|
|
||||||||


*
Third Department of Internal Medicine, Nagoya University School of Medicine, Nagoya, Japan; and
Department of Molecular Biology, Nagoya City University School of Medicine, Nagoya, Japan
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
Certain activated complement fragments, particularly C5a, which is known to be a potent anaphylatoxin, are thought to play a role in lowering blood pressure in patients with endotoxemia (4, 5, 6, 7, 8, 9, 10). However, fatal shock does not occur via systemic complement activation following the i.v. administration of cobra venom factor (CVF),3 which forms a stable C3 convertase causing such extensive activation of the alternative complement pathway as to exhaust all complement capacity (11). Therefore, the generation of complement fragments cannot adequately account for the manifestation of endotoxin shock.
On cell surfaces, species-specific membrane inhibitors of complement
are present, and cells are protected from homologous complement (12, 13). Previously, we generated a mAb against rat Crry, 5I2, which causes
hemolysis of rat erythrocytes by homologous rat complement (14). This
mAb suppresses the function of a membrane inhibitor of complement, rat
Crry (14, 15), which inhibits the generation of C3 convertases on cell
membranes. Administration of the F(ab')2 fragment of mAb
5I2 (anti-Crry) causes complement activation, resulting in C3
deposition on endothelial cells with certain pathological events
(16, 17, 18, 19). These events include a mild decrease in blood pressure that
was observed following an i.v. injection of the mAb, although the
reaction to a single injection of anti-Crry was transient and all
rats survived (16). However, when the F(ab')2 fragment of
5I2 (anti-Crry) was administered to rats pretreated with as little
as 0.01 mg/kg LPS (
1/1000 of the lethal dose), the result was acute
lethal shock.
| Materials and Methods |
|---|
|
|
|---|
Male Wistar rats weighing
280 g were purchased from Chubu
Kagaku Shizai (Nagoya, Japan). The rats were allowed free access to
food and water and were maintained on a 12-h light/dark cycle.
Reagents and mAbs
LPS, prepared from phenol extracts of Salmonella
typhosa, and cyclophosphamide were obtained from Sigma (St. Louis,
MO). Recombinant mouse TNF-
was kindly donated by Drs. Denichi
Mizuno and Genichiro Soma (Takano Hospital, Kumamoto, Japan).
CVF was purified from lyophilized cobra venom Naja Naja (Sigma) as
described previously (16).
DL-2-mercaptomethyl-3-guanidinoethylthiopropanoic acid, a
carboxypeptidase inhibitor used to inhibit the inactivation of
generated C5a by carboxypeptidases, was obtained from Calbiochem (La
Jolla, CA). The production and characterization of 5I2 (anti-rat
Crry mAb) and 6D1 (anti-rat CD59 mAb) have been described
previously (14, 20). 6D1 was kindly donated by Dr. B. Paul Morgan
(University of Wales, Cardiff, U.K.). These mAbs belong to the
same subclass (IgG1) and showed similar binding in rat organs (16).
Bacterial endotoxin contamination of the Ab fragments used in the
present study was assessed using a specific detection reagent
(Pregel-M, Seikagaku Kougyou, Tokyo, Japan) according to the
manufacturers instructions. Escherichia coli (055:B5 200
pg/1 European unit) was used as a control. The sensitivity of
Pregel-M was <50 pg/ml (or 0.25 European units/ml). All stock
solutions of F(ab')2 fragments of 5I2 and 6D1 tested were
negative for endotoxins. The levels of endotoxin in serum samples
obtained from rats after injection with 0.01 mg/kg LPS were also
measured.
