V. . VV . . .
VV. V V V . .VVV V V . .
. V . V V ..V .. .
V V, .V
4 VV v . .
V V V V. _ .. V V .
. . V _. V . . . .
V .VV 3. .. . . .V V V V V
V V . ._ .V . V V V .V .V . VV. V V .
. V . V .y . .. . V . V
V .V. V. V V . . V . .
V V V V ._ VV. . .
. . V . V . V . . V .. V. V V . V
V V v n V . . - . V . . . .. .
V V . V. . V V V V . V V.
.V V V V . V V V V V. v V
.. V V . V V VV V . V
VV V . .. V. V . V
V .. V . . .
.. .V. . .V . V . V V
. v VV .. V a V
V . .V V V . ...V . . V . V
V V V V V ._ . .
V .. .u .- . V .
V V .V V V. V
V V. .VV. V . .
V V V V . . VV V
V V. . . . . V V V
V. V V V V V. V V V V . V V . V
V V V .
V . . a
. V . V .
.V .
V V , . .
V . . . .. V n
V . V V. V V.. V
. V V V V V .
V V V I .v 4
. »V 7 VV V V
V . V
V. . V V Vy . . V
. V .V
V. ,
V . V V VV , V
. V V V . VVV

 

 

 

Q

THESlS

IV

1" . uV
' I.)
5: it ‘

/

 

lllllllllllllwllllzlwlylllllllllgllll

 

LIBRARY

Michigan State
Unlversity

 

 

 

This is to certify that the

thesis entitled

HEMOSTATIC EFFECTS OF ENDOTOXIN ON ENDOTHELIAL CELLS

presented by
Jing Zuo

has been accepted towards fulfillment
of the requirements for

MS Medical Technology

degree in

@m /
v_‘ {,- v 63am}

Major professor

 

 

Date 11/3199

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

PLACE IN RETURN Box to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

moo chIM.p65-p.t4

 

HEMOSTATIC EFFECTS OF ENDOTOXIN ON ENDOTHELIAL CELLS
BY

J ing Zuo

A THESIS
Submitted to
Michigan State University
In partial fulfillment of the requirements
For the degree of

MASTER OF SCIENCE

Department of Medical Technology

1999

ABSTRACT
HEMOSTATIC EFFECTS OF ENDOTONXIN ON ENDOTHELIAL CELLS
BY

J ing Zuo

Endotoxin research started about a century ago and has attracted the attention of many
investigators. Our understanding of many pivotal mechanisms in the biology of endotoxin
has increased dramatically during the past ten years. This includes the notion that specific
endotoxin receptors exist. Complexes of endotoxin and endotoxin (LPS)-binding protein
are recognized by CD14 on myeloid cells, and this clearly constitutes an important
endotoxin recognition pathway. Endotoxin signal transduction in myeloid cells involves a
sphingomyelin pathway and a mitogen-activated protein kinase (MAP kinase) pathway.
Exposure to endotoxin often results in septic shock, which is complicated by activation of
coagulation. The human coagulation system can be activated via two pathways, extrinsic
and intrinsic route. It has long been thought that endotoxin-induced coagulation results
from the contact system (intrinsic pathway). However identification of tissue factor and its
precise molecular and biological characterization has dramatically altered this view. It is
widely accepted that endotoxin activates coagulation by perturbing endothelial cells, which
then produce tissue factor to activate the extrinsic pathway of coagulation. Meanwhile
endotoxin-perturbed endothelial cells also increase their secretion of plasminogen activator
inhibitor 1(PAI-1), which inhibits fibrinolysis. Activation of coagulation and suppression
of fibrinolysis result in microthrombosis, which contributes significantly to

pathophysiology of Gram-negative bacterial sepsis and endotoxin shock.

ACKNOWLEDGEMENTS

I wish to thank the many individuals who have made this thesis possible. I especially want
to thank my major professor Dr. Gerald Davis for his wise and detailed advice. He also
provided encouragement and direction along the way. I would like to thank my committee
Dr. Robert Roth, Dr. Rawle I. Hollingsworth, and Dr. David P. Thome. Dr. Roth and Dr.
Hollingsworth both gave me excellent guidance and advice. Their efforts greatly
improved the clarity and accuracy of this thesis. Dr. Thorne, who participated in the
original development of my thesis, helped me enormously not only in this work but

throughout my school years.

Finally, I want to thank my husband Dawei Qian for his patience, support and
encouragement during the many long weekends and evenings that went into the writing

and reading.

Thank you all.

TABLE OF CONTENTS

LIST OF FIGURES ............................................................................ iii
CHAPTER
1. Review of endotoxin ............................................................ 1
1.1 Historical perspective ................................................. 2
1.2 Structure of endotoxin ................................................. 2
1.2.1 The O-antigen structure and oligosaccharide core 3
1.2.2 Chemical structure of lipid A ............................... 4
1.2.3 Physical structure of lipid A ............................... 10
1.2.4 Structure-activity relationships of lipid A .............. 12
1.3 Mechanism of endotoxin action ................................... 14
1.3.1 Receptor-mediated mechanism ........................... 14
1.3.1.1 Role of serum protein ......... . .................... 16
1.3.1.2 Receptor of LPS - CD14 and Toll-like receptor
............ 17
1.3.2 Nonspecific hydrophobic interaction .................... 20
1.4 Transmembrane signaling mechanism .......................... 23
1.4.1 Signal transduction in myeloid cells ...................... 23
1.4.2 Signal transduction in endothelial cells .................. 32
2. Functions of endothelial cells in hemostasis ............................ .. 35
2.1 Procoagulant properties of endothelium ........................ .. 35
2.1.1 Anticoagulant properties of endothelium ............. .. 36
2.1.2 Synthesis of heparan sulfate proteoglycans ............. 37
2.1.3 Synthesis of dermatan sulfate proteoglycans .......... .. 37
2.1.4 Protein Cl protein S/ thrombomodulin system ........ .. 38
2.1.5 Tissue factor pathway inhibitor (TFPI) ................. 41
2.2 Fibrinolytic properties of endothelium .......................... .. 42
3. LPS-perturbed endothelial cells in hemostasis ........................... 44
3.1 Expression of tissue factor and TFPI ............................. 44
3.2 Suppression of thrombomodulin ................................... 46
3.3 Up-regulation of tissue plasminogen activator inhibitor 1 .. .. 46
4. Summary and final remarks ................................................. 48
4.1 Summary ............................................................... 48
4.2 Final remarks .......................................................... 51
BIBLIOGRAPHY .............................................................................. 54

ii

LIST OF FIGURES

1. Chemical structure of E. Coli lipopolysaccharide

2. The structure of RsDPLA is different from E. Coli DPLA
by one less acyl groups (five versus six) and the presence
of a 3-Keto moiety and a double bond.

3. Correlation between the molecular shape of lipid A and the
three-dimensional suprarnolecular structure formed.

4. Interactions of LPS leading to cell activation or LPS
clearance.

5. Proposed model of cell activation by endotoxin.

6. Signaling role of MAP kinase in LPS-stimulated
macrophages.

7. Proposed mechanism for signal transduction initiated by
LPS, TNF-Ot, and IL-1.

8. Structural comparison of ceramide and lipopolysaccharide
(LPS).

9. The protein C-protein S system.

iii

.................. 29

.................. 31

................... 40

Haemostatic Effects of Endotoxin on Endothelial Cells

Endotoxin, a component of the Gram-negative bacterial cell wall, is a lethal molecule
inducing irreversible septic shock and death (1,2). Endotoxin stimulates many kinds of
host cells to release mediators that trigger septic shock (3,4). Endothelial cells (ECs) are
one of these cells. Activation and stimulation of ECs by endotoxin contributes
significantly to the vascular complications of septic shock, such as hypotension and
disseminated intravascular coagulation (DIC) (5). In these conditions, ECs is disturbed
following entry of pathogenic bacterial compounds into the blood stream. These bacterial
compounds bind to ECs with the initiation of cellular responses after signaling.
Elucidating the mechanisms of EC activation may not only improve our understanding of
certain diseases, but also contribute to the development of new therapeutic intervention

strategies.

Chapter 1 Review of Endotoxin

1.1 Historical Perspective

In the 18708, Robert Koch, a German microbiologist, proposed that each infectious
disease was derived from a specific microorganism. After his proposal, many investigators
showed that bacteria often made people sick by producing toxins. The first toxin isolated
was exotoxin, secreted by gram positive and gram negative bacteria. Exotoxins are
generally heat labile proteins and can be neutralized by antitoxin. They have restricted
biological activities and affect specific cells or tissues. In 1892, Richard Pfeiffer, one of
Koch’s students, described poisons that did not fit the exotoxin profile. Assuming these
substances were sequestered within bacteria, he named them as endotoxins. Pfeiffer’s
endotoxins are now known to reside on the surface of bacteria, not in the interior.
Endotoxins are produced only by gram negative bacteria. They are a major component of
gram negative bacterial envelope and are released when bacteria are killed or lysed.
Endotoxins are heat stable and are mainly composed of polysaccharide and lipid. They
were therefore termed lipopolysaccharide (LPS). Today endotoxin and LPS are used
synonymously for the same molecule. Endotoxins are not neutralized by either antitoxin or
antibodies. They have diverse activities and affect many cells and tissue types.
Endotoxins have been implicated as a major agent responsible for the pathophysiological
manifestations of gram negative bacterial sepsis.

1.2 Structure of endotoxin

Over the past two decades significant progress has been made in understanding the

2

chemistry (primary structure) and physics (mono and multimolecular conformation) of the
endotoxin molecules. Knowledge of chemical and physical LPS structure also allows the

establishment of a structure-activity relationship of LPS.

When they are released from the cell wall, they exist as high molecular weight
complexes bound with various amounts of proteins and phospholipids (6). Isolated,
purified, protein-free LPS is responsible for endotoxin's biological activities. LPS is an
amphipathic molecule consisting of a hydrophilic region (o-antigen structure and
oligosaccharide core) and a Endotoxins are produced only by gram negative bacteria and

are embedded in the cell wall. hydrophobic region (lipid A).

1.2.1 The o-antigen structure and oligosaccharide core

LPS Molecule consists of three regions, the o—antigen region, oligosaccharide core and
lipid A.

The o-antigen structure is a polysaccharide consisting of a complex, repeating
oligosaccharide whose composition varies from different species and even strains of the
same species. It is the o-polysaccharide that is responsible for determining the bacterial

o—antigen serotypes (7,8).

The oligosaccharide core serves as a linker between the O-specific polysaccharide and

lipid A and could be subdivided into an inner and an outer portion (Figure I A ). The

outer core contains the common D-glucose, D-galactose and N -acetyl-D—glucosamine.
The inner core is composed of characteristic LPS-specific components : heptose and 2-
keto- 3-deoxy-D-manno—octulosonic acid (KDO). The heptose and KDO residues are
substituted with charged groups such as phosphate, pyrophosphate, 2-
arninoethylphosphate and 2-aminoethylpyrophosphate, leading to an agglomeration of
charged residues in the inner part of the core region. The high density of negatively
charged residues is likely of physiological significance. The KDO-containing inner core
has been considered to be essential for biological activity of LPS, microbial growth and
multiplication, since bacteria with a defect in KDO biosynthesis are not viable and some
LPS consists of only KDO and lipid A. KDO appears not to be present in mammalian
cells and plays a vital role for bacteria. The structural variation within the core region is

low. For example, only one core structure is found in all of the salmonella serotypes (8).

1.2.2 Chemical structure of lipid A

The structure of lipid A is highly conserved in LPS of diverse gram negative bacteria.
Endotoxin’s biological activities are associated with LPS. More specifically, lipid A is
the active stimulatory or inflammatory moiety of LPS and is responsible for the

activation of many kinds of host cells to initiate septic shock.

Lipid A consists mainly of a disaccharide glucosamine backbone with two phosphates

and long chain fatty acids (Figure l B ). For instance, the lipid A of Escherichia coli

consists of a B-l', 6-linked glucosamine disaccharide backbone with phosphate ester
substituted at the 1 and 4' carbon. On the disaccharide backbone position 3', 2', 3, 2
carry four fatty acids - 3-hydroxytetradecanoic acids [l4:(3-OH)]. The 3-hydroxyl
groups of the two fatty acids bound to the 2', 3' positions are acylated by dodecanoic
acid [12:0] and tetradecanoic acid [14:0] respectively (Figure 2 B). In the following,
the structures of lipid A derived from three different nonenterobacterial LPS are
discussed in relation to the structure of E. coli. These examples are selected in order to
illustrate some of the lipid A components which may be important for biological

activities.

Figure 1.A

Figure 1.8

Schematic structure of E. Coli lipopolysaccharide. Lipolysaccharide
(LPS) can be distinguished into three regions: O-specific chain, core,
and lipid A.

O-specific chain consists of a complex and repeating oligosaccharide.
Core can sub—divided into outer core and inner core.

Lipid A is responsible for the endotoxic activities. Its structure is shown
and explained in Figure l.B.

Chemical structure of lipid A. Lipid A consists mainly of a disaccharide
glucosarnine backbone with two phosphates and long chain fatty acids.

