 5 ng/pl o f digoxigenin-labeled probe. Slides were incubated with the hybridization solution overnight in a humidified chamber at 42®C. Slides wane then washed 2X SSC for 1 hour, IX SSC for 1 hour, both at 25°C, 0.5X SSC at 37®C for 30 minutes, and 0.5X SSC at 25°C for 30 minutes. Immunological detection o f the digoxigenin-labeled probes was performed as indicated by the manufacturer. Briefly, slides were incubated overnight with anti- digoxigenin-alkaline phosphatase conjugate antibody, diluted 1:500 in buffer 1 (100 mmol/L Tris-HCl, 150 mmol/L NaCl, pH 7.5). Slides were washed for 10 minutes 3 times in buffer 1, then rinsed briefly in buffer 3 (100 mmol/L Tris-HCl, 100 nunol/L NaCl, 50 mmol/L MgClz, pH 9.5). Slides were incubated in 0.45 % (v/v) nitroblue tétrazolium and 0.35 % X-phosphate (v/v) in buffer 3 for 1-5 hours. The color reaction was stopped with buffer 4 (10 mmol/L Tris-HCl, 1 mmol/L EDTA, pH 8.0). Y-dutam yl Transpeptidase Staining Co-cultures of astrocytes with CS, as well as endothelial cell and astrocyte mono cultures were stained histochemically for the presence of GGTP. The assay is based on the transfer o f the glutamyl group from the substrate, y-glutamyl-4-methoxy-2- napthylamide, to glycylglycine catalyzed by GGTP, using Fast Blue BB as the chromagen (Rutenburg et al., 1969). The slides were fixed with 80 % ethanol, then incubated at 37^C for 90 minutes in a saline solution containing 0.25 % DMSO, 1.2 mmol/L Fast Blue BB (Sigma, St Louis, MO), 0.125 mg/ml y-glutamyi-4-methoxy-2- napthylamide (Vega Biotechnologies, Tucson, AZ), 20 mmol/L glycylglycine (Sigma, St Louis, MO), 2.5 mmol/L NaOH, 25 mmol/L phosphate buffer. After incubation, the 31 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. slides were washed in saline for 2 minutes, rinsed in 0.1 mol/L CUSO4 for 2 minutes, washed again in saline for 2 minutes and mounted. Polymerase Chain Reaction Total RNA was isolated with the Glassmax DNA Spin Cartridge Isolation System (Gibco BRL, Gaithersburg, MD). Total RNA from each preparation was resuspended in 40 1 1 1 o f diethyl pyrocarbonate (DEPC) treated water. cDNA was synthesized from equal volumes o f total RNA in a total volume of 20 ^1. RNA was incubated in 5 pi o f DEPC- treated water at 65^C for 3 minutes and quickly placed on ice. The RNA was then added to the transcription solution; 1.5 pmol/L oligo dT primers, 50 mmol/L Tris-HCl, 75 nunol/L KCl, 3 mmol/L MgClj, 0.5 nunol/L dNTP, I unit/pl RNase inhibitor, and 13.3 units/pl AMV reverse transcriptase. The reaction was carried out at 42®C for 1 hour and terminated at 52°C for 40 minutes. The cDNA was stored at -20°C until use. The polymerase chain reaction (PCR) reaction mixture contained 0.2-1.0 pg cDNA, 10 mmol/L Tris-HCl, pH 8.3, 50 mmol/L KCl, 0.1 nunol/L dNTP, 1.0 nunol/L M gCk, 1.0 unit Taq Polymerase and 0.5 pmol/L forward and reverse primers. The bovine thrombomodulin primers (National Biosciences Inc., Plymouth, MN) extending from bases 228-256 (forward primer, 5CTCGGCAACTACACGTGCATCTGCGAG3 ) and 907-935 (reverse primer, 5 GCCACCACCAGAGACAGGCTTGCAATGG3') were chosen from coding regions o f the mRNA (Jackman et al., 1986). P-actin primers (Stratagene, LaJoUa, CA; fijrward primer, 5TGACGGGGTCACCCACACTGTGCC CATCTA3'; reverse primer, 5CTAGAAGCATTTGCGGTGGACGATGGAGGG3') were used to amplify P-actin mRNA as a control gene. Amplification was carried out in 32 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. a DNA Thermal Cycler (Perkin-Elmer Corp., Norwalk, CT): initial dénaturation at 94°C, each cycle consists o f dénaturation at 94°C for 30 seconds, annealing at 60® C for 30 seconds, and extension at 72°C for 1 minute. PCR products were visualized by electrophoresis on a 2 % agarose gel and stained with ethidium bromide. Quantitative competitive PCR (QC-PCR) tubes contained all the amplification reagents (described above), a constant amount o f target thrombomodulin cDNA from each preparation, and serial dilutions of known concentrations o f a competitor thrombomodulin cDNA template. The reaction mixture was co-amplified, as described above. The competitor cDNA template was prepared by site-directed mutagenesis (Higuchi et al., 1988). A single base change o f A to G at base pair 346 created a unique Sail restriction site. Following co-amplification, the PCR products were digested with Sail: 10 pi o f the PCR product, 1 unit Sail restriction enzyme, and 2 pi enzyme buffer incubated at 37®C for 2 hours. The digested competitor (591 bp) and target (707 bp) cDNAs were separated by electrophoresis on a 2 % agarose gel and visualized by ethidium bromide staining and UV trans-illumination. Negatives were prepared with a Polaroid camera (Polaroid Corp., Cambridge, MA) and scanned by optical densitometry (Hoefer Instruments, San Francisco, CA). Density readings o f the target cDNA were multiplied by 591/707 to correct for differences in molecular weight. The ratio o f amplified target versus competitor cDNA optical densities was plotted as a function of competitor template concentration. The initial concentration o f target cDNA was derived from the point at which the ratio of target and competitor cDNA optical density equaled 1 (Fig. 2.2). TM mRNA concentrations were adjusted to 1x10^ endothelial cells, i.e., dividing 33 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TM concentration by cell count; data are expressed as means ± SD. For some experiments, mRNA levels were determined by standard denshometric analysis o f PCR products and subsequent calculations o f relative abundance; the latter was determined by arbitrarily characterizing the larger o f the two values (mono- or co-culture) as 100 %. Statistical comparisons between groups were pafbrm ed using unpaired Student t-tests and Pearson's correlation coefficients. Differences were considered significant for p<0.05. 34 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Results W e utilized PCR and in situ hybridization in order to demonstrate selective and specific expression o f TM mRNA by endothelial cells in our culture preparations. PCR performed on monolayer endothelial cell cultures demonstrated TM mRNA expression (Fig. 2.3). There was no PCR amplification o f TM from preparations o f either astrocytes (Fig. 2.3) or liver cells (data not shown). In situ hybridization revealed TM mRNA expression by endothelial cells (Fig. 2.4). We added primary culture murine astrocytes to CS cultures 3 days following the addition o f the second collagen matrix. The composition o f co-cultures (endothelial cells vs. astrocytes) showed little change over time (Table 2.1). After 7 days o f co-culture, nearly all o f the astrocytes were associated with CS, with astrocytic processes completely enveloping the CS (Fig 2.5). Thus, after 7 days, astrocyte-endothelial interactions in co culture mimicked morphological features o f the BBB. We performed GGTP staining to further define astrocyte-endothelial interactions in our model. We analyzed GGTP activity after I and 7 days following the addition o f astrocytes to CS. GGTP activity was not detectable at 1 day co-culture (Fig 2.6). After 7 days co-culture, there was GGTP activity present along the entire length o f CS (Fig. 2.6). Endothelial cell mono-cultures did not rachibit detectable levels o f GGTP. Thus, after 7 days co-culture, astrocytes induced further manifestations o f the BBB. We then used quantitative-competitive PCR to determine levels o f TM mRNA in both endothelial mono-cultures and astrocyte-endothelial co-cultures. After 1 day o f co- culture, there was no significant difference in levels o f TM mRNA between co-cultures 35 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (144 ± 1 1 2 p^m l) and mono-cultures (171 ± 100 pg/ml) (Fig. 2.7). However, after 7 days o f co-culture TM mRNA levels were 9 ± 2 pg/ml; mono-cultures grown in parallel had TM mRNA levels o f 189 ± 62 pg/ml. There was no significant association between TM mRNA concentration and CS length for 7 day mono- or co-culture preparations (p>0.4 and 0.6, respectively). There were no significant differences in g-actin mRNA between mono- and co-cultures after 1 and 7 days (Fig. 2.7). In order to determine the specificity o f the astrocyte findings, we established liver-endothelial co-culture preparations under the same conditions used for astrocyte-endothelial co-cultures. After 7 days, liver-endothelial co-cultures vs. endothelial mono-cultures showed no significant differences in relative abundance o f TM mRNA levels: TM mRNA levels for co-cultures were 93.4 ± 19.0 % o f mono-cultures (p>0.6), while P-actin mRNA levels for mono cultures were 94.4 ± 9.2 % o f co-cultures (p>0.5). 36 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 2.1. Confluent monolayers o f brain capillary endothelial cells demonstrate cobblestone-Uke morphology. 37 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ID 707 5D X g O I " 1 2 0 . S SERIES 1 s e r i e s : s e r i e s : icr* icr* 10* COMP g rrrO U t e m p l a t e < k U ng> Figure 2.2. Quantitative-competitive analysis o f TM mRNA. A constant amount o f unknown target cDNA was added to PCR tubes containing 10-6)Id serial dilutions o f a competitor cDNA template (LD=DNA ladder; lanes 1-6: 8.5 x 1 0® , 8.5 x 1 0 * ‘, 8.5 x 10'^, 8.5 X 10'^, 8.5 X 10"*, 8.5 x 10'^ ng o f initial competitor template, respectively (A). With decreasing amounts o f conq)etitor cDNA added , there is a decrease in competitor PCR products (591 bp) and a concommitant increase in unknown target PCR products (707 bp). Plot o f the ratio o f unknown target vs. conq)etitor PCR products (B). The same cDNA was analyzed in triplicate (series 1-3). 38 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. L D 1 707 bB_ Figure 2.3. PCR analysis o f TM expression. PCR showed amplification o f a 707 bp TM fragment from endothelial cells (lane 2) and lack o f amplification from astrocytes (lane 1); LD=DNA ladder. 39 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 2.5. Glial fibrillary acidic protein (GFAP) immunocytochemistry. Immunocytochemistry for GFAP showed astrocytes associated with capillary-like structures at I day (A), 3 days (B), and 7 days (C). Original magnifications: 50X (.A). l25X (B ),and I OX (C). 41 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 2.6. Gamma-glutamyl transpeptidase (GGT?) activity. GGTP histochemical staining o f capillary-like structures co-cultured with astrocytes after 1 day (B) and after 7 days (D,F,G); also endothelial mono-cultures after 1 day (A) and after 7 days (C,E). Note the lack o f GGTP on 1 day mono- (A) and co-cultures (B) and on 7 day endothelial mono-cultures (C,E). Original magnifications: 25X (G), 50X (A3»C4>) and 125X (E,F). 42 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. IGO-CULTIJRE a MONOCULTURE I ec E I 250 200 150 100 - 1 Day 7 Days I I E ? ja > ( S 1 100 - - 1 Day 7 Days Figure 2.7. Thrombomodulin mRNA concentration and relative percentage o f p-actin in co-culture and mono-culture preparations. Quantitative-competitive PCR was used to quantify TM mRNA isolated from 1 day (A) and 7 day (B) astrocyte-endothelial co cultures and endothelial mono-cultures. TM mRNA concentrations were adjusted to 1x10^ endothelial cells. Seven day co-culture preparations showed significant decrease (p<0.025) compared to 7 day mono-cultures. PCR amplification showed no significant difference in relative abundance o f P-actin mRNA betw eei co- and mono-cultures at 1 day (C ) and 7 days (D). Each set o f experiments was performed in quadruplicate. 43 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 Day Co-Culture 1 Day Mono-Cutture 7 Days Co-Culture 7 Days Mono-Culture vWF positive (% ) 75 100 82 100 GFAP positive (% ) 25 0 18 0 Astrocyte association with CS (% ) 35 95 GGTP staining absent absent present absent Table 2.1. After I and 7 days, astrocyte-endothelial co-culture and endothelial mono culture preparations were examined by a) immunohistochemistry for von W illebrand Factor (vWF) and glial fibrillary acidic protein (GFAP) and b) histochemical staining for gamma-glutamyl transpeptidase (GGTP). Data indicate percentage o f vWF and GFAP immunoreactive cells, percentage o f astrocytes associated with capillary-like structures (CS), and presence o f GGTP staining. 44 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Discussion We demonstrated TM mRNA in both endothelial cell mono-cultures and astrocyte-endothelial co-cultures. After 7 days o f culture, there was more than a twenty- fold decrease in TM mRNA concentration in astrocyte-endothelial co-cultures, compared to endothelial cell mono-cultures. At the same time (i.e., after seven days), elements o f the BBB phenotype were demonstrable by near-complete association o f astrocytes with CS and by GGTP expression. These findings indicate that astrocytes regulate brain capillary endothelial TM mRNA expression in this model. These data also suggest that TM mRNA expression is reduced in endothelial cells expressing the BBB phenotype. The reduction in TM mRNA in our BBB co-culture model is consistent with prior reports. The initial study o f TM in the human brain reported absence o f TM protein in brain capillaries; TM was demonstrable in all other organs investigated (Ishii et al., 1986). Later studies reported TM protein present in human brain microvessels (W ong et al., 1991; Bofta et al., 1991); however, human brain TM expression was found to be substantially reduced in several subcortical regions where infarction is common (Wong et al., 1991). Further investigations reported very limited expression o f TM protein in capillaries o f normal human brain (Isaka et al., 1994; Maruno et al., 1994). Our study examined only expression o f TM mRNA in vitro, which limits the implications o f our findings. Moreover, TM expression by arterioles and venules was not directly addressed by our study. Nevertheless, the results of the current study are consistent with the contention that expression o f TM in brain capillaries is restricted. These data also suggest that astrocytes contribute to this restricted expression. 45 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. We have used an in vitro BBB model, comprising of bovine brain micro vascular endothelial cells and murine astrocytes, to examine TM mRNA expression by endothelial cells. We deliberately used later passage endothelial cells, dedifferentiated from a BBB perspective, in order to examine the effects o f inducing elements o f the BBB. Note, however, that not all features o f the BBB have beer demonstrated in this model. For example, electrical resistance has not been studied in our co-culture system. Our usage o f endothelial cells and astro<^es from different species allowed us to selectively examine mRNA from eithw cell type by taking advantage o f differences in genetic composition between species. Nevertheless, our finding o f TM mRNA regulation by astrocytes needs to be interpreted with some caution. Our in vitro system contains only astrocytes and endothelial ceils; the effects o f pericytes, another component o f the BBB (Risau et al., 1990), on brain endothelial hemostasis factors remain undefined. Moreover, despite our astrocyte cultures having virtually complete immunoreactivity for the astrocyte marker GFAP, we cannot entirely rule out the potential contribution o f non- astrocytic cells in these co-culture preparations. Similarly, despite the presence o f the requisite morphologic, immunocytochemicai, and fimctional features o f our endothelial cell preparations, a potential contribution o f non-endothelial cells in our mono- and co cultures cannot be completed «(eluded. Astrocyte effects o f TM may be mediated by direct contact between astrocytes and endothelial cells, by the production o f a diffiisible factor(s) by astrocytes, or both. Our findings cannot distinguish between these potential mechanisms. Prior work has also demonstrated astrocyte enhancement o f angiogenesis in vitro (Laterra et al., 1990; Laterra & Goldstein, 1991; Laterra et al., 1994). However, it seems unlikely that 46 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. astrcK^e regulation o f TM was mediated by angiogenesis in our model; 1) while there was extensive CS development in both endothelial mono-cultures and astrocyte- endothelial co-cultures, TM mRNA expression declined more than twenty-fold in the co- cultures; 2) there was no significant association between CS length and TM concentration; and 3) increased TM expression has been found in highly vascular malignant gliomas (Isaka et al., 1994; Maruno et al., 1994), suggesting a potential direct (rather than inverse) relationship between angiogenesis and TM expression. However, this latter in vivo finding is not supported by our in vitro data o f astrocyte downregulation o f TM mRNA. Nevertheless, we cannot entirely rule out a possible inverse relationship between TM expression and angiogenesis in our model. Astrocytes have the capacity to secrete the cytokines tum or necrosis factor (Chung et al., 1990) and interleukin-1 (Fontana et al., 1982). Cell culture studies using non-CNS endothelium have shown that these same cytokines downregulate TM mRNA (Conway & Rosenberg, 1988; Krokawa & Aoki, 1991). Therefore, further study is warranted to investigate the role o f these and other cytokines and growth factors in astrocyte regulation o f TM expression. In conclusion, we have demonstrated that astrocytes negatively regulate brain capillary endothelial cell expression o f TM mRNA in vitro. These results suggest that astrocyte-endothelial interactions at the BBB play an important role in the expression of TM by the brain. Further delineation of BBB regulation o f hemostasis may prove helpful for the prevention o f thrombotic disorders o f the brain, including ischemic stroke. 47 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. References 1. B offa MC, Jackman RW, Peyri N, Bo£fa JF, George B (1991) Thrombomodulin in the central nervous system. Nouv R ev F r H em atol 33,423-9 2. Carson MP, Haudenschild CC (1986) Microvascular endothelium and pericytes: high yield, low passage cultures. In Vitro C ell D ev B io l 22, 344-54 3. Chung I, Norris G, Benveniste E (1990) Tumor necrosis 6ctor-alpha production by astrocytes. Induction by lipopolysaccharide, IFN-gamma, and IL-1 beta. J Inam m ol 144, 2999-3007 4. Conway EM, Rosenberg RD (1988) Tumor necrosis factor suppresses transcription o f the thrombomodulin gene in endottelial cells. M ol C ell B iol 8, 5588-92 5. DeBault LE, Cancilla PA (1980) Ganuna-ghitamyi transpeptidase in isolated brain endothelial cells: induction by glial cells in vitro. Science 207, 653-5 6. Dehouck MP, Meresse S, Delorme P, Fruchart JC, Cecchelli R (1990) An easier, reproducible, and mass-production method to study the blood-brain barrier in vitro. J Neurochem 54, 1798-1801 7. Dehouck B, Dehouck MP, Fruchart JC, Cecchelli R (1994) Upregulation of the low density lipoprotein at the blood-brain barrier: intercommunications between brain capillary endothelial cells and astrocytes. J C ell B io l 126,465-73 8. Esmon CT, Owen WG (1981) Identification o f an endothelial cell cofactor for the thrombin-catalyzed activation o f protein C. Proc N a tl A cad Sci USA 78,2249-52 9. Fisher M (1995) Inununologic mechanisms in stroke, in Brain ischemia: Basic concepts and clinical relevance (Caplan, L.R., ed), chapter 8. Springer-Verlag, New York 10. Fontana A, Kristensen F, Dubs R, Gemsa D, W eber E (1982) Production o f prostaglandin E and an interleukin-1 like factor by cultured astrocytes and C6 gliom a cells. J Imm unol 129, 2413-9 11. Fulcher CA, Gardiner JE, GrifiSn JH, Zimmerman TS (1984) Proteolytic inactivation o f human factor V m procoagulant protein by activated human protein C and its analogy with factor V. B lood 63, 486-9 12. Goldstein GW (1988) Endothelial cell-astrocyte interactions. A nnals N Y A cad S ci 529, 31-9 48 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 13. Goini K, Zushi M, Honda G, Kawafaara S, Matsuzaki O, Kanabayashi T, Yamamato S, Maniyama I, Suzuki K (1990) Antithrombotic effect o f recombinant human thrombomodulin on thrombin-induced thromboembolism in mice. B lood 75, 1396-9 14. Higuchi R, Knunmel B, Saiki K (1988) A general method o f in vitro preparation and mutagenesis o f DNA fragments; study o f protein and DNA interactions. N ucleic A cids R es 16, 7351-67 15. Hrokawa K, Aoki N (1991) Regulatory mechanisms for thrombomodulin expression in human umbilical vein endothelial cells in vitro. J C ell P hysiol 147, 157-65 16. Isaka T, Yoshimine T, Motokiko M, Kurode R, Ishii H, Hayakawa T (1994) Altered expression o f antithrombotic molecules in human glioma vessels. A cta Neuropathol 87, 81-5 17. Ishii H, Salem HH, Bell CE, Laporata EA, Majerus PW (1986) Thrombomodulin, an anticoagulant protein, is absent in the human brairL B lood 67, 362-5 18. Jackman RW, Beeler DL, DeWaters L, Rosenberg RD (1986) Characterization o f a thrombomodulin cDNA reveals structural similarity to the low density lipoprotein receptor. P roc N atl A cad Sci USA 83, 8834-8 19. Kittaka Wang L, Sun N, Schreiber SS, Seeds NW, Fisher M, Zlokovic BV (1996) Brain capillary tissue plasminogen activator in a diabetes stroke model. Stroke 27, 712-9 20. Kumada J, Dittman WA, Majems PW (1987) A role for thrombomodulin in the pathogenesis o f thrombin-induced thromboembolism in mice. B lood 71, 728-33 21. Laterra J, Guerin C, Goldstein GW (1990) Astrocytes induce neural microvascular endothelial