Effects of C5a receptor (C5aR) antagonist
The C5aR antagonist used was a hexapeptide, NMePhe-Lys-Pro-dCha-Trp-dArg (synthesized by the Peptide Institute, Osaka, Japan). This hexapeptide was originally reported to be an antagonist of C5aR on human peripheral neutrophils (21). The effect of this peptide on rat C5aR was confirmed by antagonized myeloperoxidase (MPO) release from rat polymorphonuclear leukocytes (PMNs) by zymosan-activated rat serum using a previously described MPO assay (21, 22). Briefly, zymosan-activated serum, which was used as a source of rat C5a, was obtained by incubating fresh rat serum and zymosan A (Sigma) with 1 mg/ml of DL-2-mercaptomethyl-3-guanidinoethylthiopropanoic acid at 37°C for 2 h and then pooling at -75°C until used. Rat PMNs were obtained from normal rats using Polymorphprep (Nycomed Pharma AS, Oslo, Norway). Rat PMNs were incubated with cytochalasin B (Sigma) for 5 min at 37°C, followed by the addition of the C5aR antagonist or vehicle and 5 min later by either rat zymosan-activated serum or fMLP (Sigma). Inhibition assays were performed at a concentration of zymosan-activated serum or fMLP that caused 60% MPO release from rat PMNs compared with the release from PMNs treated with Triton X-100 (Sigma). Measurements were made of triplicate samples for each concentration of C5aR antagonist used. Inhibition assays were repeated five times using different samples of rat PMNs. MPO release was calculated according to the following formula: MPO release (%) = ([OD value at the selected dose of C5aR antagonist]/[OD value without C5aR antagonist]) x 100.
The C5aR antagonist inhibited MPO release from rat PMNs
stimulated with zymosan-activated serum in a dose-dependent manner, but
did not inhibit release from rat PMNs stimulated with fMLP (see Fig. 5
B).
|
Samples of peripheral blood from the tail vein of rats were collected in tubes containing EDTA. Total leukocytes were counted using a hemocytometer. For the determination of serum complement activity, CH50 (50% complement hemolytic unit) was measured according to the manufacturers instructions using sensitized SRBCs (Ishizu Pharmaceutical, Osaka, Japan).
Arterial blood pressure
Animals were anesthetized by an i.p. injection of 50 mg/kg pentobarbital sodium (Abbott Laboratories, North Chicago, IL). Before insertion of a catheter, the baseline of each radiotransmitter (TA11PA-C40, Data Sciences, St. Paul, MN) was verified to be 4 mm Hg. A fluid-filled sensor catheter was placed into the left femoral artery slightly above the aortic bifurcation and was used for the measurement of arterial pressure, which was monitored with a transducer for 60 min after the injection of anti-Crry. Data were expressed as the percent change in mean arterial blood pressure (MBP) compared with the pretreatment value. Values at selected timepoints were calculated using the following formula: MBP (%) = ([MBP value at a selected timepoint - MBP value before anti-Crry injection]/[value of MBP before anti-Crry injection]) x 100. The treatments administered to animals in experimental groups are shown in Tables I and II.
lp;&2qHistology and immunohistology
Samples of heart, lung, and liver were obtained from selected rats from groups I-1, I-24, II-1, II-24, IV, V, VI, and VII at 5 min after the injection of the F(ab')2 fragments of 5I2 (anti-Crry) or 6D1 (anti-CD59). The organs of group III rats were obtained 5 min after the injection of CVF instead of mAb. For histological analysis, the samples were fixed in methacarn and embedded in paraffin. Sections of 3 µm thickness were stained with hematoxylin and eosin. The binding of injected mAbs, deposition of C3, and leukocyte common Ag-positive cells were detected by direct immunofluorescence. Briefly, samples were snap-frozen in liquid nitrogen. Sections of 2 µm thickness were cut on a cryostat and fixed in acetone at room temperature for 10 min. They were then incubated with fluorescein-labeled rabbit Ab against mouse IgG1 (Zymed Laboratories, San Francisco, CA) or fluorescein-labeled goat Ab against rat C3. To detect leukocyte infiltration, sections were stained with fluorescein-labeled mAb against rat leukocyte common Ag (OX-1, Dainippon Pharmaceutical, Osaka, Japan).