 

L—-—'[O-Speciiie Chaifi —“ W'

 

 

 

 

 

‘ {Poiyeaccharitie} —'|—-—’
Q: Monouccharide o: Phosphate vvvv : Fatty Acid

8
m

A/ \/% a

00 P-O

 

 

Figure 1. Chemical Structure of E. Coli lipopolysaccharide from Ulevitch, R.J.
Advance in Immunology 53:270, 1992

 

 

 

("1) (R3) (R4)
(8,)

Figure 2. The structure of RsDPLA (Figure 2A) differents from E. Coli DPLA
(Figure 2B) by one less acyl groups (five versus six) and the presence of a 3-
Keto moiety (R2 in Figure 2A) and a double bond (R4 in figure 2A).

Figure 2A. Structure of R. Sphaeroides lipid A.
Figure 2B. Structure of E. Coli lipid A.

Rhizobium meliloti 2011 (R. meliloti) and Rhizobium Ieguminosarum bv. trifolii
ANU 843 (R. trifolii) are gram negative bacteria in soil which form nitrogen-fixing
symbiotic relationships with legume plants. The lipid A of these bacteria are known to
possess some of the classical endotoxin sequences, but they have different lipid A
structures. Lipid A of R. meliloti contains a classical glucosamine disaccharide
backbone but carries two sulfate groups rather than the phosphates at positions 1 and 4’.
The backbone of R.trifolii lipid A is a disaccharide composed of glucosamine as the
reducing end sugar, linked glucosidically to the anomeric position of galacturonic acid
and contains carboxylic groups instead of phosphates at positions 1, 4’. Both lipid A’s
carry a unique and very long chain fatty acid, 27-hydroxyoctacosanoid acid [27-0H-28:]
(9,10). Rhodobater Sphaeroides (R. Sphaeroides) is a photosynthetic bacterium.
Lipid A from its LPS has little activity on human cells(11). Actually, it can effectively
block other LPS activation on host cells. This nontoxic lipid A of R. Sphaeroides
possesses a phosphorylated glucosamine disaccharide backbone that is identical in
structure to that found in E. coli. The small differences between E. coli and R.
Sphaeroides lipid A are limited to the fatty acids(figure l). R. Sphaeroides lipid A has
a hydroxy decanoic acid at positions 3'(Rr) and 3(R3), a delta-
tetradecanoyloxytetradecanoic acid at position 2(R4), and ketotetradecanoic acid at
position 2’(R2). The differences noted are one less acyl group (five versus six) and the
presence of a 3-keto moiety and a double-bond in R. Sphaeroides lipid A compared

with E. coli lipid. The chains of fatty acids in R.sphaeroieds lipid A are also shorter

than that of E.coli lipid (Figure 2 A ).

1.2.3 Physical structure of lipid A

Lipid A is amphipathic and therefore aggregates in aqueous medium to three-
dimensional supramolecular structure. The three-dimensional structure adopted by
aggregates of free lipid A or LPS depends on the primary chemical structure and
ambient conditions. In more recent works, analytical physical techniques such as small-
angle X-ray and neutron diffraction for the determination of long-range order
(supramolecular structure) and wide-angle X-ray diffraction for the determination of
short-range order (arrangement of the acyl chains) were mainly used. Many studies
agree upon the existence of lamellar and nonlamellar structures for LPS and free lipid
A. According to present understanding, biologically active lipid A adopts, at
physiological ambient condition [37°C, pH7, presence of Mg-I-i-, high water content
(90%)], exclusively nonlamellar structures (12). These nonlamellar structures are either
cubic or hexagonal (Figure 3). In contrast, biologically nonactive lipid A adopts lamellar
structures (13). This suggests that endotoxicity can be determined by a defined
conformation of lipid A. More likely, endotoxicity is expressed by individual endotoxin
or lipid A molecules possessing a conformation, which at higher concentration, leads to

the observed three-dimensional nonlamellar structures (14).

10

Lomellcr {Li

Cubic (0i

Hexagonal (Hnl

 

Figure 3. Correlation between the molecular shape of lipid A and the three-dimensional
supramolecular structure formed.

Lipid A is amphipathic and therefore aggregates in aquous sOlution to three dimensional
supramulecular structure. Aggregates of lipid A exist in lamellar(L) and non-lamellar

structures. Biologically active lipid A adopts exclusively non-lamellar structures.
These non-lamrllar structures are either cubic(Q) or hexagonal(HII).

From Seydel U. Irnmunobiology. 187-193, 1993

ll

1.2.4 Structure-activity relationships of Lipid A

While it is known that lipid A is responsible for endotoxic activities, it is still unclear

which complex of the molecule is required.

For a long time, it has been postulated that three major structural features of
lipopolysaccharide are necessary for induction of endotoxin activities. These features
are the presence of two D-glucosamine residues (which are B-l', 6-linked), two
phosphoryl groups and at least five, but not more than six fatty acids including one or
two 3-acyloxyacyl groups in the defined location as it is present in the classical
endotoxin E. coli lipid A. The toxicity of molecules lacking only one component, or
with a different distribution of components is either less or no toxic (15). It seems that
the number of acyl groups (six fatty acids being optimal) and the chain length of fatty
acids are two decisional factors for endotoxic activities. Charged functional groups in
the polar region of the lipid A moiety also appear to be essential for its full effectiveness
(15). Furthermore, one should consider that lipid A is amphipathic in aqueous solutions
and forms a supramolecular structure (16). This supramolecular structure of Lipid A is
associated with the activities of Lipid A. Actually, a highly specific three-dimensional
arrangement of hydrophobic hydrocarbon chains in lipid A is a prerequisite for the

induction of endotoxicity (16).

l2

Based on these rules of structure-activity relationship, the reason that R. Sphaeroides is
nontoxic may relate to its unusual fatty acid composition and few fatty acids. The lipid A
molecules from R. molilote and R. trifolii do not contain two phosphate groups and the
R. Trifolii lipid A carries only five fatty acids. However both behave like a typical
endotoxin molecule. They produce a positive local Swartzman reaction in rabbits and a
positive response in the lirnulus amebocyte lysate assay (17), and they stimulate the
release of plasminogen activator inhibitor-1 (PAI-l) from human umbilical vein
endothelial cells. Despite the lack of phosphate groups and the fewer number of fatty
acids, the endotoxicity of these two lipid A molecules is comparable to or higher than that

of E. coli lipid A.

If the structure of the lipid A is important for its activity, then how does one explain the
finding that these two lipid A molecules, which differ from E. coli, are still toxic? A
possible answer is that they both carry a unique long fatty acid chain, 27-
hydroxyoctacosanoic acid [28:0(27-0H)]. The 27-hydroctacosanoic acid is twice the
length of usual 14:0 (3-0H) acid. After Lipid A fuses with the plasma membrane of the
host, the 27-hydroctacosanoic acid may stretch through the entire outer membrane.
Hollingsworth et al. have suggested this structure could result in a more stable
anchorage of LPS with an increased stability of the bilayer (17). They further proposed
that signal transduction was induced by an increase in rigidity or stability of the
membrane (17). They also noticed that both lipid As contain two carboxyl or sulfate

groups, which seem to function as surrogates of the phosphates to produce a negatively

l3

charged backbone. Some investigators believe this negative charge is essential for the
endotoxin to bind to positively charged groups such as proteins on the host cell

membrane (18).

1.3 Mechanism of endotoxin action

The mechanism of LPS interaction with host cells has attracted medical researchers for
many years. There are at least two proposed mechanisms that explain how LPS stimulates
host cells. One mechanism involves the LPS stimulation of host cells via a specific
membrane receptor that initiates transmembrane signaling following receptor
occupancy(l9). Another is that LPS stimulates host cells via the “nonspecific
hydrophobic interaction” between lipid A and the phospholipid bilayer of the cell
membrane, resulting in membrane structure changes to initiate transmembrane signaling

(20).

1.3.1 Receptor-mediated mechanism

A number of experimental observations have established the existence of a receptor for
LPS (21). A major advance in our understanding of mammalian response to LPS is that
CD14, a macrophage/polymorphonuclear leukocyte cell surface differentiation antigen,
plays a role in recognition of LPS (22). The model is that LPS initially binds an acute-

phase serum protein known as lipopolysaccharide binding protein (LBP) (23,24) to form a

14

LPS-LBP complex. The LPS-LBP complex then docks with a receptor (CD14) on the

surface of a cell membrane (23) (figure 4).

15

LBP + LPS

L3?

L8? - LPS /

 

 

 

 

 

sCDl4
LBP - LPS 1'
LPS - CDl4
v (0.0, s
”ILL” 53 mCDi4 Q3.
Q
3 Receptor
Mtb/PMN / ?
. . Co-receptor ’
Lipoprotein ., ..-—---
Endothelial Cell
——-> Epithelial Cell

 

Etc.

LPS Uptake l

TNF
lL-l ——-J

 

etc.

IL - 10

IL] - RA
etc.

Anti-inflammatory Pro-Inflammatory

A
V

V

Figure 4. Interactions of LPS leading to cell activation or LPS clearance.
From Han J. Progress in Clinical and Biological Research. Vol. 397:158

16

1.3.1.1 Role of serum protein

LPS is an amphipathic molecule. It is poorly soluble in aqueous solution. The majority of
the circulating LPS is bound to serum proteins. The two best-characterized serum proteins
are high-density lipoprotein (HDL) and LPS-binding protein (LBP). HDL is known to
partially neutralize the toxic effects of LPS (25). LBP is a 60KD glycoprotein synthesized
in hepatocytes. It is present in normal serum at < 0.5 ug/rnl but can increase up to 50ug/rnl
after an acute-phase response (26). LBP binds with high affinity to the lipid A of LPS (27)
to form a LPS-LBP complex. The main function of LBP is to transfer LPS to either
membrane CD14 (28) or soluble CD14 (29). Another function of LBP is to opsonize LPS-
bearing particles such as intact gram negative bacteria (30). The resulting LPS-bearing
particles can in turn be avidly bound by monocytes, macrophages and neutrophils. The
opsonic activity of LBP suggests that it may function like the classical opsonins, Ig G and
C3, in the clearance of bacteria. LBP has two distinct binding sites for LPS and CD14.
The amino-terminus of LBP is responsible for the binding to LPS while the carboxyl-
terminal portion of the molecule involves the CD14 interaction (31). The DNA sequence
of LBP shows high homology to that of bacterial permeability increasing protein (BPI) and
cholesterol ester transport protein (CETP) (32). It appears that LBP facilitates the
interaction of LPS with CD14 by acting like a lipid transfer protein. This is consistent with
its homology to CETP, which redistributes cholesterol ester among lipoprotein particles

(32).

17

1.3.1.2 Receptor of LPS - CD14 and Toll-like receptor

Several proteins are considered as candidates for a receptor for LPS (33,34,35). However
only one protein, CD14, has been proven to be a cellular receptor (19,36). CD14 exists in
two forms, a membrane form on the cell surface and a soluble form in serum. The
membrane-form, CD14 molecule (mCDl4) constitutes a 55KD glycoprotein anchored to a
myeloid cell membrane such as monocyte/macrophage or polymorphonuclear leukocyte
via a glycerophosphatidylinositol (GPI) anchor (37). mCDl4 is not merely a receptor for
LPS. Recently Pugin et al proposed a “ CD14 is a pattern recognition receptor”(38). They
found that mCDl4 was a common receptor of myeloid cells for a wide variety of bacterial
envelope components such as LPS from gram-negative bacteria, lipoarabinomannan
(LAM) from Mycobacterium tuberculosis and molecules from gram-positive cell walls.
Interactions of these components with myeloid cells through mCD14 can lead to cell
activation. However mCD14 does not contain a transmembrane domain, it is unlikely that
it could transmit a signal from the surface to the interior of the cells. Ulevitch and Tobias
suggested that mCD14 did not play a primary role in signal transduction but it might be
the principal ligand-binding unit of a membrane-bound receptor complex (39). It was
hypothesized that LPS initially bound LBP to form a LPS-LBP complex, which shuttled
LPS to mCDl4 on membrane of myeloid cells. The LPS-mCD14 complexed with an
additional membrane component (second receptor) could initiate transmembrane signaling

(figure 4). Identification of second receptor is crucial for a full understanding of LPS

activation.

l8

Recent studies have established that the Toll-like receptor 2 (TLR2) may serves as a

secondary receptor to mediate LPS-induced transduction signal (40).

The generation of dorsoventral pattern in the early embryo of Drosophila relies on the
expression of 12 maternal-effect genes termed the dosal groups. One of these genes is Toll
gene (41). The product of toll gene is a transmembrane protein consisting of extracellular
and cytoplasmic domains. The extracelluar section of Toll protein has sequences with
leucine-rich repeats and the cytoplasmic domain is homology to human lL-l receptor (42).
Activation of both Toll protein and IL-1 receptor leads to induce NF-KB pathway (42,43).
Recently, five human Toll-like receptors (TLR 1-5) are identified (44). One human TLR,
TLR2, is found in all myeloid cells. The expression of TLR2 mRNA in myeloid cells is
upregulated after stimulation by LPS, paralleling to mCD14 mRNA expression (40,45).
Yang et al further showed that TLR2 is a LPS-signalling receptor, which transmits the
information across the plasma membrane, resulting in activation of NF-xB pathway. They
isolated a clone of human embryonic kindny 293 cells that were transfected with TLR 2
and tested the response of these cells to E.coli LPS or to a combination of E.coli LPS and
LBP by monitoring expression of E-selectin gene through NF-xB pathway. They found
that treatment of both LPS and LBP caused significant E-selectin gene activation. But
parental 293 cells that did not express TLR 2 had no response to a combination of LPS
and LBP stimulation. They further discovered that LPS and LBP induced NF-kB activity

in TLR 2-expressing cells with kinetics similar to those in myeloid and non-myeloid cells.

19

Both mCD14 and sCDl4 enhanced the TLR2-mediated LPS responsiveness. The
intracellular domain of TLR2 with homology to a portion of the IL-1 receptor is important
in signal transduction, since the truncation of carboxyl terminus in either 13 or 141 amino
acids diminishes the response to LPS (40). Although identification of TLR2 as second
receptor makes considerable progress in defining signaling mechanism, how mCD14 and

TLR2 interacting and mediating cell activation remains to be determined.

CD14 also exists in soluble form (sCDl4). Several types of sCD14 (48,53,55-kDa) are
present in normal serum at concentration of about 2.6ug/ml. (46). The different sCDl4s
can result from either shedding of membrane-bound mCD14 or from cellular production
of a GPI-free CD14 (47,48). SCD14 plays a key role in the LPS activation of cells that do
not express mCD14 on their surfaces, such as endothelial and epithelial cells. LPS-
induced endothelial cell activation requires serum containing a soluble form of CD14
(49,50,). Depleting sCD14 from serum diminishes LPS effects on endothelial cells, while
reconstitution of sCD14-depleted serum with imrnunopurified sCD14 restores LPS
activity on endothelial cells (51,52). Although the requirement for sCD14 in LPS-induced
activation of endothelial cells is established, little is known about the interaction of sCDl4

with endothelial cells and LPS.

Many experiments have confirmed that the LBP/CD14 pathway contributes to monocyte/
macrophage and polymorphonuclear leukocyte stimulation by LPS. However, the

mechanisms involved in LPS- induced endothelial cell activation are not well understood.

20

Recently two different pathways (direct pathway and indirect pathway) for endothelial
cell activation by LPS have been proposed by Ulevitch and Tobias (53). In the direct
pathway the LPS-binding protein binds LPS and transfers it to the soluble form of CD14
(sCDl4). The LPS-sCD14 complex in turn activates endothelial cells via an unknown
surface receptor (figure 4). The indirect pathway involved stimulation of myeloid cells by
LPS. The stimulated cells secrete proinflammatory mediators such as TNF and lL-l ,
which in turn efficiently activate endothelial cells (54). LPS, TNF and IL—1 are three very
different molecules that are able to evoke an identical response in endothelial cells. High-
affinity receptors for TNF and IL-1 have been identified on human endothelial cells

(55,56), but a receptor for LPS has not yet been identified.

1.3.2 Nonspecific hydrophobic interaction

In addition to these receptor-mediated processes, Schromm and his colleagues suggested
that a CD14-independent, direct activation of host cells might be present, particularly at
high endotoxin concentrations in the range of jig/ml (57). The CD14-independent
activation mechanism is probably based on nonspecific hydrophobic interaction of the
endotoxin molecule with the phospholipid bilayer of host cells when high concentrations
of endotoxin are present. LPS has a high affinity for the phospholipid monolayer and
bilayer (58,59). It has been demonstrated that LPS binds and inserts into the lipid bilayer
(60,61,62). Jacob et al proposed in 1986 that LPS interacted with cell membranes in a two

- step process (63). The first step, adherence, was reversible, temperature independent, and

21

inhibited by polyions. The second step, coalescence, was the incorporation of LPS into
the cell membrane bilayer. This process is irreversible, temperature dependent, and can not
be inhibited by polyions. Endotoxin, as an amphipathic molecule, tends to form clusters
in an aqueous environment. Thus, under normal conditions endotoxins should present in
aggregated forms. Evidence indicates that endotoxin monomers are biologically more
active than aggregates (64). Shromm and his colleague showed that LBP broke down LPS
aggregates and transported LPS to a host cell, then the small units of LPS inserted into
phospholipid bilayer. These processes did not require CD14 or any other membrane-
associated protein. The intercalation of endotoxin into phospholipid bilayer was not
enough for cell activation. They proposed a model in which the phospholipid bilayer
contained a transmembrane signal transducin g protein activated by binding of endotoxin.
This binding was facilitated via hydrogen bonds that required the presence of a sufficient
number of hydroxy fatty acids in the LPS molecules. It is noteworthy that only LPS with
the conical shaped lipid A could activate the signal protein. Lipid A with a cylindrical
conformation could not do it since it easily formed lamellar aggregates (65). After binding
of LPS to transmembrane signal transducting protein, the signaling cascade would be

provoked (Figure 5).

22

 

  
 
   
  

  

Signal
transducing
protein

 
 
 

no
activation activation .

Figure 5. Proposed model of cell activation by endotoxin. The endotoxin molecules binds
with their lipid A moiety to a transmembrane signal transducing molecule. Binding is
facilitated via hydrogen bonding. This requires the existence of hydroxy fatty acids in the
lipid A part. A further prerequisite for activation is a particular conformation of lipid A.

From Ulrich Seydel. FEBS Letters 399:267-271. 1996

23

1.4 Transmembrane signaling mechanism

The downstream pathway of signal transduction by which LPS causes cell activation is
less well understood. Transmembrane signaling mechanisms such as calcium fluxes or
intracellular pH changes do not appear to be involved in LPS stimulation via the
LBP/CD14 pathway (67). It is now recognized that enzyme mediated phosphorylation and
dephosphorylation of cellular proteins may be the principal mechanism by which an

external signal can regulate intracellular responses.
1.4.1 Signal transduction in myeloid cells

Two reports by Weinstein and his colleagues have shown that E. coli and Salmonella
Minnesota LPS induced tyrosine phosphorylation of several proteins in macrophages
(68,69). Isoforms of the mitogen activated protein kinases (i.e.MAP kinase 1 and MAP
kinase 2) were prominent in these proteins. Treatment of macrophages with a tyrosine
kinase inhibitor, herbimycin A, inhibited LPS-induced tyrosine phosphorylation and
blocked activation of macrophages by LPS. They proposed that increased protein tyrosine
phosphorylation was a rapid, LPS-mediated signal transduction event. Several other
groups extended these observations by a using mouse B-lymphoma cell line 70Z/3
(70,71). The this lymphoma cell line is LPS-responsive, even thought it normally does not

express CD14 on its surface. Legrand et al demonstrated that wild- type 702/3 cells, which

24

did not express CD14, did not undergo any detectable tyrosine kinase—mediated signaling
induced by LPS stimulation. In contrast, 702/3 cells transfected with human CD14
showed that LPS rapidly induced the tyrosine phosphorylation of a 4 lkDa protein.

Pretreatment of 702/3—hCD14 cells with antibody to CD14 inhibited tyrosine

phosphorylation of a 41 kDa protein (MAP kinase 2). This report suggested that

induction of protein tyrosine phosphorylation by LPS was CDlMependent (71).

The MAP kinase pathway has been studied extensively in yeast and mammals. MAP
kinases are proline-directed, serine/threonine protein kinases which currently comprise
three major subfamilies, namely the classical MAP kinase ( MAP kinase 1[ERK 1]/ MAP
kinase 2[ERK 2]), the c-Jun N—terminal kinase (JNK) and P38 MAP kinase (72). These
distinct sets of MAP kinases can be activated by a variety of extracellular stimuli in many
types of cells. In human neutrophils, LPS can activate only P38 MAP kinase (72,73). In
murine macrophages, LPS can activate all three of the MAP kinases (73, 74). Some
investigators suggested that LPS signal transduction in macrophages might occur via

activation of multiple MAP kinase pathways (75) (Figure 6).

It has long been recognized that IL-1, TNF and LPS all initiate similar responses in cells.
More resent studies have shown that these agents induce a nearly identical pattern of early
signaling events, including activation and phosphorylation of MAP kinases which serve as
intermediates in numerous signaling cascades from the cell surface (76). TNF, lL-l and
LPS also activate NF-kB, a factor that promotes transcription of a large number of genes.

25

NF-kB exists in the cytoplasm of many cells complexed to an inhibitor, lkB. Treatment of
cells with TNF, lL-l and LPS leads to proteolytic degradation of lkB (77) and the release
of NF-kB. NF—kB then translocates to the nucleus where it binds to its cognate DNA

sequence on responsive genes.

26

 

Figure 6. Signaling role of MAP kinase in LPS-stimulated macrophages. LPS, acting via
CD14 and a putative low affinity receptor (X) activates one or more intracellular protein
tyrosine kinase which, through a series of events, leads to activation of the three types of
mammalian MAP kinases (Erk V2, IN K, p38) and to activation of NF-KB. From a variety
of studies, TNF-a production appears to depend at least upon NF-ch, Erk 1/2 and p38.

From Weinstein SL. Progress in Clinical and Biological Research. Vol. 397: 131, 1996

27

Many studies suggested that IL-1 and TNF both utilize a sphingomyelin pathway and a
MAP kinase pathway as their signal transduction (78). The sphingomyelin pathway is
initiated by hydrolysis of sphingomyelin to ceramide by the action of sphingomyelinase
(SMase) (79). Ceramide functions as the second messenger molecule and stimulates a
membrane-bound, serine/threonine kinase, termed ceramide-activated protein kinase
(CAPK). Some investigators have demonstrated that CAPK in turn phosphorylates Raf 1
at threonine 269 (80). Raf l, a product of the proto-oncogene c-raf, is a serine threonine
kinase that phosphorylates and activates MAP kinase kinases (MEK). Raf also binds,
phosphorylates and inactivates IxB, which releases NF-KB (Figure 7). MEK activates
MAP kinases by phosphorylating adjacent threonine and tyrosine residues. MAP kinases
transduce the signal to the nucleus where they phosphorylate and thereby activate the
transcription factor EH(-l, one of the nuclear targets of MAP kinases. The activated BIK-
l with another transcription factor serum response factor (SRF) binds to serum response
element (SRE), a significant component of the c-fos promoter, to form a ternary complex
that regulates c-fos gene induction (76). In addition, MAP kinases phosphorylate and
activate cytosolic phospholipase A2 (PLA2), which leads to the production of

inflammatory mediators such as leukotrienes and platelet-activation factor(81, 82).

The similarity of actions of TNF, IL] and LPS suggests that some effects of LPS may
also be mediated through the ceramide-activated protein cascade. LPS stimulates cells by

binding to the mCDl4 of myeloid cells, followed by activation of MAP kinase (79,83) and

28

NF-kB (84). LPS was found to activate CAPK in a manner that was CD14 dependent and
enhanced by LBP but independent of sphingomyelinase activation and ceramide
production (85). Recently a structural similarity between the reducing glucosanrine of lipid
A and ceramide was revealed (85,86) (figure 8). Because LPS shows strong functional
and structural resemblance to ceramide (87), some investigators suggested that endotoxin
may stimulate cells directly by mimicking ceramide as a second messenger, via its
interaction with CAPK, rather than through the activation of SMase (85).Barber et al.
further demonstrated that sphingomyelin-pathway was only a subset of LPS-induced

signaling pathway because only some effects of LPS could be induced by ceramide (88).

Figure 7. Proposed mechanism for signal transduction initiated by LPS, TNF-a and IL- 1.
LPS, TNF-a and IL-1 activate sphingomyelinase (SMase) in the plasma membrane
resulting in hydrolysis of sphingomyelin (SM) to ceramide. Ceramide acts as a second
messenger and stimulates ceramide-activated protein kinase (CAPK). Immediate down-
stream response include activation of MAP kinases and nuclear translocation of NF-xB.
Raf kinase may play a prominent role in these processes.

30

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5.... .. .. ..... .. ...-. .. V. V... ~ $118.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

IKB ePLA
\VVVV.
acid
a
I
III-l
SRF /
SRF

 

 

 

 

 

 

 

 

Figure 7. Proposed mechanism for signal transduction initiated by LPS, TNF-a and lL-I.

Modified from Kolesnick R. Cell 77:325, 199

31

 

 

 

 

 

/O
' 5’ °” 0H
1 Od-IP\OH 3 N 1
0 on o
0 OH
OH
0 O
LPS"WAP°“‘°") Ceramide

Figure 8. Structural comparison of ceramide and lipopolysaccharide (LPS). LPS
comprises two acylated, phosphorylated glucosamine (GlcN) residues designated GlcN-I
and GlcN-II. Naturally occuning LPS also bears a large and variable amount of
polysaccharide attached to carbon 5 of GlcN-II. Since this carbohydrate is not necessary
for inducing cellular responses, it is omitted from this figure; only the lipid A portion of
the LPS molecule is shown. The boxed portion of LPS bears string structural resemblance
to ceramide.

From Wright SD. Immunology Today 162297, 1995.

32

1.4.2 Signal transduction in endothelial cells

A signaling pathway that follows occupation of LPS receptor mCD14 has been
characterized in myeloid cells by demonstrating the tyrosine phosphorylation of a number
of target proteins. However, cellular molecules initiating signaling in ECs is poorly
understood and little information is available on the signal transduction pathway triggered
by LPS in this particular cell type. So far, LPS-mediated activation of ECs and signal
transduction has two main points. One is that protein kinase C (PKC) may be involved in
stimulation of ECs by LPS. The other is that tyrosine phosphorylation triggers signal

transduction in ECs.