cells to form capillary-like structures in vitro. J C ell Physiol 144, 204-15 22. Laterra J, Goldstein GW (1991) Astroglial induced in vitro angiogenesis: requirements for RNA and protein synthesis. J Neurochem 57, 1231-9 23. Laterra J, Indurti RR, Goldstein GW (1994) R%ulation o f in vitro glia-induced microvessel morphogenesis by urokinase. J C ell Physiol 158, 317-24 24. Maruno M, Yoshimine T, Isaka T, Kurode R, Ishii H, Mayakawa T (1994) Expression o f thrombomodulin in astrocytomas o f various malignancy and in gliotic and normal brain. J Neuro-Oncology 19, 155-60 25. McCarthy KD, DeVellis J (1980) Preparation o f separate astroglial and oligodend- roglial cell cultures from rat cwebral tissue. J C ell B iol 85, 890-902 49 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 26. Minakawa T, Bready J, Berliner J, Fisher M, Cancilla PA (1991) In vitro interactions o f astrocytes and pericytes with capillary-like structures o f brain microvessel endothelium. Lab Invest 65, 32-40 27. Okada Y, Copeland BR, Fhridge R, Koziol JA, del Zoppo GJ (1994) Fibrin contributes to microvascular obstructions and parenchymal changes during early focal cerebral ischemia and repafiision. Stroke 25, 1847-54 28. Pardridge WM (1991) Advances in cell biology o f blood-brain barrier transport. Sem C ell B iol 1, 419-26 29. Raub TJ, Kuentzel SL, Sawada GA (1992) Permeability o f bovine brain microvessels endothelial cells in vitro; barrier tightening by a Victor released from astroglioma cells. E j q > C ell R es 199, 330-40 30. Risau W, Wolburg H (1990) Differentiation o f the blood-brain barrier. TINS 13, 174-8 31. Rutenburg AM, Kim H, Fischbein JW, Hanker JS, Wasserkrug HL, Seligman AM (1969) Hstochemical and ultrastructural demonstration o f gamma-glutamyl transpeptidase activity. J H istochem Cytochem \1 , 517-26 32. Salem H, Maniyama I, Ishii H, Nfojerus PW (1984) Isolation and characterization o f thrombomodulin from human placenta. J B iol Chem 259, 12246-51 33. Snow TR, Deal MT, Dickey DT, Esmon CT (1991) Protein C activation following coronary artery occlusion in the in situ porcine heart. Circulation 84, 293-9 34. Stewart PA, Wiley MJ (1981) Developing nervous tissues induces formation of blood-brain barrier characteristics in invading endothelial cells: a study using quail- chick transplantation chimeras. D ev B iol 84, 183-92 35. Svensson PJ, Dahlback B (1994) Resistance to activated protein C as a basis for venous thrombosis. N E nglJM ed'i2Q , 517-22 36. Tao-Cheng JH, Nagy Z, Brightman MW (1987) Tight junctions to brain endothelium in vitro are enhanced by astroglia. JN eurosci 7, 3293-9 37. Taylor FB, Chang A, Esmon CT, D'Angelo A, Vigano-D'Angelo S, Blick KE (1987) Protein C prevents coagulopathic and lethal effects o f Escherichia coli injection in the baboon. JC lin Invest 79,918-25 38. W alker FJ, Sexton PW, Esmon CT (1979) The inhibition o f blood coagulation by activated protein C through the selective inactivation o f activated factor V. Biochim B iophysA cta 571, 333-42 50 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 39. Wang L, Kittaka M, Sun N, Schreiber SS, Zlokovic BV (1996) Chronic nicotine treatment enhances focal ischemic brain injury and depletes fiee pool o f brain microvascular tissue plasminogen activator in rats. J Cereb B lood Flow M etab 17, 136-46 40. Wong VLY, Hofinan FM, Ishii H, Fisher M (1991) Regional distribution o f thrombomothilin in the human brain. Brain R es 556, 1-5 41. Zlokovic BV, Wang L, Sun N, HaSke S, Verrall S, Seeds NW, Fisher MJ, Schreiber SS (1995) Expression o f tissue plasminogen activator in the cerebral capillaries: Possible fibrinolytic function o f the blood-brain barrier. Neurosurg 37, 955-61 51 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 3 Astrocyte Regulation of Endothelial Tissue Plasminogen Activator in a Blood-Brain Barrier Model Abstract Expression o f tissue plasminogen activator (tPA) substantially determines endothelial-dependent fibrinolysis. We used a blood-brain barrier (BBB) model to analyze regulation o f brain capillary endothelial tPA and its inhibitor, plasminogen activator inhibitor-1 (PAI-1). This model consists o f co-culture o f murine astrocytes with bovine brain capillary endothelial cells grown as capillary-like structures (CS); after one week, astrocytes become extensively associated with CS and the BBB-associated enzyme y-glutamyl transpeptidase is present. We measured tPA and PAI-1 mRNA and tPA activity in this model. Reverse transcription-polymerase chain reaction (RT-PCR) studies showed similar tPA and PAI-1 mRNA levels after one day mono-culture (endothelial cells only) vs. astrocyte-endothelial co-culture preparations. After seven days (i.e., when elements o f the BBB are present), astrocyte-endothelial co-cultures (compared to endothelial mono-cultures) showed a 51 ± 27 % (mean ± SD) reduction in tPA mRNA (p<0.03) and a 183 ± 87 % increase in PAI-1 mRNA expression (p<0.02). 52 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Moreover, seven day co-cultures demonstrated reduced tPA activity compared to mono cultures (14.6 ± 2.9 lU/ml vs. 30.2 ± 7.7 lU/ml, p<0.01); one day co- and mono-cultures had sim ilar tPA activity. These findings demonstrate that astrocytes regulate brain capillary endothelial expression o f tP A when elements o f the BBB phenotype are present in this model. These data suggest an important role for astrocytes in the regulation of brain capillary endothelial fibrinolysis. 53 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Introduction Endothelial cells play a critical role in the regulation o f the fibrinolytic system by synthesis and secretion o f tissue plasminogen activator (tPA) and its inhibitor, plasminogen activator inhibitor-1 (PAI-1). tPA proteolytically activates plasminogen to form plasmin, a protease critically involved in fibrinolysis (Collen and Lijnen, 1995). PAI-1 rapidly forms stable 1:1 complexes with tPA and functions as the most important plasma inhibitor o f tPA (Wiman, 1995). Expression o f tPA has been demonstrated in the brain microvasculature (Levin and del Zoppo, 1994; Zlokovic et ai., 1995) and in cultured brain microvascular endothelium (Saksela et al., 1990; KoUros et al, 1994; Shatos et al., 1995); these findings suggest a role for brain endothelium in the regulation o f fibrinolysis. Moreover, fibrinolytic drugs used in stroke models reduce infarct size in both thromboembolic and non-thrombotic occlusion o f intracranial vessels (Zivin et al., 1985; del Zoppo et al., 1986). The importance o f brain expression o f tPA has been further emphasized by the demonstration that exogenous tPA can improve outcome in acute ischemic stroke (NINDS rt-PA Stroke Study Group, 1995). Nevertheless, little is known o f the mechanisms r%ulating endothelial fibrinolysis within the brain. Brain capillary endothelial cells are unique in that they demonstrate features o f the blood-brain barrier (BBB), including tight junctions, increased electrical resistance, lack o f pinocytic vesicles, and expression o f highly selective transport systems to deliver essential nutrients across the BBB (eg, GLUT-1) (Pardridge, 1984, 1991). The induction and maintenance o f the endothelial BBB phenotype is thought to be largely under the regulatory control o f perivascular astrocytes (DeBault and Cancilla, 1980; Stewart and 54 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. W il^ , 1981; Tao-Cheng et al., 1987; Dehoudc et al., 1990; Raub et al., 1992), ix*ose processes ensheathe capillaries (Goldstein, 1988; Risau and Wolburg, 1990). Astrocytes also play a r% ulatory role in other endothelial functions, including upregulating endothelial expression o f low-density lipoprotein receptor (Dehouck et al., 1994) and Na- K-Cl C O transporter (Sun et al., 1995). Moreover, endothelial cells expressing the BBB phenotype exhibit enhanced expression o f the GLUT-1 glucose transporter (Hayashi et al., 1997) and downr%ulation o f MECA-32 (Hallman et a!., 1995) and thrombomodulin (Tran et al., 1996). Several lines o f investigation show limited expression o f tPA by brain microvascular endothelium (Levin and del Zoppo, 1994; Shatos et al., 1995, Grau et al., 1997). These observations suggest brain-specific tPA expression and led us to hypothesize that astrocytes r^;ulate brain capillary endothelial expression o f genes important in the fibrinolytic pathway. 55 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Materials & Methods Cell culture We isolated bovine brain c^ illary endothelial cells by modification o f techniques o f Carson and Haudenschild (1986). Following transportation at 4°C from a local meat processing company, bovine brains were rinsed in a medium containing Dulbecco's M odified Eagle's Medium (DMEM), 1 % bovine albumin serum, 100 U/ml penicillin, 100 mg/ml amphotericin B, and mmol/L L-glutamine (Irvine Scientific, Santa Ana, CA). Under sterile conditions, the pial membrane was removed, cortical grey matter was aspirated with a pasteur pipette and caitrifuged at lOOX g for ten minutes. Following a rinse with medium, the tissue was homogenized and serially passed through nylon meshes o f 149, 74, and 20 pm. The tissue retained by the 74 and 20 pm meshes was digested at 3'fC overnight by 1 mg/ml coUagenase (Sigma, St Louis, MO). Following the overnight digestion, the tissue was incubated with trypsin-EDTA (2.5 and 0.2 mg/ml, respectively) for 30 minutes. The tissue was resuspended in medium containing DMEM, 15 % plasma derived serum (Cocalico Biological, PA), 100 U/ml penicillin, 100 mg/ml streptomycin, 2 mmol/L L-glutamine and plated on dishes coated with I % gelatin (Sigma, St Louis, MO). Twenty-four hours following plating, adherent cells were washed and fed fresh medium. Pancreatin (2.5 mg/ml for three to five minutes at 25®C) was used to passage subconfluent cells three to four days following plating. This step is followed by a trypsin-EDTA treatment for two to three minutes, resulting in selective release o f endothelial cells. Pancreatin-trypsin-EDT A treatment was repeated for four to five passages to obtain a pure bovine brain capillary endothelial cell population. 56 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Endothelial cells were maintained on uncoated culture dishes in DMEM supplemented with 2.5 % equine serum (Hyclone Labs, Logan, UT), 100 U/ml penicillin, 100 mg/ml streptomycin, and 2 mmol/L L-glutamine in a humidified 5 % CCh 95 % air incubator at 37®C. Endothelial cells, passaged twice weekly using trypsin-EDTA at a split ratio of 1 ;2, were characterized by cobblestone-like morphology, uptake o f acetylated low density lipoprotein labeled with 1-1 '-dioctacecyl-1-1 -3-3-3'-3'-tetramethy 1-indocarbocyanine perchlorate (Biomedical Technologies, Stoughton, MA), and immunoreactivity for von Willebrand factor (vWF), as previously described (Tran et al., 1996). Experiments were performed on endothelial cells between passages 10 and 25. Neonatal mouse astrocytes were isolated according to the methods o f McCarthy and deVellis (1980), and performed within institutional guidelines. Briefly, cerebral hemispheres were removed from one to two day old Swiss-Webster pups, cleaned of meninges and choroid plexus and serially sieved through meshes o f 230 and 140 pm. The filtrate was centrifuged at 200X g for five minutes at room temperatures and resuspended in DMEM supplemented with 10 % fetal bovine serum (Hyclone lab, Santa Ana, CA), 100 U/ml penicillin, 100 mg/ml streptomycin, and 2 mmol/L L-glutamine. Cells were plated at a density o f 30 cm % rain and maintained on DMEM with 10 % fetal bovine serum in a humidified 5 % CO2 95 % air incubator at 37®C. At confluence, oligodendroglia were removed by orbital shaking at 37°C. Astrocytes were characterized by >99 % immunoreactivity for glial fibrillary acidic protein (GFAP). Experiments were performed using primary culture astrocytes between 20-40 days postnatal (P20-P40). Mouse liver cells (CCL 9.1, ATCC, Rockville, MD) were maintained under conditions similar to astrocytes. 57 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Dnmumocytocbembtry Methanol-fixed fiozen slides were rehydrated with phosphate bufibred saline (PBS, pH 7.4), incubated with 0.3 % H2O2 fi)r five minutes, and washed with PBS. After a IS minute incubation with 5 % normal goat serum, the slides were incubated for one hour at room temperature with rabbit anti-bovine GFAP antibody (Dako, Carpinteria, CA) or rabbit anti-human vWF antibody (Dako) at a dilution o f 1:200 and 1:100, respectively. Control slides were incubated with PBS in place o f the primary antibody. Slides were then incubated with biotinylated goat anti-rabbit inununoglobulin (Vector, Burlingame, CA) at 1:200 dilution for 30 minutes, avidin-biotin peroxidase complex (Vector, Burlingame, CA) for 15 minutes, and amino-ethyl carbazole for ten minutes. Slides were counterstained in Mayer's hematoxylin and mounted in glycerol. Blood-brain barrier model We prepared our BBB model as previously described (Minakawa et al., 1991; Tran et al., 1996). To prepare capillary-like structures (CS) we first coated two 2x2 cm chamber Lab-Tek glass slides with 1 % gelatin (Sigma, St Louis, MO) and then added 4x10^ endothelial cells/chamber in 1.0 ml DMEM with 2.5 % equine serum. After incubation for 24 hours, the cells were washed with cold PBS, then 0.4 ml o f a collagen solution (pH 7.4) containing 80 % type I collagen (Vitrogen, Celtrix Lab, Palo Alto, CA), 10 % lOX Minimum Essential Medium (Gibco, Gaithersburg, MD), and 10 % 0.1 mol/L NaOH was added to the subconfluent monolayer and excess Vitrogen solution was then aspirated. The slides were incubated for ten minutes at 37°C and culture medium (2.5 % equine serum-supplemented DMEM) was then added to the slides. Endothelial cells 58 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. elongated and formed CS within 24 hours (Fig. I). We established astrocyte-endothelial or liver-endothelial co-cultures three days after the addition o f the second collagen layer. Prior to the addition o f astrocytes (4x10^ cells/chamber) or liver cells (4x10* cells/chamber) to the capillary preparations, serum-supplemented media from astrocyte or liver prqiarations were removed, the cells were treated with trypsin-EDTA, and resuspended in endothelial culture medium. One day and seven days following the addition o f astrocytes, the cultures were fixed with 80 % ethanol for ten minutes and stained for GFAP to demonstrate association with CS. After one day o f co-cxilture, only 35 % o f the astrocytes were associated with CS. After seven days, nearly all o f the astrocytes were associated with CS and their processes frequently enveloped the entire CS (Fig. 2). Thus, astrocyte-endothelial cell co-cultures exhibited morphological features o f the BBB after seven days o f co-culture. We performed histochemical staining for the presence o f y-glutamyl transpeptidase (GGTP), a putative marker o f the BBB, in co culture and mono-cmlture preparations (Tran et al., 1996). At the end o f one day in co- culture GGTP activity was not detectable. However, after seven days extensive GGTP activity was detectable along the length o f CS. GGTP activity was neither detectable when CS were cultured without astrocytes nor in astrocyte mono-cultures. Additional work has described tight junctions in this model (Minakawa et al., 1991). We monitored and photographed (lOX magnification) cultures with an Olympus CK-2 phase-cx)ntrast microscope. The extent o f CS formation for both mono- and co cultures was determined by computer assisted image analysis o f photomicrographs using a Quantimet 970 Image Analysis System and the Quips software package (Cambridge Instruments Limited, U.K.). Photomicrographs from CS preparations were digitized 59 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. using a video camera and stored as a 512x512 pixel matrix. The image was displayed as a combination o f a gray image and a binary overlay, representing the detected region. An Amend computer algorithm was used to measure CS length. Digitized images o f CS photomicrographs were calibrated against a digitized image of a 10 cm ruler. Quantifications were performed in duplicate. Cells from co-culture and mono-culture preparations were treated with trypsin and counted. Cytopreps (Shannon Inc., Pittsburgh, PA) were prepared from co-culture and mono-culture preparations. Slides were fixed in 80 % ethanol and stained for vWF and GFAP. The number o f endothelial cells in culture was derived from the cell count and the percentage o f cells showing immunoreactivity for vWF. The relative composition of endothelial cells and astrocytes showed only minor differences for one vs. seven days co- culture (Tran et al., 1996). Conditioned-media experiments We performed conditioned-media experiments according to the method of Maxwell et al. (1987). Briefly, astrocyte-conditioned media were prepared by first aspirating the growth media from these cultures, washing with PBS, and replacing media with 5 % equine serum-supplemented media. Twenty-four and forty-eight hours later the astrocyte-conditioned media (ACM-24 and ACM-48, respectively) were collected, centrifuged to remove cellular debris and stored at -80°C until use. Bovine brain endothelial cells were plated in 24-well tissue culture plates (4x10^ cells/well) in 1 ml 5 % equine serum supplemented media. At confluence the cells were incubated for 24 60 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. hours with ACM-24, ACM-48 or non-condhioned 5 % equine serum-supplemented media. Plasminogen activator assay We assayed cultured media from mono- and co-culture preparations for total plasminogen activator activity by amidolytic assay (American Diagnostica, Greenwich, CT). Amiloride (0.5 mmol/L, Sigma) was used to inhibit urokinase plasminogen activator (uPA) and allow for the determination o f net tPA activity (Vassalli and Belin, 1987). Polymerase chain reaction Because our culture preparations contained a relatively small number of cells, we used reverse transcription-polymerase chain reaction (RT-PCR) to measure the abundance o f tPA mRNA transcripts. Total RNA was isolated with the Glassmax DNA Spin Cartridge Isolation System (Gibco BRL, Gaithersburg, MD). Total RNA from each preparation was resuspended in 40 pJ o f diethyl pyrocarbonate (DEPC)-treated water. cDNA was synthesized from equal volumes of total RNA in a total volume of 20 jil. RNA was incubated in 5 pi o f DEPC-treated water at 65°C for three minutes and quickly placed on ice. The RNA was then added to the transcription solution: 1.5 pmol/L oligo dT primers, 50 mmol/L Tris-HCl, 75 mmol/L KCl, 3 mmol/L MgClz, 0.5 mmol/L dNTP, 1 unit/pl RNase inhibitor, and 13.3 units/pl AMV reverse transcriptase. The reaction was carried out at 42® C for one hour and terminated at 52°C for 40 minutes. The cDNA was stored at -20®C until use. PCR primers (National Biosciences Inc., Plymouth, MN) for bovine tPA (forward primer, BTPA-1: 5'AAGGTTGCAGAAGAAGATGG3' and reverse primer, BTPA-2: 61 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5'GTGAGGCGGGTACCTCTCCTGGAA3') and PA I-l (forward primer, BPAI-A: 5’GCCTCCTACCAGCCCCAGTCT3’ and reverse primer, BPAI-B: 5' AATTCCAGGATGTCGTAGTAACGG3 ') were chosen from coding regions o f the mRNA (Mimuro et al., 1989; Prendergast GC et al., 1990; Ravn et al., 1995; Rickies et al., 1988). In order to elucidate the contribution o f endothelial cell transcription, we chose primers that were specific and selective for bovine endothelial tPA and PAI-1 mRNA (Fig. 3). Primers for tPA (BTPA-1 and BTPA-2) and PAI-1 (BPAI-A and BPAI- B) share 100 % homology with sequences for bovine tPA and PAI-1 mRNA, respectively; these primers have limited homology with sequences for murine tPA (40 and 54 %) and PAI-1 mRNA (57 and 75 %, respectively) (Mimuro et al., 1989; Prendergast GC et al., 1990; Ravn et al., 1995; Rickies et al., 1988). (B-Actin primers (Stratagene, La JoUa, CA; forward primer, 5TGACGGGGTCACCCACACTGTGCCCATCTA3’; reverse primer, 5CTAGAAGCATTTGCGGTGGACGATGGAGGG3') were used to amplify 3-actin mRNA as a hous^eeping gene control. The PCR mixture contained 0.2-1.0 jag cDNA, 10 mmol/L Tris-HCl, pH 8.3, 50 mmol/L KCl, 0.1 mmol/L dNTP, 1.0 mmol/L MgClz, 1.0 unit Taq Polymerase and 0.5 |imol/L forward and reverse primers. Amplification was carried out in a DNA Thermal Cycler (Perkin-Elmer Corp., Norwalk, CT): initial dénaturation at 94®C, each cycle consists of one minute dénaturation at 94°C, one minute annealing at 54®C, and two minutes extension at 72°C. AH RT-PCR assays were performed within the linear range o f amplification. PCR products were visualized by electrophoresis on a 2 % agarose gel and stained with ethidium bromide. Negatives were prepared with a Polaroid camera (Polaroid Corp., Cambridge, MA) and scanned by 62 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. optical densitometry (Hoefer Instniments, San Francisco, CA). Optical densitometric (CD) measurements o f tPA and PAI-1 levels were normalized to either endothelial cell counts cdls (i.e., dividing mRNA concentration by cell count), or P-actin levels (i.e., dividing by P-actin CD measurements). Subsequent calculations o f relative mRNA used the mean values of the control group (aidothelial mono-culture or unstimulated endothelial monolayers) as 100% . We validated our RT-PCR results by comparing results o f same samples (n=27) analyzed by quantitative-competitive PCR (QC-PCR). QC-PCR tubes contained all the amplification regents, a constant amount o f target cDNA from each preparation, and serial dilutions o f known concentrations o f a competitor tPA cDNA template. The competitor tPA cDNA template was generated by RT-PCR (described above), using forward primer BTPA-3; 5’AAGGTTGCAGAAGAAGATGGGAAGCACAACCACT GCA3’ and reverse primer BTPA-2: S GTGAGGCGGGTACCTCTCCTGGAA3' Both the wild-type tPA template (479 bp) and competitor tPA template (332 bp) contain primers bindings sites for primers BTPA-1 and BTPA-2. The QC-PCR reaction mixture was co-amplified, as described above. The target and competitor cDNAs were separated by electrophoresis on a 2 % agarose gel and visualized by ethidium bromide staining and UV trans-illumination. OD readings o f the target cDNA were multiplied by 332/479 to correct for differences in molecular weight. OD measurements o f the target and competitor cDNAs were plotted as a function o f competitor template concentration. The concentration o f target cDNA was derived from the point at which the OD o f the target and competitor cDNAs were equal. QC-PCR analyses o f mRNA concentrations were 63 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. normalized to 3-actin levels. Thaïe was a high correlation between results &om standard densitometric analyses o f RT-PCR products and results by QC-PCR (r=0.91, p<0.001). Statistical analysis All data are expressed as means ± SD. For the conditioned-media experiments, statistical analysis was performed on pooled data from three independent experiments. Statistical comparisons between groups were performed using unpaired Student t-tests and Pearson's correlation coefGcients. Differences were considered significant for p<0.05. 