Statistical analysis
Statistical analysis was performed in a nonparametric manner using the Mann-Whitney U test for unpaired data. A difference between two groups was considered significant when the p value was <0.05 (5%).
| Results |
|---|
|
|
|---|
The MBP of rats injected with 10 mg/kg LPS dropped by >40%
within 1 h of injection, and all of the animals died between 24
and 48 h postinjection. Rats injected i.v. with 0.01 or 1.0 mg/kg
LPS showed a minimal change in MBP, and all survived. In the following
experiments, rats were sensitized with an i.v. injection of 0.01 mg/kg
LPS. This was followed by an i.v. injection 0, 0.5, 1, 4, 24, or
48 h later of 1 mg/kg of the F(ab')2 fraction of
anti-Crry. When rats were injected with anti-Crry 1 h
(group I-1), 4 h (group I-4), or 24 h (group I-24) after LPS
treatment, MBP progressively decreased; all rats died within 20 min
(Table I
and Fig. 1
A).
However, the change in the MBP of rats injected with anti-Crry
immediately (group I-0) after LPS administration, was practically the
same as that of rats injected with anti-Crry alone (16) (both
showing only a transient decrease without any fatal reaction) (Table I
and Fig. 1
A). The rats of group I-48, which were challenged
with anti-Crry 48 h after LPS treatment, showed a prolonged
reduction in MBP, although all rats survived (Table I
). Among group
I-0.5 rats, three of eight survived, and the changes in their MBP were
similar to those of group I-0 rats (Table I
). An injection of 1 mg/kg
of the F(ab')2 fraction of mAb 6D1 (anti-CD59), which
suppresses the function of CD59, did not affect the MBP of rats
pretreated with LPS (groups II-1 and II-24; Fig. 1
B).
|
|
To determine the effect of acute complement activation in the fluid
phase, as opposed to on cell membranes, the rats of group III were
injected i.v. with a bolus of 25 U of CVF at 1 h after the
administration of LPS, instead of with anti-Crry. The MBP of the
group III rats decreased immediately following the CVF injection and
was restored to baseline level within 15 min (Fig. 2
). Rats simultaneously injected with
100,000 U/rat of recombinant mouse TNF-
and anti-Crry showed
only a mild and transient decrease in MBP (group IV; Fig. 3
).
|
|
A previous study showed that when rats were injected i.v. with 25
U of CVF, complement activity, as determined by the CH50 level,
was undetectable from 24 to 48 h postinjection; no reduction in
the number of peripheral leukocytes was observed (23). CVF-treated,
complement-depleted rats were injected with 0.01 mg/kg LPS at 1 and
24 h before the i.v. injection of 1 mg/kg anti-Crry. The MBP
of these rats did not change after the anti-Crry injection, and all
rats survived (group V; Table II
and Fig. 4
). Deposition of C3 was
undetectable in the heart, lung, or liver of the
complement-depleted rats, although binding of anti-Crry to these
tissues was evident.
|
|
At 45 days after an i.p. injection of 200 mg/kg
cyclophosphamide, the total number of peripheral leukocytes dropped
to <25/mm3, with no change in complement activity (19, 23). When leukocyte-depleted rats received the same LPS and
anti-Crry injections as the rats of group I-1, they showed only
a mild and transient decrease in MBP after anti-Crry injection, and
all survived (group VI; Table II
, Fig. 4
A). The binding of
the injected anti-Crry and deposition of C3 in the
leukocyte-depleted animals appeared similar to the results obtained
with normal rats (group I-1). However, when the interval between LPS
and anti-Crry administration was 24 h, five of six
leukocyte-depleted rats died. Four of the rats died within 20 min after
anti-Crry injection, one died within 1 h, and the last rat
recovered from a severe decrease in MBP and survived.
Effects of C5aR antagonist in acute lethal shock
To determine a possible involvement of C5a in the
anti-Crry-induced lethal shock in LPS-treated rats, 3 mg/kg of C5aR
antagonist was administered i.v. 30 min before the anti-Crry
challenge. When the interval between the administration of LPS and
anti-Crry was 1 h, one-forth of the rats were rescued from
lethal shock by C5aR antagonist (group VII; Table II
). Furthermore,
when the interval was 24 h, the change in the MBP was slight and
all rats survived (Fig. 5
A),
although C3 deposition in the heart, lung, and liver was evident.
| Discussion |
|---|
|
|
|---|
production
could be induced by 0.01 mg/kg LPS, serum TNF-
levels were only
detectable at 30 min, peaked at 60 min, and were again undetectable at
4 h (data not shown). Furthermore, rats injected simultaneously
with a large amount of recombinant mouse TNF-
together with
anti-Crry showed only a mild and transient decrease in MBP.
Therefore, the serum level of TNF-
at the time of anti-Crry
injection might not be directly correlated with the severity of the
decrease in MBP. We were interested in determining whether pretreatment
with TNF-
would sensitize an animal, making it susceptible to lethal
shock by the mAb. However, this experiment could not be performed
because all of the TNF-
preparations available were slightly
contaminated with LPS, which would interfere with the results. Because there was no MBP change in complement-depleted rats (group V) following anti-Crry injection after priming with LPS, complement activation was considered indispensable for inducing the lethal reaction. When LPS-primed rats were challenged with CVF instead of anti-Crry (group III), only a mild and transient decrease in MBP was observed. Although the complement activation induced by CVF is so profound as to exhaust complement capacity in a body, the reaction occurs in the fluid phase; in contrast, anti-Crry blocks the function of Crry on cell membranes, resulting in complement activation on cell surfaces. Therefore, for induction of the acute lethal reaction, complement activation on cell membranes must be essential. Activation of complement on cell membranes in vivo results in the deposition of complement components such as C3 (16). The deposition of C3 may facilitate cell adhesion to or aggregation of PMNs and macrophages as well as platelets, which have C3 receptors such as CR1, CR2, CR3, and/or CR4 on their cell membranes. This type of cell aggregation might facilitate the release of the factors responsible for the lethal shock reaction.
With a 1-h interval between LPS sensitization and anti-Crry
administration, the decrease in the MBP of leukocyte-depleted rats
(group VI) was mild and transient after an injection of anti-Crry
(Fig. 4
A); the value obtained was essentially the same as
that for rats injected with anti-Crry alone. These results indicate
that the involvement of LPS-sensitized leukocytes is required for
severe MBP decrease and for acute death in addition to complement
activation on cell membranes. Only one of six rats pretreated with
cyclophosphamide survived lethal shock when the interval between LPS
sensitization and anti-Crry administration was 24 h, and
significant MBP decrease was observed in all six animals (Fig. 4
B). It has been reported that a trace amount of LPS primes
and causes an increase in chemoattractant-induced cell surface
expression (25, 26). Therefore, it may be possible that during the
24 h, LPS may have stimulated the few remaining leukocytes that
survived cyclophosphamide treatment, and these cells may have generated
granules in sufficient amounts to cause the lethal shock after
anti-Crry administration. Among this group, four rats died within
20 min after anti-Crry injection, one died within 1 h, and one
survived. This variation might be a reflection of the amount and/or
character of the leukocytes remaining after cyclophosphamide treatment.
C5a has been shown to participate in the mediation of endotoxin shock
(5, 6), and its receptors are known to be mainly expressed on
leukocytes (22, 27). Because a C5aR antagonist administered 30 min
before anti-Crry administration prevented the lethal reaction in
rats that had been pretreated for 24 h with LPS (Table II
, Fig. 5
A), the C5a generated by complement activation on cell
membranes by anti-Crry may be playing a role in the lethal shock.