Evidence indicates that transmembrane protein kinase C may be involved in stimulation of
endothelial cells by LPS (89-91). Prydz et a] demonstrated that two different inhibitors of
protein kinase C (PKC), H7 and staurosporine (STS), could prevent the LPS- induced
expression of tissue factor(TF) by human umbilical vein endothelial cells (HUVEC).
Human endothelial cells exposed to LPS, IL-1 or TNF-a acquire a cell surface property by
the synthesis of surface proteins such as the endothelial intercellular adhesion molecule 1
(ICAM-l) that promotes the adherence of neutrophils. Lane et al also showed the H7 and
STS blocked ICAM-l expression of HUVEC in response to LPS, IL] and TNF-a.

However, they noticed that neither H7 nor STS was a specific inhibitor of PKC. At a high

concentration STS also inhibited the CAMP / cGMP- dependent protein kinases (92). H7

33

has been shown to inhibit calmodulin activity (93). To exclude these effects of H7 and
STS as an explanation for the observation, Lane et al used K252a, an STS analogue that is
a potent inhibitor of cAMP/cGMP- dependent protein kinase and a weak inhibitor of PKC,
and W7, an analogue of H7 that is a potent inhibitor of calmodulin and a weak inhibitor
PKC. They found that K252a and W7, which were less potent inhibitors of IL-1- induced
up- regulation of ICAM-1 expression than either STS or H7. They suggested that the
surface expression of ICAM-1 on human umbilical vein endothelial cells by the
stimulation of LPS and cytokines was PKC-dependent, and not mediated by

cAMPchMP-depentent kinases or calmodulin.

However, Hissner et al found that LPS and TNF-or induced ICAM-l expression on
HUVEC were unaffected by PKC inhibition (94). This observation was in contrast with
Lane’s data (92), which suggested a PKC—dependence. HisSner suggested that the use of
different PKC-inhibitors could explaine this contradiction in findings. They used
bisindolylmaleimide GF 109203 X, a selective inhibitor of PKC (95), whereas Lane used
H7, which is an inhibitor with a broad specificity for serine-threonine kinases (96). Myers
et al also supported the conclusion that PKC is not essential for the induction of ICAM-1

by LPS and TNF-ot(97).

Meanwhile, several investigators demonstrated that MAP kinase phosphorylation is
involved in mediating LPS-induced human EC activation (98-100). It is known that

binding of LPS to mCD 14 on monocyte/macrophage induces the tyrosine phosphorylation

of several cellular proteins. A family of enzymes, the MAP kinases have been shown to be
phosphorylated rapidly on tyrosine residues following cellular activation and to be
involved in the regulation of many intracellular signaling pathways. Studies have shown
that LPS can also rapidly induced the tyrosine phosphorylation of several proteins in ECs.
Three proteins, the molecular weights of which are 44, 42 and 41 KD, were predominant
among the LPS-induced phosphoproteins. These proteins have been identified as MAP
kinase 1, MAP kinase 2 and P38, respectively. LPS-induced protein phosphorylation in
ECs was sCDl4 dependent, since pretreatment of ECs with an antiCD14 mAb inhibited
the LPS-induced tyrosine phosphorylation. Moreover, tyrosine kinase inhibitors such as
genistein partially inhibited LPS-induced kinase activity and ICAM-1 expression (100). It
seems that tyrosine phosphorylation of MAP kinase is a general pattern of LPS action
towards host cells. However, in ECs involvement of this pathway appears to be weaker
than in myeloid cells (100). Thus combined PKC and phosphorylation of MAP kinase

may mediate LPS responses in ECs (100).

Involvement of the ceramide pathway in signal transduction initiated by LPS in
macrophages has recently been suggested (85,86). It has also been reported that, in vivo
LPS-induced endothelial cell apoptosis is mediated by TNF release and subsequent
ceramide generation (101). However ceramide elevation appears to be dependent on TNF
action since TNF binding proteins blocks the LPS-induced increase in ceramide. The role
of the ceramide pathway in LPS-induced endothelial cell activation remains to be

elucidated.

35

Chapter 2 Functions of endothelial cells in hemostasis

Hemostasis can be defined as the property of the circulation that involves both a rapid,
effective response to vascular injury preventing excessive blood loss and a mechanism to
limit this response ensuring blood in the fluid state within the blood vessels (102).
Hemostasis depends on the balanced interaction among vascular endothelial cells, platelets
and coagulation factors. A key factor in the development of endotoxin-induced shock is
disturbed hemostatic equilibrium between blood and the endothelial cells of the vessels

(103).

For many years, endothelial cells have been considered to be homogeneous cells, which
form the luminal vascular surface and serve exclusively as a selective filter and barrier.
With the introduction of endothelial cell culturing technique by J affe (104), unexpected
functions of these cells were discovered. It is well known that vascular endothelium plays
a key role in regulation of vessel tone, platelet activation, clot formation and dissolution
(103). This chapter will focus on the functions of endothelium in clot formation and

dissolution.

2.1 Procoagulant properties of endothelium

Blood coagulation, which is important in the arrest of hemorrhage at the site of blood

36

vessel injury, results from the conversion of a soluble plasma protein, fibrinogen, to
insoluble fibrin. The conversion is catalyzed by the enzyme, thrombin. Thrombin is not
normally present in circulating blood but exists as an inert precursor, prothrombin.

Activation of the extrinsic and/or intrinsic pathways of coagulation can lead to the

activation of prothrombin to thrombin (102).

The intact endothelium can not initiate coagulation but can actively express a number of
procoagulant properties. When vascular injury occurs, the presence of these noninitiating
procoagulant activities enhance the hemostatic response. The major procoagulant activities
of the endothelium include synthesis and expression of cellular receptors for many
coagulation factors such as IX, lXa, X, Xa, high molecular weight kallikrein and thrombin
(105-109), and synthesis coagulation factor V which serves as an essential cofactor to
accelerate the activation of prothrombin by factor Xa (109). In addition to synthesizing

factor V, endothelium also provides binding site for this protein ( 109,200).

2.2 Anticoagulant properties of endothelium

Endothelial cells, which form the interface between blood and vessel wall, provide a
thromboresistant surface under normal conditions. This “anticoagulant behavior”
promotes blood flow.(103) The effective anticoagulant properties of the endothelial cells
depend in part on the expression of surface-bound proteoglycans such as heparan sulfate

proteoglycans, derrnatan sulfate proteoglycans and thrombomodulin, as well as the release

37

of coagulant factor inhibitors, such as protein S and tissue factor pathway inhibitor (TFPD.

2.2.1 Synthesis of Heparan Sulfate Proteoglycan (HSPG)

It is known that antithrombin III (AT 111), synthesized by the liver, serves as a major
inhibitor of thrombin and factor Xa. It is also able to inactive factors D(a, XIa, XIIa and
kallikrein. However ATIII does not display full anticoagulant activity without the
presence of heparin or endogenous heparin-like molecules (111). Many investigators have
shown that HSPG, a heparin-like molecule, is specifically synthesized by endothelium
(112-114) and positions on the luminal surface of endothelium where circulating ATIII
can be bound and activated. The amount of HSPG produced by the microvascular
endothelium is five times greater than by the macrovascular endothelium (113,114) and
the number of ATIII binding sites per micro- and macrovascular endothelium are about
500,000 and 50,000 respectively. The distribution of HSPG and binding sites for ATIII in
vessels is consistent with the primary need for providing the rnicrocirculation with

anticoagulant mechanisms to protect against thrombosis.

2.2.2 Synthesis of Dermatan Sulfate Proteoglycans(DSPG)

Heparin cofactor II, also synthesized by the liver, selectively inhibits thrombin in the

presence of heparin or derrnatan sulfate proteoglycans (DSPG), which is synthesized by

38

endothelium (113). However the clinical significance of heparin cofactor II and DSPG
remain unclear, since a decrease in the level of either of these does not appear to be related

to thrombotic events.

2.2.3 Protein C/protein S/thrombomodulin system

In the past two decades, it has been shown that protein C/protein S and thrombomodulin

form a major anticoagulant pathway (115).

Protein C, a vitamin K-dependent glycoprotein, is synthesized by the liver. Protein C is
activated to Protein Ca (APC) by thrombin bound to thrombomoduhn on the surface of
endothelial cells. The activity of APC depends on the presence of its cofactor, protein S
(116). Protein S, a vitamin K—dependent glycoprotein, is produced by both the liver and
the endothelial cells (117). It has been presumed that the endothelial cell is a major source
of protein S, since patients with severe liver disease do not demonstrate a significant
protein S deficiency (117). As a cofactor of APC, protein S promotes the binding of APC
to biological membranes such as platelets and endothelial cells. In plasma, about 40% of
the total protein S are free and 60% form a non-covalent complex with C4BP, a
component of complement and an acute-phase plasma glycoprotein. Only free protein S

can serve as a cofactor of activated APC (118).

39

Thrombomodulin (TM) is an acidic integral membrane protein that is synthesized by

endothelial cells (119). Thrombin in circulating blood binds to TM forming a thrombin-
thrombomodulin complex. This complex activates protein C to protein Ca (APC). APC,
with its cofactor protein S, in turn inactivates factors Va and VIIIa and neutralizes
plasminogen activator inhibitor 1 (PAI-l) (figure 9)(119). When thrombin binds to TM, it
is no longer able to cleave fibrinogen to fibrin and activate factors V, VIII and XIII (120-

122).

Comparative studies have demonstrated both qualitative and quantitative differences in
hemostatic properties between macro- and microvascular ECs. Most of the vascular TM is
contained in capillaries, which comprises 99% of the endothelial surface area. As
mentioned above macrovascular ECs secret 5-fold more HSPG and contain 10- fold more
ATIII binding sites than macrovascular ECs. These studies indicate that anticoagulant
activities of rrricrovascular ECs are stronger than that of macrovascular ECs, which is

consistent with the primary need of microcirculation for the prevention of thrombosis.

40

Thrombin
+
Thrombomodulin

Factor V

Activated /
Protein C
Protein C + g
Protein S \
(Free) Factor VIII

A

 

V

 

 

v
(Protein S + 04b Binding Protein)

Figure 9. The protein C-protein S system. Activated protein C and its cofactor, free
protein S, are potent inhibitors of coagulation factors V and VIII.

4i

2.2.4 Tissue factor pathway inhibitor (TFPI)

Until last two decades, people understood that tissue factor pathway (extrinsic pathway)
plays a dominant role in initiating blood coagulation in vivo. This pathway begins the
exposure of blood to tissue factor (TF) at an injury site and formation of the complex
between TF and plasma factor VII/VIIa. The TF/VIIa complex activates both factor IX and
X, leading to thrombin generation and fibrin formation. The primary physiologic inhibitor
of TFNIIa, termed as tissue factor pathway inhibitor (TFPI), was first identified in the
1950’s, but it was not fully characterized until the last decade. TFPI plays a pivotal role in
the regulation of tissue factor - initiated blood coagulation (extrinsic pathway) through its
ability to inhibit TF/VIIa and Xa. It is a single chain glycoprotein present in plasma in a
trace amount and synthesized primarily by endothelial cells (123). It directly inhibits factor
Xa activity and inhibits the factor VIIa-TF complex activity by forming a quaternary factor
Xa—TFPI-VIIa-TF complex. TFPI is a 42 KD glycoprotein and contains a negatively
charged amino-terminal region, three tandem Kunitz—type inhibitor domains and basic
carboxyl-terminal tail (124). In plasma, about 80% of the TFPI are bound to a low density
lipoprotein (125). TFPI inactivates Xa primarily involving the second inhibitor domain
while the inactivation of TF/VIIa is dependent on the first domain (126). In the presence of
a sulfate glycosaminoglycan such as heparin, HSPG, and DSPG, the inhibitory activities
of TFPI are greatly enhanced(127. 128). These findings suggested that TFPI may be bound

to and activated by endogenous sulfate glycosaminoglycans.

42

In 1994, a second human tissue factor pathway inhibitor (TFPIZ) was identified by Foster
et. al (129). TFPI 2, also known as placenta protein 5(130), consists of three domains
which are similar to TFPI (130). TFPI 2 is heparin independent and potently inhibits
TFNIIa, Xa, kallikrein and plasmin. The majority of the TFPI 2 synthesized by ECs

exists mainly in the extracellular matrix (131). The role of TFPI2 as a natural

anticoagulant is still unknown.

2.3 Fibrinolytic properties of endothelium

The coagulation and fibrinolysis pathways are two separate but interlinked enzyme
cascades that regulate the production and breakdown of fibrin. Fibrinolysis is the

dissolution of fibrin clots and functions to remove the fibrin deposition in blood vessels.

Fibrinolysis is primarily initiated by the incorporation of the enzyme tissue type
plasminogen activator (t-PA) into the fibrin clot, where it converts fibrin-bound
plasminogen to plasmin. The rate of endogenous fibrinolysis is largely determined by the
activity of t-PA, The activity of t-PA is regulated by plasminogen activator inhibitor (PAI),

a serine protease inhibitor.

Tissue plasminogen activator (tPA), synthesized by ECs, is the major activator of

plasminogen. The major inhibitor of tPA, plasminogen activator inhibitor 1 (PAI-l) is

43

also synthesized by ECs (132,133). ECs play a central role in fibrinolysis due to their
production of both t-PA and PAI-1. PAI-1, which was first recognized by Loskutoff in
1983(134), is the major PAI in the plasma. Other PAIs that are not present in normal
plasma include PAI-2 which is synthesized in the placenta, and PAI-3 which is originally
detected in urine and functions as an inhibitor of urokinase and APC. The PAI-l gene has
nine exons and eight introns and is located on chromosome 7 (135,136). PAI-l exists in
two forms in plasma — active and latent form (137). The latent form comprises more than
98% of the total concentration of PAI-1. These two forms are identical in molecular
weight and immunological activity but different in conformation. PAI-1 is synthesized as
an active molecule that is rapidly converted into the latent form. PAI-1 is mainly
synthesized by ECs. In addition to ECs, other cells such as hepatocytes, cultured granulosa
and glomerular epithelial cells also synthesize PAI-l (138,139,140). PAI-l is a
glycoprotein of 50,000 MW and inhibits both tissue and urokinase-like plasminogen
activators. PAI-l activity is higher in men than in women and increases with age (141).
PAI-l activity is correlated with triglycerides (142,143) and obesity (143) as well as
exercise status (144), and is inversely correlated with levels of high-density lipoprotein
and with t-PA activity (142,143). It is generally assumed that fibrinolysis is mainly
controlled by production and release of tissue type plasminogen activator (t-PA) and PAI-

1 from the endothelium.

Chapter 3 LPS-perturbed endothelial cells in coagulation and fibrinolysis

Recent studies have begun to provide a molecular basis for the observed link between
endotoxin-perturbed endothelial cells and a shift in hemostatic properties. It is (now clear
that both procoagulant and anticoagulant properties of endothelial cells are altered by
endotoxin and inflammatory mediators released in response to it such as lL—l and TNF
(145). Under normal physiological conditions the anticoagulant properties of ECs are
predominant. However, after ECs are exposed to endotoxin or cytokines, anticoagulant
properties of ECs diminish and procoagulant properties increase. ECs synthesize TF,
suppress thrombomodulin and increase PAI-l release when stimulated with endotoxin or
cytokines in vitro. This shift in the hemostatic properties of ECs favors intravascular

coagulation and may contribute to pathogenesis of DIC in sepsis. (146)

3.1 Expression of TF and TFPI in LPS-perturbed ECs

Tissue factor, thromboplastin, is a 47KD membrane-bond glycoprotein that triggers the
extrinsic pathway of blood coagulation (147). TF on the cell surface forms a complex with
coagulation factor VII/VIIa. The activation of factor X and IX by this complex leads to
thrombin generation and fibrin formation (148). In studies in vitro, TF is not detectable on
resting ECs or on macrophages (149). However, shortly after stimulation with either LPS,

IL-1 or TNF, TF is synthesized and expressed on the surface of ECs and macrophages

45

(150-152). A significant elevation of TF antigen in the plasma has been observed in LPS-
induced DIC (153,154). Previous studies have demonstrated that the infusion of
recombinant TF causes DIC (153), and inhibition of TF activity by monoclonal antibodies
reduces the incidence of DIC and protects animals from a lethal dose of LPS (155). These
results indicate that the TF-dependent coagulation alteration plays a major role in the
pathogenesis of DIC in sepsis. The onset of coagulation in sepsis results primarily from

the induction of TF synthesis in ECs and monocytes by endotoxin.

TF can be found both in plasma and isolated monocytes from patients with DIC, but TF
synthesis in ECs is more difficult to demonstrate, probably because of the presence of
TFPI (156). TFPI is the most important physiological inhibitor of the TF—dependent
coagulation. Recently, clinical studies have shown that administration of a large dose of
TFPI can reduce the number of fibrin thrombi and the incidence of mortality in the DIC
model (157). TFPI-depleted animals develop DIC after the administration of even small
amounts of TF or LPS, which did not have any significant effects in normal animals (158-
160). The results of these studies indicated that TFPI played an important inhibitory role
against TF-dependent coagulation in DIC. Ameri et al have demonstrated that neither LPS
nor cytokines can significantly influence the synthesis of TFPI by cultured ECs, and that
the normal level of TFPI did not prevent the development of LPS or TF-induced DIC
( 161). In contrast, Hera and his coworkers found that in this DIC model the number of
TFPI-positive pulmonary ECs and mRN A expression of TFPI, and TFPI activity in lung

tissue were remarkably decreased after LPS injection. They assumed over production of

46

TF and down expression of TFPI by ECs may contribute to thrombus formation in DIC

(162).

3.2 Suppression of thrombomodulin in LPS-perturbed ECs

Thrombomodulin(T M) is a specific endothelial cell receptor that forms a complex with
thrombin. This complex rapidly converts protein C to activated protein C, which
proteolytically destroys factor Va and VIIIa and thereby suppresses thrombin generation.
TM is primarily synthesized by ECs. Irnmunohistochemical analysis of human tissues
demonstrated TM on the endothelium of blood vessels and lymphatics except for in the
central nervous system (CNS) (163,164). Many stimuli that induced tissue factor also
reduced TM density. In particular, TM density decreased after exposure of ECs to
LPS(165), TNF(166) or IL-1(166). Such agents also induced expression of TF on EC8.
This conversion of EC properties from anticoagulant to procoagulant may relate to the

hypercoagulable state followed by thrombosis and/or DIC in sepsis.

Although it is confirmed that LPS, TNF and IL-1 suppresses TM synthesis of ECs, the
precise cellular inhibitory mechanism of TM expression is not yet fully understood.

Several reports indicated that TNF reduced TM density due to the
intemalization/degradation of TM and down-regulation of TM mRNA level (166,167).
The mechanism involved in LPS and lL-l is not known, but presumably represents

increased endocytosis and degradation.

47

3.3 Up-regulation of PAI-l by perturbed ECs

Fibrinolysis is achieved by an enzyme, plasmin. Plasmin is formed by the activation of
plasminogen which is normally present in the circulation. Plasminogen is converted to
plasmin by the specific enzymes known as plasminogen activators. The conversion is
inhibited by plasminogen activator inhibitor 1 (PAI-1). PAI-1 is mainly synthesized by
endothelial cells. The antigen level of PAI-lin platelet-poor plasma of healthy subjects is

18 i 10 ng/ml (168).

PAI-1 synthesis in endothelial cells can be stimulated directly by LPS and cytokines
(169,170). For example, the addition of endotoxin to bovine aortic and HUVEC cultures
stimulates the synthesis PAI-1 by 3 to 40 fold (170). In studies in vivo, normal volunteers
infused with a small amount of endotoxin (4ng/ml) increased PAI-l antigen level 5 fold
(171). In another study of 39 patients with a systemic meningococcal disease, PAI-l levels
were more than 360ng/ml in 12 of the 13 patients who developed severe septic shock and
renal impairment. Levels greater than 1850ng/ml were associated with 100 % fatality (5 of
5 patients) (172). These studies indicated that an elevated PAI-1 level is a risk factor for

thrombosis (173,174).

As previously described, LPS-perturbed ECs facilitate thrombosis by inducing EC

procoagulant activity, inhibiting the thrombomodulin/protein C anticoagulant pathway and

48

blocking fibrin dissolution via stimulation of PAI-l.

49

Chapter 4 Summary and final remarks

4.1 Summary

LPS, an integral part of the bacterial cell wall, is one of the strongest stimulators for many
cell types in the body and responsible for tissue and organ damage during gram-negative
septicemia and septic shock. The endothelium, although initially envisioned as a passive
inert vascular cell lining, is now considered as an important player in the regulation of
vascular tone, coagulation and fibrinolysis, cellular growth and differentiation, and

immune and inflammatory response.

It is now well established that the mechanism of LPS-induced cell activation involves an
initial interaction with a specific LPS binding protein (LBP). The LBP-LPS complexes
are recognized by CD14, present either anchored to the myeloid cell membrane by GPI
tail, or free as soluble plasma protein in serum. Membrane-bond CD14 (mCDl4) does not
contain a transmembrane domain and therefore can not transmit a signal from the surface
to the inside of cell. Existence of a second receptor in myeloid cells was postulated by
Ulevitch and Tobias (39). Recently, Yang et al identified Toll-like receptor 2 as a
secondary receptor(40). They suggested that the binding of the LPS-LBP complex to
mCDl4 and Toll-like receptor 2 leads to signal transduction. Host cells that do not

express membrane-bond mCD14 (e.g endothelial cells and epithelial cells) respond to LPS

via a soluble CD14 (sCDl4), in which LPS-sCD14 complexes are recognized by an

unknown component on the cell surface.

The intracellular pathways that mediate LPS-induced effects are not yet completely
understood. Recent studies suggest that the ceramide signaling pathway is involved in the
LPS effects. Ceramide is a lipid second messenger produced by the action of
sphingomyelinase (SMase). However LPS does not activate sphingomyelinase or cause
accumulation of ceramide. Some laboratories also found a similarity between ceramide
and the reducing glucosamine of lipid A. Based on these observations, researchers have
suggested that LPS may simply mimic ceramide. The LPS interaction with LPS receptor,
perhaps through mimicking ceramide, stimulates ceramide-activated protein kinase
(CAPK) to phosphorylate Raf, which in turn phosphorylates rrritogen—activated protein
kinases (MAPKs). MAPKs transduce the signal to the nucleus where it activates a

transcription factor.

The mechanisms involved in LPS-EC interactions are less well known and have only
recently been explored. Binding studies with fluorescent-labeled anti-CD14 failed to
identify the presence of a mCD14 receptor on human ECs. Nevertheless, LPS has proven
effects on EC5 despite the lack of a mCDl4. Recent studies from several laboratories have
shown that sCDl4 is involved in LPS-induced EC responses. The sCD14-mediated LPS
effects on EC8 are independent of LBP, although LBP enhances the binding of LPS to

sCD14. The signal transduction events of the LPS-induced EC activation are still

51

controversial.

DIC is a serious and frequent complication of gram-negative bacterial sepsis. Despite the
use of potent antibiotics and intensive supportive care, mortality among patients with
sepsis-induced DIC remains close to 60%(3). Sepsis-induced DIC appears to be related to
endotoxin-perturbed vascular ECs. Endotoxin stimulates ECs to express TF which
activates coagulation through the extrinsic pathway. As a result, large amounts of
thrombin are generated. Thrombin transforms fibrinogen into fibrin. In the meantime,
endotoxin suppresses ECs to express TM. Down-regulation of TM decreases the binding
of thrombin to TM. When thrombin is bound to TM its functional properties are changed
from a procoagulant protein to an anticoagulant protein. Thrombin bound to TM can
activate Protein C to Protein Ca (PCA), which inhibits activated coagulation factors Va
and VIIIa. The decrease in binding of thrombin to TM leads to a decrease in PCA and a
prolongation in the half-life of factors Va and VIIIa. An increase in the half-life of factors
Va and VIIIa is associated with an increased risk of thrombosis and DIC. The fibrinolytic
system, which lyses fibrin polymers, also plays an important role in the pathogenesis of
DIC. Plasmin, the key enzyme in fibrinolysis, is generated through the activation of
plasminogen by t-PA. The activity of t-PA is controlled by PAI-l. A remarkable increase
in plasma levels of PAI-1 and PAI-1 activity subsequent to endotoxin infusion has been
reported in healthy human subjects as well as in animal models and cultured endothelial
cells. The increased PAI-1 plasma levels and activity potently suppress fibrinolytic system,

which may contribute to the persistence of fibrin deposits in DIC.

52

4.2 Final remarks

In the past ten years tremendous progress has been made in characterizing the structure of
LPS, mechanisms of LPS interaction with host cells and signal transduction. However, to
date, many aspects of actions of LPS remain unanswered. For example, regarding the
structural requirements governing endotoxicity, we do not know whether phosphates are
actually required for the expression of bioactivity or whether sulfate, nitrate could replace
phosphate, or whether just negative charge, perhaps at a specific location, is necessary. As
is shown by many reports (15), biological activity of lipid A is extremely sensitive to even
slight modification of its chemical structure. This indicates that a peculiar molecular
conformation determines bioactivity. Although considerable progress has been made in
the characterization of this conformation, we are far from knowing the submolecular

details essential for endotoxin’s effects ( 16).

A great deal of data supports the role of ceramide and MAP kinase in cytokines and LPS
signal transduction (74,75,76). However the precise mechanism by which the
sphingomyelin pathway induces tyrosine phosphorylation and activation of MAP kinase
remains to be unidentified. Some investigators suggested that both Res and Raf may be

involved in linking two pathways but there are still unsolved controversies in this area.

It is known that endotoxin can potently activate blood coagulation. Activated ECs by

 

endotoxin play a central role in development of DIC in sepsis. In order to treat DIC
effectively, many investigators have extensively studied the ECs. One of the unsolved
issues in this area is the degree of EC heterogeneity in terms of morphology, marker
expression and function in the analysis of macrovascular versus microvascular ECs in
the vascular beds of different organs. ECs also show heterogeneity in the response to
and production of cytokines and endotoxin. The molecular and cellular basis for the

phenomenology of EC heterogeneity currently remains elusive.

BIBLIOGRAPHY

BIBLIRGRAPHY

. Glauser MP. G Zanetti. JD Baumgrtner. J cohen. Septic shook; pathogenesis. Lancet
338:732-736, 1991.