64 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Results We examined astrocyte effects on endothelial tPA transcription in our BBB model. RT-PCR showed no significant difference in levels o f tPA transcripts for astroq^e-endothelial co-cultures vs. endothelial mono-cultures after one day (Fig. 4). After seven days, however, astitxyte-endothelial co-cultures showed a 51 ± 27 % (mean ± SD) reduction in tPA mRNA (p < 0.03) compared with endothelial mono-cultures. This reduction thus occurred when elements o f the BBB were demonstrable in our model, i.e., astrocytes extensively associated with CS and GGTP detectable. To determine the specificity o f the changes in tPA mRNA, we measured the levels o f 3-actin mRNA, a housekeeping gene. Astrocyte-endothelial co-cultures showed no significant difference in 3-actin mRNA compared with mono-cultures after one day (94 ± 5 % o f mono-cultures; p>0.2) and seven days (108 ± 21 % of mono-cultures; p>0.7). To investigate the cellular specificity o f astrocyte regulation o f endothelial tPA mRNA, we analyzed the expression o f tPA mRNA in liver-endothelial co-cultures. After seven days, liver-endothelial co-cultures mchibited no significant differences in levels of tPA mRNA (88 ± 30 % o f mono-cultures; p<0.6). We analyzed our data for a potential association between CS length and mRNA levels. There was no significant correlation between CS length and tPA mRNA concentration in mono-cultures (r=0.41, p>0.9) or co-cultures (r=0.43, p>0.6). We then performed additional experiments using astrocyte-conditioned media in a non-angiogenic model in order to further assess a potential relationship between tPA mRNA regulation and angiogenesis. Bovine brain capillary endothelial cells were grown to confluence and 65 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. incubated with media that had been conditioned by astrocytes for 24 (ACM-24) or 48 hours (ACM-48). Examination o f tPA transcripts twenty-four hours following incubation with ACM-48 revealed a significant reduction (65 ± 15 % o f controls, p<0.02) in endothelial expression o f tPA mRNA as compared to endothelial cells incubated with non-conditioned media (Fig. 5). Thus, astnx^e-dependent tPA r^ulation was unrelated to the angiogenesis. Moreover, these studies suggested that this regulation was mediated in part by a soluble factor secreted by astrocytes. To investigate the functional consequences o f the downregulation o f tPA mRNA, we examined tPA activity in culture media from our BBB model, ie, from mono- and co cultures o f CS. There was no significant difference in tPA activity between mono- and co-cultures after one day o f culture (2.7 ±0.1 lU/ml vs. 2.3 ± 0.6 lU/ml, p<0.4). However, after seven days o f culture astrocyte-endothelial co-cultures echibited a substantial reduction in tPA activity (14.6 ± 2.9 lU/ml) as compared to endothelial mono cultures (30.2 ± 7.7 lU/ml, p<0.01) (Fig. 6). Concurrently, endothelial PAI-1 mRNA expression was significantly enhanced in co-culture 183 ± 87 % (p<0.02) compared with mono-culture preparations (Fig. 7). These findings suggested that astrocytes produced reduced tPA activity by dow nr^vlation o f endothelial tPA mRNA and increasing expression o f the tPA inhibitor PAI-1. 66 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 3.1. Phase-contrast photomicrograph o f capillary-like structures. Original magnification: 25X. 67 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 3.2. Glial fibrillary acidic protein (GFAP) immunocytochemistry. Immunocytochemistry for GFAP, a marker for astrocytes, showed astrocytes associated with capiUary-like structures at 7 days. Original magnification: 125X. 68 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LD I 2 3 4 Figure 3.3. PCR analysis o f tPA and PAI-1 mRNA e7q)ressioa RT-PCR showed an^lification o f tPA (kme 1) and PAI-1 (lane 3) from endothelial cells, and the lack of an^lification o f either tPA (lane 2) or PAI-1 (lane 4) from astrocytes. LD= 1 kb marker. 69 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. □MONO-CULTURE BCO-CULTURE 140 c o 5 1 2 0 - I H s S . I & 100 - 1 Day 7 Days Figure 3.4. Densitrometric analyses o f PCR products. RT-PCR analysis showed significant decrease in tPA mRNA levels in 7 day astrocyte-endothelial co-cultures compared with endothelial mono-cuhures (* p < 0.03). Data are presented as mean ± SD. Data are from 1 «(périment (performed in quadruplicate) and representative o f 3 independent experiments. No significant differences were observed in tPA in I day cultures. 70 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. g r 1 1 0 S 100 0 I 9 0 - 80 -- I I 60 -- 50 Control A C M -2 4 ACM-48 Figure 3.5. tPA mRNA expression in endothelial monolayer preparations. RT-PCR analysis o f tPA transcripts after 24 hours incubation with media conditioned by astrocytes for 24 hours (ACM-24) and 48 hours (ACM-48) revealed reduction in endothelial expression o f tPA mRNA as compared to endothelial cells incubated with non conditioned media (* p < 0.02). Data represent 3 pooled experiments and are presented as mean ± SD. 71 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. □MONOCULTURE ■CO-CULTURE 40 > 35 -- 30 -- 25- 20 - - S 1 5 - S 1 0 - 5 -- 0 7 Days Figure 3.6. tPA activity. Assay o f tPA activity in culture media from mono- and co culture preparations showed a significant reduction in tPA activity in 7 day astrocyte- endothelial co-cultures compared with endothelial mono-cultures (* p< 0.01). Data are presented as mean ± SD. Data are from 1 experiment (performed in quadruplicate) and representative o f 3 independent experiments. 72 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. o MONO-CULTURE ■GO-CULTURE § 1 I O iê s s 1 400 350 300 250 200 150 100 50 0 7 Days Figure 3.7. PAI-1 mRNA expression. RT-PCR analysis o f PAI-1 transcripts showed significant increase in 7 day astrocyte-endothelial co-cultures compared with endothelial mono-cultures (* p < 0.02). Data are presented as mean ± SD. Data are from 1 experiment (performed in quadruplicate) and representative o f 3 independent experiments. 73 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Discussion Studies o f fibrinolysis and the BBB are confounded by potential contribution o f both endothelial and glial plasminogen activators and plasminogen activator inhibitors (Kalderon et al., 1990). Because endothelial-derived fibrinolytic factors can directly contribute to circulating tPA and PAI-1 in plasma and participate in intravascular thrombolysis, we have chosen to concentrate on brain endothelial tPA and PAI-1 expression. Our BBB model involves direct co-culture o f endothelial cells with astioi^es, but the differing species (bovine and mouse, respectively) allow for the selective analysis o f endothelial tPA and PAI-1 transcripts. Moreover, our use o f later passage and dedifferentiated brain endothelial cells allows for determination o f the relationship between induction o f BBB features and regulation of tPA and PAI-1 expression. We have shown that astrocytes negatively regulate brain capillary endothelial cell 0 q)ression o f tPA mRNA and tPA function in vitro. Direct astrocyte co-culture with endothelial cell capillary-like structures showed substantial downr%ulation o f tPA mRNA and activity after seven days of co-culture, a time at which BBB characteristics are demonstrable in this model. Therefore, these results suggest that astrocytes at the BBB regulate tPA expression. Moreover, the increased endothelial PAI-1 mRNA after seven days co-culture suggests that the decreased tPA function is also due to astrocyte- dependent enhanced expression o f endothelial PAI-1. Vascular endothelial expression o f tPA in vitro is dependent on both the location o f the vessel in question as well as its size (Levin et al., 1997). In primate brain, 74 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. capillaries, representing 40 % o f all vessels less than 100 |im in diameter, account for only 3% o f those vessels that are immimoreactive for tPA (Levin and del Zoppo, 1994). Moreover, bronchial capillaries are inununoreactive for tPA in the mouse, while pulmonary capillaries do not express tPA (Levin et al., 1997). These obs^vations suggest organ-specific r% ulation o f tPA expression. Our findings further suggest that the BBB, and specifically, BBB astrocytes, provide brain-specific regulation o f tPA expression. In vitro studies have shown that brain capillary endothelial cells have restricted expression o f tPA, in comparison with systemic microvascular endothelial cells. Shatos et al. (1995) showed that human systemic, but not brain, microvascular endothelial cells secreted tPA in response to a-thrombin. Grau et al. (1997) demonstrated that systemic microvascular endothelial cells produced three-fold higher levels o f tPA compared to the brain microvascular endothelium. Our findings, consistem with these earlier studies, demonstrate distinct differences in tPA «pression between capillary endothelial cells with and without the BBB phenotype. Prior studies have shown a substantial relationship between plasminogen activators and angiogenesis in vitro (Yasunaga et al., 1989; van Hinsbeigh et al., 1991). Endothelial cells can produce two molecular forms o f plasminogen activators, urokinase plasminogen activator (uPA) as well as tPA. Endothelial uPA is primarily involved with extracellular matrix degradation during cellular proliferation and angiogenesis (van Hinsbeigh et al., 1991; Laterra et al., 1994; Rao et al., 1996). Astrocyte-derived uPA is known to induce endothelial cell differentiation into capillaries (Laterra et al., 1994; Rao et al., 1996). However, prior work has shown that tPA has no substantial role in 75 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. angiogenesis o f bovine capillary endothelial cells after seven days in culture (Yasunaga et al., 1989). Our data are consistent w ith these latter findings, and indicate that astrocyte- dependent tPA regulation in vitro occurs independent o f the angiogenic process. Murine astrocytes express plasminogen activator activity (Kalderon et al., 1990). This activity is developmentally regulated, peaking at 8-14 days postnatal (P8-P14) and then declining two to threefold to a low level after P16. We used primary culture astrocytes after P I6, which minimizes astrocyte plasminogen activity. Despite potential astrocytic contributions o f tPA, our seven day astrocyte-endothelial co-cultures exhibited significantly decreased tPA activity compared to endothelial mono-cultures. It is well established that astrocytes can modulate a variety o f endothelial functions (DeBauh and Cancilla, 1980; Dehouck et al., 1990; Dehouck et al., 1994; Murphy et al., 1994; O'Donnell et al., 1995; Raub et al , 1992; Stanimirovic et al., 1995; Stewart and Wiley, 1981; Sun et al., 1995; Takakura et al., 1991; Tao-Cheng et al., 1987; Tran et al., 1996). A number o f these studies demonstrated that the effects can be induced by astrocyte-conditioned media; thus, implicating a role for astrocyte-derived soluble factors in the regulation o f endothelial function (Murphy et al., 1994; O'Donnell et al., 1995; Raub et al., 1992; Stanimirovic et al., 1995; Takakura et al., 1991). The astrocyte regulation o f endothelial fibrinolysis may require direct contact between astrocytes and endothelial cells, or may be mediated via difiusable substances. Our data show significant regulation o f tPA expression by endothelial cells incubated with media conditioned by astrocytes for two days. These findings suggest that an astrocyte-derived soluble factor is at least partially responsible for the regulation o f endothelial tPA e;q}ression by astrocytes. Astrocytes can e}q>ress cytokines known to transcriptionally 76 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. regulate tPA and PAI-1, including interleukin-13 and tumor necrosis factor-a (Schleef et al., 1988; Lieberman et al., 1989; Sharif et al., 1993). Moreover, astrocytes express transforming growth 6ctor-P (Constam et al., 1992), also known to r% ulate fibrinolysis (Saksela et al., 1987; Fujii et al, 1991). These factors may play a role in astn x ^ e- dependent regulation o f endothelial tPA and PAI-1 transcription and, ultimately, fibrinolytic activity. The importance o f brain capillary expression o f tPA in cerebrovascular disease has been emphasized in recent work analyzing tPA expression in two important stroke models. In a diabetic stroke model, reduced microvascular expression o f tPA mRNA and protein was associated with substantially increased infarct size following middle cerebral artery occlusion (Kittaka et al., 1996). Moreover, in a rat model utilizing nicotine infusion to mimic smoking, reduced microvascular tPA protein was also associated with increased infarct size after occlusion o f the middle cerebral artery (Wang et al., 1996). These observations suggest a link between brain capillary tPA expression and outcome following large artery occlusion in the brain. Our findings add to the emerging picture o f the astrocyte as a key mediator for hemostasis within the central nervous system. Tissue 6ctor, the primary generator o f the coagulation cascade, is abundantly expressed by astrocytes (Eddleston et al., 1993). Astrocytes can also express tPA (Kalderon et al., 1990; Tranque et al., 1992), PAI-1 (Tranque et al., 1992), thrombomodulin (Pindon et al., 1997) and the tPA inhibitor protease nexin-1 (Scott et al., 1983; Gloor et al., 1986; Choi et al., 1990; Festoff et al., 1996). Moreover, astrocytes regulate endothelial thrombomodulin expression (Tran et al., 1996), as well as endothelial tPA and PAI-1 expression (as described herein). Thus, 77 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. astrocytes critically participate in hemostasis by both expression and regulation o f multiple hemostasis Actors. The abundance o f astrocyte tissue factor, as well as astrocyte n ^ ativ e r% ulation o f endothelial thrombomodulin and tPA, suggest a prothrombotic milieu within the brain miCTOvasculature. We speculate that such a prothrombotic milieu provides protection from neonatal intracranial hemorrhage. This potential evolutionary advantage could become counter-productive with aging and development o f stroke risk factors, predisposing to brain infarction. In conclusion, astrocytes am atively regulate brain capillary endothelial expression o f tPA mRNA and function. This regulation is associated with development o f features o f the BBB in vitro. These findings support an important role for the BBB in the regulation o f hemostasis by the brain. Modulation o f astrocyte-dependent tPA r%ulation may lead to new strategies for the treatment o f ischemic stroke. 78 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. References 1. Carson MP, Haudoischild CC (1986) M icrovascular endothelium and pericytes; high yield, low passage cultures. In Vitro C ell D ev B io l 22,344-54 2. Choi BH, Suzuki M, Kim T, Wagner SL, Cunningham DD (1990) Protease nexin-1. Localization in the human brain suggests a protective role against extravassated serine proteases. Am JP a th o l 137, 741-7 3. Collen D, Lijnen HR (1995) Molecular basis of fibrinolysis, as relevant for thrombolytic therapy. Throm bH aem ostlA , 167-71 4. Constam DB, Plilipp J, Malipiero UV, ten Dijke P, Schachner M, Fontana A (1992) Differential expression o f transforming growth factor-|31, -P2, -P3 by glioblastoma cells, astrocytes, and microglia. Jlm tnunol 148, 1404-10 5. DeBault LE, Cancilla PA (1980) Gamma-glutamyl transpeptidase in isolated brain endothelial cells: induction by glial cells in vitro. Science 207, 653-5 6. Dehouck B, Dehouck MP, Fruchart JC, Cecchelli R (1994) Upregulation o f the low density lipoprotein receptor at the blood-brain barrier: intercommunications between brain capUlary cells and astrocytes. J C ell B io l 126, 465-73 7. Dehouck M P, Meresse S, Delorme P, Fruchart JC, Cecchelli R (1990) An easier, reproducible, and mass-production method to study the blood-brain barrier in vitro. J Neurochem 54, 1798-1801 8. del Zoppo GJ, Copeland BR, Waltz TA, Zyrofif J, Plow EF, Haricer LA (1986) The beneficial effect o f intracarotid urokinase on acute stroke in a baboon model. Stroke 17, 638-43 9. Eddleston M, de la Torre JC, Oldstone MBA, Loskutoflf DJ, Edington TS, Mackman N (1993) Astrocytes are the primary source o f tissue factor in the murine central nervous system: a role for astrocytes in cerebral hemostasis. J C lin Invest 92, 349-58 10. Festoff BW, Smirnova IV, Ma J, Citron BA (1996) Thrombin, its receptor and protease nexin I, its potent serpine in the nervous system. Sem Thromb H em ost 22, 267-71 11. Fujii S, Hopkins W, Sobel BE (1991) Mechanisms contributing to increased synthesis of plasminogen activator inhibitor-1 in endothelial cells by constituents o f platelets and their implications fi)r thrombolysis. C irculation 83, 645-51 79 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 12. Gloor S, Odink K, Guenther J, Nick H, Monard D (1986) A glial derived neurite promoting factor with protease inhibitory activity belongs to the protease nexins. C ell 47,687-93 13. Goldstein GW (1988) Endothelial cell-astrocyte interactions. Arm N Y A cad Sci 529, 31-9 14. Grau GE, de Moerloose P, Bulla O, Lou J, Lei Z, Reber G, \fill N, Ricou B, Morel DR, Suter PM (1997) Haemostatic properties o f human pulmonary and cerebral microvascular endothelial cells. Thromb H aem ost 77, 585-90 15. Hallman R, Mayer ON, Berg EL, Broermann R, Butcher EC (1995) Novel mouse endothelial cell surface marker is suppressed during differentiation o f the blood-brain barrier. D ev D ynam ics 202,325-32 16. Hayashi Y, Nomura M, Yamagishi S, Harada S, Yamashita J, Yamamoto H (1997) Induction o f various blood-brain barrier properties in non-neural endothelial cells by close opposition to co-cultured astrocytes. G lia 19,13-26 17. Kalderon N, Ahonen K, Federofif S (1990) Developmental transition in plasticity properties o f differentiating astrocytes; age-related biochemical profile o f plasminogen activators in astroglial cultures. G lia 3, 413-26 18. Kittaka M, Wang L, Sun N, Schreiber SS, Seeds NW, Fisher M, Zlokovic BV (1996) Brain capillary tissue plasminogen activator in a diabetes stroke model. Stroke 27, 712-9 19. KoUros PR, Konkle BA, Ambarian AP, Henrikson P (1994) Plasminogen activator inhibitor-1 expression by brain microvessel endothelial cells is inhibited by elevated glucose. J Neurochem 63, 903-9 20. Laterra J, Indurti RR, Goldstein GW (1994) Regulation o f in vitro glia-induced microvessel morphogenesis by urokinase. J C ell P l^ s 158, 317-24 21. Levin EG, del Zoppo GJ (1994) Localization o f tissue plasminogen activator in endothelium o f a limited number o f vessels. Am JP a th 144, 855-61 22. Levin EG, Santell L, Osborn KG (1997) The expression o f endothelial tissue plasminogen activator in vitro: a fimction defined by vessel size and anatomic location. J C ell Sci 110, 139-48 23. Lieberman AP, Pitha PM, Shin SH, Shin ML (1989) Production o f tumor necrosis factor and other cytokines by astrocytes stimulated with lipopolysaccharide or a neurotrophic virus. Proc N atl A cad S ci USA 86, 6348-52 80 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 24. Maxwell K, Berliner JA, Cancilla PA (1987) Induction of y-glutamyl transpeptidase in cultured cerd)ral endothelial cells by a product released by astrocytes. Brain R es 410, 309-14 25. McCarthy KD, DeVellis J (1980) Preparation of separate astroglial and oligodendroglial cell cultures from rat cerebral tissue. J C ell B io l 85, 890-902 26. Mimuro J, Sawdey M, Hattori M, Luskutofif DJ (1989) cDNA for bovine type 1 plasminogen activator inhibitor (PAI-1). Nucleic A cids Res 