Yersinia intestinal
enterocolitica longitudinal
enteritis affects rabbit smooth muscle function
R. B. SCOTT AND D. T. M. TAN Department of Pediatrics and Gastroenterology Calgary, Alberta TZN 1N4, Canada Scott, R. EL, and D. T. M. Tan. Yersinia enterocolitica enteritis affects rabbit intestinal longitudinal smooth muscle function. Am. J. Physiol. 262 (Gastrointest. Liver Physiol. 25): G278-G284, 1992.-To determine whether Yersinia enterocoLitica (YE) enteritis has an effect on the biomechanical properties of intestinal smooth muscle, New Zealand White rabbits (600-900 g) were divided into an infected group (n = 9) and sham-infected animals fed ad libitum (n = 9), or pair fed with the infected group (n = 9). Animals were inoculated with lOlo organisms of YE in 10 ml NaHC03 (infected group) or 10 ml NaHC03 (sham-infected control and pair-fed groups) at time 0. Daily food intake, weight gain, and YE excretion were noted. Six days later animals were killed and longitudinal smooth muscle strips prepared from proximal (P), medial (M), and distal (D) segments of intestine in each treatment group. Isometric tension was recorded in tissue baths perfused with oxygenated Krebs solution and low6 M tetrodotoxin. Basal and active (the response to 10m5 M carbachol) length-tension curves were generated. Then, with the muscle strips stretched to their optimum length for tension development, the dose response to carbachol and to graded depolarization with KC1 was determined. Infected animals had a significantly reduced food intake and weight gain compared with controls. The development of basal tension with stretch was not significantly different in infected compared with control or pair-fed tissues from the same site. The dose-response curves for tissues exposed to carbachol showed that even though the doses at which the contractile response is half-maximal were not significantly different, the response of infected tissues was significantly impaired compared with that of the pair-fed group, whereas the responsiveness of pair-fed tissues was significantly greater than the response of infected or control tissues at any specific site. Graded depolarization with KC1 reproduced this pattern of response. Thus longitudinal smooth muscle of infected animals (that are also undernourished because of reduced food intake) exhibits a significantly reduced ability to develop tension in response to carbachol compared with pair-fed animals; whereas pair-fed tissues (from animals that are only undernourished) exhibit a significantly increased response to carbachol compared with control or infected animals. The mechanism is unknown, but because the response can be reproduced with KC1 depolarization, it probably reflects receptor-independent changes in smooth muscle function. small intestine; nutrition
contraction;
infection;
inflammation;
under-
YERSINIA ENTEROCOLITICA (YE)is arecognizedcause of bacterial enteritis (13, 15, 16, 24). The spectrum of clinical illness includes both acute gastroenteritis and a chronic relapsing ileocolitis simulating Crohn’s disease of the terminal ileum. Affected patients commonly present with diarrhea and abdominal cramping, symptoms that presumably reflect alterations in the patterns of altered intestinal motor activity and perhaps abnormalities of intestinal smooth muscle contractility. We have previously shown that rabbits infected with a human G278
0193-1857/92
$2.00
Copyright
Research
Group, University
of Calgary,
pathogenic strain of YE develop a clinical illness similar to that in humans, an illness characterized by diarrhea and significantly decreased food intake, weight gain, and survival. Furthermore, infection induces changes in the patterns of intestinal motor activity that are a specific result of infection (they are not related to decreased food intake and weight gain) and are associated with an increased rate of aboral transit (22). The present studies were performed to determine whether YE enteritis and/ or associated undernutrition affects the contractile properties of intestinal smooth muscle. MATERIALS
AND
METHODS
Model New Zealand White rabbits weighing 600-900 g were studied. Animals were initially quarantined for a 3- to &day observation period to ensure the absence of diarrhea. Because YE enteritis in the rabbit is associated with a significantly decreased food intake and weight gain, the experiments were designed to separate the potential contribution of undernutrition to any observed effect of YE enteritis on small intestinal longitudinal smooth muscle function. Experimental
Design
Rabbits were divided into three groups: infected animals (n = 9) and sham-infected control animals fed ad libitum (n = 9) or pair fed with the infected group (n = 9). At time 0 animals received an intragastric inoculation (via nasogastric tube) of lOlo organisms of YE in 10 ml NaHC03 (infected group) or 10 ml NaHC03 (sham-infected control and pair-fed groups). YE strain MCH 700s (serotype 0:3), originally isolated from a patient with diarrhea, was used. The in vivo and in vitro virulence properties of the strain, which possesses the 42-MDa plasmid, have previously been described (14, 20, 21). Daily food intake, weight gain, and YE excretion were noted. The time of study of the infected and sham-infected pair-fed groups was staggered such that the daily food intake of the infected animal was measured and an identical quantity was then fed to its sham-infected pair-fed control. Daily rectal swabs were plated on Salmonella-Shigella media for YE isolation. O’Loughlin et al. (19) documented the clinical, morphological, and biochemical alterations in acute YE infection in rabbits. They showed that mucosal injury is maximal 6 days postinfection, at which time there is extensive abscess formation, villus atrophy, and diffuse brush-border injury, with decreased disaccharidase activity. For this reason animals were killed on day 6 after inoculation and longitudinal muscle strips were obtained for contractility studies from the proximal (P), medial (M), and distal (D) segments of intestine in infected and sham-infected control and pair-fed groups. Contractility
Studies
On day 6, animals anesthetic (halothane)
0 1992 the American
Physiological
were anesthetized with an inhalational and through a midline abdominal inciSociety
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 20, 2019.