. Bone, RC. The pathogenesis of sepsis. Ann Intern Med. 115: 457-465, 1991.

. Hinshaw LB. Pathophysiology of endotoxin. Vol.2 of handbook of endotoxin.(
Proctor RA., Ed. ) Amsterdam : Elsevier, 1985.

. Cybulsky, MI. MKW Chan. HZ Movat. Acute inflammation and microthrombosis
induced by endotoxin, interleukin - l and tumor necrosis factor and their implication
in Gram - negative infection. Lab Invest. 58: 365-378, 1988.

. Morrison, DC and RJ Ulevitch. The effects of bacterial endotoxins on host mediation

systems. Am J Pathol. 93 : 525-531, 1978

. Leive, L. V Shovlin. S Mergenhagen. Physical, Chemical, and immunological
properties of lipopolysaccharide Released from Escherichia coli by
Ethylenediarninetetraacetate. J Biol Chem 243: 6348-6391, 1986. -

. N ghiem,H0. A Staub. C Galanos. O Ludertz. Distribution and antigenic properties
of the O-deterrninants of Salmonella Zuerich(l,9,27,46) Eur J Biochem 125 : 431-
436, 1982.

Jansson, PE. AA Lindberg. B Lindberg. B Wollin. Structural studies on the hexose
region of the core in lipopolysaccharide from Enterobacteriaceae. Eur J Biochem 115
:571-575,1981.

. Hollingsworth, RI. RW Carlson. 27-hydroxyoctacosanoic acid is a major structural
fatty acyl component of the lipopolysaccharide of Rhizobium trifolii ANU 843. J Bio
Chem 264 : 9300-9303, 1989.

10. Bhat, UR. H Mayer. A Yokota. RI Hollingsworth. RW Carlson. Occurrence of lipid

A variants with 27-Hydroxyoctacosanoic acid in lipopolysaccharides from member of
the family Rhizobiaceae. J Bacteriology 173 : 2155-2159, 1991.

11. Strittmatter R. J Weckesser. PV Salirnath. C Galanos. Nontoxic lipopolysaccharide

from Rhodopseudomonas Sphaeroides ATCC 17023. J Bacteriol 135: 153-158,1993.

12. seydel U. H Labischinski. M Kastowsky. K Brandenburg. Phase behaviour,

56

13.

14.

15.

l6.

17.

18.

19.

20.

21.

22.

23.

supramolecular structure and molecular conformation of Lipopolysacchride.
Irnmunobiology 187:191-211, 1993.

Brandenburg K. H Mayer. MHJ Koch. J Weckesser. ET Rietschel. U Seydel.
Influence of the supramolecular structure of free lipid A on its biological activity. Eur
J Biochem 218:555-563,1993.

Schromm AB. K Brandenburg. ET Rietshel. U Sedyel. Do endotoxin aggregates
intercalate into phospholipid membranes in a nonspecific, hydrophobic manner? J
Endotoxin Res 2:313-323, 1995.

Riecschel E Th. L Brade. O Holst. VA Kulshin. B Lindner. AP Moran. UF Shade. U
Zahringer. H Brade. Molecular structure of bacterial endotoxin in relation to
bioactivity. PIS-32. In Nowotny A, J Spitzer, EJ Ziegler. Cellular and molecular
aspects of endotoxin reactions. Elsevier science publishers B.V., Amsterdam 1990.

Sedel U. K Brandenburg. Conformation of endotoxins and their relationship to
biological activity. P6l-7l. In Nowotny A, J Spitzer, EJ Ziegler. Cellular and
molecular structure aspects of endotoxin reactions. Elsevier science publishers B.V.,
Amsterdam 1990.

Hollingsworth RI. D Lili-Elghanin. Endotoxin structure and activity: An old problem
from a new perspective. P73-84. In Nowotny A, Spitzer 1, EJ Ziegler. Cellular and
molecular aspects of endotoxin reactions. Elsevier science publishers B.V.,
Amsterdam 1990.

Rietschel ET. T Kirikae. F Ulrich Schade. U Mamat. G Schmidt. H Loppnow. A
Ulmer. U Zahringer. U Seydrl. F Dipadova. M Schreier and H Brade. The FASEB
Journal 8:217-225, 1994.

Wright, SD. RA Ramos. PS Tobias. RJ Ulevitch. JC Mathison. CD14, a receptor for
complexes of lipopolysaccharide (LPS) and LPS binding proteih ? Science 249 :
1431- 3,1990.

Ray MP. and DM Jacob. Flurecsent detection of lipopolysaccharide interaction with
model membranes. Biochim Biophysica Acta 859 : 26—32,l986

Richard J .Ulevitch: Recognition of bacterial endotoxins by receptor-dependent
mechanism. Advences in immunology 53 : 267-289,1993.

Tobias,PS. K Soldau. and RJ Ulevitch. Identification of a lipid A binding site in the
acute phase reactant lipopolysaccharide binding protein. J Biol Chem 264 : 10867-
10871,1989.

Tobias,PS. K Soldau. JA Gegner. D Mintz. and RJ Ulevitch. Lipopolysaccharide
57

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

binding protein-mediated complextion of lipopolysaccharide with soluble CD14. J
Bio Chem 270: 10482-10988, 1995.

Yu,B.and SD Wright. Catalytic properties of lipopolysaccharide (LPS) binding
protein. J Bio Chem 270: 4100-4105, 1996.

Ulevitch RJ. AR Johnson. New function for hight density lipoprotein. J clin Invest 64
:1516—152, 1979.

Tobias,PS. KSoldeu. RJ Ulevitch. Isolation of a lipopolysaccharide-binding acute
phase reactant from rabbit serum. J Exp Med 164: 777-793,1986.

Fried,VA. and LI Rothfiled. Interactions between lipopolysaccharide to and
phosphatidylethanolamine in molecular monolayers. Biochim Biophy Acata 298 :
155- 147.1978.

Tobias,PS. K Soldeu. I Kline. JD Lee. K Koto. TP Martin. R] Ulevitch.
Crosslinking of lipopolysaccharide to CD114 on TI-IP-l cells mediated by
lipopolysaccharide binding protein. J immunol 151 : 3011-3021,1993.

Jerome pugin. C Cornelia. M Schurer. L Dider. A moriarty. R I Ulevitch and PS
Tobias. Lipopolysaccharide activation of human endothelial and epithelial cells in
mediated by lipopolysaccharide-binding protein and soluble CD14. Pros Natl Acad
Sci USA 90 : 2744- 2748,1993

Wright,SD. PSTobias. RJ Ulevitch. R Ramos. Lipopolysaccharide binding protein
opsonize LPS-bearing particles for recognition by a novel receptor on macropjages. J
Exp Med 170: 1231-1241,1989.

Schumann RR. SR Leong. GW Flaggs. PW Gray. SD Wright. JC Mathison. PS
Tobias. RJ Ulevith. Structure and function of lipopolysaccharide binding protein.
Science 249:1429-31, 1990.

Su GL. PD Freswick. DA Geller. Q Wang. RA Shapiro. Y Wan. TR Billiar. DJ
Tweardy. RL Simmons. SC Wang. Molecular cloning, characterization, and tissue
distribution of rat lipopolysaccharide binding protein: evidence for extrahepatic
expression. J Irnmunol 153: 743-752, 1994.

Lei, MG. SA Stimpson. DC Monison. Specific endotoxin lipopolysaccharide -
binding proteins on murine splenocyte : membrane localization and binding
characteristics. J Irnmunol 141: 1006-11,]988.

34. Wright,SD. MTC J ong. Adhesion-promoting receptors on human macrophages

recognize Escherichia E. coli by binding to lipopolysaccharide. J Exp Med 164 :
1876-8,1986.
58

-.| In 9.1-3“ *1

 

m:

35.

36.

37.

38.

39.

40.

41.

42.

43.

45.

46.

47.

Kinkland, TN. GD Vica. K Reichel. Identification of lipopolysaccharide -binding
protein in 702/3 cells by photoaffinity cross-linking. J Bio Chem 265 : 9520-5,1990.

Vogel SN. PY Perera GR Doctore. N Bhat. TM Carboni. A Haziot. SM Goyert. Cell
activation mediated by glycosylphosphatidylinositol-anchored or transmembrane
forms of CD14. Prog Clin Bio Res 397: 137-46, 1998.

Haziot, A. S Chen. E Ferrero. MG Low. R Silber. SM Goyert. The monocyte
diffirentiation antigen, CD14, is anchored to the cell membrane by a
phosphatidylinosital linkage. J Irnmunol 141: 547- 552,1988.

Pugin J. D Heumann. A Tomasz. VV Kravchenko. Y Akamatsu. M Nishijima. PS
Tobias. RJ Ulevith. CD14 is a pattern recognition receptor. Immunity 1:509-
516,1994.

Ulevitch RJ. PS Tobias. Recognition of endotoxin by cells leading to transmembrane
signalling. Current Opinion In hnmunology 6: 125-130,1994.

Yang RB. MR Mark. A Gray. A Huang. MH Xie. M Zhang. A Goddard. WI Wood.
AL Gurney. PJ Godowski. Toll-like receptor-2 mediated lipopolysaccharide-induced
cellular signalling. Nature 395:284—297,1998.

Belvin MP. KV Andersion. Conserved signalling pathway: the Dorsophila toll-dosal
pathway. Annu.Rev.cell Dev.Biol. 12:393-416,1996.

Gay NJ. and FJ Keith. Drosophila T011 and IL-1 receptor. Nature 351 :355-356,1991.

Baeuerle PA. and T Henkel. Funcation and activation of NF-KB in the immune
system. Annu Rev immunol. 12: 141-179,1994.

. Rock FL. G Hardiman. JC Tirnans. RA Kastelein. JE Bazan. A family of human

receptors structurally related to Drosophila Toll. Pro.Natl Acad Sci.USA. 95:588-
593,1998.

Marchant A. J Duchow. JP Delville. M Goldman. Lipopolysaccharide induces up-
regulation of CD14 molecule on monocytes in human whole blood. Eur J Irmnunol.
22:1663-1665, 1992.

Schutt C. T Schilling. U Grunwald. W Schonfeld. C Kruger. Endotoxin-neutralizing
capacity of soluble CD 14. Res Irnmunol 143:71-78, 1992.

Bazil U. and JL Strominger. Schedding as a mechanism of down-modulation of CD
14 on stimulated human monocytes. J Irnmunol 147: 1557-1574, 1991.

59

48. Haziot A. B Chen. E Terrero. MG Low. R Silber. SM Goyert. The monocyte
differentiation antigen, CD14, is anchored to the cell membrane by a
phosphatidylinositol linkage. J Irnmunol 141:547-552, 1988.

49. Tapping RI. and PS Tobias. Cellular binding of soluble CD14 requires
lipopolysaccharide (LPS) and LPS- binding protein. J Biol Chem 272:23157-64,
1997.

50. Ardit M. J Zhou. R Dorio. GW Rong. SM Goyert and KS Kim. Endotoxin-mediated
endothelial cells injury and activation: Role of soluble CD14. Infect Irnmun 61:3149-
3156, 1993.

51. Frey EA. DS Miller. TB Jahr. A Sundan. V Bazil. T Espevik. BB Finlay. SD Wright.
Soluble CD14 participates in the response of cells to lipopolysaccharide. J Exp Med
176:1665-1671, 1992.

52. Stelter F. Berbheiden M. Menzel R. Witt S. Jack RS. Grunwald U. Fan X. Schutt C.
Molecular basis for therapeutical concepts utilizing CD14. Prog Clin Bio Res
397:301-13, 1998.

53. Jerome Pugin. Richard J Ulevitch and Peters Tobias in “ Bacteria endotoxin :
Lipopolysaccharides from gene to therapy” P 369-373. Wiley-Liss Inc, 1995.

54. Pugin, J. RJ Ulevitch. PA Tobias. A critical role for monocytes and CD14 in
endotoxin-induced endothelial cell activation. J Exp Med 178: 2193-2200, 1993.

55. Nawrith, PP. I Bank. D Handley. J Dassimeris. L Chess. D Stern. Tumor necrosis
factor, cathectin interaction with endothelial cell receptors to induce release
ofinterleukin-l. J Exp Med 162: 1363-75, 1986.

56. Thereme,TR. SH Hefeneider. CR Wagner. DR Burger. Recombinant murine and
human IL-1 alpha bind to human endothelial cells with equal affinity, but have
unequal ability to induce endothelia cell adherence of lymphocytes. J Irnmunol 139:
1173-8, 1987.

57. Schromm AB. K Brandenburg. ET Rietschel. I-ID Flad. SF Carroll. U Seydel.
Lipopolysaccharide-binding protein mediates CD14-independent intercalation of
lipopolysaccharide into phospholipid membrane. FEBS Letter 399:267-71, 1996.

58. Benedeffo,DA. J w Shanda. The interaction of bacterial lipopolysaccharide with
phospholipid bilayer and monolayer. Biochim Biophy Acta 298 :145-147, 1973

59. Fried,VA. and LI.Rothfiled. Interactions between lipopolysaccharide and
phosphatidylethanolamine in molecular monolayers. Biochim Biophy Acta 398: 156,
197 8.

60

60.

61.

62.

63.

65.

66.

67.

68.

69.

70.

71.

Carr,C. D Monison. Lipopolysaccharide interaction with rabbit erythrocyte
membrane. Infect Irnmun 43: 600-606, 1984.

Carr,C. and D Monison. A two-step mechanism for the interaction of
lipopolysaccharide with erythrocyte membrane. Rev Infect Dis 6: 497-500, 1984.

Carr C. and D Monison. Mechanism of polymyxin B- Mediated lysis of
lipopolysaccharide treated erythrocytes. Infect Irnmun 49: 84—87, 1985.

Ray MP. and DM Jacob. Flurecesent detection of lipopolysaccharide interaction with
model membrane. Biochim Biophysica Acta 859:26—32, 1986.

. Takayarna K. DH Mitchell. 22 Din. P Mukerjee. C Li. DL Colemen. Monomeric Re

lipopolysaccharide from Escherichia coli is more active than the aggregated from in
the limelus Amebocyte Lysate assay and inducing Egr-1 mRNA in murine peritoneal.
J Biol Chem 269: 2241—44, 1994.

Brandenburg K. H Mayer. MHJ Koch. J Weckesser. ET Rietscel. U Seydel. Influence
of the supramolecular structure of free lipid A on its biological activity. Eur J
Biochem 218: 555-563, 1993.

Maryyarna N. K Yasunori. K Yamauchi. T Aizwa. T Ohrui. M Nara. T Ohno. G
Tanura. S Shimura. H Saschi. T Tahishima. K Shorato. . Quinine inhibits production
of tumor necrosis factor-a from human alveolar macrophages. Cell Mol Biol 10: 514-
20, 1994.

Martin, TR. JC Mathison. PS Tobias. RJ Maudner. RJ Ulevitch. Lipopolysaccharide
binding protein enhance the responces of alveolar macrophages to bacterial
lipopolysaccharide : implications for cytokine production in normal and injured lungs.
J Clin Invest 90 : 2209-2219, 1992.

Weinstein,SL. JS Sanghear. K Lemke. AJ Defranco. SL Dell. Bacterial
lipopolysaccharide induce tyrosine phosphorylation and activation of mitogen-
activated protein kinase in macrophages. J Biol Chem 267 : 14955-14962, 1992.

Weinstein,SL. MR Gold. AJ Defranco. Bacteria lipopolysaccharide stimulating
protein tyrosine phosphorylation in macrophages. Proc Natl Acad Sci USA 88 : 4148-
4152, 1991.

Han,J. JD Lee. PS Tobias. RJ Ulevitch. Bacterial lipopolysaccharide induces rapid
tyrosine phosphorylation in 70Z/3 cells expressing CD14. J Bio Chem 268 : 25009-
25014, 1993.

Legrand,CB. R Thieringer. CD14-dependent induction of protein tyrosine
61

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

84.

phosphorylation by lipopolysaccharide in murine B-lymphoma cells.Biochi et Biophy
Acta 1223 : 36-46, 1993.

Nick JA. NJ Avdi. P Gerwins. GL Johnson. GS Worthen. Activation of a P38
mitogen-activated protein kinase in human neutrophils J immunol 156(2) : 4867-
4875, 1996.

Lee JC. and Yang PR. Role of CSBP/p38/RK stress responses kinase in LPS and
cytokine signal mechanism. J Leuko Biol 59:152-159, 1996.

Han J. JB Lee. L Bibbs. RJ Ulevitch. A MAP kinase targeted by endotoxin and
hyperosmolarity in mammalian cells. Science 265: 808-811, 1994.

Anthony LD. M Crowley. A Finn. J Hambleton. SL Weinstein. The role of tyrosine
kinases and MAP kinases in LPS- induced signaling. Progree in clinical and
biological research. Vol 397:119-136, 1996.

Davis R]. The mitogen-activated protein kinase signal transduction pathway. J Biol
Chem 268:14553-14556, 1993.

Henkel T. T Machleidt. I Alkalay. M Kronke. Y Ben-Neriah. PA Baeuerle. Rapid
proteolysis of IkB-a is necessary for activation of transcription factor NF-kB. Nature
365:182-185, 1993.

Kolesnick RN. and DW Brolde. The sphingomyeline pathway in tumor necrosis
factor and interleukin-1 signaling. Cell 77:325-328, 1994.

Weinstein SL. JS Sanghera. K Lemke. AL DeFranco. SL Pelech. J Biol Chem
267:14955-14962, 1992.

Yao B. YH Zhang. S Delikat. S Mathias. S Basu. R Kolesnick. Phosphorylation of
Raf by ceramide-activated protein kinase. Nature 378:307-311, 1995.

Hanahan DJ. Platelet-activating factor: Biologically activate phosphoglyceride. Annu
Rev Biochem 55:483-487. 1986.

Samulesson B. SE Dahlen. J A ,Lindgren. CA Rouzer and CN Serhan. leukotrienes
and lipoxins: Structures, biosynthesis, and biological effects. Science 237: 1 171-1 128,
1987.

Weinstein SL. CH June. A DeFranco. J Irnmunol 15123829-3838, 1993.

Bagasra 0. SD wright. S Thikkavaragu. TW Oakes. R Pomerantz. CD14 is involved
in control of human immunodeficiency virus type 1 expression in latently infected
cells by lipopolysaccharide. Proc Natl Acad Sci USA 89:6285-89, 1992.

62

85.

86.

87.

88.

89.

91.

92.

93.

94.

95.

Joseph CK. SD Wright. WG Bommann. JT Randolph. ER Kamor. R Bittman. J Liu.
RN Kolesnick. Bacterial lipopolysaccharide has structural similarity to ceramide and
stimulates ceramid-activated protein kinase in myeloid cells. J Biol Chem 269: 17606,
1994.

Wright SD. and RN Kolesnick. Does endotoxin stimulates cells by mimicking
ceramide? Immunology today 162297, 1995.

Pfau TC. Walker EB. Card GL. A comparison of the effects of lipopolysaccharide
and ceramide on arachidonic acid metabolism in THP-l monocytic cells. Cell
immunol 1862147-53, 1998.

Barber SA. G Detore. R McNally. SN Vogel. Stimulation of the ceramide pathway
partially mimics lipopolysacchride-induced responses in murine peritoneal
macrophages. Infect Immune 64:3397-3400, 1996.

Magnuson,DK. RV Maier. TH Pohlman. Protein kinase C : A protential pathway of
endothelial cell ativation by endotoxin, tumor necrosis factor and interleukin-1.
Surgery 106 : 216-23, 1989.

. Pettersen, KS. MT Wiiger. N N arahara. K Andoh. G Gaudemack.H Prydz. Induction

of tissue factor synthesis in human umbilical vein endothelial cells involves protein
kinase C. Thomb Haemostas 67: 473-477, 1992.

Tamaoki,T. H Nomoto. I Takahashi. Y Kato. M Morimoto. F Tornito. A potent
inhibitor of phospholipid/ca-H- dependent protein kinase. Biochem Biophys Res
Commun 135: 397-402, 1986.

Shibanuma,M. T Kuroke. N Kiyoshe. Effects of the protein kinase C inhibitor H-7
and ccalmodulin antagonist W-7 on superoxide production in growing and resting
human histiocytic leukemia cells (U937). Biochem Biophys Res Commun 144:1371-
1323, 1987.

Lane, TA. GE Lamkin. EV Wancewicz. Protein kinase C inhibitor block the
enhanced expression of intercellular adhesion molecule-1 on endothelial cells
activated by interleukin-1, lipopolysaccharide and tumor necrosis factor. Biochem]
Biophys Res Commun 172:1273-81, 1990.

Eissner,G. W Kolch. H Mischak. GM Bomkarnm. E Moller. Different role of protein
kinase C in cytokine induced lymphocyte-endothelium interaction in Vitro. Scand J
immunol 40 : 395-402, 1994.

Toullec,D. P Pianetti. H Coste. The bisindolylmaleimide GF 109203 X is a potent
and selective inhibitor of protein kinase C. J Biol Chem 266: 15771-81, 1991.
63

96. Hidaka,H. M inagnki. S kawwamoto. Y sasaki. Isoquinoline-sulphonamides, novel

and potent inhibitors of cyclic necleotide dependent protein kinase and protein kinase
C. Biochemistry 23: 5036-41, 1984.

97. Myers,CL. J Stanley. J schembri-king. T Parks. RW Wellace. Induction of ICAM-1
by TNF-a, IL-lB and LPS in human endothelial cells after downregulation of PKC.
Am J Physi01262: C367-73, 1992.

98. Adamson P. M Tight. JD Pearson. Protein tyrosine kinase inhibitors act downstream
of lL—la and LPS stimulated MAP-kinase phosphorylation to inhibit expression of E-
selectin on human umbilical vein endothelial cells. Cell Adhe Commu 3:511-525,
1996.

99. Schumann R. D Pfeil. N Lamping. C Kirschning. G Scherzinger. P scherzinger. P
schlag. L Karanajew. F Hermann. Lipopolysaccharide induces the rapid tyrosine
phosphorylation of the mitogen-activated protein kinases erk-l and P38 in cultured
human vascular endothelila cells requiring the presence of soluble CD14. Blood
87:2805-2814, 1996.

100. Arditi M. J Zhou. M Torres. DL Durden. M Stins. KS Kim. Lipopolysaccharide
stimulates the tyrosine phosphorylation of mitogen-activated protein kinases P44, P42

and P41 in vascular endothelial] cells in a soluble CD14-dependent manner. J of
Irnmunol 185:3994—4203, 1995.

101. Adriana HF. C Cordon-Cardo. S Bayoumy. M Garzotto. M McLoughlin. R
Gallily. CK EdwardsIII. EH Schuchman. Z Fuks. R Kolesnick. Lipopolysaccharide

induces disseminated endothelial apoptosis requiring ceramide generation. J Exp Med
186:1831-41, 1997.

102. Gibbons GH. and VJ Dzau. Molecular therapies for vascular disease [Review].
Science 272:689—693, 1996.

103. Savvas CM and US Ryan. Overview of the endothelium. In Joseph Loscalzo and
Andrew Schafer. Thrombosis and hemorrage P295, Second Edition.
William&Wilkins press, Baltimore, 1998.

104. J aff E. R Nachman. G Becker. C Minick. Culture of human endothelial cells
derived from umbilical vein. J clin invest 52:2754-2756, 1973.

105. Heirnark RL. S Schwartz. Binding of coagulation factors of IX and X to the
endothelial cell surface. Biochem Biophys Res Commen 111:723-731, 1983.

106. Stern DM. PP Nawroth. W Kisiel. D Handley. M Drillings. J Bartos. A
coagulation pathway on boving aortic segments leading to generation of factor Xa and
64

thnrombin. J Clin Invest 74: 1910-1921, 1980.

107. Rodgers GM. MA Shuman. Characterization of the interaction between factor Xa
and bovine aortic endothelial cells. Biochim Biophys Acta 8442320-329, 1985.

108. Schmaier AU. A Kuo. D Cines. High molecular weight kininogen expression on
human endothelial cells. Clin Res 35: 1467-1474, 1984.

109. Cerveny TJ. DN Fass. KM Mann. Synthesis of coagulation factor Va binds to
human endothelial cells. Clin Res 35:432-37, 1987.

l 10. Maruyana 1. HH Salem. PW Majerus. Coagulation factor Va binds to human
umbilical vein endothelial cells and accelerate protein C activation. J Clin Invest 74:
224-230, 1984.

111. Darnus PS. M Hicks. RW Rosenberg. Anticoagulatant action of heparin. Nature
246 :355-357, 1973.

112. Colbum P. V Buonassisi. Anticlotting activaity of endothelial cells and heparan
surfate proteoglycans. Biochem Biophys Res Commun 104:220-7, 1982.

1 l3. Marcum J A. RD Rosenberg. Heparin-like molecules with anticoagulant activity
are synthesized by cultural endotheliall cells. Biochem Biophys Res Commun
126:365-72, 1985.

114. Marcum J A. DH Atha . IMS Fritze. P N awroth. D Stern. RD Rosenberg. Cloned
bovine aortic endothelial cells synthesize anticoagulantly active heparan sulfate
proteoglycans. J Bio Chem 261 :7507-17, 1986.

115. Esmon CI‘. The roles of protein C and thrombomodulin in the regulation of blood
coagulation. J Bio Chem 264:4743-4746, 1989.

116. Stern D. J Brett. K Harris. P Nawroth. Participation of endothelial cells in the
protein C-protein S anticoagulant pathway: the syntheis and release of protein S. J
Cell Biol 102: 1971-78, 1986.

117. Angel DA. D Vigano. S Angelo. CT Esmon. Comp PC. Aaquired deficiences of
protein S: Protein S activity during oral anticoagulation, in liver disease and in
disseminated intravascular coagulation. J Clin Invest 81:1445-53, 1988.

118. Dahlback B. Inhibition of proteinCa cofactor function of human and bovine
proteins by C4b-binging protein. J Bio Chem 261 : 12022-12027, 1986.

119. Esmon CT. The regulation of natural anticoagulant pathways. Science 235: 1348-
52, 1987.
65

120. Esmon CT. NC Esmon. KW Harris. Complex formation between thrombin and

121.

thrombomodulin inhibits both catalyzed fibrin formation and factor V activation. J
Bio Chem 258: 12238-44, 1983.