17,8872 27. W nakaw a T, Bready J, Berliner J, Fisher M, Cancilla PA (1991) In vitro interactions o f astrocytes and pericytes with capillary-like structures o f brain microvessel endothelium. Lab Invest 65, 32-40 28. Murphy S, Rich G, Orgren KI, Moore SA, Faraci FM (1994) Astrocyte-derived lipoxygenase product evokes endothelium-dependent relaxation o f the basilar artery. JN eu ro sci R es 38, 314-8 29. The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group (1995) Tissue plasminogen activator for acute ischemic stroke. N E ngl J M ed 333, 1581-7 30. O'Donnell ME, Martinez A, Sun D (1995) Cerebral microvascular endothelial cell Na-K-Cl cotransport; regulation by astrocyte-conditioned media. Am J P hysiol 268, C747-54 ' 31. Pardridge WM (1984) Transport o f nutrients and hormones through the blood-brain barrier. F ed Proc 43, 201-4 32. Pardridge WM (1991) Advances in cell biology o f blood-brain barrier transport. Sem C ell B io l 2,419-26 33. Pindon A, Hantai D, Jandrot-Perrus M, Festoff BW (1997) Novel expression and localization o f active thrombomodulin on the surfrice of mouse brain astrocytes. G lia 19, 259-68 34. Prendergast GC, Diamond LE, Dahl D, Cole MD (1990) The c-myc regulated gene nwl encodes plasminogen activator inhibitor 1. M ol C ell B iol 10, 1265-9 35. Rao JS, Sawaya R, Gokaslan ZL, Yung WK, Goldstein GW, Laterra J (1996) M odulation o f serine proteinases and metalloproteinases during morphogenic glial- endothelial interactions. J Neurochem 66, 1557-64 36. Raub TJ, Kuentzel SL, Sawada GA (1992) Permeability of bovine brain microvessels endothelial cells in vitro: barrier tightening by a factor released from astroglioma cells. Ejqj C ell R es 199, 330-40 81 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 37. Ravn P, Bergtund L, Petersen TE, Larsen LB (1995) Cloning and characterization o f the bovine plasminogen activator uPA and tPA In t D airy J 5,605-17 38. Rickies RJ, Darrow AL, Striddand S (1988) Molecular cloning o f complementary DNA to mouse tissue plasminogen activator mRNA and its expression during F9 teratocardnom a cell differentiatioa J B io lC h e m 2 6 i, 1563-9 39. Risau W, Wolburg H (1990) EKfferentiation o f the blood-brain barrier. TINS 13, 174-8 40. Saksela O, Moscatelli D, Rifldn DB (1990) The opposing effects o f basic fibroblast growth Actor and transforming growth factor beta on the regulation o f plasminogen activator activity in capillary endothelial cells. J C ell B io l 105, 957-63 41. Schleef RR, Bevilacqua MP, Sawdey M, Gimbrone Jr. MA, Loskutoff DJ (1988) Cytokine activation o f vascular endothelium: Effects on tissue-plasminogen activator and type 1 plasminogen activator inhibitor. J B io l Ghent 263, 5797-803 42. Scott RW, Eaton DL, Duran N, Baker JB (1983) Regulation o f extracellular plasminogen activator by human fibroblasts. The role o f protease nexin. J B iol Ghent 258, 4397-403 43. Sharif SF, Hariri RJ, Chang V A Barie PS, Wang RS, Gfaajar JB (1993) Human astrocyte production o f tumor necrosis factor-alpha, interleukin-1 beta, and interlaikin-6 following exposure to lipopolysaccharide endotoxin. Newrol Res 15, 109-12 44. Shatos M A Orfeo T, Doherty JM, Penar PL, Collen D, Mann KG (1995) (a- Thrombin stimulates urokinase production and DNA synthesis in cultured human cerd)ral microvascular endothelial cells. A rterioscler Thromb Vase B io l 15, 903-11 45. Stanimirovic DB, Ball R, Durkin JP (1995) Evidence for the role o f protein kinase C in astrocyte-induced proliferation o f rat cerd)romicrovascular endothelial cells. N eurosci Letters 197, 219-22 46. Stewart PA Wiley MJ (1981) Developing nervous tissues induces formation o f blood-brain barrier characteristics in invading endothelial cells: a study using quail- chick transplantation chimeras. D ev B io l 84, 183-92 47. Sun D, Lytle C, O'Donnell ME (1995) Astroglial cell-induced expression o f Na-K-Cl cotransporter in brain microvascular endothelial cells. Am J P hysiol 269, C l 506-12 82 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 48. Takakura Y, Trammel AM, Kuentzel SL, Raub TJ, Davies A, Baldwin SA, Borchardt RT (1991) Hexose uptake in primary cultures o f bovine brain microvessel endothelial cells. Effects o f conditioned media from astroglial and glioma cells. Biochim B iophysA cta 1070, 11-9 49. Tao-Cheng JH, Nagy Z, Brightman MW (1987) Tight junctions to brain endothelium in vitro are enhanced by astroglia. JN eurosci 7, 3293-9 50. Tran ND, Wong VLY, Bready J, Schreiber SS, Fisher M (1996) R ^ulation o f brain capillary endothelial thrombomodulin mRNA expression. Strolæ 27, 2304-11 51. Tranque P, Robbins R, Naftolin F, Andrade-Gordon P (1992) Regulation of plasminogen activator and type 1 plasminogen activator inhibitor by cyclic AMP and phorbol ester in rat astrocytes. G lia 6,163-71 52. van Hinsbergh VWM, Kooistra T, Emeis JJ, Koolwijk P (1991) Regulation of plasminogen activator production by endothelial cells: role in fibrinolysis and local proteolysis. In t J R adiai B io l 60^261-12 53. Vassalli JD, Belin D (1987) Amiloride selectively inhibits the urokinase-type plasminogen activator. FEBS L etter 2\A, 187-91 54. Wang L, Kittaka M, Sun N, Schreiber SS, Zlokovic BV (1996) Chronic nicotine treatment enhances focal ischemic brain injury and depletes free pool o f brain microvascular tissue plasminogen activator in rats. J Cereb B lood Flaw M etab 17, 136-46 55. Wiman B (1995) Plasminogen activator inhibitor 1 (PAI-1) in plasma: its role in thrombotic disease. Thromb H aem ost 74, 71-6 56. Yasunaga C, Nakashima Y, Sueishi K (1989) A role o f fibrinolytic activity in angiogenesis. Lab Invest 61, 698-704 57. Zivin AJ, Fisher M, deGirolami U, Hemenway CC, Stashak JA (1985) Tissue plasminogen activator reduces nojrological damage after c er^ ral embolism. Science 230, 1289-93 58. Zlokovic BV, Wang L, Sun N, Haffke S, Verrall S, Seeds NW, Fisher MJ, Schreiber SS (1995) Expression o f tissue plasminogen activator in cerebral capillaries: possible fibrinolytic function of the blood-brain barrier. Neurosurg 37, 955-61 83 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 4 Measurement of Thrombomodulin mRNA Expression in Brain Capillaries by Polymerase Chain Reaction Abstract Thrombomodulin (TM), an endothelial int%ral membrane protein, is a potent activator o f the protein C anticoagulant pathway. TM protein expression is limited and regionally distributed in the brain. Recent investigations have demonstrated low TM mRNA expression by brain endothelium, corresponding to its distribution at the protein level. To facilitate the study o f TM expression at the transcriptional level, we measured TM mRNA by quantitative-competitive PCR (QC-PCR) and by standard denshometric analysis o f reverse transcriptase-PCR products (RT-PCR) in different regions o f bovine brain. QC-PCR demonstrated differential TM mRNA expression in the pons (100 ± 9 %), cerebellum (359 ± 103 %) and cortex (441 ± 24 %). We compared these results with those o f RT-PCR and found similar differences in relative TM mRNA expression in the pons (100 ± 44 %), cerebellum (343 ± 8 %) and cortex (404 ± 62 %). Data derived by QC-PCR and RT-PCR were highly correlated (r= 0.99, p<0.03). These findings indicate that either QC-PCR or RT-PCR can be used to accurately quantify TM mRNA. 84 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Introduction Endothelial cells play an active role in hemostasis regulation and participate in the protan C anticoagulant pathway through the expression o f the glycoprotein thrombomodulin (TM). TM forms a high affinity complex with thrombin and together this complex activates protein C as much as 20,000 times 6 ster than thrombin alone (Esmon, 1987; Esmon, 1989; Dittman & Majerus, 1990). Activated protein C is a circulating antico% ulant that inhibits factors Va and V nia (Walker et al., 1979; Fulcher et al., 1984). In vivo studies demonstrate that TM plays a protective role against a variety o f thrombotic events (Kumada et al., 1987; Taylor et al., 1987; Gomi et al., 1990). Resistance to the effects of activated protein C is closely linked to venous thrombosis in humans (Svensson & Dahlback, 1994). TM abundance in brain endothelium is limited compared to microvessel endothelia from other vascular beds (Ishii et al., 1986; Wong et al., 1991; Bofra et al., 1991; Isaka et al., 1994; Maruno et al., 1994). An increased prevalence of cerebral infarction occurs in subcortical areas with reduced TM levels (Wong et al., 1991). TM mRNA also has restricted expression in brain endothelium, probably due to astrocytic influences (Tran et al., 1996; Wang et al., 1997). The regional distribution o f brain TM mRNA is consistent with its distribution at a protein level (Wong et al., 1991; Wang et al., 1997). These findings indicate that transcriptional processes play a critical role in brain TM expression. Thus, it is important to understand the mechanisms regulating TM mRNA expression within the brain. 85 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In recent years, the reverse transcriptase-polymerase chain reaction (RT-PCR) has become the method o f choice for detection o f low abundance mRNA transcripts (Tran et al., 1996; Zlokovic et al., 1995; Kittaka et al., 1996). Because PCR amplification is an exponential process, small differences in amplification efficiency will often result in substantial differences in PCR products. We previously demonstrated a quantitative- competitive PCR (QC-PCR) method for analysis o f mRNA levels (Tran et al., 1996). This relatively labor intensive method overcomes the variations in efficiency o f PCR amplification by using an internal competitor template that coamplifies with the gene o f interest. Moreover, using the same primers allows for accurate and reproducible measures o f low abundance transcripts. The goal o f the present study was to measure mRNA levels in different regions o f bovine brain by using both QC-PCR and RT-PCR in order to assess whether these techniques yield similar results. 86 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Materials & Methods Capillary {solation Bovine brain capillaries were isolated according to the method o f Zlokovic et al. (1993). Briefly, bovine brains were rinsed in ice cold buffer B containing 103 mmol/L NaCl, 4.7 mmol/L KCl, 2.5 mmol/L CaCh, 1.2 mmol/L KH2PO4, 1 2 mmol/L MgS0 4 , 15 mmol/L Hepes, pH 7.4. The cerd>ral cortical mantles, cerd)ellum, and pons were rapidly freed o f meninges (i.e., pial vessels) under sterile conditions, and the arteries o f the circle o f Willis, veins and choroid plexi discarded. Cerdiral and cerd)ellar cortices and pons were used to isolate the capillaries. Homogenization o f brain tissue in buffer A containing 103 mmol/L NaCl, 4.7 mmol/L KCl, 2.5 mmol/L CaClz, 1.2 mmol/L KH2PO4, 1.2 mmol/L MgS0 4 , 15 mmol/L Hepes, pH 7.4, 25 mmol/L HCO3, 10 mmol/L glucose, 1 mmol/L sodium pyruvate, and 0.156 mmol/L dextran, Mr=64,000, was followed by dextran density centrifugation at 5,800X g at 4°C. The pellet was resuspended in buffer A and passed over an 85 pm nylon mesh. Arterioles and venules remained on top o f the mesh, while the capillaries, red cells, nuclei, and other debris were collected in the filtrate passing through the mesh. This filtrate was then passed over a 3x4 cm glass bead column (0.45 mm glass beads) with a 44 pm nylon mesh at the bottom, and the column washed with buffer B. The brain capillaries adhered to the glass beads while the other contaminants passed unimpeded. Capillaries were recovered by repeated gentle agitation o f the glass beads in buffer A, and the supernatant with the capillaries was decanted and spun at 500X g fisr 5 min to obtain the final pellet. The purity o f the cerebral capillaries was checked by light and phase microscopy. The cerebral capillaries were free o f 87 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. adjoining brain tissue, and preparations consisted primarily o f capillaries but also contained minor amounts (S-10 %) of arterioles. The presence o f no more than minimal detectable activity o f a specific cerebral vascular marker, y-glutamyl transpeptidase, was used to confirm the absence o f contamination o f capillary-depleted brain parenchyma with microvessels. Polymerase Chain Reaction Total RNA was isolated with the Glassmax DNA Spin Cartridge Isolation System (Gibco BRL, Gaithersburg, MD). Total RNA from each preparation was resuspended in 40 nl o f diethyl pyrocarbonate (DEPC)-treated water. cDNA was synthesized from equal volumes o f total RNA in a total volume o f 20 |il. RNA was incubated in 5 |il o f DEPC- treated water at 65^C for 3 minutes and quickly placed on ice. The RNA was then added to the transcription solution: 1.5 iimol/L oligo dT primers, 50 mmol/L Tris-HCl, pH 8.3, 75 mmol/L KCl, 3 mmol/L MgClj, 0.5 mmol/L dNTP, 1 unit/|ii RNase inhibitor, and 13.3 units/p.1 Avian Myeloblastosis Virus reverse transcriptase. The reaction was carried out at 42°C for 1 hour and terminated at 52°C for 40 minutes. The cDNA was stored at - 20°C until use. The RT-PCR reaction mixture contained 0.2-1.0 ng cDNA, 10 mmol/L Tris-HCl, pH 8.3, 50 mmol/L KCl, 0.1 mmol/L dNTP, 1.0 mmol/L MgClz, 1.0 unit Taq Polymerase and 0.5 pmol/L forward and reverse primers. The bovine thrombomodulin primers (National Biosciences Inc., Plymouth, MN) extending from bases 228-256 (forward primer, 5'CTCGGCAACTACACGTGCATCTGCGAG3’ ) and 907-935 (reverse primer, S GCCACCACCAGAGACAGGCTTGCAATGG3') were chosen from coding regions o f 88 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the mRNA (Jackman et al., 1986). p-actin primers (Stratagene, La Jolla, CA; forward primer, 5TGACGGGGTCACCCACACTGTGCCCATCTA3'; reverse primer, 5CTAGAAGCATTTGCGGTGGACGATGGAGGG3') were used to amplify P-actin mRNA as a control gene. Amplification was carried out in a DNA Thermal Cycler (Peiidn-Elmer Corp., Norwalk, CT): initial dénaturation at 94® C, each cycle consisting o f dénaturation at 94*’ C for 30 seconds, aimealing at 54°C for 30 seconds, and extension at 72°C for 1 minute. All RT-PCR amplifications were performed within the linear range o f the amplification curve (Fig 4.1). PCR products were visualized by electrophoresis on a 2 % agarose gel and stained with ethidium bromide. mRNA levels were determined by standard densitometric analysis o f PCR products and subsequent calculations o f relative abundance. The latter was determined by arbitrarily characterizing the values firom the pons as 100 %. The QC-PCR assay was performed according to the methods o f Tran et al. (1996). Briefly, QC-PCR tubes contained all the amplification reagents (described above), a constant amount o f target thrombomodulin cDNA firom each preparation, and serial dilutions of known concentrations o f a competitor thrombomodulin cDNA template. The reaction mixture was co-amplified, as described above. The competitor cDNA template was prepared by site-directed mutagenesis (Tran et al., 1996; Higuchi et al., 1988). A single base change o f A to G at base pair 346 created a unique Sail restriction she. Following co-amplification, the PCR products were digested with Sail: 10 pi o f the PCR product, 1 unit Sail restriction enzyme, and 2 jii enzyme buffer incubated at 37^C for 2 hours. The digested competitor (591 bp) and target (707 bp) cDNAs were separated by electrophoresis on a 2 % agarose gel and visualized by ethidium bromide staining and S9 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UV trans-illuminadon. Negatives prepared with a Polaroid camera (Polaroid Corp., Cambridge, MA) and scanned by optical densitometry (Hoefer Instruments, San Francisco, CA). Optical density (CD) readings o f the target cDNA were multiplied by S91/707 to correct for differences in molecular weight. The ratio o f amplified target versus competitor cDNA CD was plotted as a function o f competitor template concentration. The initial concentration o f target cDNA was derived from the point at which the ratio o f target and competitor cDNA CD equaled 1 (Fig 4.2). All data were normalized to P-actin and expressed as mean ± SD. Experiments were performed in triplicate. Statistical comparisons between groups were performed using Student's t-tests and Pearson correlation coefBcients. Differences were considered significant for p<0.05. 90 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Results Brain capillaries were isolated from three different r% ions: pons, cerebellum and cortex. QC-PCR analysis demonstrated differential expression o f TM mRNA in these regions (Fig 4.3). The lowest concentration o f TM transcripts was found in the pons (5.5 ± 0.5 pg/ml) compared with the cerebellum (19.8 ± 5.7 pg/ml) and cortex (24.3 ± 1.3 pg/ml). Thus, cerebellar and cortical TM mRNA abundance represent 359 ± 103 % (p<0.02) and 441 ± 24 % (p<0.001) o f the pons, respectively. The RT-PCR analysis performed on same samples was consistent with the QC-PCR findings; TM mRNA expression in the cerd)ellum and cortex relative to pons was 343 ± 8 % (p<0.001) and 404 ± 62 % (p<0.003), respectively (Fig 4.4). There were no significant differences between ()C-PCR and RT-PCR derived data from the three brain regions. Moreover, there was a highly significant correlation between QC-PCR and RT-PCR derived TM mRNA concentrations (r= 0.99, p<0.03). 91 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 60000 S 50000 40000 lU 30000 h- 20000 10000 45 40 30 Cycle Number Figure 4.1. Kinetics o f PCR amplification. Plot o f PCR product yield as a function of number o f amplification cycles. 92 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 2 3 4 5 6 ^ Tirgrt ^ Conpditor -TAR3ET-•-OCM’EnTOR c 16000 5 8000 O 4000 10 10 CoiiyefitorTenipiate(pglirrt) Figure 4.2. Quantitative-competitive PCR analysis o f TM mRNA. A constant amount o f unknown target cDNA was added to PCR tubes containing 10-fold serial dilutions o f a conçetitor cDNA template (lanes 1-6: 10^, 10^, 10% 10 , 10'* , 10'^ pg o f initial competitor template, respectively) (A). With decreasing amounts o f competitor cDNA added, there was a decrease in conqietitor PCR products (591 bp) and a concommhant increase in unknown taiget PCR products (707 bp). Plot o f the ratio o f unknown target vs. competitor PCR products (B). 93 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. !i K E 3 0 25 - 2 0 - 15-- 1 0 - - J -500 -400 -300 -200 2 . I I O -100 Pons Cerebellum Cortex Figure 4.3. Thrombomodulin mRNA concentration. Quantitative-competitive PCR was used to quantify TM mRNA isolated from the pons, cerdiellum, and cortec. Both cerebellar and cortical capillaries showed significant increases (p<0.02 and p<0.001, respectively) compared to pontine capillaries. 94 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. - TM C 0 1 iS s i 1 500 - 40 0 - 300 - 200 - Pons C e r e b e O u m Cortex Figure 4.4. RT-PCR analysis o f tfarombcmodulin mRNA expression. (A) Agarose gel showed amplification o f brain capillary thrombomodulin isolated from the pons (lane 1), cerebellum (lane 2) and cortex (lane 3). (B) Optical densitometric analysis showed increases in TM mRNA concentration in the cerebellum and cortex compared to the pons (p<0.001 and p<0.003, respectively). 