Main

Gastrointestinal dysfunction and intestinal injury are important complications of septic shock syndrome in infants. This intestinal damage may in turn exacerbate the consequences of septic shock and thereby contribute to overall morbidity and mortality. In earlier work, we reported that infusion of endotoxin in newborn piglets produced intestinal injury and diarrhea as well as hemodynamic disturbances(1). The diarrhea noted during endotoxemia could be due to either excessive fluid secretion, induced by increased permeability, or to impaired fluid and electrolyte absorption, or both(2, 3). An increase in propulsive contractions in the gut will reduce the contact time of contents with the absorptive epithelium, causing diminished absorption. However, the role of gastrointestinal motility in diarrhea is still not completely understood(4).

Regular, repeating patterns of small intestinal motility have been recognized in humans and a number of other mammalian species. IMA tracings observed during fasting demonstrate organized phasic, neuromotor patterns known as MMC(5). Each MMC cycle is composed of three phases (Fig. 1A): a quiescent phase I, along with irregular (phase II), and regular spiking activity (phase III). Phase III, which propagates down the intestine, is a propulsive phase associated with intense rhythmic smooth muscle contractions. Intravenous administration of endotoxin in adult rats has been reported to induce a disruption of MMC patterns with a prolongation of phase II(6). The duration of this disrupted period increased with higher endotoxin doses. Normal MMC activity in newborn piglets has been described by Groner et al.(7), but the MMC changes in response to endotoxemia in newborn swine have not been reported. Endotoxin-induced intestinal hemorrhagic necrosis is more commonly seen in newborns(1). Thus, it is important to understand the alterations of IMA and MMC patterns in neonatal endotoxemia. In the present study, we developed a chronically instrumented newborn piglet model and tested the hypothesis that endotoxin infusion alters IMA. We found a previously unreported hyperactive phase that followed an initial period of gut hypoactivity.

Figure 1
figure 1

(A) Normal fasting IMA pattern showing MMC cycle duration (D), length of each phase (PI, PII, andPIII), and time (T) of phase III propagation.(B) Typical IMA alterations induced by endotoxin infusion consisting of an initial prolonged phase I (PI), delayed onset of the first phase III, and two subsequent, shortened MMC cycles. The arrow denotes the start of endotoxin infusion. J5, J25, and J45, jejunum at a distance 5, 25, and 45 cm from the duodenojejunal junction.

Full size image

METHODS

Animal preparation. Newborn piglets 5-12 d old, weighing 3.0-4.7 kg, were anesthetized with halothane and ventilated with oxygen through an endotracheal tube using standard anesthesia equipment. After a sterile midline laparotomy, three electrodes, each consisting of a pair of 0.005-inch diameter Ag/AgCl2 wires (A-M Systems, Inc., Everett, WA), were sewn onto the jejunum at distances of 5, 25, and 45 cm from the duodenojejunal junction. The stripped end of each wire was implanted between the serosa and smooth muscle layers. Next, 3½ or 5 Fr polyurethane catheters (Courtesy of Ray Bodicky, Sherwood Medical, St. Louis, MO) were inserted into the aorta and inferior vena cava through the left femoral artery and vein via a left groin incision. The electrode wires and catheters were passed through a s.c. tunnel to a left flank incision and were stored in a pouch secured to the skin. The peritoneum and all surgical incisions were closed.

Piglets were fasted for 24 h after completion of surgery, during which 80 mL/kg 5% dextrose in 0.225% sodium chloride solution (Abbott Laboratories, North Chicago, IL) was infused. Animals were then fed SPF-Lac sterile milk replacer (Pet-Ag Inc., Hampshire, IL) plus Purina Hog Grower chow (Purina Mills, Inc., St. Louis, MO) mixed with water. Catheters were flushed daily with normal saline containing 10 U/mL heparin, and 5 mg/kg gentamicin and 50 mg/kg of ampicillin were injected through the venous catheter for infection prophylaxis. Antibiotics were withheld for at least 16 h before the start of experiment.

Experimental procedure. After 5-7 d of recovery, animals were fasted overnight but allowed free access to water before each experiment. The conscious piglet was placed in a sling restraint during experiments for 6-8 h. The aortic catheter was connected to a precalibrated pressure transducer(Spectramed P23XL, Oxnard, CA) for monitoring Psa and heart rate. The electrode wires were connected to P511K A.C. preamplifiers (Grass Instrument Co., Quincy, MA) with lower and higher cutoff frequencies set at 30 and 100 Hz, respectively. IMA from each of the three electrodes was continuously recorded using a physiologic recorder (Gould 3000S, Oxnard, CA) with a paper speed of 0.25 mm/s and was simultaneously monitored using a 20-ms/s oscilloscope (Gould 1640S) with a horizontal sweep time of 2 s.

A solution of 5% dextrose in 0.225% sodium chloride was continuously infused at 5 mL/kg/h during experiments. No drug was given on the first day(control) to observe the normal fasting IMA pattern. On the next day, the animals were monitored for at least one full MMC cycle and were then given 30μg/kg Escherichia coli endotoxin (O111:B4 polysaccharide, Sigma Chemical Co.) over 10 min starting at the beginning of phase I.

Data management and analysis. Psa and heart rate were read from the strip charts at baseline, just before and 15, 30, 60, 90, 120, 180, and 240 min after the start of endotoxin or saline infusion. mPsa was calculated. IMA data were read from the same electrode tracing for each animal. IMA variables included the following (Fig. 1A): 1) MMC cycle duration, calculated as the time from the start of phase III activity to the start of the next phase III front; 2) length of each phase in minutes; 3) percentage of phase III activity in the MMC cycle; and 4) migration velocity of phase III, the speed of propagation of the leading edge of phase III, calculated as the anatomic distance (40 cm) from the jujenal electrodes 5 and 45 cm from the duodenojejunal junction divided by the number of minutes required for phase III to traverse this distance.

Normal IMA values for each animal on the control day were calculated from at least four successive MMC cycles. Because the alterations in IMA induced by endotoxin varied with time, the analysis of IMA alterations on the endotoxin day was separated into three periods. An entire cycle was observed before endotoxin infusion. The early period started at the beginning of endotoxin infusion and ended with the onset of the first phase III complex after endotoxin, whereas the later period began with the onset of this first phase III activity. IMA variables for each animal were directly measured and calculated in the preendotoxin cycle, early period, and the first cycle of the later period. In the subsequent MMC cycles, the mean IMA variables were calculated from the successive two or three cycles. Data are expressed as mean± 1 SD. The paired t test was used to evaluate the changes in IMA values induced by endotoxin. One-way repeated measures analysis of variance was used to evaluate the effect of endotoxin on mPsa and heart rate. A p value less than 0.05 was considered to be statistically significant.

RESULTS

General manifestations. mPsa and heart rate were monitored in seven of eight piglets and were stable on the control day. The sublethal doses of endotoxin caused typical biphasic responses characterized by an initial, moderate decrease in mPsa followed by a later, larger decrease(Fig. 2). The initial decrease in mPsa was observed about 15 min after beginning endotoxin infusion and lasted for 5-10 min, followed by a temporal rise to a value slightly higher than baseline. The later hypotension was observed about 60 min after endotoxin infusion and persisted for 30-60 min; mPsa then gradually returned to baseline. Mean heart rate was 124 ± 27 beats/min at baseline, which was not different from the 134± 37 beats/min observed on the control day, and did not change significantly until 1 h after endotoxin infusion. Tachycardia occurred about 1 h after endotoxin infusion with heart rate above 200 beats/min, followed by a gradual return to the normal value by 2-4 h. No diarrhea was noted on the control day. Five of eight piglets developed loose or watery stools which started about 1 h after endotoxin infusion and persisted for 1-4 h, then gradually diminished. The other three animals passed larger amounts of soft stools about 1-2 h after endotoxin administration.

Figure 2
figure 2

The mPsa values observed in the first 4 h of experiments on the control (○) and endotoxin (▴) days in seven piglets. The mPsa was stable on the control day and was significantly affected by endotoxin infusion. *p < 0.05 vs baseline value just before endotoxin infusion (BL) by one-way repeated measures analysis of variance.

Full size image

Characteristics of normal IMA. IMA recorded on the control day in fasted newborn piglets consisted of repeated MMC cycles with a mean cycle duration of 67.0 ± 18.7 min. Each MMC cycle was composed of the classic three phases (Fig. 1A). Phase I, a quiescent period, lasted for 23.4 ± 7.9 min with no to occasional IMA spikes. Phase II, characterized by irregular frequent single spikes and intermittent short bursts of several spikes, was the longest phase, averaging 37.9 ± 13.0 min. Phase III consisted of continuous spiking activity lasting more than 4 min. The phase III complex, which is associated with peristaltic activity, propagated down the intestine with a mean migration velocity of 11.6 ± 3.2 cm/min. The IMA pattern observed in the preendotoxin cycle on the endotoxin day was similar to those seen on the control day(Table 1).

Table 1 IMA alterations induced by endotoxin infusion
Full size table

IMA alterations induced by endotoxin. Seven of eight animals showed obvious changes in IMA after endotoxin infusion, whereas one animal with mild hypotension and without frank diarrhea showed only slight IMA changes. The statistical analysis includes all eight animals. Endotoxin-induced alterations in IMA were biphasic and were characterized by an initial, prolonged phase I, followed by a period of irregular spiking activity and several shortened MMC cycles (Fig. 1B). In two animals, a premature phase III complex was noted 10 and 20 min, respectively, after the start of endotoxin infusion, followed by a prolonged period of quiescence. In these animals, the initial period was calculated as beginning from the end of this premature phase III.

The initial endotoxin effect was a significantly prolonged quiescent phase I, the length of which was double that of control (Table 1). After this, there was a period with irregular spiking activity, which increased with longer bursts of many spikes near the delayed onset of phase III. Animals began passage of stools or diarrhea in conjunction with this increased irregular spiking activity. In the first MMC cycle after endotoxin, phase II was much longer than phase I. In subsequent cycles, the length of phase II shortened significantly (Fig. 1B,Table 1). Compared with control IMA, the mean MMC cycle duration in the later cycles was shorter, whereas the actual duration of phase III did not change. Thus, the calculated percentage of phase III activity in the MMC cycle increased significantly. After three or four cycles, the MMC cycle duration gradually lengthened and returned to normal; however, phase I remained longer than phase II throughout the observation period (Fig. 1B). With normalization of MMC cycling duration and prolongation of phase I length, diarrhea gradually diminished. No significant difference in MMC migration velocity was noted between endotoxin and control days.

DISCUSSION

We recorded normal IMA in chronically instrumented newborn piglets and observed the alterations induced by endotoxin. To minimize interanimal variability, each animal was studied on two separate days and served as its own control. The IMA patterns on the control day consisted of the previously described, repeating phases of MMC cycling. The same IMA patterns were also noted in the preendotoxin cycle on the endotoxin day. The presence of normal IMA patterns on the second experiment day indicates that the first day's observation period did not alter IMA patterns and thus should not have affected the experiments on the second day. The values on the control day were used as normal controls. Because these values were calculated from several successive MMC cycles, they should be closer to the actual values than those observed from the one cycle before endotoxin. On the second study day, sublethal doses of endotoxin induced biphasic alterations in IMA with initial hypoactivity followed by hyperactivity with increased irregular spiking bursts, shortened MMC cycle duration and an increased percentage of phase III activity.

MMC cycling varies considerably among humans and other animal species. The normal MMC cycle duration in the adult human is 84-112 min with min with phase I accounting for 40-60% of total cycle length(5). Fentonet al.(8) recorded small intestinal migrating motor complexes using manometric techniques in six children aged between 8 mo and 11 y, noting a mean cycle duration of 99.5 ± 19.4 min. In a later study, Berseth(9) observed intestinal motor activity in mechanically ventilated, sick newborn infants. Nineteen of 23 preterm infants had no phase III activity, whereas all eight term infants had typical fasting motor activity with a mean MMC cycle duration of 61 ± 10 min (SE). All three phases noted in the term infants were similar to those described in adults. This observation indicates that small intestinal motor activity matures with gestational age. In the present study, the piglets were 10-17 d old and showed typical MMC cycles. The normal MMC cycle duration on the control day and in the preendotoxin cycle on day two was 67.0 ± 18.7 and 73.3 ± 7.3 min, respectively, which corresponds to the 72 ± 8 min reported by Hebra et al.(10) in newborn piglets. Thus, MMC cycle duration in newborn piglets is close to the value noted in humans. This similarity, along with the well known similarities in cardiovascular physiology, supports our use of newborn piglets as a model to investigate both intestinal motility and hemodynamic changes induced by endotoxin infusion.

Endotoxin-induced IMA alterations have been reported in adult animals. Ponset al.(6) showed that i.v. infusion of 10μg/kg endotoxin in adult rats did not significantly alter intestinal MMC, whereas 25 μg/kg E. coli endotoxin caused a period of disappearance of phase III with prolongation of irregular spiking activity lasting 36.2 ± 8.1 min. The duration of this effect was increased to 114.7 ± 19.9 and 190.9 ± 29.7 min with higher doses of 50 and 100 μg/kg of endotoxin, respectively(6). Such endotoxin dose-related IMA responses were also observed in our preliminary studies. No obvious changes were noted with infusion of 10 μg/kg endotoxin, whereas 100 and 500 μg/kg of doses caused severe diarrhea, shock, and increased mortality (our unpublished data). The sublethal dose of 30 μg/kg chosen for the present study caused significant IMA alterations with diarrhea, along with moderate hemodynamic changes. The animals recovered within several hours after endotoxin challenge. Thus, this dose gave us an opportunity to observe the entire course of endotoxin-induced IMA alterations and mPsa responses during a limited experimental period.

After the initial quiescent phase I after endotoxin, we noted increased irregular spiking bursts followed by several shortened MMC cycles with an increased percentage of phase III activity. Thus, endotoxin infusion induced biphasic alterations in IMA; an initial IMA inhibition was followed by a secondary excitation. To our knowledge, no such biphasic alterations in IMA induced by i.v. endotoxin infusion have been previously described. Shortened MMC cycle duration has been reported in fasted dogs after intraluminal injection of cholera toxin(11). Roussel et al.(12) found an increased minute rhythm, defined as the number of clusters of spiking activity per min in phase II, along with diarrhea after oral administration of E. coli enterotoxin in calves. However, intraluminal or oral administration of toxin, although directly affecting the intestine, did not produce toxemia. Sjogren et al.(13) reported that intragastric inoculation of E. coli caused enteroadherent bacterial infection in rabbits and found that an increase in MMC frequency preceded the onset of diarrhea. In our model, i.v. infusion of endotoxin induced systemic hypotension and diarrhea in temporal association with increased spiking activity. Taken together, these observations suggest a possible role of altered intestinal smooth muscle motility in endotoxin-induced diarrhea.

Infusions of sublethal doses of endotoxin caused significant hemodynamic changes in our animals. We previously reported that larger doses of 5-10 mg/kg of endotoxin produced typical biphasic hemodynamic responses in juvenile piglets(14). In the present study, piglets developed similar but less intense systemic arterial hypotension after small doses of endotoxin. The initial systemic hypotension was observed during the quiescent period of IMA, whereas the secondary hypotension with tachycardia started about 1 h after endotoxin infusion, at a time when spiking activity was increasing. This observation leads us to hypothesize that endotoxin-induced IMA alterations may be related to hemodynamic changes, because the intestine is known to be among the organs which are quite sensitive to ischemia. Mesenteric ischemia induced by reversible cardiac tamponade significantly prolonged MMC cycle duration and decreased propagation velocity in a piglet model(10). In contrast, blood reinfusion shortened MMC cycle duration and increased phase III length in opossums with hemorrhagic shock(15). These studies demonstrate that intestinal ischemia inhibits IMA and that reperfusion enhances the activity. Simultaneous measurement of mesenteric blood flow and IMA in our future work will shed further light on the relationship between IMA changes and hemodynamic responses in endotoxemia.

Endotoxemia or endotoxic shock is a complex disease, in which cardiovascular, nervous, and humoral factors are all involved. IMA is primarily controlled by the enteric nervous system(16) and is further regulated by a number of humoral mediators. The combined effects of these factors not only result in hemodynamic disturbances, but may also be responsible for development of the biphasic IMA alterations. For example, IMA is enhanced by parasympathetic activity and decreased by sympathetic stimulation(16, 17). Release of large amounts of prostaglandin E2(18), platelet activating factor(19), and nitric oxide(20) during endotoxemia may cause MMC disruption, leading to a period of phase III disappearance. Prostaglandin F2α has been reported to increase IMA(18). In the present study, we found a temporal association between endotoxin-induced biphasic IMA alterations and biphasic mPsa changes. At present, however, our knowledge of the complex sequence of hemodynamic and IMA disturbances which accompany endotoxemia is rudimentary. Systematic study and identification of the key factor(s) involved in endotoxin-induced intestinal dysfunction and injury will be important in developing treatments for improving the outcome of endotoxemia and septic shock syndrome in humans.