Concomitant Increase in Neutrophil Adhesion to Inflammatory Peritoneum and Remote Organs during Peritonitis

Journal of Surgical Research 81, 156 –163 (1999) Article ID jsre.1998.5454, available online at http://www.idealibrary.com on

Concomitant Increase in Neutrophil Adhesion to Inflammatory Peritoneum and Remote Organs during Peritonitis Kazuhiko Fukatsu, M.D.,* Hideaki Saito, M.D.,† Ilsoo Han, M.D.,* Tomomi Inoue, M.D.,* Satoshi Furukawa, M.D.,* Takeaki Matsuda, M.D.,* Shigeo Ikeda, M.D.,* Hiroshi Yasuhara, M.D.,* and Tetsuichiro Muto, M.D.* *Department of Surgery and †Surgical Center, The University of Tokyo, 731 Hongo, Bunkyo-ku, Tokyo, 113, Japan Submitted for publication February 18, 1998

Background. Neutrophils contribute to the host defense mechanism, but they can cause remote organ injury in peritonitis. The purpose of this study was to examine neutrophil adhesion to the peritoneum and remote organs simultaneously in peritonitis using a fluorescence microscopic method. Study design. Experiment 1: Sprague–Dawley rats (n 5 16) were injected intraperitoneally (ip) with saline solution or 10 5, 10 7, or 10 9 Escherichia coli. Five hours after challenge, 1 3 10 6 fluorescein-labeled neutrophils were infused. Two minutes after neutrophil injection, five peritoneal samples (the greater omentum, mesentery, parietal peritoneum, colon, and ileum), both lungs, the liver, and the right kidney were harvested for counting of labeled neutrophils under epifluorescent microscopy. Lung myeloperoxidase (MPO) activity was also determined. Experiment 2: Rats (n 5 23) were given 10 9 E. coli ip. Before challenge (0 h) or at 1, 5, or 10 h after challenge, labeled neutrophils were infused. Then, the labeled neutrophil numbers in organs and lung MPO activities were assessed as described for Experiment 1. Hemodynamic and arterial blood gas data were also obtained in another set of rats before and at 1, 5, 8 and 10 h after 10 9 E. coli ip challenge. Results. Experiment 1: The labeled neutrophil numbers in the peritoneum, lungs, and kidney showed significant positive correlations with the injected bacterial numbers. Lung MPO also positively correlated with E. coli number and labeled neutrophil number in the lungs. Experiment 2: Labeled neutrophil numbers in the peritoneum and kidney peaked at 5 h. The pulmonary labeled neutrophil number rose, reaching a plateau at 5 h. No remarkable change was observed in the hepatic labeled neutrophil number. There was a positive correlation between lung MPO activity and

0022-4804/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

pulmonary labeled neutrophil number. Hemodynamic and blood gas data reflected a hyperdynamic state. Conclusions. Concomitant dose-dependent increases in neutrophil adhesion in the peritoneum, lungs, and kidney were observed in this peritonitis model. Increased neutrophil adhesion was transient in the peritoneum and kidney but persistent in the lungs. Strategies modulating neutrophil adhesion in organs are anticipated to be useful for the treatment of peritonitis. © 1999 Academic Press Key Words: peritonitis; neutrophil adhesion; fluorescent microscopic method; myeloperoxidase. INTRODUCTION

Despite the many recent advances in surgical intervention and antimicrobial therapy, overall mortality associated with peritonitis remains high [1]. Neutrophils have been implicated in nonspecific host defense and tissue damage during peritonitis. Neutrophils are the major effector of the nonspecific immune response in host resistance to infection [2– 4]. Peritoneal exudative neutrophils phagocytose and kill bacteria during bacterial peritonitis. However, neutrophils can injure vascular endothelial cells through a variety of mechanisms involving oxygen products or proteolytic enzymes [5, 6]. Thus, the neutrophil is presumably central to remote organ failure in critical illness [7, 8]. Neutrophil adhesion to endothelial cells is required for migration into the infectious focus [9, 10] and tissue injury in remote organs [11]. Therefore, the investigation of neutrophil adhesion to both the inflamed peritoneum and remote organs may provide useful information for treatment of patients with intraabdominal sepsis. However, to our knowledge, no studies have examined the pattern of neutrophil adhesion to the