95 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Discussion RT-PCR allows for the sensitive detection o f low abundance mRNAs. However, a number o f A ctors can affect the efficiency o f PCR amplification, including the annealing temperature, cycle length, cycle number, and decreasing enzymatic activity and concentrations o f reactants with time. RT-PCR products accumulate exponentially up to a concentration o f approximately 1 0 * ^ M, after which point the PCR product concentration plateaus and accumulates linearly due to depletion o f reaction components and production o f inhibitors (Gilliland et al., 1990). During the exponential phase o f amplification, a linear relationship is maintained between the starting mRNA concentration and the RT-PCR product. Thus, under optimal conditions the relative amounts o f RT-PCR products can accurately reflect the relative concentrations o f initial mRNA transcripts. Another approach to determining mRNA concentration has been the use o f quantitative competitive PCR using an internal standard o f known concentration that is co-amplified w ith the target cDNA (Tran et al., 1996). By titrating the unknown target cDNA against serial dilutions o f known concentrations o f the competitor cDNA, the target cDNA concentration can be reliably determined. A steady relationship is maintained between the target and competitor throughout the amplification that is independent o f the kinetics o f amplification. This QC-PCR assay is both accurate and highly reproducible. In the present study we used QC-PCR to examine brain capillary TM mRNA expression and compared these findings with those obtained from the standard RT-PCR 96 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. analysis. QC-PCR analysis demonstrated higher levels o f TM mRNA expression in cerebellar and cortical capillaries compared to pontine capillaries. These data correlated strongly with RT-PCR findings o f increased TM mRNA expression in the cerebellum and cortex. The similar results derived from RT-PCR and QC-PCR analyses supports the use o f RT-PCR in quantifying TM mRNA expression. This finding is in agreement with previous woric comparing data derived from RT-PCR and northern blot analyses (Boado & Pardridge, 1994). We furtho' conclude that either assay can be useful, depending upon the information required. The RT-PCR analysis has the major advantage o f simplicity when data regarding relative changes in gene expression are desired. On the other hand, QC-PCR analysis is able to provide absolute amount o f mRNA concentrations, allowing inter-experimental comparisons. The reliability o f RT-PCR is limited to the exponential phase of amplification, whereas QC-PCR can be used over a wider range o f cycles. However, QC-PCR requires the time-consuming preparation o f an internal standard. Moreover, this assay requires the use of more RNA and PCR regents. Thus, time, cost, availability o f RNA, and the types o f data desired are all critical Actors to evaluate in determining whether to employ the RT-PCR or QC-PCR 97 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. References 1. Boado RJ, Pardridge WM (1994) Measurement of blood-brain barrier GLUT-1 glucose transporter and actin mRNA by a quantitative polymerase chain reaction assay. JN eurochem 62, 2085-90 2. Bofifa MC, Jackman RW, Peyri N, BofTa JF, George B (1991) Thrombomodulin in the central nervous system. Nouv Rev F r H em atol 33, 423-9 3. Dittman WA, Majerus PW (1990) Structure and function o f thrombomodulin: a natural anticoagulant. B lood 75, 329-36 4. Esmon CT (1987) The regulation of natural anticoagulant pathways. Science 235, 1348-52 5. Esmon CT (1989) The roles o f protein C and thrombomodulin in the regulation o f blood coagulation. J B io l Chem 264, 4743-6 6. Fulcher CA, Gardiner JE, Griffin JH, Zimmerman TS (1984) Proteolytic inactivation o f human factor VTH procoagulant protein by activated human protein C and its analogy with factor V. B lood 63, 486-9 7. Gilliland G, Perrin S, Blanchard K, Bunn HF (1990) Analysis o f cytokine mRNA and DNA: Detection and quantitation by competitive polymerase chain reaction. Proc N atl A cad Set USA 87, 2725-9 8. Gomi K, Zushi M, Honda G, Kawahara S, Matsuzaki O, Kanabayshi T, Yamamato S, Maruyama I, Suzuki K (1990) Antithrombotic effect o f recombinant human thrombomodulin on thrombin-induced thromboembolism in mice. B lood 75, 1396-9 9. Jfiguchi R, Krummel B, Saiki K (1988) A general method o f in vitro preparation and mutagenesis o f DNA fragments: study o f protein and DNA interactions. Nucleic A cids R es 16, 7351-67 10. Isaka T, Yoshimine T, Motokiko M, Kurode R, Ishii H, Hayakawa T (1994) Altered e)q)ression of antithrombotic molecules in human glioma vessels. A cta N europathol 87, 81-5 11. Ishii H, Salem HH, Bell CE, Laporata EA, Majerus PW (1986) Thrombomodulin, an anticoagulant protein, is absent in the human brain. Blood 67, 362-5 12. Jackman RW, Beeler DL, DeWaters L, Rosenberg RD (1986) Characterization o f a thrombomodulin cDNA reveals structural similarity to the low density lipoprotein receptor. Proc N atl A cad Sci USA 83,8834-8 98 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 13. Khtaka Wang L, Sun N, Schreiber SS, Seeds NW, Fisher M, Zlokovic BV (1996) Brain capillary tissue plasminogen activator in a diabetes stroke model. Stroke 27, 712-9 14. Kumada J, Dittman WA, M ajous PW (1987) A role for thrombomodulin in the pathogenesis o f thrombin-induced thromboembolism in mice. B lo o d 71, 728-33 15. Maruno M, Yoshimine T, Isaka T, Kurode R, Ishii H, Mayakawa T (1994) Expression o f thrombomodulin in astrocytomas o f various malignancy and in gliotic and normal brain. J Neuro-O ncology 19, 155-60 16. Svensson PJ, Dahlback B (1994) Resistance to activated protein C as a basis for venous thrombosis. N E nglJM ed22Q ,SV J-12 17. Taylor FB, Chang A, Esmon CT, D'Angelo A, Vigano-D*Angelo S, Blick KE (1987) Protein C prevents coagulopathic and lethal effects o f Escherichia coli injection in the baboon. J C lin Invest 19, 9\Z-25 18. Tran ND, Wong VLY, Bready J, Schreiber SS, Fisher M (1996) Regulation o f brain capillary endothelial thrombomodulin mRNA expression. Stroke 27, 2304-11 19. W alker FI, Sexton PW, Esmon CT (1979) The inhibition o f blood coagulation by activated protein C through the selective inactivation of activated factor V. Biochim B iophysA cta 571, 333-42 20. Wang L, Tran ND, Kittaka M, Fisher M, Schreiber SS, Zlokovic BV (1997) Thrombomodulin expression in bovine brain capillaries: anticoagulant function o f the blood-brain barrier, r% ional differences and regulatory mechanisms. A rterioscler Thromb Vase B io l 17,3139-46 21. Wong VLY, Hofinan FM, Ishii H, Fisher M (1991) Regional distribution of thrombomodulin in the human brain. Brain R es 556, 1-5 22. Zlokovic BV, Mackic IB, Wang L, McComb JG, McDonough A (1993) Differential expression o f Na, K-ATPase a and P subunits isoforms at the blood-brain barrier and the choroid plexus. J B io l Chem 268, 8019-25 23. Zlokovic BV, Wang L, Sun N, HafOce S, Verrail S, Seeds NW, Fisher M X , Schreiber SS (1995) Expression o f tissue plasminogen activator in the cerebral capillaries: Possible fibrinolytic function o f the blood-brain barrier. N eurosurg 37, 955-61 99 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter S Transforming Growth Factor-p Mediates Astrocyte<Specific Regulation of Brain Capillary Endothelial Anticoagulant Factors Abstract Astrocytes are potent regulators o f brain capillary endothelial cell function. Recently, astrocytes were shown to regulate brain capillary endothelial expression o f the fibrinolytic enzyme tissue plasminogen activator (tPA) and the anticoagulant thrombomodulin (TM) in vitro. In order to study the mechanism o f this process, we examined the hypothesis that astrocyte regulation o f endothelial tPA and TM is mediated by transforming growth 6ctor-P (TGF-P). Astrocyte-endothelial co-cultures and media conditioned by astrocytes (ACM) exhibited significantly higher levels o f active TGF-P compared to a) brain capillary mono-cultures and b) endothelial cells grown in non conditioned media, respectively. Brain capillary endothelial cells incubated with ACM exhibited reduced tPA and TM mRNA and activity, compared to cells grown in non- 100 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. conditioned media. Treatment with exogenous TGF-P produced dose-dependent reductions in tPA and TM that w o o comparable in magnitude to that o f ACM. The effects o f ACM on both tPA and TM were blocked by TGF-P neutralizing antibody. These data indicate that TGF-P mediates astrocyte r%ulation o f brain capillary endothelial expression o f tP A and TM. 101 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Introduction Hemostasis is critically important in the pathogenesis o f stroke. Acute stroke therapies and secondary stroke prevention strat%ies involve manipulation o f coagulation pathways. Essential elements o f the coagulation cascade are endothelial-dependent, including secretion o f the fibrinolytic Actor tissue plasminogen activator (tPA) and expression of the antithrombotic integral membrane protein thrombomodulin (TM). tPA is a critical circulating fibrinolytic enzyme that proteolytically activates plasminogen to plasmin (CoUen and Lijnen, 1995). Intravenous tPA improves neurological outcome in clinical stroke (NINDS rt-PA Stroke Study Group, 1995). Reduced expression o f brain capillary tPA is associated with increased infarct size following transient middle cerebral artery occlusion in diabetic and nicotine stroke models (Kittaka et al., 1996; Wang et al., 1996). Brain capillary endothelial expression of tPA is limited (Levin and Del Zoppo, 1994; Shatos et al., 1995; Grau et al., 1997), and understanding mechanisms underlying this limited expression has potential th e r^ a itic value. TM, an important antithrombotic protein, functions as a cofactor for the activation o f circulating protein C (Esmon and Owen, 1981; Salem et al., 1984). Both TM and activated protein C (AFC) offer protection against a variety o f thrombotic events. Treatment with purified or recombinant TM protects against thromboembolism in animal models (Kumada et al., 1987; Gomi et al., 1990; Solis et al., 1991; Aoki et al., 1994; Ono et al., 1994), while TM neutralizing antibodies potentiate thrombin-induced thromboembolism (Kumada et al., 1987). Treatment with AFC also provides protection against thromboembolic events (Taylor et al., 1987; Gresele et al., 1998). Resistance to 102 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the efifects o f APC is closely linked to venous thrombosis in humans (Svensson & Dahlback, 1994), including cerebral venous thrombosis (Zuber et al., 1996). Low levels o f circulating APC are present in certain forms o f ischemic stroke (Macko et al., 1996). Moreover, brain specific protein C activation has been demonstrated in humans in vivo (Macko et ai., 1997); brain capillaries ex vivo also exhibit protein C activation (Wang et al., 1997). Brain TM expression is limited (Ishii et al., 1986; Wong et al., 1991; Boffa et al., 1991; Isaka et al., 1994; Maruno et al., 1994), and, like tPA, identification of mechanisms underlying this limited expression would be expected to have therapeutic value. Astrocytes are regulators for a wide variety o f brain capillary endothelial Amctions. Astrocytes are responsible for the induction o f the blood-brain barrier (BBB), contributing to tight jim ction formation (Tao-Cheng et al., 1987; Hayashi et al., 1997), increased electrical resistance (Dehouck et al., 1990), and expression o f highly selective transport systems delivering essential nutrients to the brain (eg., GLUT-1, y-GTP) (DeBault and Cancilla, 1980; Bauer et al., 1990; Meyer et al., 1991; Tontsch and Bauer, 1991; Hayashi et al., 1997). Astrocytes also modulate endothelial expression o f the low- density lipoprotein receptor (Dehouck et al., 1994) and Na-K-Cl cotransporter (Sun et al., 1995). We have recently reported that astrocytes have an important role in the regulation of endothelial expression o f critical hemostatic factors. We have demonstrated that astrocytes regulate endothelial expression o f both tPA and TM in vitro (Tran et al., 1996, 1998a). The mechanism o f this action has been uncertain. Astrocytes can secrete a number o f cytokines and growth Actors known to modulate endothelial function, 10 3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. including transforming growth foctor-P (TGF-P) (Constam et ai., 1992). TGF-P is known to r% ulate endothelial plasminogen activator activity (Saksela et al, 1990) and TM expression (Ohji et al., 199S). We therefore hypothesized that astrocytic regulation o f endothelial tPA and TM expression is mediated by TGF-p. 104 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Materials & Methods C d l culture We isolated bovine brain capillary endothelial cells by modification o f techniques o f Carson and Haudenschild (1986). Following transportation at 4°C from a local meat processing company, bovine brains were rinsed in a medium containing Dulbecco's Modified Eagle's Medium (DMEM), 1% bovine albumin serum, 100 U/ml penicillin, 100 mg/ml amphotericin B, and 2 mmol/L L-glutamine (Irvine Scientific, Santa Ana, CA). Under sterile conditions, the pial membrane was removed, cortical grey matter was aspirated with a pasteur pipette and centrifuged at lOOX g for 10 minutes. Following a rinse with medium, the tissue was homogenized and serially passed through nylon meshes o f 149, 74, and 20 pm. The tissue retained by the 74 and 20 pm meshes was digested at 37® C overnight by 1 mg/ml coUagenase (Sigma, St Louis, MO). Following the overnight digestion, the tissue was incubated with trypsin-EDTA (2.5 and 0.2 mg/ml, respectively) for 30 minutes. The tissue was resuspended in medium containing DMEM, 10% fetal calf serum (FCS) (Omega Scientific, Tarzana, CA), 100 U/ml penicillin, 100 mg/ml streptomycin, 2 mmol/L L-glutamine and plated on dishes coated with 1% gelatin (Sigma, St Louis, MO). Twenty-four hours following plating, adherent cells were washed and fed fresh medium. Pancreatin (2.5 mg/ml for 3-5 minutes at 25°C) was used to passage subconfluent cells 3-4 days following plating. This step is followed by a trypsin-EDTA treatment for 2-3 minutes, resulting in selective release o f endothelial cells. Pancreatin-trypsin-EDTA treatment was repeated for 4-5 passages to obtain a pure bovine brain capillary endothelial cell population. Endothelial cells were maintained on 105 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. uncoated culture dishes in DMEM supplemented with 2.5 % equine serum or 5.0 % FCS, 100 U/ml penicillin, 100 mg/ml streptomycin, and 2 mmol/L L-glutamine in a humidified 5% CO2 95% air incubator at 37°C. Endothelial cells, passaged twice weekly using trypsin-EDTA at a split ratio o f 1:2, were characterized by cobblestone-like morphology, uptake of acetylated low density lipoprotein labeled with l-l'-dioctacecyl-1-1-3-3-3-3- tetramethyl-indocarbocyanine perchlorate (Biomedical Technologies, Stoughton, MA), and immunoreactivity for von Willebrand factor (vWF), as previously described (Tran et al., 1996). Experiments were performed on endothelial cells between passages 10 and 25. Neonatal mouse astrocytes were isolated according to the methods o f McCarthy and deVellis (1980), and performed within institutional guidelines. Briefly, cerebral hemispheres were removed from Swiss-Webster pups 1-2 days postnatal (P1-P2), cleaned o f meninges and choroid plexus and serially sieved through meshes o f 230 and 140 nm. The filtrate was centrifuged at 200X g for 5 minutes at room temperatures and resuspended in DMEM supplemented with 10% FCS, 100 U/ml penicillin, 100 mg/ml streptomycin, and 2 mmol/L L-glutamine. Cells were plated at a density o f 30 cm^/brain and maintained on DMEM with 10% FCS in a humidified 5% CO2 95% air incubator at 37®C. At confluence, oligodendrocytes and microglia were removed by orbital shaking at 37®C. Astrocytes were characterized by >99% immunoreactivity for glial fibrillary acidic protein (GFAP). Astrocytes used for these experiments were taken from primary cultures for establishment o f astrocyte-endothelial co-cultures. Experiments were performed using primary culture astrocytes between 20-40 days postnatal (P20-P40). Mouse liver 1 0 6 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cells (CCL 9.1, ATCC, Rockville, MD) were maintained under conditions sim ilar to astnxgtes. Blood-Brain Barrier Model We prepared our BBB model as previously described (Minakawa et al., 1991; Tran et al., 1996, 1998a). To prepare capillary-like structures (CS) we first coated two 2x2 cm chamber Lab-Tek glass slides with 1% gelatin (Sigma, St Louis, MO) and then added 4X10^ endothelial cells/chamber in 1.0 ml DMEM with 2.5% equine serum. After incubation for 24 hours, the cells were washed with cold PBS, then 0.4 ml o f a second collagen solution (pH 7.4) containing 80% type I collagen (Vitrogen, Celtrix Lab, Palo Alto, CA), 10% 1 OX Minimum Essential Medium (Gibco, Gaithersburg, MD), and 10% 0.1 mol/L NaOH was added to the subconfluent monolayer and excess solution was then aspirated. The slides were incubated for 10 minutes at 37°C and culture medium (2.5% equine serum-supplemented DMEM) was then added to the slides. Endothelial cells elongated and formed CS within 24 hours. We established astrocyte-endothelial or liver- endothelial co-cultures 3 days after the addition o f the second collagen layer. Prior to the addition o f astrocytes (4X10'* cells/chamber) or liver cells (4X10^ cells/chamber) to the capillary preparations, serum-supplemented media from astrocyte or liver preparations were removed, the cells were treated with trypsin-EDTA, and resuspended in endothelial culture medium. After 7 days, nearly all of the astrocytes were associated with CS, and extensive GGTP activity was present along the length o f CS. Moreover, tight junctions are also present in this model (Minakawa et al., 1991). Thus, astrocyte-endothelial cell co-cultures exhibited morphological and functional features of the BBB. 107 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Conditioned-Media Experiments We performed conditioned-media experiments according to the method o f Maxwell et al. (1987). Briefly, astrocyte-conditioned media (ACM) and liver cell- conditioned media (LV) were prepared by first aspirating the growth media fi'om these cultures, washing with PBS, and replacing media with 5% FCS-supplemented media. Forty-eight hours later the ACM or LV were collected, centrifuged to remove cellular debris and stored at -80®C until use. Bovine brain endothelial cells were plated in 24-well tissue culture plates (4X10* cells/well) in 1 ml 5% FCS-supplemented growth media. At confluence the cells were incubated for 24 hours with media containing 50% ACM (or LV) and 50% fresh growth media or fi'esh growth media alone in the presence or absence o f nmitralizing anti-TGF-3 monoclonal antibody (mAb) (20 p.g, Genzyme Diagnostics, Cambridge, MA). Anti-gp-120 mAb (obtained through the AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NTH fi-om Dr. Bruce Chesebro) was used as a control irrelevant antibody. Transforming Growth Factor-3 Studies Bovine brain endothelial cells were plated in 24-well tissue culture plates (4X10* cells/well) as described above. At confluence, the cells were incubated with human re c o m binant TGF-31 and TGF-32 (Sigma) for 24 hours. In separate studies, total TGF-3 activity was determined as a measure o f proliferation inhibition using the mink lung epithelial cell line ATCC CCL-64, as previously described (Danielpour et al., 1989; Correale et al., 1995). Briefly, CCL-64 cells were seeded in 96-well microtiter plates at a 108 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. density o f 2X10^ cells/well in DMEM supplemented with 100 U/ml penicillin, 100 pg/ml streptomycin, 1% nonessential amino acids (JRH Bioscience, Lenexa, KS) and 10% FCS (Omega Scientific). Culture supernatants were added in triplicate to the wells for 18 hours. During the last 6 hours o f the incubation, 1 pCi/well o f [3H] TdR (ICN Biomedicals, Irvine, CA) was added. Cells were then harvested on glass fiber filters (Whatman, Maidstone, UK) using an automated cell harvester (Cambridge Technology, Cambridge, MA) and [3H] TdR incorporation measured in a scintillation counter (Pharmacia LKB Biotechnology Inc, Gaithersburg, MD). A standard was generated using human recombinant TGF-P (R&D Systems, Mirmeapolis, MN), and unknown values were interpolated fi’ om the standard curve. Specificity o f the assay was determined using a noitralizing mAb that recognizes TGF-31, TGF-32, and TGF-33 (Genzyme Diagnostics). A total o f 75 ng/ml o f anti-TGF-3 mAB effectively naitralized 5 ng/ml o f human recombinant TGF-3 The sensitivity o f the assay was 100 pg/ml. Isoforms o f TGF-3 (TGF-31 and TGF-32) fi'om conditioned media were measured by ELISA assay (Promega, Madison, WI). TGF-3 concentrations were adjusted to endothelial cell count. Polym erase C hain Reaction We used reverse transcription-polyraerase chain reaction (RT-PCR) to measure the relative abundance o f mRNA transcripts, as previously described. Total RNA was isolated with the Glassmax DNA Spin Cartridge Isolation System (Gibco BRL, Gaithersburg, MD). Total RNA from each preparation was resuspended in 40 pi o f 109 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. diethyl pyrocarbonate (DEPC) treated water. cDNA was synthesized from equal volumes o f total RNA in a total volume o f 20 |il. RNA was incubated in S pi o f DEPC-treated w ater at 6S"C for 3 minutes and quickly placed on ice. The RNA was then added to the transcription solution: 1.5 pmol/L oligo dT primers, 50 mmol/L Tris-HCl, 75 mmol/L KCl, 3 mmoI/L MgClj, 0.5 mmol/L dNTP, 1 unit/pl RNase inhibitor, and 13.3 units/pl AMV reverse transcriptase. The reaction was carried out at 42°C fbr I hour and terminated at 52°C for 40 minutes. The cDNA was stored at -20°C until use. PCR primers for bovine tPA (forward primer, BTPA-1: 5'AAGGTTGCAGAAGAAGATGG3' and reverse primer, BTPA-2: 5GTGAGGCGGGTACCTCTCCTGGAA3') and TM (forward primer, BTM-1: 5'CTCGGCAACTACACGTGCATCTGCGAG3' and reverse primer, BTM-2: 5'GCCACCACCAGAGACAGGCTTGCAAT(Kj3') were chosen from coding regions of the mRNA (Ravn et al., 1995; Jackman et ai., 1986). g-Actin primers (Stratagene, LaJolIa, CA; forward primer, 5TGACGGGGTCACCCACACTGTGCCCATCTA3’ ; reverse primer, 5'CTAGAAGCATTTGCGGTGGACGATGGAGGG3 ') were used to amplify P-actin mRNA as a housekeeping gene control. The PCR mixture contained 0.2- 1.0 pg cDNA, 10 mmol/L Tris-HCl, pH 8.3, 50 mmol/L KCl, 0.1 mmol/L dNTP, 1.0 mmol/L MgClz, 1.0 unit Taq Polymerase and 0.5 pmol/L forward and reverse primers. Amplification was carried out in a Geneamp PCR System 2400 (Perkin-Elmer Corp., Norwalk, CT): initial dénaturation at 94® C, each cycle consists o f 30 seconds dénaturation at 94°C, 30 seconds annealing at 56°C, and one minutes extension at 72°C. All RT-PCR assays were perfr>rmed within the linear range o f amplification. PCR products were visualized by electrophoresis on a 2% agarose gel and stained with 110 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ethidium bromide. N ^ativ es wa% prepared with a Polaroid cam era (Polaroid Corp., Cambridge, MA) and scanned by optical densitometry (Hoefer Instruments, San Francisco, CA). Optical densitometric (CD) measurements o f were normalized to P-actin levels (i.e., dividing by P-actin CD measurements). Quantitative competitive PCR (QC-PCR) tubes contained all the amplification reagents (described above), a constant amount o f target cDNA from each preparation, and serial dilutions o f known concentrations o f a competitor tPA or TM cDNA template. Competitor templates were generated according to the techniques o f Tran et al. (1996 and 1998a, 1998b). Following co-amplification, the PCR products were separated by electrophoresis on a 2% agarose gel and visualized by ethidium bromide staining and UV trans-illumination. Negatives were prepared and scanned by optical densitometry, as described above. Density readings o f the targ ^ tPA and TM cDNA were multiplied by 332/479 and 591/707, respectively, to correct for differences in molecular weight. The ratio o f amplifred target versus competitor cDNA optical densities was plotted as a function o f competitor template concentration. The initial concentration o f target cDNA was derived from the point at which the ratio o f target and competitor cDNA optical density equaled 1. Concentrations o f tPA and TM mRNA were adjusted to P-actin levels, as described above. Plasminogen Activator Assay We assayed cultured media from mono- and co-culture preparations for total plasminogen activator activity by amidolytic assay (American Diagnostica, Greenwich, 111 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CT). Amiloride (0.5 mmol/L, Sigma) was used to inhibit urokinase plasminogen activator (uPA) and allow for the determination o f net tPA activity (Vassalli and Belin, 1987). Protein C Assay Endothelial TM activity was assayed by measuring the increase in protein C activation according to the methods o f Tsiang et al. (1990). Cells are collected using a rubber policeman and resuspended in a 100 ^1 solution containing 50 mmol/L Tris-HCl, 2 mmol/L CaClj, 0.1 mol/L NaCI, 0.1% BSA, 0.1 |ig bovine thrombin, 2 ng bovine protein C, pH 8.0. The solution is incubated at 37®C for 30 min. The reaction is terminated by adding anti-thrombin HI (5 |ig) and hq)arin (5 U) and the resulting mixture is centrifuged at 200X g for 5 min. Equal volumes o f the reaction mixture and a chromogenic substrate (S-2366, Kabi) are added to a 96-well microther plate. Optical densities are measured with a spectrophotometer (EL-311 SX, BioTek Instruments, Burlington, VT) at 405 nm. Endothelial cell lysates were assayed for protein content by the method o f Lowry et al. (1951). Endothelial TM activity was adjusted to protein content. Statistical Analysis All data are expressed as means ± SD. Statistical comparisons between groups were performed using unpaired Student t-tests and Pearson's correlation coefficients. Dififerences were considered significant for p<0.05. 