YERSINIA
ENTEROCOLITICA
sion the small intestine between ligament of Treitz and ileocecal valve was removed and cleared of feces by gently flushing with normal saline. Lengths (2 cm) of intestine were obtained from the proximal jejunum beginning at the ligament of Treitz, in the midgut, and in the distal ileum 10 cm proximal to the ileocecal valve. The gut segments were opened longitudinally at the mesenteric attachment, pinned out in oxygenated Krebs solution, and a 0.5 x 2 cm longitudinally oriented intestinal segment was obtained from the antimesenteric wall. Intestinal segments, mucosa intact, were anchored to the base of a 20-ml tissue bath, suspended from an isometric force transducer (Harvard Apparatus model 50-7905, O-50 g, Kent, UK) and bathed in Krebs solution. The components of the Krebs were (in mM) 120.3 NaCl, 5.9 CaCl, 2.5 KClz, 1.2 MgC1,, 15.4 NaHC03, 1.2 NaH2P04, and 11.5 glucose. Temperature was maintained at 37OC, and the bath was bubbled continuously with 95% Oa--5% CO,. Mechanical activity detected by the isometric force transducer was amplified by a transducer amplifier (Harvard Apparatus model 50-7970), input to a bioelectric amplifier (HewlettPackard model 8811A), and recorded on an eight-channel chart recorder (Hewlett-Packard model 7858A). Tetrodotoxin ( 10W6 M; Sigma, St. Louis, MO) was present in the buffer throughout. The basal and active length-tension relationships of the tissues were determined as follows. Segments were allowed to stabilize in the tissue bath for 1 h and were then stretched by increments of 1 mm until muscle tension began to increase with any further increase in muscle length. This length was noted as Lie From this initial length (Li), the muscle strip was then progressively stretched to 110, 120, 125, 130, 135, and 140% of Lie At each increment of stretch (100~140% of Li), the muscle was allowed to reach a level of tension that was stable (the basal tension at that level of stretch) and was then maximally contracted using 10m5 M carbachol. The carbachol-stimulated contraction was measured from the stable level of resting tension to the peak tension recorded after carbachol administration. The optimal length of the muscle (L,) was the muscle length at which peak active tension developed. All subsequent experiments evaluating the tissue response to agonists were performed with the segments equilibrated at L,. The dose response to carbachol was established by exposing tissues to 0.5 log M increments of carbachol in a noncumulative fashion with two washes and a 30-min equilibration period between successive doses. The effect of graded KC1 depolarization was established by exposing tissues to 5.9, IO, 20, 30, 40, 60, and 80 mM KC1 in a noncumulative fashion with at least two washes and a 30-min equilibration between successive exposures. The increment in potassium ion was balanced by an equivalent decrement in sodium within the challenge Krebs solution. To minimize variability when data derived from a specific treatment and intestinal site were analyzed, the basal and active tension measurements were normalized to cross-sectional area (i.e., expressed as a stress with units mN/mm2). At the end of each experiment the segment was removed and blotted dry. The mucosa was scraped from the muscularis propria with a glass slide before each strip was weighed. The cross-sectional area (mm”) of each muscle strip was determined using the following equation cross-sectional = length
area (mm2) mass of wet muscle strip (mg) of muscle strip (mm) x density (mg/mm”)
where the density of smooth mg/mm? Histological studies wet muscle strip attributable smooth muscle lavers. and all
(1)
muscle was assumed to be 1.05 determined the thickness of the to the longitudinal and circular calculations of mass have been
IN THE
G279
RABBIT
corrected such that tension is normalized area of longitudinal smooth muscle alone. Statistical
to cross-sectional
Analysis
Results are expressed as means t SE, where the number of rabbits (n) was nine. Incremental stretch of a muscular tissue generates basal tension as an exponential function of the stretch. Therefore the basal tension (T, in newtons) at any length (L, expressed as percentage increment of Li) is given by T = AetmL) + C where A, m, and C are constants. For the purposes of this paper, A = 1, m defines the rate of exponential increase in tension per percentage increment in Lip and C is the y-axis intercept if tension is plotted against length on Cartesian coordinates. Because by definition tension equals zero when the tissue is at its initial resting length (percent increment of Li = 0), then by substitution C must = -I. Exposure of muscle strips to the agonist carbachol (lo-” M) at each increment of stretch generates a length-active tension curve. Tissue strips develop considerable active tension in response to the agonist, even at their initial length. If active tension is plotted against percentage increment of Li on Cartesian coordinates a hyperbolic function with a positive y- intercept (tension at 0 stretch) is generated. The tension (T, in newtons) in response to agonist at any stretched length (L, expressed as percentage increment of Li) is given by T=
T max L EDs0 + L
where Tmax represents the maximal tension developed and EDso represents the dose that produces half-maximal tension, and tensions are all normalized so that T = 0 when the percentage increment of stretch equals zero. Normalized tension in any given condition equals T - mean tension at Lie Once curve parameters were determined, the true Tmax was calculated as the normalized Tmax + mean tension at Lie The muscle length at which peak active tension developed (L,) was determined from the data for each of the nine animals in each treatment group. The difference between the means of the L, for each treatment group was compared using analysis of variance (ANOVA) with a Tukey’s test for post hoc comparison of pairs of means where multiple comparisons were possible. Dose-response experiments generate tension as a hyperbolic function of stretch or agonist dose. Therefore the active tension (T, in newtons) at any concentration of agonist (C) is given by T=
Tmax
C
EDsa + C
In this equation Tmax represents the maximal tension developed and EDs0 represents the dose that produces half-maximal tension. To determine the parameters defining the basal length tension, active length tension, and dose-response parameters of the data, nonlinear least-squares regression analysis was performed using the statistical software package Systat (Systat, Evanston, IL) and a microcomputer (Zenith 386). To define the differences that might exist between curves and to define confidence intervals for comparisons of parameter estimates (m for length-basal tension data, Tmax and EDso for lengthactive tension and dose-response data), we utilized a recently described nonlinear least-squares computer analysis (18). Comparisons were made of stress developed in response to different treatments (control, pair feeding, or infection) at a given site along the intestine and for stress developed at different sites along the intestine (proximal ieiunum. midgut. or distal ileum)
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 20, 2019.
G280
YERSINIA
ENTEROCOLITICA
within a single treatment. When multiple comparisons were made between more than two curves, a Bonferroni adjustment was used so that the adjusted P value was equal to the unadjusted P value divided by the number of multiple comparisons. Comparisons were considered significantly different whenever the adjusted P < 0.05. To test for the presence of a potential interaction between the effects of treatment and location, we modeled (using multivariate least-squares regression analysis and the statistical package Systat) the stress developed in response to the classification variables level of applied stimulus, treatment (control, pair feeding, or infection), the location (distal ileum, midgut, and proximal jejunum), and the interaction term (treatment x location) stress = constant
+ level of stimulus + location
+ treatment + (treatment
X
location)
Separate analyses were made for the data defining the active length-stress response (7 levels of stimulus between 100 and 140% of Li), the dose response to carbachol(9 levels of stimulus between 10v8 and 10m4 M carbachol), and the dose-response to KC1 (7 levels of stimulus for concentrations of KC1 between 0 and 80 mM). A squared correlation coefficient ( r2, which defines the proportion of the variance of the data that is predicted by the model) and the significance level for the contribution of each of the parameters to the predictive capability of the model are reported. Histological
Technique
Muscle strips from control, infected, and pair-fed animals were fixed in formaldehyde solution, embedded in paraffin, sectioned, and stained with hematoxylin and eosin. By use of a Leitz Wetzlar microscope (FRG) at 250 power, the thickness of the longitudinal and circular muscle layers was calculated with an ocular micrometer and standard image-splitting micrometry.
IN
THE
RABBIT
of the muscularis propria attributable to the longitudinal smooth muscle layer in tissues from control, infected, and pair-fed groups at the level of the proximal jejunum, midgut, or distal ileum (Table 2). Basal length-stress response. Figure 1 shows the lengthbasal stress curve for longitudinally oriented tissue strips from the proximal jejunum of control, pair-fed, and infected groups of animals. Similar data were obtained but are not shown for midgut and distal ileal tissue strips from the same animals. In each treatment group, at each site, there was an exponential increase in basal stress with incremental stretch. Estimates of the curve parameter wl for the different treatments at each location were not significantly different from one another. Active length-stress response. The length vs. active stress in response to lo-” M carbachol for longitudinally oriented tissue strips from the midgut of control, pairfed, and infected animals are plotted in Fig. 2. Visual inspection of these length-active stress curves shows that the y-intercept (active tension at Li), slope (rate of increase in stress with incremental stretch), and the T,,, (maximal stress generated at L,) are diminished for infected compared with control or pair-fed tissues and increased for pair-fed compared with control or infected tissues. The muscle length at which peak active stress developed (L,) was similar for all three treatments at all three sites. The T,,, and L, for each treatment at all three sites are summarized in Table 3. There was no significant difference in L, (which averaged 125% of Li) among control, pair-fed, or infected tissues at any site. However, there was a consistent and significant (P < 0.05) decrease in the maximum active stress generated Table 2. Proportion of thickness of muscularis propria attributable to longitudinal smooth muscle layer
RESULTS
Control
Clinical Nine animals were inoculated with YE on day 0. All excreted YE in the stool in follow-up compared with none in the control or pair-fed groups-Infected animals demonstrated a significantly (P < 0.001) decreased food intake and weight gain compared with controls but not pair-fed animals (Table 1).
Proximal jejunum Midgut Distal ileum
_T1
Values
are means
Pair
Fed
Infected
P
0.22t0.07 0.17t0.03
0.13t0.02 0.13t0.01
0.14t0.02 0.12t0.01
0.19
0.17&0.03
0.15t0.02
0.13t0.02
0.57
c(-
0.21
t SE.
lo.0
. CONTROL A PAIR-FED q INFECTED
8.0 --
Contractility
of Intestinal Longitudinal Smooth Muscle
Longitudinal smooth muscle thickness. There was no significant difference in the proportion of the thickness Table 1. Effect of Y. enterocolitica enteritis on food intake and weight gain Control
Protocol Avg daily food intake, g Body wt, g Day 0 before inoculation Day 6 Values are means P < 0.01.
Pair
Fed
Infected
ys
4*o-
1
41.1t10.6*
41.4t10.6*
818*54
835225
831t22
1,060+54
A
m
70.0t3.1
+: SE. * Significantly
EL-+ w cr OJE 6.0-&E \ az
870t30* different
100 LENGTH
789t41* from
control
group,
120 140 (% of Li)
Fig. 1. Length-basal stress curves for strips from proximal jejunum of control, (Protocol 1). Li, initial length.
longitudinally oriented tissue pair-fed, and infected animals
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 20, 2019.
YERSINIA 20.0
. CONTROL A PAIR-FED w INFECTED T 1
ENTEROCOLITICA
IN
THE
25.0
-8
120 140 LENGTH (% of Li)
Control
fed
LO Infected
Proximal 8.0tl.O
14.4k1.7"
Control
Pair
fed
131k1.9
126t3.8
15.1*1.2*
5.7+0.2*"r
Distal 10.3tl.O
16.4tl.l*
7.8+0.5*?
124t3.7
120t3.8
126t3.5
fed
118t5.0
tension generated; L,, optimal Values are means t SE. T,,,, maximal length. * Significantly different from control, P < 0.05. 7 Significantly different from pair fed, P < 0.05.
by infected compared with control tissues and a significant increase in the maximum active stress generated by tissues from the pair-fed compared with control groups at all three sites in the intestine. Dose-response to carbachol. Figure 3 is a plot of the active stress generated in response to carbachol over the range 10m8 to 10B4 M b y 1ongitudinally oriented tissue strips from the midgut of control, pair-fed, and infected animals. All tissues were studied at L, (125% of Li). Similar plots were obtained for the proximal jejunum and distal ileum from the same animals but are not shown. The T,,, and EDs0 for each treatment at all three sites have been summarized in Table 4. At each site there was a consistent and significant (P < 0.01) decrease in the maximum active stress generated in infected compared with control tissues and a significant (P < 0.01) increase in the maximum active stress generated by tissues from the pair-fed compared with control group. There were no significant differences in EDs0 among control, pair-fed, or infected tissues. Dose response to KCI. Figure 4 graphs the active stress generated in response to graded depolarization with KC1 over the range 5.9-80 mM. All tissues were studied at L, (125% of Li). Data for tissue strips from the distal ileum
Control
Control
Pair
Infected
8.1t0.4
11.7rtO.5"
3.8+0.2*j-
13.2k0.7
17.6t0.5*
4.5+0.2*-F
13.4t0.5
15.5tl.O*
Proximal
E&o,
nM
Pair
fed
Infected
jejunum 8t3
8&3
6k2
13t4
8t2
lOk4
3t2
6t4
6t2
Midgut
Distal
123t3.3
ileum 127t4.3
mN/mm2
122t4.3
Midgut 9.9t0.8
max9
Infected
jejunum
3.8&0.2*-f-
-4 ’ M
Fig. 3. Active stress developed in response to carbachol by longitudinally oriented tissue strips from midgut of control, pair-fed, and infected animals. Where no error bars are visible, they are obscured by symbol showing mean.
T
mN/mm2
Pair
-6 LOG[CARBACHOL],
Table 4. Dose response to carbachol
Table 3. Active length-stress response to carbachol (low5 M) max9
CONTROL A PAIR-FED
0.0’
Fig. 2. Length vs. active stress in response to 10m5 M carbachol for longitudinally oriented tissue strips from midgut of control, pair-fed, and infected animals. Where no error bars are visible, they are obscured by symbol showing mean.
T
-0
T
0.0 I. 100
G281
RABBIT
ileum
11.3~0.5""f
Values are means t SE. T,,,,, maximal tension; EDSO, dose producing half-maximal response. * Significantly different from control, P < 0.01. T Significantly different from pair fed, P < 0.01.
1 0.0
. CONTROL A PAIRFED n INFECTED
T
q
60
80
8.0 6.0
0
20
40
[KCI],
mM
Fig. 4. Active stress generated in response to graded depolarization with KC1 by longitudinally oriented tissue strips from distal ileum of control, pair-fed, and infected animals. Where no error bars are visible, they are obscured by symbol showing mean.
of control, pair-fed, and infected animals are shown. Similar plots were obtained for the proximal jejunum and midgut. The T,,, and EDso for each treatment at all three sites are summarized in Table 5. With increasing
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 20, 2019.
G282 Table
YERSINIA
5. Dose response
to KC1 at L,
T Ill&IX, mN/mm2 Control
Pair
0.9kO.3
3.2tl.l"
fed
ENTEROCOLITICA
Infected
Control
Proximal 2.OkO.6
E&0,
nM
Pair
fed
Infected
jejunum 20.4t10.5
36t18.8
10.9k10.5
Midgut l.lIfrO.4
7.4t3.6"
2.2+0.6-j-
2.OkO.5
9.7*1.1*
2.5+0.5-j-
22.4t18.1
Distal
19.8t14.6
20.1t20.9
20*9
26S3
ileum 25tlO
Values are means t SE. L,, optimal length; TmBX, maximal EDso, dose producing half-maximal response. * Significantly from control, P < 0.001. t Significantly different from pair
tension; different fed, P
midgut > proximal jejunum) (23), we next tested the validity of this relationship on the data reported in these experiments. It is apparent that T,,,, in distal ileal tissues was consistently greater than in midgut, which was greater than in proximal jejunum for any treatment (control, pairfeeding, or infection) in response to stimulation with carbachol (Tables 3 and 4) or KC1 (Table 5). Although the EDSOs were all similar, the differences in T,,, were statistically significant (P < 0.05) for comparisons made between the most widely separated segments: distal ileum and proximal jejunum. Thus stress development is dependent on both treatment (control, pair feeding, or infection) and intestinal location (distal ileum, midgut, or proximal jejunum). If the T,,, responses of infected animals are expressed as a percentage of the pair-fed response, there appears to be a diminution of the effect of infection along an aboral gradient (26 vs. 76% of pair-fed controls for the proximal jejunum and distal ileum, respectively, for the data in Table 3; 32 vs. 72% of pair-fed controls for the proximal jejunum and distal ileum, respectively, for the data in Table 4). This trend suggested an interaction of aboral location with the effect of treatment. To test for any significant interaction or a local influence of intestinal location on the effect of treatment we utilized the data summarized in Tables 3-5 and for each set of data modelled the stress developed in response to stimulation against the level of stimulus, treatment, location and the interactive component (treatment X location). The model predicted 61% of the variance in the stress
IN
THE
RABBIT
developed (r2 = 0.61) for the length-active stress (Table 3) and the dose-response to carbachol (Table 4) data but only 43% of the variance in the dose-response to KC1 (Table 5) data. In each case the separate effects of level of stimulus, treatment, location, and the interactive component (treatment x location) were highly significant (P < 0.001). Thus stress development depends on the level of stimulation but is significantly affected by treatment and intestinal location. In addition to the separate effects of treatment and location, these results suggest that local factors that depend on aboral location along the gut modify the effect of treatment. DISCUSSION
Our findings demonstrate that in the rabbit animal model YE enteritis and undernutrition induce separate and contrasting alterations of the contractile properties of longitudinal smooth muscle in the small intestine. Tissues from infected animals, which manifested a significantly decreased food intake and weight gain compared with controls, had a significantly reduced responsiveness to carbachol compared with controls. In contrast, tissues from animals pair fed with the infected group (an undernourished but uninfected group of animals) had a significantly increased responsiveness to carbachol compared with ad libitum-fed controls. The decreased contractile response in infected animals and the increased responsiveness of undernourished animals were not neurally mediated inasmuch as they occurred in the presence of tetrodotoxin. There was no change in the EDSO of carbachol in tissues of infected or pair-fed compared with control animals, and the increased responsiveness of undernourished compared with control tissues and the decreased responsiveness of infected compared with undernourished tissues were reproduced with KC1 depolarization, suggesting that the mechanism involves an alteration of smooth muscle function that is not receptor mediated. The effect of inflammation on the function of intestinal smooth muscle has been studied in the human and several animal models (3). The decreased active contractile response to carbachol observed in response to YE enteritis is similar to that reported in colonic smooth muscle in human (25) and rabbit (2, 28) colitis or to the inhibition of rabbit ileal smooth muscle contractility that occurs in response to direct in vitro exposure to Clostridium dif’icile toxin B (7). However, it is in contrast to the response observed in animal models of intestinal parasitism. In rat animal models the nematodes Nippostrongylus brasiliensis and Trichinella spiralis induce timedependent reversible inflammatory changes in the jejunal mucosa and submucosa and marked hypertrophy of the muscularis propria. Parasitic infection is associated with a significantly increased development of tension by jejunal longitudinal smooth muscle in response to carbachol and 5’-hydroxytryptamine (4, 5, 26). Because the increased tension was expressed per unit cross-sectional area of muscle, the findings were independent of the trophic changes in smooth muscle observed in nematodeinfected rat intestine. The altered contractility of intestinal smooth muscle in the parasitized models was, as in
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 20, 2019.
YERSINIA
ENTEROCOLITICA
our experiments, thought to be related to changes in excitation-contraction coupling at steps distal to, or separate from, receptor occupancy. The pathological injury associated with YE enteritis in the rabbit animal model has been documented (19). By day 6 after infection, gross and microscopic morphological abnormalities are identified. On gross inspection the terminal ileum, cecum, and proximal colon are distended, hyperemic, and fluid filled. There is a patchy distribution of microabscesses surrounding bacteria in the lamina propria of the crypt region, sometimes causing the formation of small ulcers. These inflammatory infiltrates extend into the muscularis mucosa but not into the muscularis propria. Although these morphological changes are evident throughout the intestinal tract, they are more extensive in the ileocecal region. In addition there is a significant reduction in villus height and mucosal enzyme activity and a significant increase in cell proliferation in the ileum, but not in the proximal intestine or midgut, of infected compared with pair-fed and control animals. In contrast to all the findings showing that YE enteritis localizes and has its greatest impact on the mucosa of the terminal ileum, we found that the changes in contractility associated with infection and undernutrition were evident in the proximal jejunum, midgut, and terminal ileum. Although the response was generalized and suggested the involvement of a mediator(s) with a systemic response, we also demonstrated a significant interaction between the effects of treatment and location. The interaction supports the presence of additional local factors that depend on aboral location and modify the ability of treatment to produce an effect. It is very likely that the gross and microscopic localization of YE infection and its associated inflammatory response in the distal small intestine is one such local factor. However, the mechanisms of effect and the identity of the mediators responsible for the generalized and local effects of undernutrition and infection are unknown. The potential role of inflammatory mediators and cytokines needs further investigation. Support for this hypothesis is provided by the results of recent experiments in another animal model. In the nematodeinfected rat, jejunal muscle from worm-free segments excluded from the rest of the gut before infection manifested increased tension generation. In this model daily steroid therapy (betamethasone 3.0 mg/kg SC daily) abolished the increase in myeloperoxidase activity associated with infection and attenuated the increased stress generated in worm-free jejunal muscle from infected animals. These results suggest that alterations in intestinal smooth muscle function in the nematode-infected rat do not require the presence of the parasite in the lumen and may be mediated by systemic mechanisms that are part of the host inflammatory response (17). The finding that undernutrition resulted in a significant increase in intestinal smooth muscle contractility compared with controls was unexpected. During prolonged undernutrition protein synthesis and energy utilization in the viscera are maintained at the expense of peripheral protein degradation (18). However, the effect of undernutrition on striated muscle structure and function varies depending on the muscle’s contractile activity
IN
THE
RABBIT
G283
and oxidative capacity (6, 9, 10, 12). Effects on intestinal smooth muscle have not been well studied. Malnutrition of growing animals causes a significant reduction in total body weight and intestinal wet weight (11). Thus we had projected that undernutrition of the infected and pairfed groups of animals in our study would result in decreased muscle mass and a proportionate reduction in the total active tension developed by the tissue. Not only is this the opposite of what was observed, but the data are expressed as newtons per millimeter squared of intestinal longitudinal smooth muscle, and the differences in contractility that we describe cannot be attributed to differences in the mass of intestinal longitudinal smooth muscle between treatment groups. There is little literature on the effect of undernutrition on intestinal smooth muscle contractility. Interestingly, Weisbrodt et al. (27) examined a related topic, the functional and structural effects of nutrient exclusion in a rat model of jejunoileal bypass. In their study circular muscle strips from the proximal, mid, and distal portions of the bypassed intestine and from the distal portion of the in-continuity intestine generated greater maximal active stress than tissues from unoperated controls. However, increases in stress development were also observed in strips taken from equivalent areas of sham-operated animals compared with unoperated controls. These data suggest that systemic factors, perhaps related to laparotomy, intestinal transection, or the transient nutritional and metabolic alterations induced by surgical bypass of -70% of the pylorus-to-cecum length can alter intestinal smooth muscle contractility. Perhaps the undernutrition of the growing rabbits utilized in our study provoked a similar stress response that mediated the increase in contractility. The increased responsiveness of undernourished compared with control tissues and the decreased responsiveness of infected compared with undernourished tissues occurred in the presence of tetrodotoxin and probably represents an alteration of smooth muscle function. Carbachol induces a receptor-mediated contraction of intestinal smooth muscle, whereas KC1 depolarizes the membrane and induces contraction independently of membrane receptors. The increase in carbachol-stimulated maximum stresses observed in undernourished compared with control tissues was reproduced by KC1 depolarization. Although infection caused a decrease in carbacholstimulated maximum stress to a level significantly less than that seen in tissues from ad libitum-fed control animals, KC1 depolarization caused a reduction that was significantly less than that observed in pair-fed undernourished tissues but not different from ad libitum-fed control tissues. In tissues from infected animals, the portion of the decrease in carbachol-stimulated maximal stress that could not be reproduced by KC1 depolarization may be receptor dependent and perhaps reflect an alteration in receptor number. However, the changes in contractility produced by undernutrition and the major proportion of that induced by infection were reproduced with KC1 depolarization. In addition, the EDSOs of the control, pair-fed, and infected tissue responses to carbachol were not significantly different. This suggests that undernutrition and infection induce alterations in me-
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 20, 2019.
G284
YERSINIA
ENTEROCOLITICA
chanical efficiency of smooth muscle function at a point distal to or separate from receptor occupancy. Mechanisms might potentially include changes in 1) the intrinsic elasticity or resistance of the tissues to stretch due to co1 lagen deposition, edema, and so forth, 2) the smooth muscle cell content of contractile proteins, or 3) calciummediated excitation-contraction coupling. The first alternative seems less likely because the basal lengthtension curves of control pair-fed and infected tissues were similar. Further experiments need to be performed to test the latter two potential mechanisms. In summary, these studies have demonstrated that in a rabbit animal model of YE enteritis intestinal longitudinal smooth muscle shows contrasting alterations in contractility in response to the infectious enteritis (which is associated with undernutrition) and to undernutrition alone. Although the data suggest th .at there is a change in muscle function at a step distal to or separate from receptor occupancy, additional work will need to be done to identify the precise mechanism and the mediators involved. The authors are grateful to Wendy Weinstein for typing the manuscript. The work was supported through funding by the Canadian Foundation for Ileitis and Colitis and by Medical Research Council of Canada Grant MT-10014. Address for reprint requests: R. B. Scott, Dept. of Pediatrics, Faculty of Medicine, University of Calgary, 3330 Hospital Dr. NW, Calgary, Alberta T2N lN4, Canada. Received
5 November
1990; accepted
in final
form
9 September
1991.
REFERENCES 1. Brown, B. P., S. Anura, and D. D. Heistad. Responsiveness of longitudinal and circular muscle layers of the portal vein. Am. J. Physiol. 242 (Gastrointest. Liver Physiol. 5): G498-G503, 1982. 2. Cohen, J. D., H. W. Kao, S. T. Lechago, and W. J. Snape, Jr. Effect of acute experimental colitis on rabbit colonic smooth muscle. Am. J. Physiol. 251 (Gastrointest. Liver Physiol. 14): G538G545, 1986. 3. Collins, S. M., J. D. Huizinga, and D. L. Vermillion. Smooth muscle function in the inflamed bowel. In: Inflammatory Bowel Disease, edited by H. Freeman. Boca Raton, FL: CRC, 1989, p. 5976. 4. Farmer, S. G., J. M. Brown, and D. Pollock. Increased responsiveness of intestinal and vascular smooth muscle to agonists in rats infected with Nippostrongylus brasiliensis. Arch. Int. Pharmacodyn. Ther. 263: 217-227,1983. 5. Fox-Robichaud, A. E., and S. M. Collins. Altered calcium handling properties of jejunal smooth muscle from the nematode infected rat. Gastroenterology 91: 1462-1469, 1986. 6. Gardner, P. F., G. Montanaro, D. R. Simpson, and V. R. Edgerton. Effects of glucocorticoid treatment and food restriction on rat hindlimb muscles. Am. J. Physiol. 238: (Endocrinol. Metab. 1): El24E130, 1980. 7. Gilbert, R. J., C. Pothoulakis, and J. T. LaMont. Effect of purified Clostridium difficile toxins on intestinal smooth muscle. II. Toxin B. Am. J. Physiol. 256 (Gastrointest Liver Physiol. 19): G767-G772,1989.
IN
THE
RABBIT
8. Goldberg, A. L., and T. W. Cheng. Regulation and significance of amino acid metabolism in skeletal muscle. Federation Proc. 37: 2301-2307, 1978. 9. Goldspink, G., and P. S. Ward. Changes in rodent muscle fibre type during post-natal growth, undernutrition and exercise. J. Physiol. Lond. 296: 453-469, 1979. 10. Goodman, M. N., M. A. McElaney, and N. B. Ruderman. Adaptation to prolonged starvation in the rat: curtailment of skeletal muscle proteolysis. Am. J. Physiol. 242 (EndocrinoZ. Metab. 4): E321-E327,1981. 11. Hatch, T. F., E. Lebenthal, D. Branski, and J. Krasner. The effect of early postnatal acquired malnutrition on intestinal growth, disaccharidases and enterokinase. J. Nutr. 109: 1874-1879, 1979. 12. Kelsen, S. G., M. Ference, and S. Kapoor. Effects of undernutrition on the diaphragm. J. AppZ. Physiol. 58: 1354-1359, 1985. 13. Kohl, S. Yersinia enterocolitica infections in children. Pediatr. CZin. N. Am. 26: 433-443, 1979. 14. Lian, C.-J., and C. H. Pai. Inhibition of human neutrophil chemiluminescence by plasmid-mediated outer membrane proteins of Yersinia enterocolitica. Infect. Immun. 49: 145-151, 1985. 15. Maki, M., J. Vesikari, I. Rantala, and P. Gronroos. Yersiniosis in children. Arch. Dis. Child. 55: 861-865, 1980. 16. Marks, M., C. Pai, L. Lafleur, L. Lackman, and 0. Hammerberg. Yersinia enterocolitica gastroenteritis: a prospective study of clinical, bacteriologic and epidemologic features. J. Pediatr. 96: 26-31, 1980. 17. Marzio, L., P. Blennerhassett, S. Chiverton, D. L. Vermillion, J. Langer, and S. M. Collins. Altered smooth muscle function in worm-free gut regions of TrichineZZa-infected rats. Am. J. Physiol. 259 (Gastrointest. Liver Physiol. 22): G306-G313, 1990. 18. Meddings, J. B., R. B. Scott, and G. H. Fick. Analysis and comparison of sigmoidal curves: application to dose-response data. Am. J. Physiol. 257 (Gastrointest. Liver Physiol. 20): G982-G989, 1989. 19. O’Loughlin, E. V., G. Humphreys, I. Dunn, J. Kelly, C. J. Lian, C. Pai, and D. G. Gall. Clinical, morphological and biochemical alterations in acute intestinal yersiniosis. Pediatr. Res. 20: 602-608, 1986. 