Failure of the Neutrophil Response to Localized Infection | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
ABSTRACT |
|---|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
The major goal of this study was to investigate the mechanisms that link the host response to a local infection in the peritoneal cavity with the development of sepsis and lung injury. Rabbits were infected by intraperitoneal inoculation of fibrin clots containing Escherichia coli at 108, 109, or 1010 cfu/clot. Physiologic, bacteriologic, and inflammatory responses were monitored, and the lungs were examined postmortem. At a dose of 108 cfu/clot the animals had resolving infection, and a dose of 109 cfu/clot resulted in persistent infection at 24 h, with minimal systemic manifestations. In contrast, inoculation of 1010 cfu/clot resulted in rapidly lethal local infection, with septic shock and lung injury. The onset of septic shock was associated with a paradoxical lack of identifiable polymorphonuclear leukocytes (PMN; neutrophils) in the peritoneal cavity. The absence of PMN in the peritoneum was due in part to lysis of intraperitoneal PMN, because the peritoneal fluids contained free myeloperoxidase and induced rapid death of normal rabbit PMN in vitro. Although most animals became bacteremic, only those with a severe systemic inflammation response developed lung injury. These data show that control of an infection in the first compartment in which bacteria enter the host is a critical determinant of the systemic response. Above a threshold dose of bacteria, failure of the local neutrophil response is a key mechanism associated with deleterious systemic responses. Bacteremia alone is not sufficient to cause lung injury. Lung injury occurs only in the setting of a severe systemic inflammatory response and an inadequate leukocyte response at the primary site of infection.
| |
INTRODUCTION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Sepsis is one of the most important risk factors for the acute respiratory distress syndrome (ARDS) (1), but bacteremia alone is rarely associated with ARDS (2). Instead, the initiation of a systemic inflammatory response in the host appears to be a critical step in the pathogenesis of ARDS (3). The development of sepsis is associated with activation of complex cytokine cascades. However, strategies aimed at blocking single points in these cascades have failed to prevent ARDS or improve outcome in humans with sepsis (4). In order to develop more effective therapeutic strategies for sepsis, it is necessary to clarify the precise mechanisms that link an infection in a primary tissue compartment with systemic manifestations of inflammation and distant organ injury.
The site of primary infection influences whether or not severe sepsis develops. Pulmonary and abdominal infections are more likely to be associated with severe sepsis than are infections at other sites or bacteremia alone (9). In addition, the severity of the sepsis syndrome is related to the risk of developing ARDS (10, 11). The risk is greater among septic patients with positive blood cultures than among nonbacteremic patients (11), but severe systemic responses occur in many patients who do not have detectable bacteremia (1, 3, 12). Thus, the site of infection (e.g., lungs or peritoneum) and the ability to contain the infection at the primary site appear to be important factors affecting the severity of the sepsis syndrome and the likelihood of distant organ injury. Although there is evidence that the host inflammatory response is initially restricted to the primary site of infection (13), and that this compartmentalization is lost as the severity of inflammation increases (14), two important questions remain unanswered. (1) How do different primary sites of infection vary in their response to bacteria? (2) How do these differences influence the severity of systemic manifestations of local infection (e.g., sepsis and ARDS)?
We have addressed these questions in rabbit models of pneumonia and peritonitis. In a prior study, we induced pneumonia in rabbits by intratracheal instillation of Escherichia coli (15). There was an important association between the initial bacterial burden in the lungs, the local host response, and the development of the systemic manifestations of sepsis. In the present study we induced peritoneal infection by instilling into the peritoneal cavity a clot containing the same amounts of E. coli used in the pneumonia study. The main goal of the present study was to investigate the mechanisms that determine whether an infection remains localized to its primary site or evolves into a deleterious systemic inflammatory response. A second goal was to determine the local and systemic responses of the host to a peritoneal infection and to compare these responses with our prior findings about the responses to direct bacterial infection in the lungs. The results suggest that the effectiveness of the neutrophilic response at the site of infection is crucial in containing a local infection. Above a threshold bacterial inoculum, the local neutrophilic response is ineffective and systemic manifestations ensue. As compared with those of our prior pneumonia study, the data of the present study suggest that the lungs are more effective than the peritoneum in containing an infection and compartmentalizing the inflammatory response.
| |
METHODS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Bacteria and Clot Preparation
E. coli of serotype K-1 were originally isolated from the blood of a patient with biliary sepsis. This organism is virulent in rabbits, and the
methods used to pass and store the bacteria have been described (15).
Briefly, the bacteria were inoculated into the peritoneal cavity of a
specific-pathogen-free (SPF) New Zealand White (NZW) rabbit, and
after 24 h the animal's spleen was harvested aseptically and homogenized in 0.9% NaCl. The homogenate was diluted to 30% glycerol and
stored in separate aliquots at 
70° C. After each passage, the identity
of the bacteria in the homogenate was verified with standard microbiologic techniques, and rabbits were instilled intratracheally with the
bacteria according to the protocol in our original pneumonia study in
order to monitor any changes in bacterial virulence. For each experiment, a separate aliquot of the bacterial stock was thawed, inoculated
into Luria broth (LB; Gibco-BRL Laboratories, Gaithersburg, MD),
and incubated for 14 h at 37° C in a shaking incubator. The bacteria
were pelleted by centrifugation at 1,000 × g for 20 min, washed in
phosphate-buffered saline (PBS), and resuspended in 1.0 ml sterile
water. An infected clot was prepared by a modification of a previously
reported method (16, 17). The bacteria were diluted in 1% bovine fibrinogen (Sigma Chemical Co., St. Louis, MO) to produce a final bacterial concentration in the desired range (1 × 108 to 1 × 1010 cfu/clot).
Human thrombin (Sigma) was added at a final concentration of 2 U/ml,
and the mixture was allowed to clot for 20 min. The bacterial concentration in the suspension and the clot was confirmed in each experiment by quantitative culture with the pour-plate method, using LB
agar (Gibco-BRL) with overnight incubation at 37° C.