112 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Results We examined TGF-3 levels in conditioned media in our BBB model. We previously reported downiegulation o f tPA and TM transcripts in astrocyte-endothelial co-cultures in this model (Tran et al., 1996, 1998a). Astrocyte-endothelial co-cultures demonstrated a significant increase in active TGF-3 (3038 ± 323 pg/ml) compared to endothelial mono-cultures (821 ± 429 pg/ml; p<0.001). We analyzed our data for potential association between active TGF-3 and mRNA levels. There was a significant inverse correlation between active TGF-3 and tPA mRNA (r=-0.73, p<0.04) and TM mRNA concentrations (r=-0.90, p<0.003). To determine whether TGF-3 can downr%ulate tPA and TM mRNA, we incubated brain capillary EC with recombinant TCH^-31 and -32 (the two best characterized isofiarms). Treatment with TGF-31 resulted in dow nr^ulation o f endothelial tPA and TM mRNA in a dose-dependent manner (Fig 1). TGF-32 treatment produced dose-dependent downregulation o f TM mRNA and no significant reductions in tPA mRNA (Fig 1). We next examined media conditioned by astrocytes for 48 hours (ACM) to determine whether astrocytes can produce TCT-3- ACM had significantly higher levels o f active TGF-31 (443 ± 71 vs. 90 ± 36 pg/ml, p<0.03) and TGF-32 (641 ± 21 vs. 86 ± 21 pg/ml, p<0.002) than non-conditioned growth media. We then performed experiments using astroQfte-conditioned media to determine whether the astrocyte-induced mRNA downregulation is mediated by TGF-3- Bovine brain capillary EC were grown to 113 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. confluence and incubated for 24 hours with ACM. Examination o f endothelial transcripts 24 hours following incubation with ACM revealed reductions in tPA (13.5 ± 5.4 vs. 34.3 ± 3 .2 pg/ml, p<0.006) (Fig 2) and TM mRNA (43.0 ±2.1 vs. 68.8 ± 4.8 pg/ml, p<0.002) (Fig 3) compared to EC incubated with non-conditioned media. Incubation with liver cell-conditioned media (as control) had no effects on endothelial tPA (32.3 ± 9.9 pg/ml, p>0.6) or TM transcripts (70.2 ± 3.0 pg/ml, p>0.7). The downregulation o f tPA and TM mRNA was completely abolished with TGF-P neutralizing antibody (40.4 ±1.7 and 69.7 ± 5.5 pg/ml, respectively); the control irrelevant antibody produced no effect (Fig 3 and 4). To examine the functional consequences o f mRNA dow nr^ulation, we examined tPA activity in culture media and TM activity fiom cell lysates. There was a significant decrease in tPA activity in ACM incubated compared to control EC (15.3 ± 1.8 vs. 27.3 ± 1.0 lU/ml, p<0.02). This tPA downregulation was abolished by TGF-P neutralizing antibody (30.9 ±1.1 lU/ml); control amibody had no effect (17.7 ± 1.3 lU/ml). Concurrently, there was a significant reduction in TM activity in EC incubated with ACM compared to control (60 ± 2 vs. 100 ± 1 %, p<0.02) (Fig 5). This decrease was not observed in liver cell-conditioned media incubated cells (95 ± 2 % o f control, p>0.1). The downregulation in TM activity was attenuated by TGF-p naitralizing antibody (97 ± 4 % of control, p>0.5), but not by control antibody (68 ± 4 % o f control, p<0.05). These findings suggest that TGF-P mediates astrocyte transcriptional regulation o f tPA and TM. 114 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CD ■ D O Q . C g Q . " O CD C / ) o " 3 O 8 " O B 3. 3 " CD CD " O O Q . C a O 3 " O O CD Q . " O CD C / ) C / ) E t 120 1 110 - 2 100 z 9 0 ■ S E 8 0 F 7 0 8 60. g 5 0 « 40- 3 0 20000 200 2000 2 20 TGF-bl (pg/ml) 120 1 110 | £ , 0 0 . I | « g 5 80 F 7 0 g 6 0 O 5 0 - 20000 200 2000 2 20 120 1 110 I 8 0 3 ^ “ S!§ “ 20000 20 200 2000 2 D T0F-b2 (pg/ml) 120 1 110 0 2 1 0 0 l o “ g 5 « g 60 50 I 2000 20000 2 20 200 TGF-bl (pg/m l) TG F-b2 (pg/ml) Fig 1 . PCR analysis of tPA and TM mRNA expression. Reverse transcription-PCR showed dose dependent reductions in tPA (A,B) and TM (C,D) mRNA following stimulation with recombinant TGF-P 1 and -P2 (*p<0.02; **p<0.01; ♦♦*p<0,001). 1 0 Ab ■ Irrelevant Ab □ anti-T GF-b Ab E 60.0 S 40.0 Astrocyte-Conditioned Media C ontrol Figure 5.2. tPA mRNA concentration. Quantitative competitive-PCR analysis showed significant decrease in tPA mRNA in astrocyte-conditioned media-incubated endothelial cells compared with control. This reduction was abolished by TGF-p neutralizing antibody, whereas irrelevant antibody had no significant effect. Data are presented as mean ± SD. Data are from one experiment (performed in triplicate) and representative of three independent experiments (* p<0.006). 116 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 10 Ab ■Irrelevant Ab Danti-TGF-b Ab Astrocyte^onditloned Media Control Figure 5.3. TM mRNA concentration. Quantitative competitive-PCR analysis showed significant decrease in TM mRNA in astrocyte-conditioned media-incubated endothelial cells compared with control. This reduction was abolished by TGF-P nartralizm g antibody; irrelevant antibody had no significant effect. Data are presented as mean ± SD. Data are from one experiment (performed in triplicate) and representative o f three independent experiments (♦ P<0.002). 117 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 0 Ab ■ Control Ab □ anti-TGF-b Ab ABtrocyte-Conditicned Media Control Figure 5.4. tPA activity. Assay o f tPA activity from culture media showed a signifrcant reduction in tPA activity in astrocyte-conditioned media-incubated endothelial cells compared with control. This reduction was attenuated by TGF-P nw tralizing antibody, but not by irrelevant antibody. Data are presented as mean ± SD. D ata are from one experiment (performed in triplicate) and representative o f four independent experiments (♦ p<0.02). lis Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 10 Ab ■Control AbOantkTGF-b Ab ABtrocyle-Conditloned Media Control Figure 5,5. TM activity. Assay o f TM activity from ceil lysates showed a significant reduction in TM activity in astrocyte-conditioned media-incubated endothelial cells compared w ith control. This reduction was attenuated by TGF-P neutralizing antibody; e irrelevant antibody had no significant effect. Data are presented as mean ± SD. Data are from one experiment (performed in duplicate) and representative o f three independent experiments (*p<0.05). 119 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Discussion The present study showed that T Œ -P mediates astrocyte r% uIation o f endothelial expression o f anticoagulant Actors. We demonstrated a significant increase in active TGF-P in astrocyte-endothelial co-cultures compared to endothelial mono-cultures. We showed a strong inverse correlation between active TGF-P concentrations and tPA and TM expression. We also demonstrated that ACM negatively regulates endothelial expression o f tPA and TM transcripts. These effects were not demonstrable in EC incubated with conditioned media from control liver cells, thus indicating astrocyte- specific regulation. These changes in endothelial mRNA expression were blocked by neutralizing antibodies to TGF-P, but not by isotype-matched control antibody. The reductions in EC transcripts were present with a concommitant reduction in tPA activity and TM activity. These functional changes were also blocked by TGF-P neutralizing antibodies. Our findings demonstrated that astrocytes specifically regulate two critical endothelial anticoagulant Actors (tPA and TM), and that this regulation occurs via TGF- 3. Our data suggest that astrocytes contribute to TGF-P activation in astrocyte- endothelial co-cultures. TGF-P is synthesized as a large precursor protein consisting o f TGF-P and a latency-associated peptide (LAP) (Rifkin et al., 1993; Attisano et al., 1994). The LAP is associated with TGF-p through non-covalent interactions, and prevents TGF- P from binding to its receptor. TGF-P can be activated in vitro by a variety o f exogenous treatments, including acidification, alkalinization, heat or protease treatment (Miyazono 120 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. et al., 1993). The cellular mechanisms for TGF-P activation are not well understood; however, increases in active TGF-P have also been reported following heterotypic co cultures o f EC with other CNS derived cells (Antonelli-Orlidge et al., 1989; Sato and Rifkin, 1989; Sato et al., 1990). We have shown that both latent and active TGF-P are present in conditioned media from astrocyte mono-cultures. These data suggest that astrocytes not only secrete TGF-P, but can also r%ulate its activation. We found that astrocytes express both TGF-P 1 and TGF-P2 isoforms. Astrocytes have been reported to express TGF-Pl, TGF-P2, and TGF-p3 mRNA, but only secrete TGF-P 1 and TGF-P2 in vitro (Constam et al., 1992). TGF-P 1 and TGF-P2 are isoforms found most frequently in the CNS (Rifkin et al., 1993). TGF-P can be produced by a number o f cells in the CNS, including astrocytes and endothelial cells (Constam et al., 1992). Both astrocyte-derived and endothelial-derived TGF-P were secreted into the growth media in our co-culture preparations. Our assays for TGF-P caimot assess the relative contributions o f TGF-P by each o f these cell types. Our data showed that TGF-P2 is not as effective as TGF-P 1 in reducing tPA mRNA. TGF-P 1 has been suggested to be more biologically active than TGF-p2 (Ohta et al., 1987). However, in the same cultures we demonstrated comparable reductions in TM mRNA following exogenous treatment with TGF-P 1 and TGF-P2. Similar inhibitions o f TM transcription by TGF-P 1 and TGF-p2 have also been reported in human umbilical vein EC (Ohji et al, 1995). These data argue against the possibility that differences in bioactivity occur as a result o f differential binding afSnities. Taken 121 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. togaher, these findings suggest that tPA and TM downreguiation by TGF-P may occur through different intracellular mechanisms. The extent o f tPA downr%ulation by astrocyte-conditioned media was comparable to those previously reported by direct astrocyte-endothelial co-cultures (Tran et al., 1998a). These findings further corroborate the role o f an astrotqde-derived soluble factor, namely TGF-P, in astrocyte-mediated tPA emulation. However, the extent o f TM regulation was not as extensive as previously reported in direct astrocyte-endothelial co cultures. These data suggest that TM regulation in co-cultures may occur through multiple mechanisms; TGF-P may constitute only one component o f the astrocyte-EC interactions. Cellular contact between the two cell types may induce further changes in endothelial function. Astrocytic cellular membrane has been reported to modulate selective endothelial function in vitro (Tontsch and Bauer, 1991). Moreover, cellular contact in our astrocyte-endothelial co-cultures may function as a positive feedback mechanism by enhancing production and/or activation o f TGF-P. The regulatory roles of astrocytes on EC through direct cell contact are well defined (DeBault and Cancilla, 1980; Tao-Cheng et al., 1987; Bauer et al., 1990; Dehouck et al., 1990; Tio et al., 1990; Meyer et al., 1991; Tontsch and Bauer, 1991; Sun et al., 1995; Hayashi et al., 1997). Astrocyte-induced effects can also occur via astrocyte- conditioned media (Maxwell et al., 1987; Tao-Cheng et al., 1987; Rubin et al., 1991; Raub et al., 1992; O'Donnell et al., 1995). Our findings suggest that astrocyte regulation o f endothelial hemostasis function is mediated in part by astrocyte-derived TGF-P. Astrocytes can also elaborate cytokines and other growth Actors known to affect endothelial hemostasis function (Conway and Rosenberg, 1988; Schleef et al., 1988; 122 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Lieberman et al., 1989; Hirokawa and Aoki, 1991; Sharif et al., 1993). Our data cannot exclude the roles o f these soluble Actors in the regulation o f endothelial function. Further, we cannot rule out the potœ tial contributions o f direct cell contact on regulation o f endothelial hemostasis. Our findings provide fiirtho' support for the existence o f a unique hemostatic r% ulatory apparatus o f the brain. This r%ulation is astrocyte-dependent, occurs in the microcirculation, and is mediated by TGF-p. It is important to be cautious in «ctrapolating our in vitro findings to in vivo phenomena. Nevertheless, modification o f brain expression o f anticoagulant factors may now be possible. 123 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reference 1. Antonelli-Orlidge A, Saunders KB, Smith SR, D'Amore PA (1989) An activated form o f transforming growth foctor-P is produced by cocultures o f endothelial cells and pericytes. P roc N atl A cad Sci USA 86:4544-58 2. Aoki Y, Takei R, Mohri M, Gonda Y, Gomi K, Sugihara T, Kiyota T, Yamamoto S, Ishida Y, Maruyama I (1994) Antithrombotic effects o f recombinant human soluble thrombomodulin (ihs-TM) on arterial venous shunt thrombosis in rats. Am J H em atol 47: 162-6 3. Attisano L, W rana JL, Lopez-Cassila F, Massague J (1994) TGF-P receptors and actions. BiocM m Biophys A cta 4. Bauer HC, Tontsch U, Amberger A, Bauer H (1990) Ganuna-glutamyl-transp^mdase (GOT?) and Na-K-ATPase activities in different subpopulations o f cloned cerdjral endothelial cells: responses to glial stimulation. Biochem Biophys R es Comm 168: 358-63 5. Boffa MC, Jackman RW, Peyri N, Bof& IF, George B (1991) Thrombomodulin in the central nervous system. Nouv R ev Fr H em atol 33:423-429 6. Carson, MP, Haudenschild, CC (1986) Micro vascular aidothelium and pericytes: high yield, low passage cultures. In Vitro C ell D ev B io l 22: 344-54 7. Collen D, Lijnen HR (1995) Molecular basis o f fibrinolysis, as relevant for thrombolytic therapy. Thromb HaemostlA-. 167-71 8. Constam DB, Philipp J, Malipiero UV, Dijke P, Schachno' M, Fontana A (1992) Differential expression o f transforming growth 6ctor-pi, -P2, and -P3 by glioblastoma cells, astrocytes and microglia. J Im m unol 148: 1404-10 9 Conway EM, Rosenberg RD (1988) Tumor necrosis 6 cto r suppresses transcription o f the thrombomodulin gene in endothelial cells. M o l C ell B iol 8: 5588-92 10. Correale J, Gilmore W, McMillan M, Li S, McCarthy K, Le T, Weiner LP (1995) Patterns o f cytokine secretion by autoreactive proteolipid protein-specific T cell clones during the course o f multiple sclerosis. JIm m uno 154: 2959-68 11. Danielpour D, Dart LL, Flanders KC, Roberts AB, Spom MB (1989) Immunodetection and quantitation o f the two forms o f transforming growth firctor-P (TGF-P 1 and TGF-P2) secreted by cells in culture. J C ell P hysiol 138: 79-86 124 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 12. DeBault LE, Cancilla PA (1980) Gamma-glutamyl transpeptidase in isolated brain endothelial cells: induction by glial cells in vitro. Science 207: 653-5 13. Ddiouck MP, Meresse S, Ddorm e P, Frudiart JC, Cecchelli R (1990) An easier, reproducible, and mass-production method to study the blood-brain barrier in vitro. J Neurochem 54: 1798-1801 14. Dehouck B, Dehouck MP, Fruchait JC, Cecchelli R (1994) Upregulation o f the low density lipoprotein at the blood-brain barrier mtercommunications between brain capillary endothelial cells and astrocytes. J C ell B io l 126: 465-73 15. Esmon NL, Owen WG (1981) Identification o f an endothelial cell cofactor for the thrombin-catalyzed activation o f protein C. Proc N atl A cad Sci USA 78: 2249-52 16. Gomi K, Zushi M, Honda G, Kawahara S, Matsuzaki O, Kanabayashi T, Yamamato S, Maruyama I, Suzuki K (1990) Antithrombotic efiTect o f recombinant human thrombomodulin on thrombin-induced thromboembolism in mice. B lood 75: 1396-9 17. Grau Œ , de Mberioose P, Bulla 0 , Lou J, Lei Z, Reber G, Mill N, Ricou B, Morel DR, Suter PM (1997) Haemostatic properties o f human pulmonary and cerd)ral microvascular endothelial cells. Thromb Haem ost 77: 585-90 18. Gresele P, Momi S, Berrettini M, Nenci GG, Schwarz HP, Semeraro N, Colucci M (1998) Activated human protein C prevents thrombin-induced thromboembolism in mice. Evidence that activated protein C reduces intravascular fibrin accumulation through the inhibition o f additional thrombin generatiorL J Clin Invest 101: 667-76 19. Hayashi Y, Nomura M, Yamagishi S, Harada S, Yamashita J, Yamamoto H (1997) Induction o f various blood-brain barrier properties in non-noiral endothelial cells by close apposition to co-cuhured astrocytes. G lia 19: 13-26 20. Hirokawa K, Aoki N (1991) Regulatory mechanisms for thrombomodulin expression in human umbilical vein endothelial cells in vitro. J C ell P l^sio l 147: 157-65 21. Isaka T, Yoshimine T, Motokiko M, Kurode R, Ishii H, Hayakawa T (1994) Altered expression o f antithrombotic molecules in human glioma vessels. A cta N europathol 87: 81-5 22. Ishii H, Salem HH, Beil CE, Laporata E A, Majerus PW (1986) Thrombomodulin, an anticoagulant protein, is absent in the human brain. B lood 67: 362-5 23. Jackman, RW, Beeler, DL, DeWaters, L, Rosenberg, RD (1986) Characterization o f a thrombomodulin cDNA reveals structural similarity to the low density lipoprotein receptor. Proc N atl A cad Sci USA 83: 8834-8 125 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 24. Kittaka M, Wang L, Sun N, Schreiber SS, Seeds NW, Fisho' M, Zlokovic BV (1996) Brain capillary tissue plasminogen activator in a diabetes stroke model. Stroke 27; 712-9 25. Kumada J, Dittman WA, Majeius PW (1987) A role for thrombomodulin in the pathogenesis o f thrombin-induced thromboembolism in mice. B lood 71: 728-33 26. Levin EG, del Zoppo GJ (1994) Localization o f tissue plasminogen activator in endothelium o f a limited number o f vessels. Am J Path 144: 855-61 27. Lieberman AP, Pitha PM, Shin SH, Shin ML (1989) Production o f tumor necrosis Actor and other cytokines by astrocytes stimulated with lipopolysaccharide o r a neurotrophic virus. Proc N a tl A cad Sci USA 86:6348-52 28. Lowry OH, Rosdirough NJ, Farr AL, Randall RJ (1951) Protein measurement w ith the Folin phenol reagent. J B io l Chem 193: 265-75 29. Macko RF, Ameriso SF, Gruber A, Griffin JH, Fernandez JA, Bamdt R, Quismoro FP, W einer JM, Fisher M (1996) Impairment o f the protein C system and fibrinolysis in infection-associated stroke. Stroke 27: 2005-11 30. Macko RF, BCillewich LA, Fernandez JA, Cox DK, Gruber A, Griffin JH (1997) Brain-specific protein C activation during carotid artery occlusion in humans. Arm N eurol 42: 438. Abstract 31. Maruno M, Yoshimine T, Isaka T, Kurode R, Ishii H, Mayakawa T (1994) Expression o f thrombomodulin in astrocytomas o f various malignancy and in gliotic and normal brain. J Neuro-O ncology 19: 155-160 32. Maxwell K, Berliner JA, Cancilla PA (1987) Induction o f y-glutamyl transpeptidase in cultured cerebral endothelial cells by a product released by astrocytes. B rain R es 410: 309-1 33. McCarthy, KD, DeVellis, J (1980) Preparation o f separate astroglial and oligodend- roglial cell cultures firom rat cerebral tissue. J C ell B io l 85: 890-902 34. M ey^ J, Rauh J, Galla H (1991) The susceptibility o f cerebral endothelial cells to astroglial induction o