156

FUKATSU ET AL.: PERITONITIS INCREASES NEUTROPHIL ADHESION

157

FIG. 1. A schematic representation of the protocol used in Experiments 1 and 2. Neutrophils were removed from another set of male SD rats and labeled with 5,(6)-carboxyfluorescein diacetate.

peritoneum and remote organs concomitantly in peritonitis. We have recently used a fluorescence microscopic method for visualizing and examining neutrophil adhesion in the vasculature of various tissues [12–14]. The purpose of this study was to examine neutrophil adhesion to the inflamed peritoneum and remote organs simultaneously in a bacterial peritonitis model using this fluorescence microscopic method. In the first experiment, we determined the influence of the amount of bacteria injected on neutrophil adhesion. Second, the kinetics of neutrophil adhesion after bacterial injection were examined. MATERIALS AND METHODS

Animals Male Sprague–Dawley rats (Nippon Bio-Supp, Center, Tokyo, Japan) weighing 220 –250 g were used for all experiments. In accordance with our institutional guidelines, the rats suffered no unnecessary discomfort, pain, or injury and received proper care and maintenance. The animals were quarantined for at least 1 week to allow adaptation to the environmental conditions. During this period, they were provided standard laboratory food and water ad libitum.

Preparation of Bacteria Escherichia coli (American Type Culture Collection 25922) was used. The bacteria were cultivated in brain heart infusion broth (Nissui Pharmaceutical Co, Tokyo, Japan) for 18 h. After the bacterial suspension had been centrifuged at 4000 rpm for 10 min, the

broth was discarded and the bacteria were resuspended in saline solution. The resuspended bacteria were then washed and centrifuged twice. After the final wash, a small portion of the bacterial suspension was serially diluted, plated on a sheep blood agar plate (Nissui Pharmaceutical Co, Japan), and incubated for 24 h to determine the bacterial concentration. The remainder was stored at 4°C for approximately 20 h until use. Just before bacterial injection, the stored bacterial suspension was diluted in saline solution in order to achieve a final bacterial concentration of 5 3 10 4, 5 3 10 6, or 5 3 10 8 colony-forming units per milliliter.

Rat Neutrophil Isolation and Labeling Neutrophils were isolated from whole blood drawn using a sterile technique from another set of male donor Sprague–Dawley rats. Under pentobarbital sodium anesthesia, blood was collected by cardiac puncture and anticoagulated with heparin sodium. Platelet-rich plasma was obtained by centrifuging the blood at 1400 rpm for 10 min. After the platelet-rich plasma had been discarded, platelet-poor plasma was overlaid with Lymphosepal (Immuno-Biological Lab, Gunma, Japan) and centrifuged at 1800 rpm for 30 min. IMDM (Nikken, Kyoto, Japan) and 6-hydroxyethylated starch (Morishita, Osaka, Japan) were added to the erythrocyte–leukocyte suspension. After mixture by inversion, the erythrocytes were allowed to settle for 30 min at 37°C. The upper suspension was collected. IMDM was added to the suspension, which was then recentrifuged at 2000 rpm for 5 min. The pellet was resuspended with Tris–NH 4Cl (1 M Tris:1 M NH 4Cl, 1:9 (v/v)) (Tris, Tris(hydroxymethyl)aminomethane; Nacalai tesque, Kyoto, Japan) and lysed at 37°C for 15 min. After washing, isolated neutrophils were labeled with 0.1 mM of 5,(6)carboxyfluorescein diacetate (Sigma Chemical Co., St. Louis, MO) for 15 min in the dark. Cells were then centrifuged at 2000 rpm for 5 min and the pellet was resuspended in PBS–EDTA. Fluorescein-labeled cells were counted under an epifluorescent microscope and diluted in

158

JOURNAL OF SURGICAL RESEARCH: VOL. 81, NO. 2, FEBRUARY 1999

FIG. 2. The numbers of labeled neutrophils in the peritoneum, lungs, kidney, and liver at 5 h after bacterial challenge vs the logarithm of the number of bacteria injected in Experiment 1. The rats were given normal saline or 10 5, 10 7, or 10 9 E. coli. The labeled neutrophil numbers in the peritoneum, lungs, and kidney showed significant positive correlations with the number of bacteria injected. However, there was no significant correlation between the numbers of hepatic labeled neutrophils and bacteria injected. PBS–EDTA to achieve a final neutrophil concentration of 2 3 10 6 cells per milliliter.