Animal Protocol
All experiments were approved by the Animal Research Committee of the Veterans Affairs/Puget Sound Health Care System. Female SPF NZW rabbits, weighing 3.0 to 3.5 kg (Wester Oregon Rabbit Company, Philomath, OR), were housed in the animal facility until the day of the experiment. Each rabbit was anesthetized with xylazine (0.33 mg/kg) and ketamine (15 mg/kg), and was then intubated endotracheally and allowed to breathe room air spontaneously. Arterial and venous catheters were placed in the right common carotid artery and jugular vein. A 1.5-cm incision was made in the linea alba above the umbilicus, the peritoneum was nicked with a scalpel, and the infected clot was placed into the peritoneal cavity through a 14-gauge catheter. The peritoneum and the skin incision were closed with sutures, and the animal was returned to its cage, placed on a heating pad, and allowed to recover from anesthesia.
Heart rate, respiratory rate, arterial blood pressure, and central
venous pressure were monitored continuously with a computerized system (Mac Labs Inc., Milford, MA). Core temperature and behavioral observations were recorded at 0, 1, 2, 3, 4, 12, and 24 h. Blood samples (1.0 ml) were drawn from the arterial catheter at 0, 2, 4, and
24 h for measurement of arterial blood gases (ABG), cell and differential counts, quantitative blood cultures, and serum cytokine measurements. The alveolar-arterial oxygen difference was calculated
with the equation [150 (PaCO2/0.8] PaO2 (where PaO2 is arterial oxygen tension). Total blood leukocytes were counted in a hemacytometer, and differential counts were made on peripheral blood smears
stained with Diff-Quik (Scientific Products, McGaw Park, NJ). Quantitative blood cultures were done with the pour-plate method, using
LB agar. Nonquantitative blood cultures also were done, by inoculating 10 ml of trypticase-soy broth with 100 µl blood. The remaining serum was stored at 70° C for cytokine measurements.
All animals were treated with a continuous infusion of 0.9% NaCl
given at a rate of 4 ml/h. Additional 10 ml boluses of 0.9% NaCl were
given every 15 min when the mean arterial blood pressure was < 75 mm Hg, the central venous pressure was < 10 cm H2O, or arterial
pH was < 7.30. Fluids were withheld if there was a sustained increase
in CVP of > 10 cm H2O.
At the end of the experiment, the animal was euthanized with 150 mg of pentobarbital given intravenously, and was exsanguinated by cardiac puncture. The thorax was opened rapidly, the lungs and heart were removed en bloc, and the lungs were dissected free. The trachea was cannulated, the right mainstem bronchus was clamped with a hemostat, and the left lung was lavaged with five separate 15-ml aliquots of 0.9% NaCl containing 0.6 mM ethylenediamine tetraacetic acid (EDTA). The right lung was fixed by intrabronchial instillation of 10% neutral buffered formaline at a transpulmonary pressure of 15 cm H2O, and was embedded in paraffin and processed for histologic analyses. Tissue sections were stained with hematoxylin and eosin. Immediately after thoracotomy, the peritoneal cavity was lavaged with 20 ml of 0.9% NaCl containing 0.6 mM EDTA.
The bronchoalveolar lavage fluid (BALF) and peritoneal fluid
(PF) from each animal were cultured by the pour-plate method, and
aliquots were taken for total and differential cell counts. The remainder of the fluid was centrifuged at 200 × g to pellet cells. Cell-free aliquots of the BALF and PF were stored at 70° C.
Experimental Protocols
We studied animals treated with intraperitoneal clots containing E. coli at three different doses: 1 × 108 cfu/clot (low dose, n = 5), 1 × 109 cfu/clot (intermediate dose, n = 5), and 1 × 1010 cfu/clot (high dose, n = 9). Control animals were treated with sterile clots (n = 4). Because all of the animals in the 1 × 1010 cfu/clot group died at 6 ± 0.7 (mean ± SD) h after clot placement, we studied an additional "early response" group of rabbits, consisting of animals euthanized at 4 h (control: n = 3; low dose: n = 3; intermediate dose: n = 5; high dose: n = 3). The data generated from the animals in the 1 × 1010 cfu/clot group that died at 6 ± 0.7 h were similar to those from the animals euthanized at 4 h, and were combined with the latter for subsequent analyses.
Measurement of Total Proteins and Cytokines
The total protein concentration in the BALF was measured with the
bicinchoninic acid (BCA) method (BCA assay; Pierce, Inc., Rockford, IL). Interleukin (IL)-8, granulocyte chemoattractant (GRO),
macrophage chemotactic protein (MCP)-1, and tumor necrosis factor
(TNF)-
in BALF and plasma samples were measured with rabbit-specific immunoassays as previously described (18, 19). The assay sensitivities were < 0.1 ng/ml for IL-8, GRO, MCP-1, and TNF-.
Myeloperoxidase Assay
Myeloperoxidase (MPO) activity in the PF was measured with a colorimetric assay, using guaiacol as substrate (Sigma). Briefly, 16.6 µl of PF were added to 1 ml of reaction buffer (0.4% H2O2 and 0.429 M guaiacol in 50 mM potassium phosphate buffer, pH 7). The change in optical density (OD) at 470 nm was measured over a period of 1 min in a Becton-Dickinson spectrophotometer, and the slope of the change in OD was calculated to generate the rate of change in units/ml/min. The results were adjusted for sample dilution to generate units/ml/ min. Human MPO (Sigma) was used as the standard.
Rabbit Neutrophil Isolation
Heparinized normal rabbit blood was centrifuged at 300 × g, and the
plasma was separated and saved. The cell pellet was resuspended in
6% dextran (Pharmacia, Piscataway, NJ) and 0.9% NaCl, and was incubated for 30 min at 37° C and then centrifuged at 350 × g for 10 min. The supernatant was discarded and the cell pellet was resuspended in the plasma that was saved from the initial centrifugation.
The resulting suspension was divided into 5-ml aliquots, which were
placed into separate 15-ml conical tubes. A density gradient was created by sequentially adding to the bottom of each tube 2.5 ml of iodixanol (OptiPrep; Nycomed Pharma AS, Oslo, Norway) at 1.085 g/ml,
followed by iodixanol at 1.086 g/ml. The tubes were then centrifuged
at 500 × g for 35 min at 22° C. The lower interface, containing PMN,
was removed and washed in 50 ml of 0.9% NaCl, and was then recentrifuged at 400 × g for 10 min. The resulting cell pellet was briefly resuspended in ice-cold 0.2% NaCl to lyse red blood cells, after which
an equal volume of 1.6% NaCl was added and the mixture was centrifuged at 400 × g. The PMN were then washed with 0.9% NaCl and
resuspended at 1 × 106 cells/ml in RPMI-1640 medium without phenol red (Bio Whittaker, Walkersville, MD) and containing 10% heat-inactivated fetal calf serum (FCS) (HyClone, Logan, UT), L-glutamine
(292 µg/ml), penicillin (100 U/ml), and streptomycin (100 µg/ml). The
resulting preparations contained 
98% PMN with 98% viability.
Chemotaxis Assay
PMN chemotaxis was measured with a fluorescent chemotaxis assay as previously described (20). Briefly, normal rabbit PMN were isolated as described earlier, and were resuspended in RPMI-1640 without phenol red and supplemented with 10% FCS (RPMI-FCS). Calcein AM (Molecular Probes, Eugene, OR) was added to 5.0 ml of the cell suspension at a final concentration of 5 µg/ml, and the resulting preparation was incubated for 30 min at 21° C. The PMN were washed twice with PBS, counted, and resuspended in RPMI-FCS to 3 × 106 PMN/ml. The bottom wells of disposable 96-well chemotaxis chambers (ChemoTx; Neuro Probe Inc., Gaithersburg, MD) were filled with either 29 µl of PF diluted in PBS-human serum albumin (HSA) or with a negative control (i.e., PBS supplemented with 1% HSA). To determine the total fluorescence of the labeled PMN, 25 µl of the cell suspension was placed directly in three bottom wells of the chemotaxis chamber. Polycarbonate filters with 8-µm pores were positioned on the loaded microplate,and calcein labeled PMN (25 µl) were placed directly onto the filter above each well. The chamber was incubated for 1 h (37° C and 5% CO2), and the nonmigrating cells on the origin side (i.e., top) of the filter were removed by gently wiping the filter with a tissue. The chemotaxis chamber was then placed in a multiwell fluorescence plate reader (CytoFluor II; PerSeptive Biosystems, Framingham, MA) and the cells that migrated into the bottom chamber were measured from the calcein fluorescence signal (excitation = 485 nm, emission = 530 nm). The data for the chemotaxis assay are expressed as a percentage of the response to zymosan-activated normal rabbit serum (ZAS) in the same assay.
Effect of PF on PMN Survival
Normal rabbit PMN were resuspended at 1 × 106 cells/ml in PF at a 50% concentration in RPMI-1640 without serum and containing penicillin (100 U/ml), streptomycin (100 µg/ml), and L-glutamine (292 µg/ml). The PMN were incubated at 37° C in 5% CO2 for 0 to 4 h. At hourly intervals the cells were counted in a hemacytometer, and cytospin preparations were stained with Diff-Quik. The percentage of apoptotic cells was determined by counting 300 cells and using previously described morphologic criteria for apoptotic cells (21).
Histologic Criteria for Lung Injury
Lung Injury Score. The lung tissue in each slide preparation was evaluated without knowledge of the treatment group from which it came. To generate the lung injury score, a total of 300 alveoli were counted on each slide at a magnification of ×400. Within each field, points were assigned according to predetermined criteria (Table 1). All of the points for each category were added and weighted according to their relative importance. The injury score was calculated according to the following formula:
|
Injury score = [(alveolar hemorrhage points/number of fields) + 2 × (alveolar infiltrate points/number of fields) + 3× (fibrin points/number of fields) + (alveolar septal congestion/number of fields)]/total number of alveoli counted.
At least two separate lung tissue samples were evaluated for each rabbit. The individual scores for each sample were averaged to produce a final score for the rabbit.
Statistical Analyses
For comparisons among several groups, normally distributed data collected at the same time from multiple groups were analyzed through
factorial analysis of variance (ANOVA), and Fisher's post hoc
method was used for secondary analysis. Data that were not normally
distributed were normalized by log10 transformation, followed by
ANOVA performed as described previously. For small groups (n 
3), the Kruskall-Wallis test was performed. Data are expressed as
mean ± SEM.
| |
RESULTS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Clinical Observations
The rabbits inoculated with E. coli showed three distinct clinical syndromes that depended on the size of the intraperitoneal inoculum. The animals in the control and low-dose groups (1 × 108 cfu/clot) were alert and active after recovering from anesthesia. The animals in the intermediate-dose group (1 × 109 cfu/clot) were alert but less active. The animals in the high-dose group (1 × 1010 cfu/clot) appeared lethargic and did not move about their cages.
From 4 to 6 h after the inoculation with the E. coli-containing clot, the animals in the high-dose group developed an abrupt decrease in blood pressure that was unresponsive to additional fluids and was associated with a low CVP, severe metabolic acidosis, and hypothermia (Table 2). At the time of death, the animals in the high-dose group met clinical criteria for septic shock in humans (22). No animals died for technical reasons.
|
Bacterial Inocula and Neutrophilic Response in the Peritoneal Compartment
To our surprise, higher bacterial inocula were associated with fewer PMN in the PF (Table 3). At 4 to 6 h after inoculation there was a remarkable absence of PMN from the peritoneal lavage fluids of all animals in the high-dose group. In contrast, the peritoneal lavage fluids of the animals in the intermediate- and low-dose groups contained abundant PMN. The clot by itself produced an inflammatory response, since PMN were also present in the peritoneal lavage fluid of the control animals.
|
We investigated the mechanisms responsible for the absence of PMN in the peritoneal cavity of the animals in the high-dose group. Although these animals experienced transient neutropenia, they had abundant circulating PMN at the time of death, suggesting that the absence of peritoneal PMN was not due to reduced systemic availability of PMN (Figure 1). The PF from the animals in the high-dose group had increased chemotactic activity for PMN as compared with the PF from the animals in the intermediate-dose group, as indicated by the left shift of the chemotaxis dose-response curve (Figure 2). The curves showed that the effective concentration for 50% maximal chemotaxis (EC50) for the high-dose group was closer to the 1:1,000 dilution of PF, whereas the EC50 for the intermediate-dose group was closer to the 1:100 dilution. The PF from the animals in the high-dose group contained increased concentrations of two PMN attractants, IL-8 and GRO (Table 4). Thus, the lack of identifiable PMN was not associated with a lack of PMN chemoattractants in the PF.
|
p < 0.05 versus control group.
|
|
MPO activity was highest in the PF from the animals in the high-dose group (Figure 3), suggesting that the absence of PMN in these fluids was not due to lack of recruitment. Instead, the increased MPO activity suggests that PMN had entered the peritoneal cavity and were rapidly destroyed. To determine whether the PF from the animals in the high-dose group caused PMN death, we incubated normal rabbit PMN in PF from animals in the normal, intermediate-dose, and high-dose groups. PMN survival was significantly reduced after incubation for 3 or 4 h in PF from animals in the high-dose group as compared with those in the intermediate-dose and control groups (Figure 4A). There was a trend toward an increased percentage of apoptotic cells after incubation of normal PMN in PF from the animals in the high-dose group (Figure 4B), but accelerated apoptosis alone did not account for the destruction of PMN.
|
|
Association among Neutrophilic Response in PF, Local Control of Infection, and Systemic Response
There was a strong inverse correlation among PMN in PF,
peritoneal bacterial recovery, and severity of bacteremia.
Numbers of PMN in PF were significantly and inversely correlated with numbers of PMN in peritoneal lavage cultures
(Pearson's r = 0.81, p < 0.001). Likewise, the recovery of
bacteria in blood was inversely correlated with PMN in PF (r = 0.76, p < 0.001). The PF of animals in the high-dose group
yielded approximately 4,000-fold more bacteria than that of
animals in the intermediate-dose group, even though the initial bacterial inoculum was only 10-fold greater (Figure 5). The
animals in the high-dose group were bacteremic for the duration of the experiments, and at each time at which counts were
made, had significantly more circulating bacteria than did animals in the intermediate- and low-dose groups (p < 0.001)
(Figure 6). The animals in the high-dose group were the only
ones that showed significantly increased serum concentrations
of TNF-, IL-8, GRO, and MCP-1 at 4 to 6 h (Table 4).
|
|
The animals in the intermediate-dose group showed a trend
toward a lower percentage of PMN in their PF at 24 h (Table
3). Although this trend failed to reach statistical significance, there was a significant inverse correlation between bacterial inoculum size and percent PMN in the peritoneal lavage fluid
(r = 0.74, p < 0.001). The median number of bacteria recovered from the peritoneal fluid of the animals in the intermediate-dose group did not change significantly between 4 h and
24 h, suggesting containment of the intraperitoneal infection
without bacterial clearance (Figure 5). By 24 h, only two of
five animals in the intermediate-dose group were bacteremic,
but the serum concentrations of TNF-, IL-8, GRO, and
MCP-1 in this group remained persistently elevated (Table 4).
The low-dose group had the highest number of peritoneal
PMN among the infected animals (Table 3). In the low-dose
group, three of eight animals were bacteremic at 2 h, and only
one of eight animals was bacteremic at 4 h (Figure 6). None of
the animals in the low-dose group was bacteremic at 24 h. These
animals had cleared TNF-, IL-8, and GRO from their serum at
24 h, but retained low concentrations of MCP-1 (Table 4).
Lung Injury
To measure the effects of peritoneal sepsis on the lungs, we evaluated gas exchange, bacterial cultures of BALF, lung inflammation as evidenced by PMN and cytokine concentrations in BALF, permeability changes as measured by the total protein concentration in BALF, and tissue histopathology.
Gas Exchange. The animals in the high-dose group developed a widening alveolar-arterial oxygen tension difference ([A-a]DO2) at 4 h and premortem (p = 0.02). This was associated with severe hypoxemia. There was no significant difference in the (A-a)DO2 between either the intermediate- or low-dose group and the control group (Table 5).
|
Bacterial cultures. Bacteria were recovered from the BALF on only three of the 12 animals in the high-dose group at 4 h to 6 h. The bacteria from two of these animals grew only in culture broth; the BALF of the third animal contained 2 × 102 cfu/ml. At 24 h, bacteria were recovered from one animal in the intermediate-dose group (3 × 102 cfu/ml), and from none of the animals in the low-dose group. In all cases the bacteria were lactose-fermenting, gram-negative rods consistent with E. coli.
BALF PMN. At 4 h to 6 h, the BALF of the animals in the high-dose group showed a significant increase in PMN as compared with that of the control group (p = 0.04) (Table 3). There was no significant difference in the number or percentage of BALF PMN among the control, low-dose, and intermediate-dose group at either 4 h or 24 h.
BALF cytokines. The BALF of all animals euthanized at 4 h
contained low concentrations of IL-8 and MCP-1 (Table 4).
GRO was present only in the BALF of animals in the high-dose group and TNF- was present only in the BALF of animals in the intermediate-dose group. At 24 h, only animals in
the intermediate group had detectable concentrations of cytokines in their BALF.
BALF total protein. At 4 h to 6 h, the animals in the high-dose group showed a trend toward a higher protein concentration in their BALF, but this failed to reach statistical significance (Table 5). The total protein concentrations were similar among the other groups of animals at 24 h.
Histology. In the high-dose group, eight of 10 animals had histologic evidence of lung inflammation, characterized by the presence of alveolar septal edema, neutrophilic infiltrates, alveolar hemorrhage, or intraalveolar fibrin deposition (Figures 7G through 7H). In the intermediate-dose group, only one animal met histologic criteria for lung inflammation, and this occurred at 24 h. The remainder of the animals showed subendothelial and perivascular PMN, without evidence of abnormalities in the air spaces (Figures 7E through 7F). There were no histologic abnormalities in the lungs of animals in the low-dose or control groups (Figures 7A through 7D).
|
The histopathology slides were graded in a blinded fashion to determine a lung injury score (Table 1). The lung injury score in the control animals was 3.9 ± 1.4 (mean ± SEM); in the animals in the low-dose group it was 2.7 ± 1.1; in those in the intermediate-dose group it was 4.1 ± 2.3; and in those in the high-dose group it was 7.9 ± 3.1. This score correlated significantly with BALF total protein (r = 0.85, p < 0.0001).
| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
The main goal of this study was to investigate the mechanisms that link the host response to a local infection in the peritoneal cavity to the development of sepsis and lung injury. Fibrin clots containing increasing concentrations of bacteria were placed into the peritoneal cavities of rabbits. Although smaller bacterial inocula resulted in a strong local neutrophilic response and minimal systemic manifestations, infection with the highest concentration of bacteria was associated with a paradoxical absence of PMN in the peritoneal cavity, local bacterial proliferation, development of septic shock, and lung injury. The decrease in the number of peritoneal PMN occurred in the presence of high concentrations of IL-8 in PF and in the bloodstream, and was due in part to rapid lysis of PMN in the peritoneal space. Thus, failure of the host to recruit and maintain adequate numbers of PMN at the primary site of infection appears to be a key mechanism in determining whether an infection is controlled locally or causes severe systemic effects.
The remarkable absence of PMN in the peritoneal cavity of the animals given a high dose of E. coli was not due to a lack of intraperitoneal signals for PMN migration. The peritoneal lavage fluid from the animals in the high-dose group had chemotactic activity for normal rabbit PMN in vitro, and had the highest concentrations of the CXC chemokines IL-8, and GRO. These animals also had high concentrations of IL-8 and GRO in their systemic circulation. In rabbits, intravenous administration of human IL-8 reduces PMN recruitment to sites of inflammation (23). Thus, the high concentrations of IL-8 and GRO in the systemic circulation of the animals in the high-dose group may have resulted in decreased PMN recruitment to the peritoneal cavity, either by downregulating the CXCR1 and CXCR2 receptors on PMN, or by reducing the effective IL-8 gradient between the vascular space and the peritoneal compartment. Downregulation of CXCR2 on PMN has been found in humans with severe sepsis (24). However, circulating PMN in the animals in the high-dose group in our study were capable of migrating into tissues, as shown by their presence in the lungs.
The finding that the PF of the animals in the high-dose
group had the greatest MPO activity suggests that PMN that
were recruited into the peritoneal cavity were rapidly lysed. The
conclusion that the PF of animals in the high-dose group
caused the death of PMN in vivo is reinforced by the finding
that normal PMN incubated in PF from this group were rapidly lysed, as compared with PMN incubated in PF from normal animals or those in the intermediate-dose group (Figure
2). The mechanism of PMN death could have been necrosis or
apoptosis. Although PMN undergo apoptosis in response to
bacterial products and/or oxidant stress (25), the high MPO
activity in the PF of the animals in the high-dose group suggests that necrosis and severe degranulation had occurred,
since MPO is contained in the primary granules of PMN. It is
possible that the death of PMN was mediated by an E. coli exotoxin, such as -hemolysin or 
-hemolysin (26). The association of local bacterial proliferation with low numbers of PMN
suggests that a key determinant of the systemic response to
bacterial infection is the fate of the PMN at the primary site of infection.
The failure of the peritoneal PMN response to high doses of bacteria in the present study contrasts with prior findings in a rabbit model of pneumonia (15), in which PMN recruitment occurred in the lungs of all rabbits with pneumonia, even at the highest bacterial inocula. In the pneumonia study, the number of PMN in the lungs reached a maximum at low bacterial inocula and failed to increase in response to higher bacterial loads, despite active bacterial proliferation in the lungs. Although the two studies differed in times and techniques, their results suggest that over the range of bacterial inocula studied, the tissue population of PMN is increased and maintained more effectively in the lungs than in the peritoneal cavity.
A second major finding in the present study was that the containment of infection was poor in the peritoneal cavity. Bacteremia was detected in all of the animals in the intermediate- and high-dose groups, and in three of the eight animals in the low-dose group. These findings also contrast with the pneumonia model, in which none of the animals were bacteremic at 4 h or 24 h, even with very high inocula in the lungs (15). These findings suggest that the development and severity of bacteremia is a function of the local characteristics of the primary site of infection, and not only of the size of the bacterial inoculum. The interpretation that bacteria are contained more effectively in the lungs than in the peritoneal cavity is supported by differences in anatomic factors, since the peritoneum contains 8- to 12-µm pores that may allow bacteria access to the systemic circulation (27), whereas the alveolar epithelium forms a tight barrier (28). In addition, there may be important differences in the responses to bacteria of the local macrophage populations of the alveoli and peritoneum.
The third major finding of this study was the lack of compartmentalization of the cytokine response to intraperitoneal
infections. In the study, all of the cytokines that were assayed
(TNF-, GRO, IL-8, and MCP-1) showed a dose-based-response
to the bacterial inoculum not only in the PF but also in the
serum, and in the case of IL-8, GRO, and MCP-1, in the
BALF as well. Lack of cytokine compartmentalization following peritonitis was also observed by Walley and colleagues,
who found increased concentrations of CC and CXC chemokines in the peritoneum and serum of mice after cecal ligation
and puncture, and by Eichacker and coworkers, who found increased serum concentrations of TNF- in dogs after intraperitoneal placement of E. coli-containing clots (30, 31). This
lack of cytokine compartmentalization may be due to activation of the mononuclear phagocyte system (e.g., circulating
monocytes and Kupffer's cells) by bacteria and/or bacterial
products, with systemic production of cytokines; or to increased permeability of the peritoneum resulting from more
severe infection, with subsequent leakage of cytokines into the
systemic circulation.
The lack of cytokine compartmentalization following peritonitis contrasts with the finding in our prior study of rabbits
with E. coli pneumonia, in which the cytokine response was
restricted to the lungs, even at the highest bacterial inocula
(15). Prior studies have shown that the response of TNF- to a
pulmonary challenge with lipopolysaccharide is compartmentalized in the lungs, and that this compartmentalization is lost
when lung injury occurs (13, 14). Direct leakage of cytokines
from the lungs into the systemic circulation has been documented in rabbits with Pseudomonas aeruginosa pneumonia
(32) and in dogs with E. coli pneumonia (33, 34). When the
bacterial inoculum is high, however, pulmonary and peritoneal infections cause similar systemic responses and mortality,
as shown in a series of studies of dogs with either pneumonia
or peritonitis caused by bacterial inocula in the range of 1 × 1010 cfu/kg (17, 31, 33, 34). Taken together, the present and
previously reported studies suggest that in peritoneal infections with E. coli, a systemic cytokine response occurs even
when the inoculum is small, whereas in pulmonary infections
the cytokine response remains localized to the lungs unless the
inoculum is very high and the infection is severe.
The final major observation in the present study was that there is a direct relationship between the size of the bacterial inoculum at the primary site and the development of injury in the lungs. The lung changes were characterized by histologic evidence of inflammation and evidence of abnormalities in permeability and gas exchange. It is not clear whether lung injury occurred as a result of more severe bacteremia or higher circulating concentrations of proinflammatory cytokines. However, the finding that all of the animals in the intermediate-dose group became bacteremic yet failed to develop lung injury suggests that bacteremia per se is insufficient to cause lung injury unless it is a high grade bacteremia and/or is associated with a severe systemic inflammatory response. The findings in the study also confirm clinical observations that bacteremia alone (without a systemic inflammatory response) is seldom associated with septic shock and lung injury (2).
The combined findings of this study and the prior pneumonia study lead to the hypothesis that a three-compartment model best explains the host response to local infections. In this model, the first compartment is the site at which bacteria enter the host and proliferate, the second compartment includes the circulation and the reticuloendothelial system, and the third compartment includes distant organs such as the lung. In the first compartment, bacterial products stimulate host leukocyte responses at low doses. This is associated with systemic manifestations analogous to the systemic inflammatory response syndrome in humans (fever, leukopenia). At high bacterial inocula, the host response in the first compartment fails to contain the infection and bacteria proliferate. The failure to contain the infection is associated with inadequate local numbers of PMN. The actual number of bacteria needed to elicit these responses probably depends on specific characteristics of the local site of infection. The onset or magnitude of the response in the second (systemic) compartment depends on the initial inflammatory response in the first compartment. Damage to distant organs (the third compartment) occurs when the neutrophilic response in the first compartment fails to contain the bacterial inoculum and bacterial proliferation occurs, leading to high-grade bacteremia and a severe systemic inflammatory response.
We conclude that development of lung injury in rabbits with peritoneal sepsis occurs when the bacterial inoculum in the peritoneal cavity exceeds a threshold above which the local infection is not controlled by leukocytes in the first compartment, and bacteria proliferate. The local failure to control the infection is associated with a paradoxical absence of PMN in the peritoneum, and results in a severe systemic inflammatory response. The severity of the systemic inflammatory response, rather than the presence of circulating bacteria, determines whether septic shock and lung injury occur.
| |
Footnotes |
|---|
Correspondence and requests for reprints should be addressed to Thomas R. Martin, M.D., Seattle VA Medical Center, 151L, 1660 South Columbian Way, Seattle, WA 98108-1597. E-mail: trmartin{at}u.washington.edu
(Received in original form September 9, 1999 and in revised form July 14, 2000).
Acknowledgments: The authors thank Krystine Wynant, Venus A. Wong, Stephen Mongovin, and Lena Strait for expert technical assistance.
Supported in part by grants HL 30542, AI 29103, and GM 37696 from the National Institutes of Health, the American Heart Association of Washington, and the Medical Research Service of the U.S. Department of Veterans Affairs.
| |
References |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
1. Hudson LD, Milberg JA, Anardi D, Maunder RJ. Clinical risks for development of the acute respiratory distress syndrome. Am J Respir Crit Care Med 1995; 151: 293-301 [Abstract].
2. Fowler AA, Hamman RF, Good JT, Benson KN, Baird M, Eberle DJ, Petty TL, Hyers TM. Adult respiratory distress syndrome: risk with common predispositions. Ann Intern Med 1983; 98: 593-597 .
3. Martin TR, Matute-Bello G, Skerrett SJ, Frevert CW. Extrapulmonary sepsis and cytokines in lung injury and defense. In: Nelson S, Martin RT, editors. Cytokines and pulmonary disease. New York: Marcel Dekker; 2000. p. 403-458.
4.
Fisher CJJ,
Agosti JM,
Opal SM,
Lowry SF,
Balk RA,
Sadoff JC,
Abraham E,
Schein RM,
Benjamin E.
Treatment of septic shock with the
tumor necrosis factor receptor: Fc fusion protein.
N Engl J Med
1996;
334:
1697-1702
5. Abraham E, Glauser MP, Butler T, Garbino J, Gelmont D, Laterre PF, Kudsk K, Bruining HA, Otto C, Tobin E, et al . . p55 Tumor necrosis factor receptor fusion protein in the treatment of patients with severe sepsis and septic shock: a randomized controlled multicenter trial. JAMA 1997; 277: 1531-1538 [Abstract].
6. Reinhart K, Wiegand-Lohnert C, Grimminger F, Kaul M, Withington S, Treacher D, Eckart J, Willatts S, Bouza C, Krausch D, et al . . Assessment of the safety and efficacy of the monoclonal anti-tumor necrosis factor antibody-fragment, MAK 195F, in patients with sepsis and septic shock: a multicenter, randomized, placebo-controlled, dose-ranging study. Crit Care Med 1996; 24: 733-742 [Medline].
7. Opal SM, Fisher CJJ, Dhainaut JF, Vincent JL, Brase R, Lowry SF, Sadoff JC, Slotman GJ, Levy H, Balk RA, et al . . Confirmatory interleukin-1 receptor antagonist trial in severe sepsis: a phase III, randomized, double blind, placebo controlled, multicenter trial. Crit Care Med 1997; 25: 1115-1124 [Medline].
8.
Bernard GR,
Wheeler AP,
Russell JA,
Schein R,
Summer WR,
Steinberg KP,
Fulkerson WJ,
Wright PE,
Christman BW,
Dupont WD, et al
.
.
The effects of ibuprofen on the physiology and survival of patients
with sepsis.
N Engl J Med
1997;
336:
912-918
9. Brun-Buisson C, Doyon F, Carlet J, French Bacteremia-Sepsis Study Group. Bacteremia and severe sepsis in adults: a multicenter prospective survey in ICUs and wards of 24 hospitals. Am J Respir Crit Care Med 1996;154:617-624.
10. Bates DW, Pruess KE, Lee TH. How bad are bacteremia and sepsis? Outcomes in a cohort with suspected bacteremia. Arch Intern Med 1995; 155: 593-598 [Abstract].
11. Rangel-Frausto MS, Pittet D, Costignan M, Hwang T, David CS, Wenzel RP. The natural history of the systemic inflammatory response syndrome (SIRS). JAMA 1995; 273: 117-123 [Abstract].
12. Montgomery AB, Stager MA, Carrico CJ, Hudson LD. Causes of mortality in patients with the adult respiratory distress syndrome. Am Rev Respir Dis 1985; 132: 485-489 [Medline].
13. Nelson S, Bagby GJ, Bainton BG, Wilson LA, Thompson JJ, Summer WR. Compartmentalization of intraalveolar and systemic lipopolysaccharide-induced tumor necrosis factor and the pulmonary inflammatory response. J Infect Dis 1989; 159: 189-194 [Medline].
14. Tutor JD, Mason CM, Dobard E, Beckerman RC, Summer WR, Nelson S. Loss of compartmentalization of alveolar tumor necrosis factor after lung injury. Am J Respir Crit Care Med 1994; 149: 1107-1111 [Abstract].
15. Fox-Dewhurst R, Alberts M, Kajikawa O, Caldwell E, Johnson MC II,, Skerrett SJ, Goodman RB, Ruzinski JT, Wong VA, Chi EY, et al . . Pulmonary and systemic inflammatory responses in rabbits with gram-negative pneumonia. Am J Respir Crit Care Med 1997; 155: 2030-2040 [Abstract].
16. Natanson C, Fink MP, Ballantyne HK, Mac Vittie TJ, Conklin JJ, Parrillo JE. Gram-negative bacteremia produces both severe systolic and diastolic cardiac dysfunction in a canine model that simulates human septic shock. J Clin Invest 1986;78:259-270.
17. Eichacker PQ, Waisman Y, Natanson C, Farese A, Hoffman WD, Banks SM, Mac Vittie TJ. Cardiopulmonary effects of granulocyte colony-stimulating factor in a canine model of bacterial sepsis. J Appl Physiol 1994;77:2366-2373.
18. Kajikawa O, Goodman RB, Johnson MC II,, Martin TR. Sensitive and specific immunoassays for rabbit IL-8 and MCP-1: two important cytokines that regulate leukocyte migration in the lungs. J Immunol Methods 1996; 197: 19-29 [Medline].
19. Kajikawa O, Johnson MC II,, Goodman RB, Frevert CW, Martin TR. A sensitive immunoassay to detect the alpha-chemokine GRO in rabbit blood and lung fluids. J Immunol Methods 1997; 205: 135-143 [Medline].
20. Frevert CW, Wong VA, Goodman RB, Goodwin R, Martin TR. Rapid fluorescence-based measurement of neutrophil migration in vitro. J Immunol Methods 1998; 213: 41-52 [Medline].
21.
Matute-Bello G,
Liles WC,
Radella F,
Steinberg KP,
Ruzinski JT,
Jonas M,
Chi EY,
Hudson LD,
Martin TR.
Neutrophil apoptosis in the
acute respiratory distress syndrome.
Am J Respir Crit Care Med
1997;
156:
1969-1977
22.
Bone RC,
Balk RA,
Cerra FB,
Dellinger RP,
Fein AM,
Knaus WA,
Schein RMH,
Sibbald WJ.
Definitions for sepsis and organ failure and
guidelines for use of innovative therapies in sepsis.
Chest
1992;
101:
1644-1655
23. Hechtman DH, Cybulsky MI, Fuchs HJ, Baker JB, Gimbrone MAJ. Intravascular IL-8: inhibitor of polymorphonuclear leukocyte accumulation at sites of acute inflammation. J Immunol 1991; 147: 883-892 [Abstract].
24.
Cummings CJ,
Martin TR,
Frevert CW,
Quan JM,
Wong VA,
Mongovin SM,
Hagen TR,
Steinberg KP,
Goodman RB.
Expression and function of the chemokine receptors CXCR1 and CXCR2 in sepsis.
J Immunol
1999;
162:
2341-2346
25.
Kasahara Y,
Iwai K,
Yachie A,
Ohta K,
Konno A,
Seki H,
Miyawaki T,
Taniguchi N.
Involvement of reactive oxygen intermediates in spontaneous and CD95 (Fas/Apo-1)-mediated apoptosis of neutrophils.
Blood
1997;
89:
1748-1753
26.
Bhakdi S,
Greulich S,
Muhly M,
Eberspacher B,
Becker H,
Thiele A,
Hugo F.
Potent luekocidal action of Escherichia coli hemolysin mediated by permeabilization of target cell membranes.
J Exp Med
1989;
169:
737-754
27. Heemken RL, Gandawidjaja L, Hau T. Peritonitis: pathophysiology and local defense mechanisms. Hepatogastroenterology 1997; 44: 927-936 [Medline].
28.
Matthay MA,
Folkesson HG,
Verkman AS.
Salt and water transport
across alveolar and distal airway epithelia in the adult lung.
Am J
Physiol
1996;
270:
L487-L503
29.
Gjomarkaj M,
Pase E,
Melis M,
Spatafora M,
Profita M,
Vignola AM,
Bonsignore G,
Toews GB.
Phenotypic and functional characterization
of normal rat pleural macrophages in comparison with autologous
peritoneal and alveolar macrophages.
Am J Respir Cell Mol Biol
1999;
20:
135-142
30. Walley KR, Luckacs NW, Standiford TJ, Strieter RM, Kunkel SL. Elevated levels of macrophage inflammatory protein 2 in severe murine peritonitis increase neutrophil recruitment and mortality. Infect Immun 1997; 65: 3847-3851 [Abstract].
31.
Eichacker PQ,
Hoffman WD,
Farese A,
Danner RL,
Suffredini AF,
Waisman Y,
Banks SM,
Mouginis T,
Wilson L,
Rothlein R, et al
.
. Leukocyte
CD18 monoclonal antibody worsens endotoximia and cardiovascular
injury in canines with septic shock.
J Appl Physiol
1993;
74:
1885-1892
32. Kurahashi K, Kajikawa O, Sawa T, Ohara M, Gropper MA, Frank DW, Martin TR, Wiener-Kronish JP. Pathogenesis of septic shock in Pseudomonas aeruginosa pneumonia. J Clin Invest 1999; 104: 743-750 [Medline].
33.
Freeman BD,
Quezado Z,
Zeni F,
Natanson C,
Danner RL,
Banks S,
Quezado M,
Fitz Y,
Bacher J,
Eichacker PQ.
rG-CSF reduces endo-
toxemia and improves survival during E. coli pneumonia.
J Appl
Physiol
1997;
83:
1467-1475
34.
Quezado ZM,
Natanson C,
Karzai W,
Danner RL,
Koev CA,
Fitz Y,
Dolan DP,
Richmond S,
Banks SM,
Wilson L, et al
.
. Cardiopulmonary
effects of inhaled nitric oxide in normal dogs and during E. coli pneumonia and sepsis.
J Appl Physiol
1998;
84:
107-115
This article has been cited by other articles:
 |
|
 |
|
  G. Matute-Bello, C. W. Frevert, and T. R. Martin Animal models of acute lung injury Am J Physiol Lung Cell Mol Physiol, September 1, 2008; 295(3): L379 - L399. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X. Su, M. Johansen, M. R. Looney, E. J. Brown, and M. A. Matthay CD47 Deficiency Protects Mice from Lipopolysaccharide-Induced Acute Lung Injury and Escherichia coli Pneumonia J. Immunol., May 15, 2008; 180(10): 6947 - 6953. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Bellmeyer, J. M. Martino, N. S. Chandel, G. R. Scott Budinger, D. A. Dean, and G. M. Mutlu Leptin Resistance Protects Mice from Hyperoxia-induced Acute Lung Injury Am. J. Respir. Crit. Care Med., March 15, 2007; 175(6): 587 - 594. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. M. Mutlu, C. Snyder, A. Bellmeyer, H. Wang, K. Hawkins, S. Soberanes, L. C. Welch, A. J. Ghio, N. S. Chandel, D. Kamp, et al. Airborne Particulate Matter Inhibits Alveolar Fluid Reabsorption in Mice via Oxidant Generation Am. J. Respir. Cell Mol. Biol., June 1, 2006; 34(6): 670 - 676. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
X. Su, L. Robriquet, H. G. Folkesson, and M. A. Matthay Protective effect of endogenous beta-adrenergic tone on lung fluid balance in acute bacterial pneumonia in mice Am J Physiol Lung Cell Mol Physiol, April 1, 2006; 290(4): L769 - L776. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. S. Munford Invited review: Detoxifying endotoxin: time, place and person Innate Immunity, April 1, 2005; 11(2): 69 - 84. [Abstract] [PDF]
|
|
|
|
|
|
J. D. Lang, M. Figueroa, K. D. Sanders, M. Aslan, Y. Liu, P. Chumley, and B. A. Freeman Hypercapnia via Reduced Rate and Tidal Volume Contributes to Lipopolysaccharide-induced Lung Injury Am. J. Respir. Crit. Care Med., January 15, 2005; 171(2): 147 - 157. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. A. Altemeier, G. Matute-Bello, C. W. Frevert, Y. Kawata, O. Kajikawa, T. R. Martin, and R. W. Glenny Mechanical ventilation with moderate tidal volumes synergistically increases lung cytokine response to systemic endotoxin Am J Physiol Lung Cell Mol Physiol, September 1, 2004; 287(3): L533 - L542. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S.-M. Lin, C. W. Frevert, O. Kajikawa, M. M. Wurfel, K. Ballman, S. Mongovin, V. A. Wong, A. Selk, and T. R. Martin Differential Regulation of Membrane CD14 Expression and Endotoxin-Tolerance in Alveolar Macrophages Am. J. Respir. Cell Mol. Biol., August 1, 2004; 31(2): 162 - 170. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. F. Benjamim, C. M. Hogaboam, N. W. Lukacs, and S. L. Kunkel Septic Mice Are Susceptible to Pulmonary Aspergillosis Am. J. Pathol., December 1, 2003; 163(6): 2605 - 2617. [Abstract] [Full Text]
|
|
|
|
|
|
S. E. Sinclair, D. A. Kregenow, W. J. E. Lamm, I. R. Starr, E. Y. Chi, and M. P. Hlastala Hypercapnic Acidosis Is Protective in an In Vivo Model of Ventilator-induced Lung Injury Am. J. Respir. Crit. Care Med., August 1, 2002; 166(3): 403 - 408. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. F. Benjamim, J. S. Silva, Z. B. Fortes, M. A. Oliveira, S. H. Ferreira, and F. Q. Cunha Inhibition of Leukocyte Rolling by Nitric Oxide during Sepsis Leads to Reduced Migration of Active Microbicidal Neutrophils Infect. Immun., July 1, 2002; 70(7): 3602 - 3610. [Abstract] [Full Text] [PDF]
|
</ |