f blood-brain barrier enzymes depends on their prolifmative state. J Neurochem 57: 1971-7 35. Nfinakawa, T, Bready, J, Berliner, J, Fisher, M, Cancilla, P.A. (1991) In vitro interactions o f astrocytes and pericytes with capillary-like structures o f brain microvessel endothelium. Lab Invest 65: 32-40 126 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 36. AAyazono K, Ichijo H., Heldin CH (1993) Transforming growth factor-beta: latent forms, binding proteins and r e c to r s . Growth F actors 8; 11-22 37. The National Institute o f Neurological Disorders and Stroke rt-PA Stroke Study Group (1995) Tissue plasminogen activator for acute ischemic stroke. N E ngl J M e d 333; 1581-7 38. O'Donnell ME, Martinez A, Sun D (1995) Cerebral microvascular endothelial cell Na-K-Cl cotransport: r% ulation by astrocyte-conditioned media. Am J P hysiol 268: 0747-54 39. Ohji T, Urano H, Shirahata A, Yamagishi M, Higashi ^ Gotoh S, Karasaki Y (1995) Transforming growth foctor betal and beta2 induce down-modulation o f thrombomodulin in human umbilical vein endothelial cells. Thromb H aem ost 73: 812-8 40. Ohta M, Grenberger JS, Anklesaria P, Bassolo A, Massague J (1987) Two forms o f transforming growth factor-beta distinguished by multipotential hematopoietic progenitor cells. N ature 329: 539-41 41. Ono M, Nawa K, Marumoto Y (1994) Antithrombotic effects o f recombinant human soluble thrombomodulin in a rat model o f vascular shunt thrombosis. Thromb Haemost 72: 421-5 42. Raub TJ, Kuentzel SL, Sawada GA (1992) Permeability of bovine brain microvessels endothelial cells in vitro: barrier tightening by a factor released from astroglioma cells. Exp C ell R es 199: 330-40 43. Ravn P, Berglund L, Petersen TE, Larsen LB (1995) Cloning and characterization o f the bovine plasminogen activator uP A and tP A. In t D airy J 5: 605-17 44. Rifkin DB, Kojima S, Abe M, Harpel JG (1993) TGF-P: Structure, function, and formation. Thromb H aem ost 10: 177-9 45. Rubin LL, Hall DE, Porter S, Barbu K, Cannon C, Homer HC, Janatpour M, Liaw CW, Manning K, Morales J, Tanner LI, Tomaselli KJ, Bard F (1991) A cell culture of the blood-brain barrier. J C e llB io l 115: 1725-35 46. Saksela O, Moscatelli D, Rifkin DB (1990) The opposing effects o f basic fibroblast growth foctor and transforming growth factor beta on the regulation o f plasminogen activator activity in capillary endothelial cells. J C e ll B iol 105: 957-63 47. Salem HH, Maruyama I, Ishii H, Majems PW (1984) Isolation and characterization of thrombomodulin from human placenta. J B iol Chem 259: 12246-51 127 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 48. Sato Y, Rifidn DB (1989) Inhibition o f endothelial cell movement by pericytes and smooth muscle cells: activation o f a latent transforming growth factor-beta 1-like molecule by plasmin during co-cultures. J C e ll B iol 109: 309-15 49. Sato Y, Tsuboi R, Lyons RM, Moses H, Rifidn DB (1990) Characterization o f the activation o f latent TGF-beta by co-cultures of endothelial cells and pericytes in smooth muscle cells: a self r%ulating system. J C e ll B io l 111: 757-63 50. Schleef RR, Bevilacqua MP, Sawdey M, Gimbrone Jr. MA, Loskutoff DJ (1988) Cytokine activation o f vascular endothelium: Effects on tissue-plasminogen activator and type 1 plasminogen activator inhibitor. J B iol Chem 263: 5797-803 51. Sharif SF, Hariri RJ, Chang V A Barie PS, Wang RS, CHiajar JB (1993) Human astrocyte production o f tumor necrosis factor-alpha, interleukin-1 beta, and interleukin^ following exposure to lipopolysaccharide endotoxin. N eurol R es 15: 109-12 52. Shatos M A Orfbo T, Doherty JM, Penar PL, Collen D, Mann KG (1995) a - Thrombin stimulates urokinase production and DNA synthesis in cultured human cerebral microvascular endothelial cells. A rterioscler Thromb Vase B io l 15: 903-911 53. Solis MM, Cook C, Cook J, Q aser C, Light D, Morser J, Yu SC, Fink L, Eidt JF (1991) Intravenous recombinant soluble human thrombomodulin prevents venous thrombosis in a rat model. J Vase Surg 14: 599-604 54. Sun D, Lytle C, O'Donnell ME (1995) Astroglial cell-induced expression o f Na-K-Cl cotransporter in brain microvascular endothelial cells. Am J P hysiol 269: C1506-12 55. Svensson PJ, Dahlback B (1994) Resistance to activated protein C as a basis for venous thrombosis. N E jig l JM ed 330: S\1~H 56. Tao-Cheng JH, Nagy Z, Brightman MW (1987) Tight junctions to brain endothelium in vitro are enhanced by astroglia. JN eurosei 7: 3293-9 57. Taylor FB, Chang A Esmon CT, D*Angelo A Vigano-D'Angelo S, Blick KE (1987) Protein C prevents coagulopathic and lethal effects o f Escherichia coli injection in the baboon. J C lin Invest 19: 9\%-15 58. Tio S, Deenen M, Marani E (1990) Astrocyte-mediated induction o f alkaline phosphatase activity in human umbilical cord vein endothelium. E ur J M orphol 28: 289-300 59. Tontsch U, Bauer H (1991) Glial cells and neurons induce blood-brain barrier related aizym es in cultured cerebral endothelial cells. Brcdn R es 539: 247-53 128 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 60. Tran ND, Wong VLY, Bready J, Schreiber SS, Fisher M (1996) Regulation o f brain capillary endothelial thrombomodulin mRNA expression. Stroke 27: 2304-11 61. Tran ND, Schreiber SS, Fisher M (1998a) Astrcx^e r% ulation o f endothelial tissue plasminogen activator in a blood-brain barrier model. J Cereb B lood Flow M et (In Press) 62. Tran ND, Wang L, Schreiber SS, Zlokovic BV, Fisher M (1998b) Measurement o f thrombomodulin mRNA expression in brain capillaries by polymerase chain reaction. Thromb R es (In Press). 63. Tsiang M, Lentz SR, Dittman WA, Wen D, Scaarpati EM, Sadler JE (1990) Equilibrium binding o f thrombin to recombinant human thrombomodulin: effect o f hirudin, fibrinogen, factor Va, and peptide analogues. Biochem 29: 10602-12 64. Vassalli JD, Belin D (1987) Amiloride selectively inhibits the urokinase-type plasminogen activator. F E B SLetter 2 \4 : 187-91 65. Wang L, Kittaka M, Sun N, Schreiber SS, Zlokovic BV (1996) Chronic nicotine treatment enhances focal ischemic brain injury and depletes fi-ee pool o f brain microvascular tissue plasminogen activator in rats. J Cereb B lood Flaw M etab 17; 136-46 66. Wang L, Tran ND, Kittaka M, Fisher MJ, Schreiber SS, Zlokovic BV (1997) Thrombomodulin expression in bovine brain capillaries. Anticoagulant function o f the blood-brain barrier, regional differences, and regulatory mechanisms. A rterioscler Thromb Vase B iol 17: 3139-46 67. Wong VLY, Hofinan FM, Ishii H, Fisher M (1991) Regional distribution o f thrombomodulin in the human brain. B rain Res 556: 1-5 68. Zuber M, Toulon P, Mamet L, Mas JL (1996) Factor V Leiden mutation in cerebral venous thrombosis. Stroke 27: 1721-3 129 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 6 Discussion This woiic demonstrated that astrocytes regulate a number o f endothelial functions, including blood-brain barriers (BBB) and hemostasis function. Further, we showed that astrocyte-dependent endothelial hemostasis regulation is mediated in part via transforming growth factor-^ (TGF-P). In chapter 2, we examined the potential regulatory role o f astrocytes in astrocyte-endothelial co-cultures. After 7 days o f culture, astrocyte-endothelial co-cultures demonstrated increased expression of the BBB marker Y-glutamyl transpeptidase (GGTP) compared to endothelial cell (EC) mono-cultures. Concomitantly, co-cultures expressed a greater than twenty-fold decrease in thrombomodulin (TM) mRNA concentration. These findings suggest that astrocytes induce endothelial expression o f BBB properties and n^atively regulate endothelial TM mRNA expression. TM is found in abundance throughout the peripheral vasculature (Esmon et al., 1982; Salem et al., 1984; Kurosawa & Aoki, 1985; Maruyama & Majerus, 1985; Jakubowski & Owen, 1986; Kumada et al., 1987). However, its expression is reportedly absent or limited in brain EC (Ishii et al., 1986; DeBault et al., 1986; Wong et al., 1991; Isaka et al., 1994; Maruno et al., 1994). This differential distribution o f TM in the 130 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. vasculature suggests that TM expression is restricted at the BBB. Our findings of reduced TM mRNA in our BBB co-cultures are consistent with these prior reports. Moreover, our data support the contention that TM mRNA is reduced in EC demonstrating BBB properties. The relationship between the BBB and TM expression should be interpreted with caution. Our findings demonstrated a correlative relationship between BBB phenotypes and TM mRNA, both o f vdiich occurred in response to co-culture with astrocytes. Our data cannot elucidate whether these effects occur through similar mechanisms (ie, via the same signaling pathway), nor can we determine whether TM mRNA regulation occurs in response to BBB induction. Thus, further investigations are warranted to delineate the potential causal relationship between the BBB and TM «(pression. Our in vitro model is comprised of bovine brain microvascular EC and murine astrocytes. The small differences in genetic composition between the two species allows us to selectively examine TM mRNA expression by EC. Despite the molecular advantage of using cells fi'om two different species, our «cperimental paradigm raises questions regarding the specificity of endothelial re la tio n by astrocytes. In order to determine whether TM downregulation is specific to astrocytes or merely an artifact of culturing cells from different species, we examined TM mRNA expression in co-cultures of bovine EC and murine liver cells. After 7 days of co-culture, there wwe no significant differences between EC-liver cell co-cultures and EC mono-cultures. These findings indicate that reduced TM mRNA expression otxxirs specifically in response to astrocyte- endothelial interactions. Moreover, these findings rule out an immunological response between bovine and murine cells as a potential mechanism for TM downregulation. 131 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Astnx^e-dependent TM mRNA downr%uIation may occur as a result o f decreased transcription and/or decreased mRNA stability. Our data cannot distinguish between these potential mechansisms. Previous studies have reported that TM dow nr^ulation occurs primarily via a decrease in transcription with out attenuation o f TM mRNA stability (Conway & Rosenberg, 1988; Lentz et al., 1991). W e speculate that decreased TM transcription may be responsible for astrocyte-dependent TM mRNA downr%ulation; however, additional sutdies are required to exclude the possibility o f decreased mRNA stability. The TM gene, unlike most mammalian genes, lacks introns (Jackman et al., 1987; Dittm an & Majerus, 1990). While the implications o f these findings are not fully understood, the lack o f introns suggests that TM protein expression is directly dependent on its rate of transcription. Other post-translational processes, such as vesicle-mediated internalization and degradation, may play a role in TM protein downr%ulation. However, this mechanism o f regulation is unlikely because TM’s cytoplasmic domain lacks the consensus signal for internalization by coated pit-mediated cytolysis (Goldstein et al., 1985). Taken together, these data suggest that downregulation o f TM activity in our BBB model occurs as a result o f astrocyte-dependent TM mRNA downregulation. In chapter 3, we examined the role o f astrocytes in the regulation o f endothelial fibrinolysis. Astrocyte-endothelial co-cultures demonstrated reduced expression o f tissue plasminogen activator (tPA) mRNA and enhanced expression of plasminogen activator inhibitor-1 (PAI-1) mRNA compared to EC mono-cultures. These changes occurred after 7 days co-culture, a time when astrocyte-endothelial interactions where maximal. No changes in either tPA or PAI-1 transcripts were observed following endothelial co-culture 132 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. with liver cells. These findings indicate that astrocytes n^ativeiy r% ulate endothelial fibrinolysis by decreasing tPA mRNA and increasing PAI-1 mRNA. Studies o f fibrinolysis are also confounded by the potential contribution of astrocyte-derived tPA (Kalderon et al., 1990; Tranque et al., 1992) and PAI-1 (Tranque et al., 1992). Our usage o f murine astrocytes and bovine EC allows for the selective examination o f only endothelial transcripts by taking advantage o f the differences in genetic composition between the two species. We have shown that astrocytes modulate endothelial expression o f TM, tPA, and PAI-1 mRNA. Endothelial TM and tPA transoipts were reduced following 7 day co culture with astrocytes. This downr%ulation can occur as a result o f a global non specific reduction in endothelial transcription. However, the demonstration o f enhanced PAI-1 mRNA rules out this possibility. Thus, these data provide strong evidence for the specificity and selectivity o f astroqde-dependent mRNA r^ulation. Astrocyte-endothelial co-cultures demonstrated reduced tPA activity compared to EC mono-cultures. This reduction in fibrinolytic function is likely to occur through reduced tPA transcription and subsequently, reduced tPA production. The upr%ulation in endothelial PAI-1 mRNA in co-cultures sr%gests that the decreased fibrinolytic activity is also due to astrocyte-dependent enhanced production o f PA I-1. In chapter 4, we used quantitative competitive-PCR and reverse transcription- PCR to examine TM mRBA expression in capillaries from different brain regions. Data from both assays demonstrated significantly highw levels o f TM mRNA in cortical and cerd)ellar capillaries compared to pontine capillaries. This differential mRNA expression corresponded strongly with previously reported studies o f Tm protein 133 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. expression from the same regions (Wong et al., 1991). These data support the contention that reduced TM mRNA contributes to reduced TM protein in the brain. In chapter 5, we examined the hypothesis that TGF-P mediates astrocyte r^ ^ latio n o f endothelial hemostasis factors. Astrocyte-conditioned media (ACM) contained high levels o f active TGF-P compared to non-condhioned growth media. We demonstrated that ACM downr%ulates endothelial expression o f TM and tPA mRNA and activity. ACM-mediated effects on EC were similar to those observed in astrocyte- endothelial co-cultures. These astrocyte-specific effects were not observed with conditioned media from control liver cells. Moreover, these changes in endothelial TM and tPA transcripts and activity were blocked by neutralizing antibody to TGF-P, but not by isotype-matched control antibody. Our findings suggest that TGF-P mediates astrocyte regulation o f endothelial anticoagulant Actors, TM and tPA. TGF-P can be produced by astrocytes and EC in the CNS (Constam et al., 1992). Both astrocyte-derived and endothelial derived TGF-P were secreted imo the growth media in our co-culture preparations where they can elicit an effect. Moreover, astrocyte- derived TGF-P may act via a positive feedback mechanism to enhance endothelial production o f TGF-p. Our assays for TGF-P caimot assess the relative contributions o f TGF-P by each o these cell types. Further investigations using either transgenic astrocytes or EC that lack the TGF-P gene are warranted to delineate the relative contributions o f TGF-P by these cells. In addition to astrocytes and EC, TŒ^-P can be produced by other cell types in the CNS, including microglial and oligodendrocytes (Massague 1990; Rifidn et al., 1993; 134 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Attisano et al., 1994). Howevo', it is unlikely that these sources of TGF-P contribute to endothelial hemostasis regulation under physiological conditions. TGF-P modulates endothelial function in a dose-dq>endent manner. The juxtaposition o f astrocytes to EC at the glial limitans ensures that local concentrations of TGF-P are sufficiently high to elicit a biological response. Because microglia and ogigodendrocytes are physically separated from the endothelium by distance and by an astrocyte barrier, TGF-P concentrations derived from these cells are like to be too low to affect endothelial function. This does not necessarily exclude the role of microglia- and oligodendrocyte- derived TGF-P in endothelial r%ulation. Under pathological conditions, these glial cells may be induced to produce higher concentrations of TGF-p. Alternatively, activated gUal cells may migrate toward EC where they can elicit an effect. Our findings provides evidence for the emerging view of astrocytes as a key mediator of hemostasis within the braiiL This r%ulation occurs in the microcirculation and is mediated in part by T C H ^-P . The negative r^ulation of endothelial TM and tPA and enhancement of PAI-1 suggest a prothrombotic miliar within the brain. A prothrombotic milieu may provide an evolutionary advantage by protecting the brain from neonatal intracranial hemorrhage. With age and development of stroke risk factors, this advantage could become counter-productive. An alternative view is that this prothrombotic m iliai provides no advantage. It has managed to elude evolutionary pressures because it mainly affects individuals well beyond the child-bearing age. Nevertheless, modulation of astrocyte-dependent endothelial hemostasis regulation may lead to new strategies for the treatment and prevention of ischemic stroke. 135 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reference 1. Attisano L, Wrana JL, Lopez-Cassila F, Massague J (1994) TGF-P recq*tws and actions. B iochim B iophys A cta 1132, 2. Constam DB, Philipp J, Malipiero UV, Dijke P, Schachner M, Fontana A (1992) Diffo-ential expression o f transforming growth 6 c to r-P l, -P2, and -33 by glioblastoma cells, astrocytes and microglia. J Immunol 148, 1404-10 3. (Zonway EM, Rosenberg RD (1988) Tumor netxosis Actor suppresses transoiption of the thrombomodulin gene in endothelial cells. M ol C ell B io l 8, 5588-92 4. DeBault LE; Esmon NL; Olson JR; Esmon CT (1986) Distribution o f the thrombomodulin antigen in the rabbit vasculature. Lab In vest 54; 172-8 5. Dittman WA, Majerus PW (1990) Structure and function o f thrombomodulin: a natural anticoagulant. B lood 75, 329-36 6. Goldstein JL, Brown MS, Anderson TG, Russell DW, Schneider WJ (1985) Receptor mediated endocytosis: concepts emerging from the LDL system. Ann R ev C ell B iol 1,1 7. Esmon NL, Owen WG, Esmon CT (1982) Isolation o f a membrane-bound coActor for thrombin-catalyzed activation of protein C J B iol Chem 257, 859-64 8. Kalderon N, Ahonen K, Federoff S (1990) Developmental transition in plasticity properties o f differentiating astrocytes: age-related biochemical profile o f plasminogen activators in astroglial cultures. G lia 3, 413-26 9. Isaka T, Yodiimine T, Motokiko M, Kurode R, Ishii H, Hayakawa T (1994) Altered expression o f antithrombotic molecules in human glioma vessels. A cta N europathol 87, 81-5 10. Ahii H, Salem HH, Beil CE, Laporata EA, Majerus PW (1986) Thrombomodulin, an anticoagulant protein, is absent in the human brain. B lood 67, 362-5 11. Jackman RW, Beeler DL, VanDeWato’ L, Rosenberg RD (1986) Characterization of a thrombomodulin cDNA reveals structural similarity to the low density lipoprotdn receptor. P roc N atl A cad Sci USA 12. Jakubowski HV, Owen WG (1986) The effect of bovine thrombomodulin on the specificity o f bovine thrombirL J B io l Chem 264, 111 17-21 13. Kumada J, Dittman WA, Majerus PW (1987) A role for thrombomodulin in the pathogenesis o f thrombin-induced thromboembolism in mice. B lood 71, 728-33 136 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 14. Kurosawa S, Aoki N (1985) Preparation o f thrombomodulin from human placenta. Thromb R es 37, 353-64 15. Lentz SR, Tsiang M, Sadler £ (1991) r% ulation o f thrombomodulin by tumor necrosis & ctor-a: comparison o f transcriptional and post-transcriptional mechanisms. B lood n , 542-50 16. Massague J (1990) The transforming growth foctor-P family. Arm R ev Cell B iol 6, 597-601 17. Maruyama I, Majems PW (1985) The turnover o f thrombin-thrombomodulin complex in cultured human umbilical vein endothdial cells and A549 lung cancer cells. Endocytosis and degradation o f thrombin. J B io l Chem 260, 15432-8 18. Maruno M, Yoshimine T, Isaka T, Kurode R, Ishii H, Mayakawa T (1994) Expression o f thrombomodulin in astrocytomas o f various malignancy and in gliotic and normal brain. J Neuro-O ncology 19, 155-60 19. Rifkin DB, Kojima S, Abe M, Harpel JG (1993) TGF-p: Stmcture, fonction, and formation. Thromb Haemost 10, 177-9 20. Salem HH, Mamyama I, Ishii H, Majerus PW (1984) Isolation and characterization o f thrombomodulin from human placenta. J B io l Chem 259, 12246-51 21. Tranque P, Robbins R, Naftolin F, Andrade-Gordon P (1992) Regulation o f plasminogen activator and type 1 plasminogen activator inhibitor by cyclic AMP and phorbol ester in rat astrocytes. G lia 6, 163-71 22. Wong VLY, Hofrnan FM, Ishii H, Fisher M (1991) R%ional distribution o f thrombomodulin in the human brain. Brain R es 556, 1-5 137 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. IM AG E EV ALUATIO N T E S T T A R G E T ( Q A - 3 ) / / Vf 1 .0 i.i 1.25 Là Lâ 1*0 1.4 22 2.0 .8 1 .6 150mm 6" V f 7 V W J - / 4 P P L I E D A IIV14GE . In c 1653 East Main Street Rocliester. NY 14609 USA Phone: 716/462-0300 ------ Fax: 716/288-5989 O '993. AppAed Image. Inc.. A N Rights Reserved Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