The requirement of a
1-h interval between LPS sensitization and
anti-Crry administration for the induction of acute lethal shock
may be related to the up-regulation of C5aR by LPS (28, 29, 30). The
binding of C5a to circulating neutrophils is enhanced on leukocytes
primed with LPS and certain cytokines (27, 31, 32). However, the
binding of C5a to the up-regulated C5aR should not be sufficient to
induce lethal shock because CVF, which strongly activates complement in
vivo, could not induce such a lethal response in the LPS-treated rats.
In addition to the generation of C5a through complement activation, the
deposition of complement on the membranes of blood cells and
endothelium would likely facilitate cell adhesion and aggregation, and
would consequently appear to play an essential role in the lethal
reaction. Therefore, we further speculate that the following mechanism
may be involved in the various types of serious shock syndromes often
observed in clinical practice. LPS of bacterial origin may sensitize
patients, resulting in the up-regulation of inflammation-related
receptors including C5aR (28, 29). Serious shock, such as that
involving disseminated intravascular coagulation, could be induced when
complement is activated on cell membranes, causing cell adhesion and
aggregation via complement receptors on leukocytes and platelets, which
could then release inflammatory mediators such as IL-1, TNF-
, IL-12,
superoxides, and nitric oxide. Impairment of the membrane complement
regulatory proteins demonstrated here by the neutralizing mAb may
possibly occur in vivo by some other mechanisms, and may allow for
significant complement activity to take place on cell membranes. For
example, a transient down-regulation of complement regulatory membrane
proteins has been found to occur as a result of viral infection
(33, 34, 35, 36). Complement activation on membranes could also be induced by
LPS adsorbed onto cell surfaces or by excess deposition of immune
complexes on cell membranes of the capillary endothelium and blood
cells, overpowering the function of membrane inhibitors of complement.
| Acknowledgments |
|---|
| Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Hidechika Okada, Department of Molecular Biology, Nagoya City University School of Medicine, Mizuho-cho, Nagoya 467-8601, Japan. E-mail address: ![]()
3 Abbreviations used in this paper: CVF, cobra venom factor; Crry, a membrane regulatory protein of complement of mouse and rat; C5aR, C5a receptor; MBP, mean arterial blood pressure; MPO, myeloperoxidase; PMN, polymorphonuclear leukocyte. ![]()
Received for publication October 8, 1998. Accepted for publication February 16, 1999.
| References |
|---|
|
|
|---|
up-regulates the human C5a receptor (CD88) in myeloblastic U937 cells and related cell lines. J. Immunol. 155:4419.[Abstract]
This article has been cited by other articles:
![]() |
M. Mizuno, Y. Ito, N. Hepburn, T. Mizuno, Y. Noda, Y. Yuzawa, C. L. Harris, B. P. Morgan, and S. Matsuo Zymosan, but Not Lipopolysaccharide, Triggers Severe and Progressive Peritoneal Injury Accompanied by Complement Activation in a Rat Peritonitis Model J. Immunol., July 15, 2009; 183(2): 1403 - 1412. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Asai, T. Sato, T. Tada, T. Miyamoto, N. Kimbara, N. Motoyama, H. Okada, and N. Okada Absence of Procarboxypeptidase R Induces Complement-Mediated Lethal Inflammation in Lipopolysaccharide-Primed Mice J. Immunol., October 1, 2004; 173(7): 4669 - 4674. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. Fujita, I. Farkas, W. Campbell, L. Baranyi, H. Okada, and N. Okada Inactivation of C5a Anaphylatoxin by a Peptide That Is Complementary to a Region of C5a J. Immunol., May 15, 2004; 172(10): 6382 - 6387. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. J. Strachan, T. M. Woodruff, G. Haaima, D. P. Fairlie, and S. M. Taylor A New Small Molecule C5a Receptor Antagonist Inhibits the Reverse-Passive Arthus Reaction and Endotoxic Shock in Rats J. Immunol., June 15, 2000; 164(12): 6560 - 6565. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |