INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Intra-abdominal sepsis occurs as a serious complication in surgical intensive care units and is associated with high mortality (1). One of the major complications of intra-abdominal sepsis is acute respiratory failure (1, 2). Recent evidence has shown that respiratory muscle dysfunction, especially diaphragmatic dysfunction, plays an important role in acute respiratory failure during sepsis (3, 4).

Cecal ligation and perforation (CLP) is a typical experimental animal model of intra-abdominal sepsis and resembles the clinical situations of bowel perforation and bacterial infection of intestinal origin (5, 6). CLP animals show hyperdynamic, hypermetabolic sepsis in the early phase (within 10 h after CLP) followed by hypodynamic, hypometabolic sepsis in the late phase (12 to 30 h after CLP) (6). We recently demonstrated that CLP is closely associated with diaphragmatic dysfunction (7) and that diaphragmatic dysfunction begins to develop from 10 h after CLP (early phase). However, the mechanism of diaphragmatic dysfunction occurring after CLP is not fully understood.

Some investigators have demonstrated that oxygen-derived free radicals play an important role in the development of diaphragmatic dysfunction during endotoxemia. Free radical scavengers prevented diaphragmatic dysfunction and lipid peroxidation occurring after endotoxin administration (8, 9). Endotoxin plays a role in the development of sepsis in CLP (10). Endotoxin activates macrophages, which produce cytokines such as interleukin-1 (IL-1) and tumor necrosis factor-alpha (TNF-) (11). TNF- is known as a potent activator of polymorphonuclear leukocytes (PMN), which trigger the production of oxygen- derived free radicals (12). TNF- gene expression and production occurred in the diaphragm muscle after endotoxin administration (13). Furthermore, the diaphragm muscle is in direct contact with intra-abdominal fluid containing numerous enteric organisms that activate macrophages in CLP animals.

We therefore hypothesized that diaphragmatic dysfunction in CLP animals is mediated by oxygen-derived free radicals and that administration of free radical scavengers could prevent these alterations. Thus, the present study was designed to evaluate the effects of polyethylene glycol-adsorbed superoxide dismutase (PEG-SOD; scavengers of superoxide ions), polyethylene glycol-adsorbed catalase (PEG-CAT; scavengers of hydrogen peroxide), and dimethyl sulfoxide (DMSO; scavengers of hydroxyl radicals) on diaphragmatic contractility and malondialdehyde (MDA) concentrations, as an index of oxygen-derived free radical-mediated lipid peroxidation (14), during intra-abdominal sepsis. Because oxygen-derived free radicals cause cellular damage only when there is an increased production of free radicals or antioxidant defenses are decreased (12), we measured the activities of two main antioxidant enzymes, superoxide dismutase (SOD) and glutathione peroxidase (GPx), after CLP as an index of antioxidant defenses.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study was approved by the Committee on Animal Research of Sapporo Medical University. All experiments were conducted on male Wistar rats (weighing 250 to 350 g) between 8 to 12 wk old.

Animals and Protocol for Induction of Sepsis

Intra-abdominal sepsis was produced using the cecal ligation and perforation technique described previously (6). Briefly, under isoflurane and oxygen anesthesia, we performed a laparotomy through a midline abdominal incision. We used volatile anesthetics to anesthetized rats because volatile anesthetics have little effect on diaphragmatic contractility (15). We then ligated the cecum just below the ileocecal valve with 3-0 silk ligature, so that intestinal continuity was maintained. Using an 18-gauge needle, the cecum was perforated in two locations, 1 cm apart, on the antimesentric surface of the cecum, and the cecum was gently compressed until feces were extruded. The bowel was then returned to the abdomen, and the incision was closed with a layer of proline sutures for the muscles and 3-0 silk for the skin. Afterward, we observed the rats in a recovery cage for 2 h. Antibiotics were not administered to any of the rats. All rats were resuscitated with saline solution (5 ml per 100 g body weight) injected subcutaneously in the back at the time of the operation. The rats were deprived of food but had free access to water after the operation. The control group underwent laparotomy, and the cecum was manipulated but not ligated or punctured.

Study Design

In our preliminary study, we found that reduction in diaphragmatic contractility began to develop from 10 h after CLP (the early phase). We therefore studied the effects of PEG-SOD (Sigma Chemical Co., St. Louis, MO), PEG-CAT (Sigma Chemical Co.), and DMSO (Sigma Chemical Co.) on diaphragmatic contractility 10 h (the early phase) and 16 h (the late phase) after CLP.

Experiment 1: Effect of Free Radical Scavengers on Diaphragmatic Contractility and Concentration of MDA after CLP

One hundred twenty rats were used in the first experiment. The rats were randomly divided into the following 10 groups:

a) A sham-10-h group (n = 12), in which a sterile saline (2 ml) was administered intraperitoneally 30 min before the operation, and the left hemidiaphragm was removed 10 h after the operation;

b) A CLP-10-h group (n = 12), in which a sterile saline (2 ml) was administered intraperitoneally 30 min before CLP, and the left hemidiaphragm was removed 10 h after CLP;

c) A CLP-SOD-10-h group (n = 12), in which PEG-SOD (4,000 U/kg) dissolved in saline (2 ml) was administered intraperitoneally 30 min before CLP, and then the left hemidiaphragm was removed 10 h after CLP;

d) A CLP-CAT-10-h group (n = 12), in which PEG-CAT (15,000 U/kg) dissolved in saline (2 ml) was administered intraperitoneally 30 min before CLP, and then the left hemidiaphragm was removed 10 h after CLP;

e) A CLP-DMSO-10-h group (n = 12), in which DMSO (0.5 ml/kg; 50% solution) dissolved in saline (2 ml) was administered intraperitoneally 30 min before CLP, and then the left hemidiaphragm was removed 10 h after CLP;

f) A sham-16-h group (n = 12), in which a sterile saline (2 ml) was administered intraperitoneally 30 min before the operation, and the left hemidiaphragm was removed 16 h after the operation;

g) A CLP-16-h group (n = 12), in which a sterile saline (2 ml) was administered intraperitoneally 30 min before and 12 h after CLP, and then the left hemidiaphragm was removed 16 h after CLP;

h) A CLP-SOD-16-h group (n = 12), in which PEG-SOD dissolved in saline (2 ml) was administered intraperitoneally 30 min before and 12 h after CLP, and then the left hemidiaphragm was removed 16 h after CLP;

i) A CLP-CAT-16-h group (n = 12), in which PEG-CAT dissolved in saline (2 ml) was administered intraperitoneally 30 min before and 12 h after CLP, and then the left hemidiaphragm was removed 16 h after CLP;

j) A CLP-DMSO-16-h group (n = 12), in which DMSO dissolved in saline (2 ml) was administered intraperitoneally 30 min before and 12 h after CLP, and then the left hemidiaphragm was removed 16 h after CLP.

The hemidiaphragms were used for measuring the effects of PEG-SOD, PEG-CAT, and DMSO on diaphragmatic contractility (n = 60; n = 6 in each group) and MDA concentrations (n = 60; n = 6 in each group) after CLP. The doses and dosing schedule of PEG-SOD, PEG-CAT, and DMSO used in this study were the same as those previously reported (8).

Experiment 2: Diaphragmatic SOD and GPx Activity after CLP

Thirty-six rats were used in the second experiment. The rats were used for measuring diaphragmatic SOD (n = 18) and GPx activity (n = 18) after CLP.

The rats were randomly divided into three groups: (1) a sham group (16 h after a sham operation, n = 12), (2) CLP-10-h group (10 h after CLP, n = 12), and (3) CLP-16-h group (16 h after CLP, n = 12).

Diaphragmatic SOD and GPx activities were determined in the sham group and in the CLP-10-h and -16-h groups (n = 6 in each group).

Experiment 3: Serum and Intra-abdominal TNF- Concentrations after CLP

Thirty rats were used in the third experiment. The rats were used for measuring serum and intra-abdominal TNF- concentrations (n = 30) after CLP.

The rats were randomly divided into five groups (n = 6 in each group): (1) control group (pre-CLP), (2) CLP-2-h group (2 h after CLP), (3) CLP-4-h group (4 h after CLP), (4) CLP-10-h group (10 h after CLP), and (5) CLP-16-h group (16 h after CLP).

Blood and intra-abdominal samples were taken in each group.

In Vitro Diaphragm Muscle Strips

The rats were killed under deep anesthesia of isoflurane. Strips of approximately 10 mm in width, without the phrenic nerves, were dissected from the medial aspect of the left hemidiaphragm of each rat. All strips were removed together with the associated ribs and central tendon. The isolated strips were placed in an organ bath containing oxygenated Krebs' solution (27° C). The strips were mounted vertically in a tissue chamber, with the central tendon superiorly positioned and attached to a Grass FT-10 force transducer (Grass Instruments, Quincy, MA), which was connected to a micropositioner, and was positioned between two platinum plates. Strips were stimulated with supramaximal currents (1.2 to 1.3 times the current required to elicit maximal tension) delivered via platinum field electrodes. The current was supplied by an amplifier driven by a Grass S48 stimulator. In these experiments, d-tubocurarine (15 µm) was added in the Krebs' solution to eliminate activation of intramuscular nerve branches.

Stimulation Paradigm

Strips were allowed to equilibrate for 20 min in the organ bath. The optimal force-length (L₀) relationship was then determined by adjusting the micropositioner between intermittent stimulations of the muscle strip. All stimulations during the study were performed at L₀. Muscle contractile characteristics were then assessed by twitch characteristics and force-frequency relationship. After the 20 min force- frequency determinations, the strips were fatigued over a 4-min period (20 Hz, 330-ms train and 670-ms rest, one train/s).

Twitch Characteristics

Peak twitch tension, contraction time (CT; time to peak tension), and half relaxation time (half RT: time for tension to decay from maximum to half maximum) were calculated from single twitches (0.1 Hz stimulation), which were recorded at high speeds (100 mm/s).

Force-Frequency Relationship

Force-frequency relationships were determined by stimulating the diaphragm strips tetanically at frequencies from 10 to 100 Hz over 10-Hz increments. An interval of 10 s was used between stimuli, and pulses were of 0.2 ms in duration with a train duration of 400 ms. Force-frequency relationships were determined at 20 min after determination of L₀.

Analysis of MDA Activity

MDA concentration was determined in the sham-10-h and -16-h groups CLP-10-h and -16-h groups, CLP-SOD-10-h and -16-h groups, CLP-CAT-10-h and -16-h groups and CLP-DMSO-10-h and -16-h groups (n = 60; n = 6 in each group).

The MDA concentration was determined from the reaction of N-methyl-2-phenylindole with MDA (16). Briefly, diaphragm muscle samples were homogenized with 20 mM Tris buffer, pH 7.4, containing 5 mM butylated hydroxytoluene to make a 20% homogenate. The homogenate was centrifuged at 3,000 g and for 10 min at 4° C. N-methyl-2-phenylindole and 12 N HCl were then added to the supernatant and incubated at 45° C for 60 min. Turbid samples were centrifuged at 15,000 g for 10 min. Measurement of 586 nm absorbance was performed. Final MDA levels are reported as µM of MDA per mg protein. Protein concentrations were measured by the Lowry method (17).

Analysis of SOD Activity

Diaphragmatic SOD activity was determined in the CLP-10-h and -16-h groups and in the sham group (n = 18; n = 6 in each group). To avoid contamination by red cell SOD, diaphragm muscle was perfused through the heart with 50 ml of NaCl 0.9% containing 0.16 mg heparin. The SOD activity was determined from measurement of reaction rates that are accelerated by the SOD (Bioxytech, OR) (18). Briefly, diaphragm muscle samples were homogenized with ice-cold 0.25 M sucrose to make a 10% homogenate. The homogenate was centrifuged at 8,500 g for 10 min at 4° C. The supernatant was mixed with ice-cold ethanol/chloroform, 62.5/37.5 (vol/vol), and centrifuged at 3,000 g for 10 min at 4° C. Then 1,4,6-trimethyl-2-vinylpyridinum trifluoromethanesulfonate was added to the supernatant. Kinetic measurement of 525 nm absorbance changes was performed for 1 min after the addition of 5,6,6a,11b-tetrahydro-3,9,10-trihydroxybenzo [c]fluorene. The SOD activity was determined from the ratio of autoxidation rates measured in the presence (Vs) and in the absence (Vc) of SOD. One SOD-525 activity unit is defined as the activity that doubles the autoxidation background (Vs/Vc = 2). Final SOD activities are reported as units of SOD activities per mg protein.

Analysis of GPx Activity

Diaphragmatic GPx activity was determined in the CLP-10-h and -16-h groups and in the sham group (n = 18; n = 6 in each group). The GPx activity was determined from measurement of the oxidation of NADPH to NADP⁺ reaction rates that are accelerated by the glutathione reductase (Bioxytech) (19). Briefly, diaphragm muscle samples were homogenized with ice-cold buffer (50 mM Tris-HCl, pH 7.5, containing 5 mM ethylenediaminetetraacetic acid (EDTA) and 1 mM 2-mercaptoethanol) to make a 20% homogenate. The homogenate was centrifuged at 10,000 g and for 20 min at 4° C. Glutathione, glutathione reductase, and -nicotinamide-adenine dinucleotide phosphate were then added to the supernatant. Kinetic measurement of 340 nm absorbance changes was performed for 3 min after the addition of tert-butyl hydroperoxide. The GPx activity was determined from the rate of absorbance change/min for the sample to NADPH consumed (nmol/ min/ml). Final GPx activities are reported as units of GPx activities per mg protein.

TNF- Concentration

Intra-abdominal fluid samples were collected by adding sterile phosphate-buffered saline (PBS) solution (5 ml) into the peritoneal cavity. Blood and intra-abdominal fluid samples were centrifuged, and the plasma and the supernatant were frozen in liquid nitrogen. Plasma and supernatant TNF- concentrations were assayed with an enzyme-linked immunoabsorbant assay (Biosource, CA). According to the manufacturer, the sensitivity is 2.3 pg/ml.

Data Analysis

The cross-sectional area was calculated by dividing the muscle mass by the length in centimeters, and the density of muscle was assumed to be 1.056 g/cm³. Tension was calculated as force per unit of cross-sectional area (kg/cm²). Statistics were calculated using a software program (Macintosh StatView J 4.02; Abacus Concepts, Berkeley, CA). Comparisons of force-frequency curves among the groups were made by repeated-measures analysis of variance (ANOVA). Comparisons of the diaphragm muscle twitch characteristics, peak tetanic tension, MDA concentrations, and SOD and GPx activities among the groups were made with one-way ANOVA combined with Scheffe's procedure used for post hoc comparison of data sets. Data are presented as mean ± SEM.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Dry/Wet Ratio

The dry/wet ratio in each group was as follows: 0.256 ± 0.07 in the sham-10-h group, 0.271 ± 0.05 in the CLP-10-h group, 0.261 ± 0.07 in the CLP-SOD-10-h group, 0.258 ± 0.05 in the CLP-CAT-10-h group, 0.261 ± 0.06 in the CLP-DMSO-10-h group, 0.268 ± 0.03 in the sham-16-h group, 0.262 ± 0.06 in the CLP-16-h group, 0.272 ± 0.06 in the CLP-SOD-16-h group, 0.258 ± 0.06 in the CLP-CAT-16-h group, and 0.264 ± 0.06 in the CLP-DMSO-16-h group. There were no significant differences in the wet/dry ratio among the groups.

Oxygen-derived Free Radical Scavengers Pretreatment Twitch Characteristics

The twitch characteristics of the sham groups, CLP groups, CLP-SOD groups, CLP-CAT groups, and CLP-DMSO groups are shown in Tables 1 and 2. Peak twitch tensions in the CLP groups were significantly lower than those in the sham groups (p < 0.01). Peak twitch tensions in the CLP-SOD, CLP-CAT, and CLP-DMSO groups were significantly higher than those in the CLP groups (CLP-SOD-10-h, CLP-CAT-10-h, CLP-SOD-16-h, and CLP-DMSO-16-h groups, p < 0.01; CLP-DMSO- 10-h and CLP-CAT-16-h groups, p < 0.05).

There were no significant differences in half RT and CT between the sham-10-h and CLP-10-h groups. Half RT in the CLP-16-h group was significantly longer than that in the sham-16-h group (p < 0.01). Half RT in the CLP-SOD-16-h, CLP-CAT-16-h, and CLP-DMSO-16-h groups were significantly shorter than that in the CLP group (p < 0.01). CT in the CLP-16-h group was significantly shorter than that in the sham-16-h group (p < 0.05). There were no significant differences in CT among the CLP-16-h, CLP-SOD-16-h, CLP-CAT-16-h, and CLP-DMSO-16-h groups.

Diaphragm Force-Frequency Relationship

The force-frequency curves of the sham groups, CLP groups, CLP-SOD groups, CLP-CAT groups, and CLP-DMSO groups are shown in Figure 1. The force-frequency curves of the CLP groups were significantly lower than those of the sham groups (p < 0.01). The force-frequency curves of CLP-SOD, CLP-CAT, and CLP-DMSO groups were significantly higher than those of the CLP groups (p < 0.01).

View larger version (18K): [in this window] [in a new window]   Figure 1.   Force-frequency curves in 10 experimental groups. 10 h = 10 h after operation; 16 h = 16 h after operation; sham = sham group; CLP = CLP group; CLP-SOD = CLP-SOD group; CLP-CAT = CLP-CAT group; CLP-DMSO = CLP-DMSO group. Values are mean ± SEM.

Peak tetanic tension in each group was as follows: 1.90 ± 0.06 kg/cm² in the sham-10-h group, 1.47 ± 0.08 kg/cm² in the CLP-10-h group, 1.68 ± 0.08 kg/cm² in the CLP-SOD-10-h group, 1.65 ± 0.05 kg/cm² in the CLP-CAT-10-h group, 1.70 ± 0.13 kg/cm² in the CLP-DMSO-10-h group, 1.79 ± 0.05 kg/cm² in the sham-16-h group, 1.03 ± 0.03 kg/cm² in the CLP-16-h group, 1.72 ± 0.09 kg/cm² in the CLP-SOD-16-h group, 1.58 ± 0.05 kg/cm² in the CLP-CAT-16-h group, and 1.60 ± 0.05 kg/cm² in the CLP-DMSO-16-h group. Peak tetanic tensions in the CLP groups were significantly lower than those in the sham groups (p < 0.01). Peak tetanic tensions in the CLP-SOD, CLP-CAT, and CLP-DMSO groups were significantly higher than those in the CLP groups (p < 0.01).

Force generation over time during repetitive contraction trials is shown in Figure 2. During these trials, force generation by muscles from the sham groups was greater than the force generated by muscles from the CLP groups. The force generation by muscles from CLP-SOD groups, CLP-CAT groups, and CLP-DMSO groups was greater than the force generated by muscles from the CLP groups.

View larger version (14K): [in this window] [in a new window]   Figure 2.   Tension over time during repetitive-contraction trials for experimental groups. 10 h = 10 h after operation; 16 h = 16 h after operation; sham = sham group; CLP = CLP group; CLP-SOD = CLP-SOD group; CLP-CAT = CLP-CAT group; CLP-DMSO = CLP-DMSO group. Values are mean ± SEM. **p < 0.01 compared with the sham group; #p < 0.05 compared with the CLP group; ##p < 0.01 compared with the CLP group.

MDA Levels

Diaphragmatic MDA concentrations are shown in Figure 3. Diaphragmatic MDA concentrations in the CLP groups were significantly higher than those in the sham groups (p < 0.01). Diaphragmatic MDA levels in the CLP-SOD, CLP-CAT, and CLP-DMSO groups were lower than those in the CLP groups (p < 0.01).

View larger version (15K): [in this window] [in a new window]   Figure 3.   Diaphragmatic malondialdehyde (MDA) concentrations in 10 experimental groups. 10 h = 10 h after operation; 16 h = 16 h after operation; sham = sham group; CLP = CLP group; CLP-SOD = CLP = SOD group; CLP-CAT = CLP-CAT group; CLP-DMSO = CLP-DMSO group. Values are mean ± SEM. **p < 0.01 compared with the sham group; ##p < 0.01 compared with the CLP group.

SOD and GPx Activities

Diaphragmatic SOD and GPx activities are shown in Figure 4. Diaphragmatic SOD activities in the CLP-10-h and -16-h groups were significantly higher than those in the sham groups (p < 0.01). There were no significant differences between GPx activities in the sham groups and the CLP groups.

View larger version (21K): [in this window] [in a new window]   Figure 4.   Diaphragmatic SOD and GPx activities in three experimental groups. sham = sham group; CLP-10 h = CLP-10-h group; CLP-16 h = CLP-16-h group. Values are mean ± SEM. **p < 0.01 compared with the sham group.

TNF- Concentration

Blood and intra-abdominal TNF- concentrations are shown in Figure 5. Blood and intra-abdominal TNF- concentrations increased after CLP.

View larger version (9K): [in this window] [in a new window]   Figure 5.   TNF- concentrations after CLP. pre-pre-CLP; 2 h = 2 h after CLP; 4 h = 4 h after CLP; 10 h = 10 h after CLP; 16 h = 16 h after CLP. Values are mean ± SEM.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study demonstrated that CLP causes reduction in diaphragmatic contractility, as manifested by reductions in twitch tension and a downward shift of force-frequency curves both in the early stage (10 h after CLP) and in the late stage (16 h after CLP) of sepsis. Diaphragmatic MDA levels were elevated after CLP. Pretreatment with free radical scavengers (PEG-SOD, PEG-CAT, and DMSO) improved diaphragmatic contractility and prevented the elevation in diaphragmatic MDA concentrations after CLP. These results suggested that several types of oxygen-derived free radicals play a role in the reduction in diaphragmatic contractility after CLP.

Effects of Free Radical Scavengers on Diaphragmatic Contractility after CLP

CLP caused reductions in diaphragmatic force generation at all frequencies. Although diaphragmatic force generation was reduced without any change in twitch characteristics at 10 h after CLP, the reductions in force generation were accompanied by decreases in CT and an increase in half RT at 16 h after CLP. The pattern of observed dysfunction at 10 h after CLP is thought to be consistent with either dysfunction of the contractile proteins or altered sarcolemmal function (20, 21). Because twitch kinetics represents a function of the rate of release and reuptake of the sarcoplasmic reticulum, the changes in twitch characteristics and reductions in force generation observed at 16 h after CLP are thought to be consistent with dysfunction of the contractile proteins or altered sarcolemmal function accompanied by the dysfunction of the sarcoplasmic reticulum (20, 21).

Free radical scavengers improved reductions in force generation after CLP. However, the effects of free radical scavengers on the generation of diaphragmatic contractility at 10 h and at 16 h after CLP were different. At 10 h after CLP, free radical scavengers improved muscle force generation at all frequencies of stimulation without changing twitch characteristics. Those effects of free radical scavengers on the generation of diaphragmatic contractility 10 h after CLP were similar to those previously reported in endotoxin-induced respiratory muscle dysfunction (8, 9). Free radical scavengers might prevent dysfunction of the contractile proteins or altered sarcolemmal function at 10 h after CLP. At 16 h after CLP, free radical scavengers improved muscle force generation mainly at high frequencies of stimulation accompanied by increasing CT and decreasing half RT; however, they did not improve muscle-force generation at low frequencies. Changes in twitch characteristics suggested that free radical scavengers prevent dysfunction of the sarcoplasmic reticulum. However, failures in improvement in muscle-force generation at low frequencies suggested that there still existed a dysfunction of contractile proteins.

Effects of Free Radical Scavengers on Diaphragmatic Fatigue after CLP

CLP caused a downward shift of fatigue curves. Pretreatment with free radical scavengers reduced the downward shift of fatigue curves after CLP. The effects of free radical scavengers on the generation of diaphragmatic tension after CLP during fatigue trials were similar to those previously reported in endotoxin-induced respiratory muscle dysfunction (8, 9). During sepsis, hyperventilation, increased transpulmonary resistance, and decreased lung compliance secondary to lung disease increase the work of breathing and the energy demand of respiratory muscles (22). Respiratory muscle fatigue occurs easily under such conditions (23). Furthermore, underlying muscle weakness renders the diaphragm muscle more susceptible to fatigue (24). Several studies have demonstrated that free radicals play an important role in the development of respiratory muscle fatigue (25). Free radical scavengers reduce the rate of development of diaphragmatic fatigue and prevent lipid peroxidation (25). Free radical scavengers might prevent diaphragmatic fatigue after CLP.

Free Radicals and Antioxidants after CLP

In this study, diaphragmatic MDA levels were elevated after CLP. The elevation in MDA levels indicates that oxygen- derived free radicals promote the peroxidation of diaphragm lipids (9, 12). PEG-SOD, PEG-CAT, and DMSO, specific free radical scavengers of superoxide ions, hydrogen peroxide, and hydroxyl radicals, respectively, attenuated both the reduction in diaphragmatic contractility and elevation in MDA levels after CLP, suggesting that several types of free radical species directly damaged diaphragm muscle cells, resulting in a reduction in diaphragmatic contractility after CLP. The productions of free radicals are interrelated (28). Furthermore, the effects of free radical scavengers on diaphragmatic contractility and MDA concentrations, especially at 10 h after CLP, were almost equivalent; it seems that diaphragmatic dysfunction was induced by the interaction of several free radical species rather than by a single free radical species.

There are several sources of oxygen-derived free radicals that might affect diaphragmatic contractility during sepsis. Endotoxin produces diaphragmatic contractile impairment by inducing the production of superoxide or hydroxy radicals in the diaphragm muscle cells (8, 9). TNF- is considered to be a central mediator of inflammatory response during endotoxemia, and injection of TNF- caused deterioration of diaphragmatic contractility, similar to the effect of endotoxin injection (29). Because TNF- induces free radical production (12), the effects of endotoxin are thought to result from TNF--induced free radical production. TNF- gene expression and production occurred in the diaphragm muscle cells after injection of endotoxin, and pretreatment of TNF- antibody prevented deterioration of diaphragmatic contractility (13). Recent evidence suggests that free radical formation during contraction is markedly enhanced in endotoxin-induced septic animals (20). Free radicals generated in this manner are thought to play a role in the endotoxin-induced diaphragmatic contractile alterations. Because endotoxin plays a role in the development of sepsis after CLP, free radicals induced by endotoxin or TNF- might account for the reduction in diaphragmatic contractility after CLP.

Because free radicals cause damage to cells only when the production of free radicals is increased or endogenous antioxidant concentrations are decreased, we examined the activities of two major antioxidant enzymes. Although diaphragmatic MDA concentrations were increased, there were increases in SOD activities without significant changes in GPx activities. Because PEG-SOD prevented increases in diaphragmatic MDA concentrations, overproduction of oxygen-derived free radicals seems to play an important role in the reduction of diaphragmatic contractility after CLP. During endotoxemia, SOD and GPx activities were decreased early in livers (30). Manganese-SOD (Mn SOD) activities in the rat diaphragm were not changed after injection of endotoxin (31). In CLP animals, an increase in SOD activities was reported in hind limb skeletal muscle, although there were no significant changes in copper- zinc SOD (Cu-Zn SOD) activities (32). GPx activities transiently decreased but later increased to the control level. TNF- has been shown to increase Mn SOD activity in both in vivo and in vitro experiments (33, 34). In the present study, we found that there were high concentrations of TNF- in both the circulating plasma and intra-abdominal fluid. It is not yet clear whether the increase in SOD activity is secondary to free radical production or is a direct action of TNF-, and it is also not clear whether the changes in GPx activity represent induction of GPx enzymes.

Intra-abdominal Inflammatory Process after CLP

In CLP animals, sepsis is caused by an intra-abdominal abscess and devitalized tissue. Intra-abdominal fluid contains numerous enteric organisms. We hypothesized that an inflammatory process develops in the abdomen and that this process directly compromises diaphragm function. We measured the intra-abdominal TNF- concentration to examine the intra-abdominal inflammatory process. Although serum TNF- was increased in the early stage (during the first 4 h) after CLP, with a peak at 2 h after CLP, the peak concentration was less than the intra-abdominal TNF- concentration. Furthermore, a higher TNF- concentration was still observed in the intra-abdominal fluid at 16 h after CLP. The increase in intra-abdominal TNF- at 16 h after CLP indicates that the inflammatory process in the abdomen was still in progress. Although the precise role of intra-abdominal TNF- in the deterioration of diaphragmatic contractility is still unclear, our data suggest that oxygen-derived free radicals produced in the intra-abdominal cavity directly affect diaphragmatic contractility in CLP.

Conclusion

In conclusion, intra-abdominal sepsis is closely associated with deterioration in diaphragmatic contractility, and overproduction of oxygen-derived free radicals plays an important role in the alterations in diaphragmatic contractility during intra-abdominal sepsis.