AMPA receptor regulation in organotypic hippocampal cultures

PDF

Characterization of subcellular AMPA receptor populations: Changes with synaptic plasticity.

PDF

Career advancement in managerial hierarchies in the United States firms: A multi-theoretic model and empirical tests of the determinants of managerial promotion

PDF

Calpain-mediated alpha-amino-3-hydroxy-5-methylisoxazole-4-propionate receptor regulation.

PDF

Oxygen Free Radical Involvement In Excitotoxic Neuronal Injury

PDF

Communicating via electronic mail: Patterns and predictors of use in organizations

PDF

Apoptotic pathways involved in kainate excitotoxicity

PDF

A dynamic reformulation of perceptions of inequity: Their organizational antecedents and outcomes

PDF

A role for transcription factor Msx2 in vivo in cardiac neural crest cell development

PDF

Cellular and molecular effects of vasopressin, a mnemonic neuropeptide, on cultured hippocampal and neocortical cells derived from embryonic rat brain

PDF

An organizational culture approach to the study of individual power and its use

PDF

Analysis of the function and regulation of the 94kDa glycoprotein of the glucose-regulated protein family.

PDF

Activation of transcription by the C1 protein of bacteriophage P22.

PDF

Aging-induced expression of hsp22 in Drosophila melanogaster

PDF

Characterization Ofcis-Elements Responsible For The Eye Expression Of Glass, A Gene Required For The Drosophila Photoreceptor Development.

PDF

Charles Ives - master songwriter: The methods behind his madness

PDF

Correlates of deceit: A cross-cultural examination

PDF

A molecular, genetic and immunohistochemical characterization of the DrosophilanudC (DnudC) gene

PDF

Connections of the dorsomedial hypothalamic nucleus in the rat

PDF

A system of participation and allocation preferences in organizations

Asset Metadata

Creator Tran, Nam Duy (author)

Core Title Astrocyte regulation of endothelial haemostasis function via transforming growth factor-beta

Contributor Digitized by ProQuest (provenance)

Degree Doctor of Philosophy

Degree Program Neuroscience

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag biology, cell,biology, molecular,biology, neuroscience,OAI-PMH Harvest

Language English

Advisor Fulk, Janet (committee chair), [illegible] (committee member), Schreiber, Steven S. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c17-403786

Unique identifier UC11353918

Identifier 9919113.pdf (filename),usctheses-c17-403786 (legacy record id)

Legacy Identifier 9919113.pdf

Dmrecord 403786

Document Type Dissertation

Rights Tran, Nam Duy

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA