SHOCK, Vol. 26, No. 1, pp.

41Y49, 2006

EXOGENOUS PLATELET-ACTIVATING FACTOR ACETYLHYDROLASE

REDUCES MORTALITY IN MICE WITH SYSTEMIC INFLAMMATORY
RESPONSE SYNDROME AND SEPSIS

Rachel N. Gomes,* Fernando A. Bozza,† Rodrigo T. Amâncio,* André M. Japiassú,†

Rosa C. S. Vianna,† Andréa P. Larangeira,‡ Juliana M. Gouvêa,*
Marcela S. Bastos,* Guy A. Zimmerman,§ Diana M. Stafforini,k
Stephen M. Prescott,k Patrı́cia T. Bozza,* and Hugo C. Castro-Faria-Neto*
*Laboratório de Imunofarmacologia, Departamento de Fisiologia e Farmacodinâmica, IOC, Fundação
Oswaldo Cruz; † Centro de Tratamento Intensivo, Hospital Universitário Clementino Fraga Filho,
Universidade Federal do Rio de Janeiro; ‡ Laboratório de Tecnologia Bacteriana (LATEB), Biomanguinhos,
Fundação Oswaldo Cruz, Rio de Janeiro, Brazil; §Program in Human Molecular Biology and Genetics; and
k
Huntsman Cancer Institute, University of Utah, Salt Lake City, UT

Received 22 Nov 2005; first review completed 9 Dec 2005; accepted in final form 18 Jan 2006

ABSTRACT—Current evidence indicates that dysregulation of the host inflammatory response to infectious agents is
central to the mortality of patients with sepsis and in those with systemic inflammatory response syndrome. Strategies to
block inflammatory mediators, often with complicated outcomes, are currently being investigated as new adjuvant
therapies for sepsis. Here, we determined if administration of recombinant platelet-activating factor (rPAF)Yacetylhydrolase
(rPAF-AH), an enzyme that inactivates PAF and PAF-like lipids, protects mice from inflammatory injury and death after
administration of lipopolysaccharide (LPS) or cecal ligation and puncture (CLP). Administration of rPAF-AH increased
plasma PAF-AH activity and reduced mortality in both models. Treatment with rPAF-AH increased peritoneal fluid levels of
monocyte chemoattractant protein 1/CCL-2 and decreased interleukin 6 and macrophage migration inhibitory factor levels
after LPS administration or CLP. Administration of a broad-spectrum antibiotic together with rPAF-AH was more protective
than single treatment with either of these agents. The combined treatment was associated with reduced interleukin 6 levels
in mice subjected to CLP. We observed acute decreases in plasma PAF-AH activity in mice subjected to CLP or challenged
with LPS and in human patients with sepsis. We conclude that alterations in the endogenous PAF-AH contribute to the
pathophysiology of sepsis and that administration of exogenous rPAF-AH reduces inflammatory injury and mortality in
models relevant to the clinical syndrome. Variations in endogenous PAF-AH activity may potentially account for variable
responses to exogenous rPAF-AH in previous clinical trials. Serial measurements of plasma PAF-AH activity in murine
models demonstrate dynamic regulation of the endogenous enzyme, potentially explaining the variations in human subjects.
KEYWORDSVSepsis, PAF, PAF-AH, cytokines, SIRS

INTRODUCTION pathways, including production of mediators, during sepsis.

Sepsis is an important clinical problem. In the United States
alone, 710,000 people develop sepsis annually, and more than
210,000 of these patients have unfavorable outcomes (1).
Despite new and more effective life support measures and
more potent antibiotics, the mortality rate for patients with
sepsis is still unacceptably high.
A number of experimental studies aimed at characterizing
biochemical responses in sepsis suggest that dysregulated
inflammation and actions of innate immune effector cells play
important roles during initiation and progression of this
condition (2, 3). A consensus conference defined sepsis as
Bthe systemic inflammatory response syndrome (SIRS) that
occurs during infection^ (4). Extensive animal and patient
studies have examined the participation of inflammatory
Address reprint requests to Dr. Hugo C. Castro-Faria-Neto, Laboratório de
Imunofarmacologia, Departamento de Fisiologia e Farmacodinâmica, IOC,
Funda0ão Oswaldo Cruz. Av. Brasil 4365, Manguinhos, Rio de Janeiro, RJ,
Brazil, CEP 21045-900; E-mail: hcastro@ioc.fiocruz.br.
This work was supported in part by a Fogarty International Research Collaboration
Award (TW5531) and a Special Center of Research in ARDS (HL 50153) from the
National Institutes of Health and by the Conselho Nacional de Desenvolvimento
Cientı́fico e Tecnológico (CNPq).
DOI: 10.1097/01.shk.0000209562.00070.1a
Copyright Ó 2006 by the Shock Society
41
Copyright @ 2006 by the Shock Society. Unauthorized reproduction of this article is prohibited.
Triggering of these pathways results in pathologic activities of
proinflammatory molecules, including cytokines, chemokines,
oxygen radicals, and lipid mediators (3, 5).
There is evidence that platelet-activating factor (PAF) and
other PAF-like lipids that are recognized by the PAF receptor
on effector cells play central roles in septic syndromes (6, 7).
Increased PAF activity has been reported in patients with
septic shock and SIRS (7, 8). In addition, bacterial endotoxins
induce PAF production both in vivo and in vitro (7, 9).
Therapeutic strategies based on the use of PAF receptor
antagonists in animals generated encouraging results that led
to clinical trials in humans. Although these studies did not
identify a clearly effective treatment based on receptor
blockade, positive effects in subgroups of patients with sepsis
were reported in several studies (10).
An alternative strategy is to increase inactivation of ligands
recognized by the PAF receptor by administration of PAF
acetylhydrolase (PAF-AH). The PAF-AHs are highly specific
phospholipases that hydrolyze the sn-2 acyl group of PAF and
structurally related phospholipids to produce inactive com-
pounds (11). The plasma form of PAF-AH limits the half-life
of PAF to minutes in whole human blood (11). Plasma PAF-
AH expression is transcriptionally regulated by inflammatory
mediators such as interferon-+, by gram-negative bacterial
42 SHOCK VOL. 26, NO. 1
lipopolysaccharides (LPSs), and by PAF itself, indicating that
the PAF-AH gene contains elements that confer responsive-
ness to inflammatory challenge (12).
Cloning of the plasma form of PAF-AH identified a novel
mechanism for terminating inflammatory signals (13) that
may have advantages over the use of PAF receptor antagonists
in syndromes of dysregulated inflammation, including SIRS
and sepsis (7). Administration of rPAF-AH resulted in a
striking survival advantage and other positive effects in an
initial phase II study of patients with sepsis or multiple
traumas (14). In contrast, a follow-up phase III trial of rPAF-
AH in patients with severe sepsis was subsequently halted
because of lack of efficacy (15). The reasons for the
difference in the phase II and phase III results have not been
adequately explained and likely involved multiple variables
inherent to sepsis trials (10). Additional mechanistic inves-
tigation in relevant experimental models may provide impor-
tant insights to the pathobiology of sepsis and may be useful if
additional trials are undertaken in the future. In the present
study, we examined the activity of endogenous PAF-AH in
experimental models of sepsis and the effect of recombinant
PAF-AH (rPAF-AH) on mortality and biologic markers in
these models. We found that experimental sepsis and endotox-
emia induce significant alterations in PAF-AH activity in the
plasma of mice and that endogenous PAF-AH is dynamically
regulated under experimental septic conditions. We also found
that replacement with rPAF-AH dramatically decreased mortal-
ity and inflammation in murine models under conditions
chosen to increase their relevance to clinical sepsis in humans.
MATERIALS AND METHODS
Patients with sepsis
This study enrolled 54 patients with sepsis hospitalized in the critical care unit
of the University Hospital at the Rio de Janeiro’s Federal University or at the
Spanish Hospital, Rio de Janeiro, Brazil. Thirty-seven patients were diagnosed
with septic shock, and 17 patients experienced severe sepsis, as defined by the
Consensus Conference of the American College of Chest Physicians and Society of
Critical Care Medicine (4). Twelve healthy volunteers served as controls. The
subjects were from multiple ethnic backgrounds, reflecting the demography of the
local population. Subjects participated in this study based on informed consent
approved by the institutional ethics review boards.
Animals
Swiss and C57BL6 mice from the Oswaldo Cruz Foundation breeding unit
weighing 20 to 25 g were used for most studies. The animals were kept at con-
stant temperature (25-C) with free access to chow and water in a room with a 12-h
light/dark cycle. In one set of experiments, wild-type mice and mice deficient in
the receptor for monocyte chemoattractant protein 1 (MCP-1)/CCL-2, CCR2
(CCR2j/j) in the C57BL/6 background were used. In a preliminary experiment
at the University of Utah, female C57BL6 animals (range, 19Y21 g) from B&K
(Freemont, Calif) were used. The experiments in this study were approved by
the Oswaldo Cruz Institute’s Animal Welfare Committee and by the University of
Utah Animal Care and Use Committee.
Materials
Human rPAF-AH (Pafase) was a kind gift from ICOS Corporation (Bothell,
Wash) and was provided as an endotoxin-free solution. LPS from Escherichia coli
(serotype 0111:B4) was purchased from Sigma Chemical Company (St. Louis,
Mo). 6-(2-Chlorophenyl)-8-9-dihydro-1-methyl-8-(4-morpho-linyl-carbonyl)-
4H,7H-cyclopental (4,5) thieno (3,2-f) (1,2,4)-triazolo-(4,3) (1,4)-diazepine
(WEB 2170) was from Boehringer-Ingelheim, Ingelheim, Germany. Imipenem
was obtained from Merck Sharp & Dohme (Harlow, UK). Thiopental (Thionem-
butal) was from Abbott Laboratories (Abbott Park, Ill) do Brasil, LTDA, and
ketamine was from Cristália (Rio de Janeiro, Brazil). Anti-LPS polyclonal IgG
enriched with 12% IgM (Pentaglobin) was from Biotest Pharma GmbH (Germany).
Copyright @ 2006 by the Shock Society. Unauthorized reproduction of this article is prohibited.
GOMES ET AL.
PAF-AH activity was measured using a commercial kit (Cayman Chemicals,
Ann Arbor, Mich) according to the instructions provided by the manufacturer.
Cecal ligation and puncture
Mice were anesthetized with a mixture of thiopental (40 mg kgj1) and
ketamine (80 mg kgj1) diluted in sterile saline and administered intraperitoneally
(0.2 mL). Laparotomy was performed, and the cecum was exposed and carefully
ligated below the ileocecal junction with care to avoid bowel obstruction. The
cecum was punctured once with an 18-gauge needle and was then gently squeezed
to empty its contents through the puncture. The cecum was returned to the
peritoneal cavity, and the abdominal muscle and skin incisions were closed in
layers using a 3-0 nylon suture line. Immediately after the surgery, 0.5 mL of
sterile saline was administered subcutaneously to the animals for volume
resuscitation. Sham-operated mice were subjected to identical procedures, except
that ligation and puncture of the cecum were omitted. Animals subjected to cecal
ligation and puncture (CLP) developed early signs of sepsis, including lethargy,
piloerection, and diarrhea. Survival of mice subjected to CLP or sham injury was
determined daily for 7 days.
Murine model of endotoxemia
A murine model of endotoxemia was established by injecting mice intra-
peritoneally with 300 2g per cavity of LPS diluted in sterile saline. Control animals
received equal volumes of saline. All endotoxin-treated animals appeared acutely
ill and displayed signs of lethargy and diarrhea. Survival was assessed everyday for
a total of 7 days.
Treatments
Mice were treated with an intraperitoeal dose of rPAF-AH (1.0, 5.0 mg kgj1) or
the PAF antagonist WEB 2170 (20 mg kgj1) 15 min after surgery or LPS
administration. In another set of experiments, the animals treated with rPAF-AH
(1 mg kgj1) received an injection of the antibiotic imipenem (100 mg kgj1) 3 h
after CLP. Pentaglobin was administered to a different group of animals at
250 mg kgj1 i.p. 15 min after CLP. The same amount of Pentaglobin was
administered to sham-operated animals and showed no effect on the survival curve.
Plasma samples
Blood samples were collected from patients with sepsis in tubes containing
3.2% sodium citrate no later than 48 h after diagnosis. The plasma fraction was
separated by centrifugation at 800g for 10 min, and the samples were frozen and
stored at j20-C for assessment of PAF-AH activity. Plasma samples from
experimental animals were obtained at designated time points after either the CLP
procedure or LPS administration (at 0, 3, 6, 12, 24, 48, 72, and 96 h). Mice were
anesthetized with isoflurane, and blood (0.2 mL) was collected from the retro-
orbital venous plexus in tubes containing 20 2L of 3.2% sodium citrate and
euthanized immediately after the procedure in a carbon dioxide (CO2) chamber.
Thirty animals were either injected with LPS or subjected to CLP and divided in
groups of 10 animals. Blood sampling was performed twice in each group with at
least 12-h interval between retro-orbital venous punctures (20 2L in each
puncture). The plasma fraction was separated from the cellular components by
centrifugation at 800g for 10 min and then was stored at j20-C for PAF-AH
activity determinations.
Cytokine measurements
The magnitude of the inflammatory response was evaluated by measuring the
levels of interleukin-6 (IL-6), MCP-1/CCL2, and migration inhibitory factor (MIF) in
the peritoneal fluid, using enzyme-linked immunosorbent assay technique with
specific monoclonal antibodies, according to manufacturer instructions (Duo Set Kit
from R&D systems). Mice were killed in a CO2 chamber at designated time points,
and the peritoneal cavity was opened and rinsed with Hanks balanced saline solution
(HBSS) without calcium (Ca2+) or magnesium (Mg2+). The particulate matter was
removed by centrifugation at 800g for 10 min, and the supernatant fractions were
used for immunoassays. All the measurements were performed in duplicate.
Cell counting
At chosen time points, animals injected with LPS (300 2g) or subjected to CLP
were killed in a CO2 gas chamber, the peritoneal cavity was opened and washed
with 1 mL of Ca2+/Mg2+-free HBSS. The peritoneal wash was recovered, and the
volume was measured in a graduated syringe. Peritoneal wash aliquots were
collected and diluted in Turk solution (2% acetic acid) for total leukocyte count in
Neubauer chambers. Differential leukocyte analysis was performed on cytocentri-
fuged smears stained by the May-Grunwald-Giemsa method. Lipid body formation
in leukocytes was evaluated on cytosmears. The cells were centrifuged (550 rpm/
5 min) in a cytocentrifuge (Shandon cytospin III) onto glass slides (105 cells per
slide) and stained with osmium tetroxide (OsO4). While still moist, leukocytes on
SHOCK JULY 2006 RPAF-AH PROTECTS MICE AGAINST SEPSIS 43
TABLE 1. Characteristics of patients with septic shock and sepsis AH activities in both the patients with sepsis and septic shock
Septic shock Sepsis compared with the controls, and that in some patients, plasma
No. patients 37 17 PAF-AH activity was markedly depressed (Fig. 1A). This
Age (y) 60.7 58.5
finding suggests that there are multiple factorsVboth positive
and negativeVthat influence the level of PAF-AH activity at
Sex (% men/% women) 65/35 58.8/41.2
a given time.
APACHE II 22.3 16.7 To determine if endogenous plasma PAF-AH activity is
Observed mortality at 28 d (%) 54.0 17.6 altered in response to inflammatory insults, we examined the
Sites of infection, n (%) activity of plasma PAF-AH in well-characterized murine
Lung 20 (54.0) 9 (52.9) models of SIRS and sepsis, namely, endotoxemia induced by
Abdomen 10 (27.0) 4 (23.5)
LPS and systemic polymicrobial infection induced by CLP.
We found that plasma PAF-AH activity decreased in both
Blood 2 (5.4) 3 (17.6)
models (Fig. 1, B and C). In both, there was an initial decline
Other 5 (13.5) 1 (5.9)
Microbiological data, n (%)
Gram-negative bacteria 17 (65.4) 7 (53.8)
Gram-positive bacteria 5 (19.2) 3 (23.1)
Polybacterial 3 (11.5) 2 (15.4)
Fungi 1 (3.8) 1 (7.7)
Positive cultures 26 (70.3) 13 (76.5)
Positive blood cultures 10 (27.0) 4 (23.5)

cytospin slides were fixed in 3.7% formaldehyde in Ca2+/Mg2+-free HBSS, pH

7.4, rinsed in 0.1 mol Lj1 cacodylate buffer, stained with 1.5% OsO4 (30 min),
rinsed in distilled water, immersed in 1.0% thiocarbohydazide (5 min), rinsed in
0.1 mol Lj1 cacodylate buffer, restained in 1.5% OsO4 (3 min), rinsed in
distilled water, and then dried and mounted. The morphology of fixed cells was
observed, and lipid bodies were enumerated by light microscopy with a
100
objective lens in 50 consecutively scanned leukocytes.

Statistical analyses
The Kruskal-Wallis and Mann-Whitney U tests were used to compare
nonparametric data from multiple groups. Log-rank tests were used for comparison
of mortality rates. A P value of G0.05 was considered statistically significant. All
data are expressed as mean T SEM.

RESULTS
Plasma PAF-AH activity is reduced in patients with septic
shock and in mice undergoing experimental SIRS
or sepsis
We first examined the levels of PAF-AH activity in the
plasma of patients with sepsis in two different intensive care
units in Rio de Janeiro, Brazil, within 48 h of diagnosis.
Sepsis and septic shock were defined as described in the
recommendations of the Consensus Conference of the
American College of Chest Physicians. Demographic and
clinical characteristics are shown in Table 1.
We found (Fig. 1A) that mean values for plasma PAF-AH
activity were moderately decreased in patients with sepsis
(median, 11.7 nmol mLj1 minj1) and septic shock (median,
12.4 nmol mLj1 minj1) compared with control volunteers
(median, 16.7 nmol mLj1 minj1). This difference reached
statistical significance in patients with shock (P G 0.001).
Eighteen (49%) of 37 patients with septic shock and 10 (59%)
of 17 patients with sepsis had plasma PAF-AH activity levels
equal to or below the lowest activities assayed in control
subjects (Fig. 1A). Although the magnitude of the mean
changes was moderate, we observed a wide spectrum of PAF-
FIG. 1. Plasma PAF-AH activity is reduced in patients with septic
shock and in two models of experimental sepsis. A, Values for plasma
PAF-AH activity in patients with septic shock (n = 37) or sepsis (n = 17) and
in healthy subjects (n = 11). Median PAF-AH activity was 12.4 nmol mLj1
minj1 in patients with septic shock, 11.7 nmol mLj1 minj1 in patients with
sepsis, and 16.7 nmol mLj1 minj1 in healthy controls. The difference
between patients with septic shock and controls was statistically significant
(*P G 0.001). B, Swiss mice were subjected to the CLP procedure and PAF-
AH activity in plasma samples were measured at the time shown. C, Swiss
mice were injected intraperitoneally with a lethal dose of LPS (300 2g), and
plasma PAF-AH activity was measured. We obtained blood by cardiac
puncture at the indicated time intervals and then determined the levels of
plasma PAF-AH activity using a commercial assay as described in Materials
and Methods. Twelve animals were included in each group at the beginning
of the experiment, and three animals were still alive at the end the experiment.
Horizontal lines indicate the mean PAF-AH activity at each time point. *Statisti-
cally significant differences between activity at different experimental times and
the activity at the beginning of the experiment (time 0, P G 0.001).

Copyright @ 2006 by the Shock Society. Unauthorized reproduction of this article is prohibited.
44 SHOCK VOL. 26, NO. 1 GOMES ET AL.

FIG. 2. Recombinant PAF-AH increases the survival and inhibits neutrophil migration and lipid body formation in mice challenged with a lethal
dose of LPS. A, Swiss mice were injected intraperitoneally with rPAF-AH (1 mg kgj1, n = 10) or vehicle 15 min after LPS administration (300 2g, n = 10). The
survival rate was monitored for 7 days. At least 10 mice were included in each treatment group at the beginning of the experiment. This figure illustrates data
from one of two experiments with similar results. Vehicle-injected mice showed no mortality during the duration of the experiment (not shown). *Statistically
significant differences when compared with LPS + vehicle group (P = 0.039, log-rank test). B, In a second group of experiments, peritoneal lavage fluid was
obtained 6 h after the LPS administration, and the number of neutrophils was determined. C, The number of lipid bodies per cell was determined in OsO4-
stained cytocentrifuged smears. Each column is the mean of measurements from six animals with SEM indicated by a vertical line. *Statistically significant
differences when compared with saline + vehicle group. † Statistically significant differences when compared with LPS + vehicle group.

in the levels of PAF-AH activity that persisted up to 48 h.
There was then recovery in the activity levels over the next
48 h in animals that survived CLP (Fig. 1B). In contrast, the
plasma activity remained depressed in animals injected with
LPS (Fig. 1C). The results in the CLP model suggest that
recovery of endogenous plasma PAF-AH activity contributes
to the survival of some animals.
Administration of rPAF-AH reduces inflammation and
protects mice from death in response to systemic LPS
To further examine the role of PAF-AH in experimental SIRS,
we used rPAF-AH to supplement the endogenous activity to
normal or increased levels. To first assess the stability of the
recombinant protein in vivo, we injected mice with rPAF-AH
intravenously or intraperitoneally and then measured the activity
in the plasma at various times after injection. In both cases, the
administration of 1 mg kgj1 of rPAF-AH was followed by a
rapid 3-fold increase in enzymatic activity (from 0.025 in control
animals to 0.078 2mol mLj1 hj1) that was detectable after 5 min
and that was sustained for at least 24 h. We chose this dose based
on previous work, where the same dose was used to treat patients
with sepsis (14,15). We next challenged mice with a lethal
injection of LPS (300 2g i.p.) and, 15 min later, administered
rPAF-AH (1 mg kgj1). We observed that 1 mg kgj1 rPAF-AH
dramatically decreased mortality induced by LPS (Fig. 2).
Treatment with 1 mg kgj1 of rPAF-AH also modified the
inflammatory response to LPS. As shown in Figure 2B,
intraperitoneal injection of LPS induced a significant accu-
mulation of neutrophils that was inhibited by treatment with
rPAF-AH (1 mg kgj1), indicating an anti-inflammatory effect
of this enzyme. Morphological analysis of the cells recovered
from the peritoneal cavity revealed that LPS induced an
increase in the number of lipid bodies and that this response
was completely inhibited by treatment with rPAF-AH (Fig. 2C).
Lipid bodies are cytoplasmic inclusions that form in
leukocytes participating in an inflammatory response (16).
The ability of rPAF-AH to reduce the number of neutrophils
and lipid bodies in the peritoneum after LPS stimulation is
additional evidence that this enzyme has a major anti-
inflammatory effect.
To further explore the effects of rPAF-AH, we examined
the levels of MCP-1/CCL2 and IL-6, mediators that influence
the outcome of experimental sepsis, in the peritoneal fluid
from animals challenged with LPS 6 h earlier. Treatment with
rPAF-AH resulted in increased levels of MCP-1/CCL2 and
lower IL-6 concentrations in the peritoneal wash (Table 2).
Administration of rPAF-AH reduces inflammation and
protects mice subjected to CLP from death
Our next goal was to determine if rPAF-AH also has
beneficial effects in a more physiologically relevant model
involving polymicrobial infection in mice subjected to CLP.
In preliminary experiments using C57BL6 mice, 100% of the
control animals (9/9) died within 72 h, whereas there was 75%
survival (6/8 animals) in those given rPAF-AH immediately
before CLP (not shown). Subsequent experiments were then
done in outbred animals of the Swiss strain in which control
mortality, approximately 80% at 72 h (Fig. 3A), was high but
less than that of C57BL6 mice in the preliminary study. In
Swiss mice, administration of rPAF-AH (1 mg kgj1) 15 min
after CLP increased survival to approximately 50% from the
control value of about 20% (Fig. 3A). This result provides
clear evidence that the PAF signaling system plays key roles
in the pathogenesis of this model of sepsis. To further address
this issue using an alternative approach, we asked if blockade
of the PAF receptor alters survival in parallel. Administration
of the PAF receptor antagonist WEB 2170 (20 mg kgj1 i.p.)
at a dose previously shown to antagonize systemic effects of
PAF (17), resulted in moderate protection that did not reach
statistical significance (Fig. 3B). These results indicate that a
molecule that can modify PAF receptor ligands (i.e., PAF or
PAF-like lipids) is protective in CLP-induced sepsis. The
results also suggest that degradation of PAF ligands may be
a more efficacious therapeutic strategy than competitive
blockade of the receptor (7), although the difference between
the outcomes in animals treated with rPAF-AH and the

Copyright @ 2006 by the Shock Society. Unauthorized reproduction of this article is prohibited.
SHOCK JULY 2006 RPAF-AH PROTECTS MICE AGAINST SEPSIS 45
TABLE 2. Levels of cytokines in peritoneal fluid after PAF-AH treatment
CLP LPS
Vehicle CLP + Vehicle CLP + PAF-AH Vehicle LPS + Vehicle LPS + PAF-AH
(ng mLj1) (ng mLj1) (ng mLj1) (ng mLj1) (ng mLj1) (ng mLj1)
IL-6 0.01 T 0.00 13.35 T 7.50* 12.91 T 4.52 0.01 T 0.00 17.37 T 3.47* 7.38 T 1.22‡
†
MIF 0.01 T 0.00 59.02 T 3.37* 27.05 T 5.32 V V V
†
MCP-1/CCL-2 0.01 T 0.00 0.82 T 0.16* 1.75 T 0.06 0.01 T 0.00 0.92 T 0.03* 1.35 T 0.08‡
Swiss mice were injected intraperitoneally with rPAF-AH (1 mg kgj1) or vehicle 15 min after the CLP procedure or LPS administration. After 6 h, the
peritoneal lavage fluid was obtained and used for cytokine determination by enzyme-linked immunosorbent assay. A group of naive animals that
received an intraperitoneal injection of saline was used as control. The values are the mean T SEM from 6 animals.
*Statistically significant differences when compared with vehicle group. †Statistically different from CLP + vehicle group. ‡Statistically different from
LPS + vehicle group.

receptor antagonist did not reach statistical significance.
Importantly, the administration of rPAF-AH 6 h after CLP
showed a protective effect on mortality that was similar to that
observed when it was administered 15 min after CLP (Fig. 3C).
This finding establishes the possibility that rPAF-AH may
have utility as a therapeutic measure for some individuals
with SIRS or sepsis.
Antibiotics are central to the treatment of sepsis (3, 18).
Therefore, we asked if combined administration of rPAF-AH
and a broad-spectrum antibiotic, imipenem, further increased
survival. This yielded additional protection (70% survival,
Fig. 3B). The administration of imipenem alone resulted in
50% survival. These results indicate that exogenous rPAF-AH
increases survival under conditions in which antibiotic treat-
ment reduced mortality, a situation more relevant to clinical
management. In contrast, anti-LPS polyclonal antibodies
failed to show any beneficial effect when administered alone,
in combination with imipenem as a control and comparative
condition (not shown). Importantly, the same amount of
polyclonal antibodies was administered to controls, and
controls showed no sign of toxicity.
Treatment with rPAF-AH also significantly altered peri-
toneal inflammation in animals subjected to CLP. As with
intraperitoneal LPS, neutrophils accumulate in the peritoneal
cavity 6 h after CLP. This response was dramatically reduced
by rPAF-AH treatment (Fig. 4A). In addition, we again found
that treatment with rPAF-AH reduced the number of lipid
bodies in the cells recovered from the peritoneal cavity
(Fig. 4B). Thus, rPAF-AH altered the inflammatory response
in a complex model of microbial challenge (Figs. 3 and 4) as
it did in a model based on challenge with LPS alone (Fig. 2).
Moreover, the results with the CLP model are further
evidence for a central role for PAF in cell activation, mi-
gration, and mortality in a surrogate model of sepsis.
We examined levels of IL-6, MIF, and MCP-1/CCL2
because these mediators have been used as markers of severity
and progression of sepsis (19) and play important roles in its
pathophysiology. Furthermore, the PAF and IL-6 signaling

FIG. 3. Combined administration of recombinant PAF-AH and a broad-spectrum antibiotic affords maximal protection after CLP. A, Swiss mice
were injected intraperitoneally with rPAF-AH (1 mg kgj1, n = 30, n = 10) or vehicle (n = 30) 15 min after the CLP procedure. B, Swiss mice were injected
intraperitoneally with rPAF-AH (1 mg kgj1, n = 30), imipenem (100 mg kgj1, n = 10), PAF-AH plus imipenem (n = 10), WEB 2170 (20 mg kgj1, n = 10), or
vehicle (n = 30) 15 min after the CLP procedure. C, Swiss mice were treated intraperitoneally with rPAF-AH (1 mg kgj1, n = 10) 6 h after CLP. Sham-operated
animals (n = 10) also received intraperitoneal injection of vehicle and showed no mortality during the duration of the experiment (not shown). The survival rate
was monitored for 7 days. *Statistically significant differences when compared with CLP + vehicle group (P G 0.03, log-rank test). This figure shows results from
one of three repetitions with similar results.

Copyright @ 2006 by the Shock Society. Unauthorized reproduction of this article is prohibited.
46 SHOCK VOL. 26, NO. 1 GOMES ET AL.

significantly reduced when the animals were treated with a

combination of rPAF-AH and imipenem (Fig. 5). These
results further indicate that combination therapy with rPAF-
AH and antibiotics modulates inflammatory responses in the
surrogate CLP model.
Increased MCP-1/CCL2 levels were observed in animals
treated with rPAF-AH (Table 2) and could be associated with
the protective effect of rPAF-AH on mortality caused by
CLP (20). To further explore this possibility, we examined
the effect of rPAF-AH in animals genetically deficient in the
receptor for MCP-1/CCL2, CCR2 (21). We found that the
administration of rPAF-AH to deficient mice (CCR2j/j)
failed to reproduce the same protective effect observed in
wild-type controls (Fig. 6) consistent with a modulatory role
for signaling by MCP-1/CCL2 and engagement of CCR2 by
this ligand.

FIG. 4. Recombinant PAF-AH inhibits intraperitoneal neutrophil accu-
mulation and lipid body formation in mice subjected to CLP. A, Swiss mice
were injected intraperitoneally with rPAF-AH (1 mg kgj1) or vehicle 15 min after
the CLP procedure. After 6 h, peritoneal lavage fluid was obtained, and the
number of neutrophils was determined. B, The number of lipid bodies per cell
was determined in OsO4-stained cytocentrifuged smears. A group of naive
animals that received an intraperitoneal injection of the vehicle for PAF-AH
was used as control. Each column is the mean from six animals with SEM
indicated by a vertical line. *Statistically significant differences when compared
with control. #Statistically different from CLP + vehicle group.
systems are linked in inflammatory responses (7). CLP
induced an increase in the peritoneal fluid levels of MCP-1/
CCL2, IL-6, and MIF (Table 2). Treatment with rPAF-AH
alone resulted in decreased levels of MIF but failed to alter
IL-6 concentration in the peritoneal fluid or in the plasma (not
shown). The administration of the broad-spectrum antibiotic
imipenem alone did not reduce IL-6 levels; however, the
greatest variation in IL-6 concentrations occurred in these
animals (Fig. 5) and may have blunted our ability to detect
moderate changes. In contrast, the levels of IL-6 were
DISCUSSION
In this report, we characterized endogenous PAF-AH in
sepsis and tested the hypothesis that exogenous recombinant
PAF-AH terminates pathologic signals. Humans with sepsis
have reduced levels of endogenous PAF-AH that varies among
subjects and over time (our unpublished data) (7, 13, 22). Mice
subjected to systemic LPS or ligation and perforation of the
cecum developed septic manifestations had altered levels of
inflammatory mediators and leukocyte counts and high
mortality rates and, in parallel, had depressed levels of plasma
PAF-AH that varied with the inflammatory challenge and the
time after its initiation. The patterns of PAF-AH activity
levels over time and in response to different challenges in the
animal models (Fig. 1, B and C) clearly demonstrate that this
is a dynamic variable in experimental sepsis. Variability in
activity levels in individual patients with sepsis (14) and time-
dependent changes in PAF activity in tissue and plasma
samples from human subjects (22, 23) are consistent with
these findings and indicate that they are clinically relevant.
These observations establish a new pathophysiological feature
of experimental sepsis and add to evidence that the PAF
signaling system is altered in septic syndromes (7Y10).
Similar time-dependent changes in endogenous PAF-AH
activity that are dictated by both genetic and environmental
factors (7, 11) may be critical variables in human septic
syndromes. The CLP model is well characterized, has been

FIG. 5. IL-6 levels are reduced in the peritoneal lavage fluid of mice treated with a combination of recombinant PAF-AH and imipenem. Swiss mice
were subjected to CLP and treated with intraperitoneal injections of rPAF-AH (1 mg kgj1), imipenem (100 mg kgj1) or a combination of both after 15 min.
Sham-operated animals also received intraperitoneal injection of vehicle. Six hours after the procedure, the animals were killed in a CO2 chamber, and the
peritoneal cavities were opened and rinsed with PBS. Cells were collected by centrifugation, and IL-6 was measured in the supernatant using a commercially
available enzyme-linked immunosorbent assay protocol. Each column is the mean T SEM from six animals. This experiment was repeated a second time with
similar results. All the measurements were performed in duplicate. *Statistically significant differences when compared with control. #Statistically different from
CLP + vehicle group.

Copyright @ 2006 by the Shock Society. Unauthorized reproduction of this article is prohibited.
SHOCK JULY 2006
FIG. 6. Treatment with rPAF-AH failed to protect CCR2 knockout mice from CLP-induced mortality. CCR2j/j (n = 8) and wild-type (n = 8) mice were
injected intraperitoneally with rPAF-AH (1 mg kgj1) or vehicle (n = 10) 15 min after the CLP procedure. Sham-operated animals (n = 10) also received an
intraperitoneal injection of vehicle and showed no mortality during the duration of the experiment (not shown). The survival rate was monitored for 7 days.
widely used to study the pathophysiological mechanisms that
underlie sepsis, and is widely used in the evaluation of
potential new therapies for the treatment of sepsis (10, 24, 25).
Using this surrogate model and a second model based on
challenge with LPS (26), we demonstrated that increased
levels of plasma PAF-AH activity attained by administration
of the recombinant enzyme decrease mortality in experimental
murine SIRS and sepsis. This represents a novel approach to
therapy of sepsis and related disorders in which the strategy is
based on enzymatic degradation of pathologically generated
signaling molecules and is different from other biologic
therapies currently under investigation (7, 10, 27). Extensive
evidence supports the idea that the PAF signaling system
plays an important role in the inflammatory response in sepsis
(7, 9). Consequently, a number of strategies for altering PAF
signaling have been examined as potential treatments for
sepsis including blockade of the PAF receptor by specific
antagonists (7, 9, 28, 29). In agreement with earlier reports in
several models, we observed moderate protection by the PAF
receptor antagonist WEB 2170 in the CLP model (Fig. 4).
Although in our study, a dose-response curve was not
constructed for this antagonist, the dose used in this experi-
ment was previously shown to completely inhibit systemic
effects after the administration of high doses of PAF (17).
A new approach to block the effects of PAF and PAF-like
lipids became available when Tjoelker et al (30). cloned
plasma PAF-AH and showed an anti-inflammatory activity of
the recombinant protein (7, 11). This strategy may have
advantages over the use of receptor antagonists, including
high efficiency of PAF-AH in degrading PAF and related
oxidized phospholipids to products not recognized by the PAF
receptor (31), and lack of a requirement to achieve a favorable
stoichiometric relationship among the PAF or PAF-like lipid
ligands, the receptor, and the blocking antagonist (7). One
advantage is that a single intraperitoneal or intravenous
injection of rPAF-AH is sufficient to maintain elevated
PAF-AH activity in the plasma for at least 24 h, thereby
providing continued protection from the effects of PAF and
PAF-like lipids (see Results). Plasma PAF-AH is known to be
associated with low-density lipoprotein and, to a much lower
extent, with high-density lipoprotein (HDL) in human
blood (32). The mouse PAF-AH has an intrinsically low
Copyright @ 2006 by the Shock Society. Unauthorized reproduction of this article is prohibited.
RPAF-AH PROTECTS MICE AGAINST SEPSIS 47
affinity for apolipoprotein B100, the major lipoprotein
associated with low-density lipoprotein, and is found almost
entirely in the HDL fraction of murine plasma (32). We
speculate that rPAF-AH dominantly localizes to HDL in mice
and that this association is responsible for the stability of
activity and slow clearance rate in murine blood reported here.
Our analysis of plasma samples from animals with sepsis
demonstrated that there is heterogeneity in alterations in
endogenous PAF-AH activity levels among animals with
sepsis. These results are consistent with studies by Graham
et al. (33) who reported decreased activity in some subjects
and that the half-life of PAF was prolonged in patients who
died compared with survivors or normal volunteers. These
suggest that individuals who are unable to mount a robust
PAF-AH response may be the most likely to benefit from
administration of rPAF-AH. Consistent with this interpreta-
tion, mice that survived the CLP procedure showed a late
increase in PAF-AH levels (Fig. 2A). In addition to CLP, we
observed a marked decrease in systemic levels of PAF-AH
activity in animals injected with LPS (Fig. 2B) and in mice
infected with Neisseria meningitidis (data not shown).
Decreased systemic PAF-AH activity levels have also been
described in a number of diseases with components of
inflammation, including systemic lupus erythematosus (34),
necrotizing enterocolitis (35), and asthma (36).
Multiple factors alter the levels of plasma PAF-AH in
inflammation (11). LPS down-regulates expression of PAF-AH
in cell culture at the transcriptional (12), post-transcriptional
(12), and activity (37) levels in some, but not all, conditions (11),
and oxidation of PAF-AH results in decreased activity
(7, 11, 32). In contrast, up-regulation of PAF-AH activity has
been reported in response to inflammatory stimuli (38). In
addition, there may be compartmentalized changes in PAF-AH
activity. Recently, Grissom et al. (22) observed an elevated
expression of PAF-AH mRNA and activity in the bronchoal-
veolar lavage of patients with acute respiratory distress
syndrome, including some subjects with sepsis. Although the
precise mechanisms are yet to be determined, in the aggregate,
these observations indicate that inflammatory pathways can
have differential and time-dependent effects on PAF-AH
activity. Here, we used two independent models of systemic
inflammation and, in both instances, we documented a marked
48 SHOCK VOL. 26, NO. 1
decrease in the levels of circulating PAF-AH activity and then a
return of activity that may have influenced the outcome of
experimental septic challenge. Moreover, a significant decrease
in PAF-AH activity was also detected in a group of human
patients with septic shock. Our combined results indicate that
restoring the levels of plasma PAF-AH activity has beneficial
effects in the modulation of host responses to bacterial
components in these models.
Antimicrobial drugs are central to the management of sepsis,
and we found that administration of rPAF-AH has a protective
effect in animals receiving an efficacious antibiotic regimen that
may be additive to the beneficial effect of the antibiotic alone
(Fig. 3). The release of proinflammatory bacterial products such
as LPS occurs as a consequence of bacterial death that results
from antibiotic treatment. These products trigger a systemic
inflammatory response, including production and release of
PAF, cytokines, and other mediators (10). Thus, PAF-AH may
limit the intensity of the inflammatory response by decreasing
the availability of PAF and PAF-like lipids and thereby act in
an additive or synergistic fashion with antibiotic therapy.
Our studies also demonstrate that animals that received
rPAF-AH had altered cytokine levels and decreased leukocyte
accumulation and activation. We observed a significant
decrease in IL-6 in animals treated with rPAF-AH in
combination with antibiotic treatment (Fig. 6). Interestingly,
rPAF-AH alone did not significantly reduce IL-6 levels, which
probably indicates that although provided with clear anti-
inflammatory effects, rPAF-AH is not sufficient to completely
ablate systemic inflammation in sepsis. IL-6 is a severity
marker of sepsis and other inflammatory syndromes (39), and
this finding reinforces the possibility that a combined regimen
may have advantages over the use of either antibiotics or rPAF-
AH alone. Moreover, rPAF-AH administration was also
associated with increased levels of MCP-1/CCL2 in animals
with sepsis. Previously, MCP-1/CCL2 was reported to have a
protective effect in murine endotoxemia (20). Administration of
blocking monoclonal antibodies (40) or genetic deficiency of
MCP-1/CCL2 increases mortality rates in mice with sepsis
(Gomes et al., manuscript in preparation). In addition, admin-
istration of MCP1/CCL2 significantly decreased mortality in
these same surrogate models (20). Although the precise
molecular mechanism for this protective effect is not com-
pletely understood, the fact that MCP-1/CCL2 levels are further
increased in both LPS and CLP models after rPAF-AH
administration may contribute to the decreased mortality that
we observed under this condition. This interpretation is
reinforced by our observation that rPAF-AH failed to decrease
mortality in animals that were genetically deficient in the MCP-
1/CCL2 receptor (CCR2j/j; Fig. 6), further suggesting that
PAF signaling is central to cytokine response in sepsis.
Treatment with rPAF-AH was also associated with a
decrease of MIF levels measured in animals subjected to
CLP (Table 2). Previously, neutralization of MIF with anti-
bodies led to better survival in a murine CLP model and after
peritoneal infection with Escherichia coli (41). Recently, it has
been shown that MIF regulates the expression of toll-like
receptor 4, the signal-transducing molecule of the LPS receptor
complex (42). In addition, we have recently shown that MIF is
Copyright @ 2006 by the Shock Society. Unauthorized reproduction of this article is prohibited.
GOMES ET AL.
a better indicator of poor outcome in patients with sepsis than is
IL-6 and can be used in the management of critically ill patients
(43). Together, our findings indicate that treatment with rPAF-
AH has beneficial modulatory effects on the cytokine response
in experimental murine sepsis, and that it favorably alters
cytokine variables that are relevant to human septic syndromes.
In summary, we observed that plasma PAF-AH activity is
reduced in humans with septic shock and in two murine
models of sepsis and systemic inflammation. Increasing
plasma PAF-AH activity by administration of the recombinant
enzyme significantly protected animals from death in the
experimental sepsis conditions, alone or in combination with
an intervention commonly used in the management of clinical
sepsis condition, a broad-spectrum antibiotic.
Our studies further indicate that genetic, environmental, or
iatrogenic features yet to be completely defined alter endoge-
nous PAF-AH activity and may influence the response to
administration of rPAF-AH, and that the level of endogenous
plasma PAF-AH is a time-dependent variable in experimental
murine sepsis. Because such variability also occurs in humans
with septic syndromes (7, 23), these observations are con-
sistent with the suggestion that rPAF-AH may be a useful
agent in the treatment of subgroups of these patients.
ACKNOWLEDGMENTS
The authors thank Dr. Marcelo Bozza from the Department of Microbiology,
Federal University of Rio de Janeiro, Brazil, for helpful discussions, and Dra.
Claudia Benjamin and Dr. Willian A. Kuziel for CCR2 knockout animals. PTB is
an international scholar from Howard Hughes Medical Institute.
REFERENCES
1. Angus DC, Wax RS: Epidemiology of sepsis: an update. Crit Care Med
29:S109YS116, 2001.
2. Hotchkiss RS, Karl IE: The pathophysiology and treatment of sepsis. N Engl J
Med 348:138Y150, 2003.
3. Cohen J: The immunopathogenesis of sepsis. Nature 420:885Y891, 2002.
4. Levy MM, Fink MP, Marshall JC, et al: 2001 SCCM/ESICM/ACCP/ATS/SIS
International Sepsis Definitions Conference. Crit Care Med 31:1250Y1256,
2003.
5. Glauser MP, Zanetti G, Baumgartner JD, et al: Septic shock: pathogenesis.
Lancet 338:732Y736, 1991.
6. Lopez Diez F, Nieto ML, Fernandez-Gallardo S, et al: Occupancy of platelet
receptors for platelet-activating factor in patients with septicemia. J Clin Invest
83:1733Y1740, 1989.
7. Zimmerman GA, McIntyre TM, Prescott SM, et al: The platelet-activating
factor signaling system and its regulators in syndromes of inflammation and
thrombosis. Crit Care Med 30:S294YS301, 2002.
8. Imaizumi TA, Stafforini DM, Yamada Y, et al: Platelet-activating factor: a
mediator for clinicians. J Intern Med 238:5Y20, 1995.
9. Chang SW, Feddersen CO, Henson PM, et al: Platelet-activating factor
mediates hemodynamic changes and lung injury in endotoxin-treated rats.
J Clin Invest 79:1498Y1509, 1987.
10. Marshall JC: Such stuff as dreams are made on: mediator-directed therapy in
sepsis. Nat Rev Drug Discov 2:391Y405, 2003.
11. Stafforini DM, McIntyre TM, Zimmerman GA, et al: Platelet-activating factor,
a pleiotrophic mediator of physiological and pathological processes. Crit Rev
Clin Lab Sci 40:643Y672, 2003.
12. Cao Y, Stafforini DM, Zimmerman GA, et al: Expression of plasma platelet-
activating factor acetylhydrolase is transcriptionally regulated by mediators of
inflammation. J Biol Chem 273:4012Y4020, 1998.
13. Tjoelker LW, Eberhardt C, Unger J, et al: Plasma platelet-activating factor
acetylhydrolase is a secreted phospholipase A2 with a catalytic triad. J Biol
Chem 270:25481Y25487, 1995.
14. Schuster DP, Metzler M, Opal S, et al: Recombinant platelet-activating factor
SHOCK JULY 2006
acetylhydrolase to prevent acute respiratory distress syndrome and mortality in
severe sepsis: phase IIb, multicenter, randomized, placebo-controlled, clinical
trial. Crit Care Med 31:1612Y1619, 2003.
15. Opal S, Laterre PF, Abraham E, et al: Recombinant human platelet-activating
factor acetylhydrolase for treatment of severe sepsis: results of a phase III,
multicenter, randomized, double-blind, placebo-controlled, clinical trial. Crit
Care Med 32:332Y341, 2004.
16. Pacheco P, Bozza FA, Gomes RN, et al: Lipopolysaccharide-induced
leukocyte lipid body formation in vivo: innate immunity elicited intracellular
loci involved in eicosanoid metabolism. J Immunol 169:6498Y6506, 2002.
17. Castro-Faria-Neto HC, Bozza PT, Martins MA, et al: Platelet mobilization
induced by PAF and its role in the thrombocytosis triggered by adrenaline in
rats. Thromb Haemost 62:1107Y1111, 1989.
18. Vianna RCS, Gomes RN, Bozza FA, Amâncio RT, et al: Antibiotic treatment
in a murine model of sepsis: impact on cytokines and endotoxin release. Shock
21:115Y120, 2004.
19. Hack CE, Aarden LA, Thijs LG: Role of cytokines in sepsis. Adv Immunol
66:101Y195, 1997.
20. Zisman DA, Kunkel SL, Strieter RM, et al: MCP-1 protects mice in lethal
endotoxemia. J Clin Invest 99:2832Y2836, 1997.
21. Lu B, Rutledge BJ, Gu L, Fiorillo J, et al: Abnormalities in monocyte
recruitment and cytokine expression in monocyte chemoattractant protein
1Ydeficient mice. J Exp Med 187:601Y608, 1998.
22. Grissom CK, Orme JF Jr., Richer LD, et al: Platelet-activating factor
acetylhydrolase is increased in lung lavage fluid from patients with acute
respiratory distress syndrome. Crit Care Med 31:770Y775, 2003.
23. Claus RA, Russwurm S, Dohrn B, et al: Plasma platelet-activating factor
acetylhydrolase activity in critically ill patients. Crit Care Med 33:
1416Y1419, 2005.
24. Wichterman KA, Baue AE, Chaudry IH: Sepsis and septic shockVa review of
laboratory models and a proposal. J Surg Res 29:189Y201, 1980.
25. Hubbard WJ, Choudhry M, Schwacha MG, et al: Cecal ligation and puncture.
Shock 24(1):52Y57, 2005.
26. Remick DG, Ward PA: Evaluation of endotoxin models for the study of sepsis.
Shock 24(1):7Y11, 2005.
27. van der Poll T: Immunotherapy of sepsis. Lancet Infect Dis 1:165Y174, 2001.
28. Dhainaut JF, Tenaillon A, Le Tulzo Y, et al: Platelet-activating factor receptor
antagonist BN 52021 in the treatment of severe sepsis: a randomized, double-
blind, placebo-controlled, multicenter clinical trial. BN 52021 Sepsis Study
Group. Crit Care Med 22:1720Y1728, 1994.
Copyright @ 2006 by the Shock Society. Unauthorized reproduction of this article is prohibited.
RPAF-AH PROTECTS MICE AGAINST SEPSIS 49
29. Kruse-Elliott KT, Albert DH, Summers JB, et al: Attenuation of endotoxin-
induced pathophysiology by a new potent PAF receptor antagonist. Shock
5:265Y273, 1996.
30. Tjoelker LW, Wilder C, Eberhardt C, et al: Anti-inflammatory properties of a
platelet-activating factor acetylhydrolase. Nature 374:549Y553, 1995.
31. Marathe GK, Davies SS, Harrison KA, et al: Inflammatory platelet-activating
factor-like phospholipids in oxidized low density lipoproteins are fragmented
alkyl phosphatidylcholines. J Biol Chem 274:28395Y28404, 1999.
32. Stafforini DM, Tjoelker LW, McCormick SP, et al: Molecular basis of the
interaction between plasma platelet-activating factor acetylhydrolase and low
density lipoprotein. J Biol Chem 274:7018Y7024, 1999.
33. Graham RM, Stephens CJ, Silvester W, et al: Plasma degradation of platelet-
activating factor in severely ill patients with clinical sepsis. Crit Care Med
22:204Y212, 1994.
34. Tetta C, Bussolino F, Modena V, et al: Release of platelet-activating factor in
systemic lupus erythematosus. Int Arch Allergy Immunol 91:244Y256, 1990.
35. Caplan MS, Hsueh W: Necrotizing enterocolitis: role of platelet activating
factor, endotoxin, and tumor necrosis factor. J Pediatr 117:S47YS51, 1990.
36. Stafforini DM, Numao T, Tsodikov A, et al: Deficiency of platelet-activating
factor acetylhydrolase is a severity factor for asthma. J Clin Invest
103:989Y997, 1999.
37. Narahara H, Nishioka Y, Johnston JM: Secretion of platelet-activating factor
acetylhydrolase by human decidual macrophages. J Clin Endocrinol Metab
77:1258Y1262, 1993.
38. Memon RA, Fuller J, Moser AH, et al: In vivo regulation of plasma platelet-
activating factor acetylhydrolase during the acute phase response. Am J
Physiol 277:R94YR103, 1999.
39. Hack CE, De Groot ER, Felt-Bersma RJ, et al: Increased plasma levels of
interleukin-6 in sepsis. Blood 74:1704Y1710, 1989.
40. Matsukawa A, Hogaboam CM, Lukacs NW, et al: Endogenous monocyte
chemoattractant proteinY1 (MCP-1) protects mice in a model of acute septic
peritonitis: cross-talk between MCP-1 and leukotriene B4. J Immunol
163:6148Y6154, 1999.
41. Calandra T, Echtenacher B, Roy DL, et al: Protection from septic shock by
neutralization of macrophage migration inhibitory factor. Nat Med 6:164Y170,
2000.
42. Roger T, David J, Glauser MP, et al: MIF regulates innate immune responses
through modulation of toll-like receptor 4. Nature 414:920Y924, 2001.
43. Bozza FA, Gomes RN, Japiassu AM, et al: Macrophage migration inhibitory
factor levels correlate with fatal outcome in sepsis. Shock 22:309Y313, 2004.