Polgar J. I Lerant. L Muszberk. R Machovich. Thrombomodulin inhibits the
activation of factor XI[I by thrombin. Thromb Res 43:585-95, 1986.

122. Ye J. NL Esmon. CT Esmon. A J ophnson. The cative site of thrombin is altered

upon binding to thrombomodulin : two distinct structural changes are detected by
fluoresence, but only one correlates with protein C activation. J Bio Chem 266:
23016-22, 1991.

123. Bajaj MS. MN Kuppuswamy. H Surto. SG Spitzer. SP Bajaj. Cultured normal

124.
characterization of cDNA coding for the lipoprotein-associated coagulation inhibitor
shows that it is consists of three tendem kunitz-type inhibitory domains. I Bio Chem
263:6001-4, 1988.

125.

126.

127.

128.

129.

130.

human hepatocytes do not synthesize lipopolysaccharide-associated coagulation
inhibitor: Evidence that endothelium is the principal site of its synthesis. Proc. Natl.
Acad. Sci. USA 87 :8869-8873, 1990.

Wun TC. KK Kretzmer. TJ Girard. JP Miletich. GJ Broze. Jr. Coning and

Novotny WF. SG Brown. JP Miletich. DJ Rader. GJ Broze .J r. Plasma antigen level
of the lipoprotein-associated coagulant inhibitor from human plasma. J Bio Chem
264:18832-9, 1989.

Girard TJ. LA Warren. WF Novotny. KM Likert. SG Brown. JP Miletich. GJ
Broze.Jr. Functional significance of the Kunitz-type inhibitory domain of
lipoprotein-associated coagulant inhibitor. Nature 338:518-23, 1989.

Hamamoto K. W Kisiel. The effects of hrparin on the regulation of factoe VIIa-
tissue factor activity by tissue factor pathway inibitor. Blood Coagul Fibrinolysis
7:470—6, 1996.

Kaiser B. DA Hoppensteadt. W Jeske. TU Wun. J Fareed. Inhibitory effects of TFPI
on thrombin and factor Xa generation in vitro modulatory action of
glycosarninoglycans. Thromb Res 75:609-11, 1994.

Sprecher CA. W Kisiel. S Mathewes. DC Forster. Molecular cloning expression and
particular characterization of a second human tissue factor pathway inhibitor. Proc
Natl Acad Sci USA 91 :3353-60, 1994.

Kisiel W. CA Sprecher. DC Forster. Evidence that a second human tissue factor
pathway inhibitor (TFPI2) and human placental protein 5 are equivalent. Blood
66

131.

132.

133.

134.

135.

136.

137.

138.

139.

140.

141.

142.

8424384, 1994.

Petersen LC. CA Sprecher. DC Foster. H Blumberg. T Harnarnoto. W Kisiel.
Inhibitory properties of a novel human Kunitz-type protease inhibitor homogolous
to tissue factor pathway inhibitor. Biochemistry 35:266-72, 1996.

Levin EG. And DJ Loskutoff. Cultural bovine endothelial cells produced both
urokinase and tissue type plasminogen activators. J Cell Biol 94:631, 1982.

Erikson LA. RR Schul. T Nu. DJ Loskutoff . The fibrinolytic system of the vascular
wall. Clin Haematol 14: 513, 1984.

Loskutoff DJ. J A vanMourik. LA Erickson. D Lawerence. Detection of an unusally
stable fibrinolytic inhibitor produced by bovine endothelial cells. Proc Natl Acad Sci
USA 80:2950—2960, 1983.

klinger KW. R Winquist. A Riccio. PA Andreasen. R Sartorio. LS Nielsen. N
Stuart. P stanislovitis. P Watkins. R Douglas. KH Grzeschik. K Alitalo. F Blasi. K
Dano. Plasminogen activator inhibitor 1 gene is located at region q21.3-q 22 of
chromosome 7 and genetically linked with cystic fibrosis. Proc Natl Acad Sci USA
84:8448-52, 1987.

Follo M. D Ginsburg. Structure and expression of human gene encoding PAI-1.
Gene 842447-53, 1989.

Levin EG. Quantitation and properties of the active and latent plasminogen activator
inhibitor in curtured of human endothelial cells. Blood 67: 1309-13, 1986.

Sprengers ED. I-IMG Princen. T Koolstra. VWM vanI-Iinsberg. Inhibition of
plasminogen activators by conditional medium of human hepatocytes and hepatoma
cell line Hep G2. J Lab Clin Med 105:751-8, 1985.

Biersing L. AJM Hsueh. DJ Loskutoff. Cultured glanulosa cells produce two
plasminogen activators and an antiactivator, each regulated differently by
gonadotropins. Endocrinology 116: 1666-68, 1985.

Iwamoto T. Y Nakashitna. K Sueishi. Secretion of plasminogen activator and its
inhibitor by glomerular epithelial cells. Kindney Int 37: 1466-76, 1990.

Krishnamurti C. DB Tang. CF Ban'. BM Aliving. Plasminogen activator and
plasminogen activator inhibitor 1 activities in reference population. Am J Clin
Pathol 89:747-752, 1988.

Harnsten A. B Wiman. U deFaire. M Blomback. Increased plasma levels of a rapid
inhibitor og tissue plasminogen activator in yang survivors of myocardial infarction.
67

143.

144.

145.

146.

147.

148.

149.

150.

151.

152.

153.

154.

N Engl J Med 313:1557-1563, 1985.

Oseroff A. C K Narnurti. A hassett. D Tang. B Alving. Plasminogen activator and
plasminogen activator inhibitor activities on man with coronary artery disease. J Lab

Clin Med 113:88-93, 1988.

gris JC. JF Schved. PA Martinez. A Amaud. N Sancherz. Impact of physical
training on plasminogen activator inhibitor activity in sedentary men. Fibrinolysis
4(supp12):97-98, 1990.

Mason RG. D Sharp. HYK Chuang. SF Mohammad. The endothelium: Role in
thrombosis and hemostasis. Arch Pathol Lab Med 101 :61-63, 1977.

Pohlman TH. J M Harlan. Endotoxin-endothelial cell interactions. P459-472. In
Morrison DC, J Ryan, Bacterial endotoxic lipopolysaccharide. CRC press,Boca
Raton, Fla. 1992.

Nemerson Y. Tissue factor and hemostasis. Blod 71:1-8, 1988.

Camerer E. AB Kolsto. H Prydz. Cell biology of tissue factor, the principal initiator
of blood coagulation. Throm Res 81:1-21, 1996.

Drake TM. JH Morrissey. TS Edgington. Selective cellular expression of tissue
factor in human tissue. Am J Pathol 134: 1087-1093, 1989.

Colucci M. G Balconi. R Corenzet. A Pietra D Locati. MB Donati. N Semeraro.
Cultured human endothelial cells generates tissue factor in response to endotoxin. J
Clin Invest 71: 1893-96, 1983.

Rijburg PN. DM Stern. B Wollitzky. S Rimon. A Rimon. D Handley. P Nawroth. JJ
Sixma. PG deGroot. Activation of coagulation mechanism on tumor necrosis factor-
stimulated cultured endothelial cells and their extracellular matrix, the role of flow
and factor IX/IX a. J Biol Chem 266: 12061-64, 1984.

Bevilcqua MP. J S Pober. GR Majeau. RS Cotran. MAJ Gimbrone. Interleukin 1(IL-
1) induces biosynthesis and cell surface expression of procoagulant activity in
human vascular endothelial cells. J Exp Med 160:618-23, 1984.

Marr TA. LA Rao. SI Rapaport. Disseminated intravascular coagulation in rabbits
induced by administration of endotoxin or tissue factor: Effects of anti-tissue factor
antibodies and measurement of plasma extrinsic pathway inhibitor activity. Blood
75:1481-89, 1990.

Asakura H. Y Karnikabo. A Goto. Y Shiratori. M Yamazaki. H Jokaji. M Saito. C
Votani. I Kumubashiri. E Morishita K Aoshima. S Nakamuras. T Matsuda. Role of
68

155.

156.

157.

158.

159.

160.

161.

162.

163.

164.

165.

tissue factor in disseminated intravascular coagulation. Thromb Res 90:217-224,
1995.

Weda H. T N akaset. R Nakaya. K Minamikawa. Y Wakita. T Kaneko. M Ohiwa. K
Deguchi. S Shirakawa. Elevated plasma tissue factor antigen level in patients with
Disseminated intravascular coagulation. Am J Hematol 45:232-36, 1994.

Takahashi H. N Satoh. K Nada E Takakama. Y Seki. A Shibata. Tissue factor in
plasma of patients with disseminated intravascular coagulation. Am J Hamatol
46:333-337, 1994.

Taylor FBJ. A Chang. W Raf. JH Morrissey. L Hinshaw. R Catlett. Lethal E. coli
septic shock is prevented by blocking tissue factor with monoclonal antibody. Circ
Shock 33:127-134, 1991.

Elsayed YA. K Nakagawa Y karrrikubo. K enjyji. H Kato. K Sueishi. Effects of
recombinant human tissue factor pathway inhibitor on thrombus formation and its in
vivo distribution in a rat DIC model. Am J Clin Pathol 106:574-83, 1996. '

Sandset PM. BJ Wam-Crarner. SL Maki. SL Rapaport. Irnmunodepletion of
exrinsic pathway inhibitor sensitizes rabbits to endotoxin-induced DIC and the
generalizes shwartzman reaction. Blood 78: 1496-1502, 1991.

Sanfset PM. E] Warm-Cramer. LAM Rao. SL Maki. SL Rapaport. Depletion of
extrinsic pathway inhibitor (BPI) sensitizes rabbits to DIC induced with tissue factor
: Evidence supporting a physiological role for EPI as a natural anticoagulant. Pro

Natl Acad Sci USA 88: 708-12, 1991.

Ameri A. MN Kuppuswamy. S Basu. SP Bajaj. Expression of tissue factor pathway
inhibitor by cultured endothelial cells in response to inflammatory mediators. Blood
79:3219-26, 1992.

Hera S. Y Asada K Hatakeyama. K Marutsuka. Y Sato. A Kisanuki. A Sumiyoshi.
Expression of tissue factor and tissue factor pathway inhibitor in rat lungs with LPS-
induced disseminated intravascular coagulation. Lab Invest 77:581-89, 1997.

Maruyama 1. CE Bell. PW Majerus. Thrombomodulin found on endothelium of
arteries, veins,capillaries and lymphatics, and on cyncytiophoblast of human
placenta. J Cell Biol 100:364-9, 1985.

Ishii H. HH Salen. CE Bell. EA Laposata. PW Majerus. Thrombomodulin, an
endothelial anticoagulain protein is absent from the human brain. Blood 67:362-79,
1986.

Moore KL. NL Esmon. SP Andreoli. Cl‘ Esmon. NU Bang. Endotoxin enhances
69

166.

167.

168.

169.

170.

171.

172.

173.

174.

tissue factor and suppresses thrombomodulin expression of human vascular
endothelium in vitro. J Clin Invest 79: 124-130, 1987.

Moore 10.. CT Esmon. NL Esmon. Tumor necrosis factor leads to the
internalization and degradation of thrombomodulin from the surface of bovine
aortic endothelial cells in curture. Blood 73: 159-165, 1989.

Conway EM. RD Rosenberg. Tumor necrosis factor suppresses transcription of the
thrombomodulin gene on endothelial cells. Mol Cell Biol 8:5588-92, 1988.

Declerck PJ. MC Alessi. M Verstreken. EKO Kruithof. U-vague. D Collen.
Measurement of plasminogen activator inhibitor 1 in biological fluides with a

murine monoclonal antibody-based enzyme-linked immunosorbent assay. Blood
71:220-225, 1988.

Schleef RR. MP Bevilacqua. M Sawdey. MA Gimbrone. DJ Roskutoff. Cytokine
activation of vascular endothelium: Effects on tissue-type plasminogen activator and
type 1 plasminogen activator inhibitor. J Biol Chem 263: 5797-5803, 1988.

Colucci M. JA Paramo. D Collen. Generation in plasma of fast-acting inhibitor of
plasminogen activator in response to endotoxin stimulation. J Clin Invest 75:818-
824, 1985.

Suffredini AF. PC Harpel. JE Parrillo. Promotion and subsequent inhibition of
plasrrrinogen activator after administration of intravenous endotoxin to normal
subjects. N Engl J Med 320: 1165-1172, 1989.

Brandtzang P. GB Joe. B Brusletto. P Kierulf. Plasminogen activator inhibitor 1 and
2, alpha-2-antiplasmin, plasrrrinogen activator and endotoxin levels in systemic
meningococcal disease. Thromb Res 57:271-278, 1990.

Hamsten A. G Walldius. A Szumosi. M Blomback. U deFaire. G Dahlen. C Landou.
B Wiman. Plasminogen activator inhibitor in plasma: Risk factor for recurrent
myocardial infarction. Lancet 2:3-9, 1987.

Juhan-Vague I. J Vaiadier. M Alessi. M Aillaud. J Ansaldi. P Holvoet. A
Serradimigni. D Collen. Defficient t-PA release and elevated plasminogen activator
inhibitor levels in patients with spontaneous or recurrent deep venous thrombosis.
Thromb Haemost 57:67-72, 1987.

70