Experimental Design Experiment 1: Dose–response study. The animals (n 5 16) were randomly divided into four groups as follows: normal saline solution treatment (control group); injection of 10 5 E. coli (E. coli 10 5 group); injection of 10 7 E. coli (E. coli 10 7 group); and injection of 10 9 E. coli (E. coli 10 9 group) (Fig. 1). Five hours after the ip injection of 2 ml saline solution or E. coli solution, 1 3 10 6 fluorescein-labeled neutrophils suspended in 0.5 ml PBS–EDTA were infused over 10 –15 s. Fifteen minutes before neutrophil infusion, all animals were anesthetized with pentobarbital sodium (50 mg/kg). A polyethylene catheter was inserted into the right jugular vein for infusion of the neutrophil suspension. Two minutes after neutrophil infusion, the rats were killed by means of cervical dislocation. The omentum, the mesentery of the terminal ileum, the parietal peritoneum of the anterior abdominal wall, the ascending colon, and the terminal ileum were harvested. The lungs, liver, and right kidney were also harvested. Experiment 2: Time course study. Rats (n 5 23) were given 10 9 E. coli ip. Before challenge (0 h), or at 1, 5, or 10 h after challenge, fluorescein-labeled neutrophils were infused. Two minutes after neutrophil infusion, the rats were killed and the organs were obtained as in experiment 1 (Fig. 1). In another set of rats (n 5 4), a catheter was inserted into the right jugular artery. Before challenge (0 h), and at 1, 5, 8, and 10 h after

10 9 E. coli challenge, mean arterial blood pressure and heart rate were measured (Life Scope, Nihon Kohden, Tokyo, Japan). Arterial blood gases were also measured with a blood gas analyzer (Stat Profile 3; Nova Biomedical, Boston, MA) at each time point. Measurement of labeled neutrophil numbers in organs. Harvested organs were washed in sterile saline solution, to clear surface blood, and observed under an epifluorescent microscope (BX 40, Olympus, Tokyo, Japan) connected with an image-processing system (ARGUS-50; Hamamatsu Photonics, Hamamatsu, Japan). The colon and ileum were opened before washing and observed from the serosal side. The fluorographs of these organs were recorded digitally under epi-illumination and processed to determine the number of labeled neutrophils in selected portions of the images. In the peritoneum, three fields from each tissue sample (omentum, mesentery, parietal peritoneum, colon, and ileum) were recorded. We expressed the labeled neutrophil number in the peritoneum as the sum of the number in 15 fields, three from each of the five peritoneal samples. In the lungs, four fields (one field from the upper lobe and one field from the lower lobe of both lungs) were examined. Four liver fields and two from the right kidney also were examined. All fluorographs were recorded and stored in the memory of a computer. Subsequently, fluorescein-labeled neutrophils in the stored images were counted with an image processing system in blinded fashion. Measurement of lung myeloperoxidase activity. Lung myeloperoxidase (MPO) activity was determined in Experiments 1 and 2 using a modification of the method of Grisham et al. [15]. After the lung fluorographs had been recorded digitally, the lung surface was rinsed with

FUKATSU ET AL.: PERITONITIS INCREASES NEUTROPHIL ADHESION

159

FIG. 3. (Top) The lung MPO activity at 5 h after bacterial challenge vs the logarithm of the injected bacterial number in Experiment 1. (Bottom) The lung MPO activity vs the number of labeled neutrophils in a total of four fields from the both lungs at 5 h after bacterial challenge in Experiment 1. The rats were given normal saline or 10 5, 10 7, or 10 9 E. coli. The lung MPO activities showed significant positive correlations with the number of bacteria injected (n 5 16, P 5 0.01, r 5 0.61) and with labeled neutrophil numbers in the lungs (n 5 16, P 5 0.04, r 5 0.52). sterile normal saline and blotted dry. The lungs were weighed and frozen at 280°C. The organs were then thawed and homogenized in 5 volumes of 50 mM acetate buffer (pH 6.0) containing 0.5% cetyltrimethylammonium bromide. The homogenate was centrifuged at 40,000g for 10 min at 4°C. MPO activities of the supernatants were quantitated against a standard curve of human MPO by measuring the H 2O 2-dependent oxidation of 3,39,5,59-tetramethylbenzidine and expressed as units per gram of lung tissue. Histologic analysis. In Experiment 2, the upper lobe of the right lung, liver, and right kidney harvested from two animals at each time point were fixed in 10% formalin and embedded in paraffin. The paraffin sections were stained with hematoxylin and eosin. The slides were evaluated by two of the authors and a pathologist.

Statistical analysis. Results are presented as means 6 SE. Analysis of variance, followed by Fisher’s protected least significant difference post hoc test, was used to make statistical comparisons. P , 0.05 was considered statistically significant.

RESULTS

Experiment 1: Dose–Response Study Numbers of labeled neutrophils in organs. Five hours after bacterial injection, the fluorescein-labeled neutrophil numbers in the peritoneum, lungs, and kidney

160

JOURNAL OF SURGICAL RESEARCH: VOL. 81, NO. 2, FEBRUARY 1999

TABLE 1 Labeled Neutrophil Numbers in Organs and Lung MPO Activities after 10 9 E. coli Challenge

Peritoneum Lungs Kidney Liver Lung MPO

(/15 fields) (/4 fields) (/2 fields) (/4 fields) (U/g)

0h

1h

5h

10 h

10.0 6 1.0 337 6 13 3.2 6 1.2 53 6 12 12.0 6 0.6

17.6 6 4.1 353 6 23 3.8 6 1.2 35 6 4 57.7 6 12.4

23.3 6 3.7* 453 6 48† 6.2 6 1.9 44 6 15 93.0 6 42.0

16.0 6 3.0 425 6 49 2.8 6 0.9 45 6 2 59.5 6 27.2

Note. Values are mean 6 SE numbers of labeled neutrophils in organs and lung MPO activities. * P , 0.01 vs 0 h; †P , 0.01 vs 0 h and 1 h.

showed significant positive correlations with the numbers of bacteria injected (Fig. 2). However, there were no significant correlations between hepatic labeled neutrophil numbers and numbers of bacteria injected. Lung MPO activity. Lung MPO activities showed significant positive correlations with both the numbers of bacteria injected and the labeled neutrophil numbers in the lungs (Fig. 3). Experiment 2: Time Course Study Numbers of labeled neutrophils in organs. Two animals died before the 10-h observation point. The remaining rats survived until the final time point observed. The labeled peritoneal neutrophil numbers after challenge increased at each observation point, compared with those at 0 h. In particular, at 5 h, the labeled neutrophil number in the peritoneum was significantly higher than that at 0 h (Table 1). The labeled neutrophil number in the lungs was significantly higher at 5 h after challenge than those at 0 and 1 h

(Fig. 4). The pulmonary labeled neutrophil number at 10 h was also higher than those at 0 and 1 h, but the difference did not reach statistical significance (P 5 0.08 and 0.1, respectively). In the kidney, the neutrophil number at 5 h tended to be higher than those at 0, 1 and 10 h, while the number of labeled hepatic neutrophils was essentially unchanged. Lung MPO activity. Although the differences did not reach statistical significance, lung MPO activities appeared to increase after E. coli injection and peaked at 5 h (Table 1). There was a significant positive correlation between lung MPO activity and the pulmonary labeled neutrophil number (P , 0.0001, r 5 0.81) (Fig. 5). Histology in Peritonitis Intraperitoneal injection of 10 9 E. coli induced an increase in the number of neutrophils infiltrating the interstitial space, swelling of alveolar epithelial cells, and mild interstitial edema in the lungs at 5 and 10 h after challenge. However, significant neutrophilic infil-

FIG. 4. Epifluorographs obtained by epifluorescent microscopy in Experiment 2; original magnification, 403. (Left) Lungs from animals sacrificed before bacterial challenge (0 h). (Right) Lungs from animals sacrificed 5 h after 10 9 E. coli challenge. Note the increased numbers of fluorescein-labeled neutrophils in lungs at 5 h after challenge compared with those before challenge.

FUKATSU ET AL.: PERITONITIS INCREASES NEUTROPHIL ADHESION

FIG. 5. The lung MPO activity vs the number of labeled neutrophils in a total of four fields from both lungs before and at 1, 5, and 10 h after bacterial challenge in Experiment 2. The rats were given 10 9 E. coli ip. Lung MPO activities showed a significant positive correlation with labeled neutrophil numbers in the lungs (n 5 21, P , 0.0001, r 5 0.81).

tration or tissue damage was not seen in the kidney or liver at any time point. Hemodynamic and Arterial Blood Gas Data One rat died at 5.5 h after challenge. At 5 h, the dying rat showed marked hypotension and low P aO2. The remaining animals, which survived until 10 h, also showed hypotension at 5 h, hypocapnia at 1, 5, 8, and 10 h, and alkalosis at 8 and 10 h after challenge (Table 2). DISCUSSION

The present results show that peritonitis induced by E. coli leads to dose-dependent increases in labeled neutrophil numbers in the peritoneum, lungs, and kidney at 5 h after challenge. Moreover, neutrophil numbers in the peritoneum and kidney peaked at 5 h after an injection of 10 9 E. coli ip, while pulmonary labeled neutrophil numbers rose to reach a plateau at 5 h. Hemodynamic and arterial blood gas data reflected the presence of a hyperdynamic state.

161

Neutrophils represent an important host defense mechanism against bacterial peritonitis. The influx of neutrophils occurs in response to an inflammatory stimulus [1]. Peritoneal exudative neutrophils localize and contain infection in tissues by phagocytosing and killing bacteria [2– 4]. On the contrary, it has become increasingly clear that activation of neutrophils can also exert deleterious effects on the host. Neutrophils injure tissues distant from the site of infection by releasing enzymes or toxic oxygen products [5]. Thus, the neutrophil is considered to be a primary mediator of remote organ injury in the setting of peritonitis [7, 8]. Therefore, the role of neutrophils should be assessed in light of both host defense mechanisms and remote organ injury, in designing treatments for peritonitis. Neutrophil adherence to the endothelium is one of the sequential processes of exudation into the peritoneal cavity [9, 10]. This adhesion is also essential for endothelial cell injury [11]. Consequently, the examination of neutrophil adhesion in both the peritoneum and remote organs could provide a new approach to developing therapeutic modalities for peritonitis. However, to our knowledge, there have been no reports of the simultaneous observation of neutrophil adhesion at the primary focus of infection and in remote organs in peritonitis. Thus, we studied neutrophil adhesion in the peritoneum and remote organs concurrently in a peritonitis model using a fluorescence microscopic method. Using this method, we previously clarified the following phenomena: a positive correlation between the peritoneal labeled neutrophil number and the peritoneal exudative neutrophil number in a peritonitis model induced by injecting 10 7 E. coli ip [12] and a decrease in the number of labeled neutrophils in the lungs when the antibody to an adhesive molecule, antiICAM-1, was given before endotoxin administration (submitted data). Moreover, the present study revealed a positive correlation between lung MPO activity, a marker of neutrophil infiltration, and the labeled neutrophil number in the lungs. These results suggest that our technique is useful for detecting ICAM-1-mediated adherence of neutrophils to the endothelium. A 2-min harvest time following the injection of neu-

TABLE 2 Hemodynamic and Arterial Blood Gas Data

MAP HR pH PaO2 PaCO2

0h

1h

5h

8h

10 h

134 6 4 466 6 24 7.45 6 0.02 94 6 4 42 6 2

136 6 2 425 6 14 7.44 6 0.003 91 6 4 35 6 1

99 6 14 441 6 22 7.45 6 0.02 88 6 2 31 6 4

128 6 15 454 6 36 7.50 6 0.04 95 6 7 31 6 3

131 6 8 466 6 24 7.49 6 0.01 92 6 13 31 6 3

Note. Values are means 6 SE. The animals were administered 10 9 E. coli ip. One rat died at 5.5 h after challenge. MAP, mean arterial pressure (mmHg); HR, heart rate (beats/min); PaO2, PaCO2 (mmHg).

162

JOURNAL OF SURGICAL RESEARCH: VOL. 81, NO. 2, FEBRUARY 1999

trophils was determined based on our preliminary experiment. In a rat endotoxemia model, thoracic windows were prepared under mechanical ventilation. Then the kinetics of labeled neutrophil number in the lungs after neutrophil injection were studied. The labeled neutrophil numbers in the lungs peaked within 1 min after neutrophil infusion and reached a plateau at 2 min. Therefore, we chose a harvest time of 2 min. Previous studies have focused on alterations in neutrophil populations within the peritoneal cavity during peritonitis [4, 16, 17]. Peritoneal cells were harvested by lavage and induced to differentiate using a cytocentrifuge preparation technique in those studies. Using such a technique reportedly resulted in no significant differences in peritoneal exudative neutrophil number when various amounts of bacteria were given ip [4, 16]. However, in the present study, the labeled neutrophil number in the peritoneum increased E. coli dose dependently. The results suggest that the influx of neutrophils into the peritoneal cavity is augmented as the number of bacteria injected increases. The neutrophil number in the peritoneal cavity is determined by the difference between neutrophil influx and efflux. Taken together, our results also suggest that neutrophil efflux increases according to the injected bacterial number. As there is little evidence that exudative neutrophils return to the circulation from the inflamed site or that lymphatics provide a major disposal route, the number of exudative neutrophils undergoing cell death in the peritoneal cavity may increase when the injected bacterial number increases [18]. In fact, we revealed that coculture of neutrophils with a large number of E. coli increased neutrophil destruction in vitro (unpublished data). In the time course study, the labeled neutrophil number apparently increased within 1 h after challenge. This finding is in agreement with those of previous reports and with our earlier study [4, 12]. Host defense mechanisms involving peritoneal exudative neutrophils are likely to be enhanced at the early stage of peritonitis, i.e., within 1 h after insult. The lungs have been considered to be the remote organ system most commonly affected and the first organ to fail in peritonitis [7, 19–21]. Neutrophils reportedly mediate pulmonary failure primarily [22, 23]. Ishizaka et al. showed that 2 3 10 9 E. coli injected ip increased neutrophil number and the albumin concentration in bronchoalveolar lavage fluid in a guinea pig peritonitis model [23]. However, no reports have described the effect of ip injecting various amounts of bacteria on neutrophil adhesion in a lung during peritonitis. Our data showed that the labeled neutrophil number in the lungs increased according to the number of bacteria injected ip. Moreover, the pulmonary labeled neutrophil number began to increase within 5 h after ip challenge, reaching a plateau at 5 or 10 h. Thus, the number of bacteria injected ip influences neutrophil adhesion to the pulmonary vasculature. In-

creased and persistent neutrophil adhesion in the lung may be the first step in neutrophil mediated lung injury. It has also been reported that neutrophils mediate glomerular injury, possibly by producing toxic oxygen radicals or via myeloperoxidase-dependent reactions [24, 25]. In the present study, the labeled neutrophil number in the kidney correlated positively with the number of E. coli injected ip 5 h after challenge. Although the labeled neutrophil number in the kidney decreased again at 10 h after bacterial injection, neutrophil mediated organ injury can also occur in the kidney during peritonitis. On the contrary, there were no recognizable differences in neutrophil adhesion in the liver in either the dose–response study or the time course study. Although activated neutrophils have been shown to mediate hepatic parenchymal cell injury [26], neutrophils do not appear to affect liver injury in this type of peritonitis. The mechanism by which peritonitis increases neutrophil adhesion in organs, particularly in the lungs, cannot be determined from the present study. However, as the labeled neutrophil numbers seem to reflect ICAM-1-mediated neutrophil adhesion, increased ICAM-1 expression on endothelium may be one of the important mechanisms. Furthermore, Pane´s et al. reported that ICAM-1 expression in the lungs was far greater than that in other organs under baseline conditions, and that endotoxin induced ICAM-1 upregulation in the lungs [27]. Thus, abundant ICAM-1 expression in the lungs may lead to the persistent increase in pulmonary neutrophil adhesion. Increased neutrophil adhesion to the peritoneum may enhance nonspecific host defenses in the peritoneal cavity. However, increased neutrophil adhesion in the lungs and kidney appears to cause neutrophilmediated organ injury. Moreover, differences in the kinetics of neutrophil adhesion among various remote organs may be associated with the susceptibility of an organ to injury in peritonitis. Persistently increased neutrophil adhesion in the lungs may account, at least in part, for the clinical finding that the lung is the most commonly affected organ in peritonitis. In fact, histological studies have demonstrated that neutrophilic infiltration into the interstitial space and tissue damage occurred in the lungs, but not in the kidney or liver, after 10 9 E. coli injection. Consequently, a new approach promoting neutrophils migration toward the peritoneum while reducing neutrophil adhesion in remote organs may provide a breakthrough in the treatment of peritonitis. As the fluorescence microscopic method allows examination of neutrophil adhesion in various tissues simultaneously, the results obtained by this method may yield useful information for developing peritonitis treatments. Further studies in more clinically relevant peritonitis models such as the barium fecal capsule model or cecal

FUKATSU ET AL.: PERITONITIS INCREASES NEUTROPHIL ADHESION

ligation and puncture would also provide interesting information. It must be recognized, however, that the injected neutrophils were activated through the isolation procedures. It is possible that injected neutrophils might have adhered differently than circulating neutrophils to the endothelium. Nevertheless, neutrophil research has relied largely on studies with highly purified cells. To clarify neutrophil adhesion, isolation procedures are seemingly essential. Moreover, the recent study demonstrated that stimulation of endothelial cells is much more important for neutrophil adhesion to the endothelium than neutrophil stimulation [28]. Therefore, our experimental model may reflect the actual setting to some extent. In summary, our fluorescence microscopic method revealed that peritonitis induced by an ip injection of E. coli increased neutrophil adhesion concomitantly in the peritoneum, lungs, and kidney as the injected bacterial load increased. The increase in neutrophil adhesion was transient in the peritoneum and kidney, while being persistent in the lung. Strategies modulating neutrophil adhesion in organs are anticipated to be useful for treating peritonitis. ACKNOWLEDGMENT We thank Tsuyoshi Ishida, from the Department of Pathology, Faculty of Medicine, University of Tokyo, for assistance in performing the histologic analysis.

9. 10.

11. 12.

13.

14.

15.

16.

17.

18.

19. 20.

REFERENCES 21. 1.

2.

3.

4.

5.

6. 7.

8.

Maddaus, M., Ahrenholz, D., and Simmons, R. The biology of peritonitis and implications for treatment. Surg. Clin. N. Am. 68: 431, 1988. Joyce, L. D., Hau, T., Hoffman, R., Simmons, R. L., and Lillehei, R. C. Evaluation of the mechanism of zymosan-induced resistance to experimental peritonitis. Surgery 83(6): 717, 1978. Hau, T., Ahrenholz, D. H., and Simmons, R. L. Secondary bacterial peritonitis: The biologic basis of treatment. Curr. Prob. Surg. 16(10): 1, 1979. Dunn, D. L., Barke, R. A., Knight, N. B., Humphrey, E. W., and Simmons, R. L. Role of resident macrophages, peripheral neutrophils, and translymphatic absorption in bacterial clearance from the peritoneal cavity. Infect. Immun. 49(2): 257, 1985. Goya, T., Morisaki, T., and Torisu, M. Immunologic assessment of host defense impairment in patients with septic multiple organ failure: Relationship between complement activation and changes in neutrophil function. Surgery 115(2): 145, 1994. Varani, J., Ginsburg, I., Schuger, L., et al. Endothelial cell killing by neutrophils. Am. J. Pathol. 135: 435, 1989. Wickel, D., Cheadle, W., Mercer-Jones, M., and Garrison, R. Poor outcome from peritonitis is caused by disease acuity and organ failure, not recurrent peritoneal infection. Ann. Surg. 225: 744, 1997. Anderson, B., Brown, J., and Harken, A. Mechanisms of neutrophil-mediated tissue injury. J. Surg. Res. 51: 170, 1991.

22.

23.

24.

25.

26. 27.

28.

163

Allison, F., Smith, M., and Wood, W. Studies on the pathogenesis of acute inflammation. J. Exp. Med. 102: 655, 1955. Downey, G., Fialkow, L., and Fukushima, T. Initial interaction of leukocytes within the microvasculature: deformability, adhesion, and transmigration. New Horizons 3: 219, 1995. Marcus, B., Hynes, K., and Gewertz, B. Loss of endothelial barrier function requires neutrophil adhesion. Surgery 122: 420, 1997. Fukatsu, K., Saito, H., Han, I., et al. The greater omentum is the primary site of neutrophil exudation in peritonitis. J. Am. Coll. Surg. 183: 450, 1996. Fukatsu, K., Saito, H., Han, I., et al. Nitric oxide inhibition decreases neutrophil adhesion at the inflammatory site, while increasing adhesion in remote organs in peritonitis. J. Surg. Res. 68: 79, 1997. Fukatsu, K., Saito, H., Han, I., et al. Nitric oxide donor decreases neutrophil adhesion in both lung and peritoneum during peritonitis. J. Surg. Res. 74: 119, 1998. Grisham, M., Benoit, J., and Granger, D. Assessment of leukocyte involvement during ischemia and reperfusion of intestine. Methods Enzymol. 186: 729, 1990. Verweij, W., Namavar, F., Schouten, W., et al. In vitro activity of peritoneal cells from rats after intra-abdominal infection with Bacteroides fragilis and Escherichia coli. J. Med. Microbiol. 38: 13, 1993. Verweij, W. R., Namavar, F., Schouten, W. F., de Graaff, J., and MacLaren, D. M. Migration of rat peritoneal cells after intraabdominal infection with Bacteroides fragilis and Escherichia coli. J. Gen. Microbiol. 139: 1739, 1993. Watson, R. W. G., Redmond, H. P., Wang, J. H., Condron, C., and Bouchier-Hayes, D. Neutrophils undergo apoptosis following ingestion of Escherichia coli. J. Immunol. 156: 3986, 1996. Beal, A., and Cerra, F. Multiple organ failure syndrome in the 1990s. J. Am. Med. Assoc. 271: 226, 1994. Richardson, J., DeCamp, M., Garrison, R., and Fry, D. Pulmonary infection complicating intra-abdominal sepsis. Ann. Surg. 195: 732, 1982. Tighe, D., Moss, R., Boghossian, S., Heath, M., Chessum, B., and Bennett, E. Multi-organ damage resulting from experimental faecal peritonitis. Clin. Sci. 76: 269, 1989. Stephens, K., Ishizaka, A., Wu, Z., Larrick, J., and Raffin, T. Granulocyte depletion prevents tumor necrosis factor-mediated acute lung injury in guinea pigs. Am. Rev. Respir. Dis. 138: 1300, 1988. Ishizaka, A., Stephens, K., Tazelaar, H., Hall, E., O’Hanley, P., and Raffin, T. Pulmonary edema after Escherichia coli peritonitis correlates with thiobarbituric-acid-reactive materials in bronchoalveolar lavage fluid. Am. Rev. Respir. Dis. 137: 783, 1988. Boyce, N., and Holdsworth, S. Hydroxyl radical mediation of immune renal injury: Abrogation of renal injury and neutrophil hydroxyl radical generation by desferrioxamine. Kidney Int. 30: 813, 1986. Johnston, R., Couser, W., Chi, E., Adler, S., and Klebanoff, S. New mechanisms for glomerular injury: Myeloperoxidase– hydrogen peroxide– halide system. J. Clin. Invest. 79: 1379, 1987. Holman, J., and Saba, T. Hepatocyte injury during postoperative sepsis. J. Leukocyte Biol. 43: 193, 1988. Panes, J., Perry, M., Anderson, D., et al. Regional differences in constitutive and induced ICAM-1 expression in vivo. Am. J. Physiol. 269: H1955, 1995. Chen, X., and Christou, N. Relative contribution of endothelial cell and polymorphonuclear neutrophil activation in their interactions in systemic inflammatory response syndrome. Arch. Surg. 131: 1148, 1996.