20. Pai, C. H., and L. DeStephano. Serum resistance associated with virulence in Yersinia enterocolitica. Infect. Immun. 35: 605611, 1982. 21. Pai, C. H., V. Mors, and T. Seemayer. Experimental Yersinia enterocoZitica enteritis in rabbits. Infect. Immun. 28: 238-244, 1980. 22. Scott, R. B., D. G. Gall, and S. C. Diamant. Intestinal motility during acute Yersinia enterocolitica enteritis in rabbits. Can. J. Physiol. PharmacoZ. 67: 553-560, 1988. 23. Scott, R. B., and D. T. M. Tan. Differential smooth muscle contractility along the rabbit small intestine (Abstract). Gastroenterozogy 98: A389, 1990. 24. Simmonds, S. D., M. A. Noble, and H. J. Freeman. Gastrointestinal features of culture-positive Yersinia enterocolitica infection. Gastroenterology 92: 112-117, 1987. 25. Snape, W. J., Jr., R. Williams, E. A. Mayer, and D. Root. Disturbances in colonic contraction in colonic muscle removed from patients with ulcerative colitis (Abstract). Gastroenterology 92: Al648,1987. 26. Vermillion, D. L., and S. M. Collins. Increased responsiveness of jejunal longitudinal muscle in TrichineZZa-infected rats. Am. J. Physiol. 254 (Gastrointest. Liver Physiol. 17): Gl24-G129, 1988. 27. Weisbrodt, N. W., P. R. Nemeth, R. L. Bowers, and W. A. Weems. Functional and structural changes in intestinal smooth muscle after jejunoileal bypass in rats. Gustroenterology 88: 958963,1985. 28. Xie, Y. N., S. N. Reddy, W. Gerthoffer, and W. J. Snape. Effect of mucosal inflammation on colonic smooth muscle contraction (Abstract). GastroenteroZogY 98: A402, 1990.
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 20, 2019.
enterocolitica longitudinal
enteritis affects rabbit smooth muscle function
R. B. SCOTT AND D. T. M. TAN Department of Pediatrics and Gastroenterology Calgary, Alberta TZN 1N4, Canada Scott, R. EL, and D. T. M. Tan. Yersinia enterocolitica enteritis affects rabbit intestinal longitudinal smooth muscle function. Am. J. Physiol. 262 (Gastrointest. Liver Physiol. 25): G278-G284, 1992.-To determine whether Yersinia enterocoLitica (YE) enteritis has an effect on the biomechanical properties of intestinal smooth muscle, New Zealand White rabbits (600-900 g) were divided into an infected group (n = 9) and sham-infected animals fed ad libitum (n = 9), or pair fed with the infected group (n = 9). Animals were inoculated with lOlo organisms of YE in 10 ml NaHC03 (infected group) or 10 ml NaHC03 (sham-infected control and pair-fed groups) at time 0. Daily food intake, weight gain, and YE excretion were noted. Six days later animals were killed and longitudinal smooth muscle strips prepared from proximal (P), medial (M), and distal (D) segments of intestine in each treatment group. Isometric tension was recorded in tissue baths perfused with oxygenated Krebs solution and low6 M tetrodotoxin. Basal and active (the response to 10m5 M carbachol) length-tension curves were generated. Then, with the muscle strips stretched to their optimum length for tension development, the dose response to carbachol and to graded depolarization with KC1 was determined. Infected animals had a significantly reduced food intake and weight gain compared with controls. The development of basal tension with stretch was not significantly different in infected compared with control or pair-fed tissues from the same site. The dose-response curves for tissues exposed to carbachol showed that even though the doses at which the contractile response is half-maximal were not significantly different, the response of infected tissues was significantly impaired compared with that of the pair-fed group, whereas the responsiveness of pair-fed tissues was significantly greater than the response of infected or control tissues at any specific site. Graded depolarization with KC1 reproduced this pattern of response. Thus longitudinal smooth muscle of infected animals (that are also undernourished because of reduced food intake) exhibits a significantly reduced ability to develop tension in response to carbachol compared with pair-fed animals; whereas pair-fed tissues (from animals that are only undernourished) exhibit a significantly increased response to carbachol compared with control or infected animals. The mechanism is unknown, but because the response can be reproduced with KC1 depolarization, it probably reflects receptor-independent changes in smooth muscle function. small intestine; nutrition
contraction;
infection;
inflammation;
under-
YERSINIA ENTEROCOLITICA (YE)is arecognizedcause of bacterial enteritis (13, 15, 16, 24). The spectrum of clinical illness includes both acute gastroenteritis and a chronic relapsing ileocolitis simulating Crohn’s disease of the terminal ileum. Affected patients commonly present with diarrhea and abdominal cramping, symptoms that presumably reflect alterations in the patterns of altered intestinal motor activity and perhaps abnormalities of intestinal smooth muscle contractility. We have previously shown that rabbits infected with a human G278
0193-1857/92
$2.00
Copyright
Research
Group, University
of Calgary,
pathogenic strain of YE develop a clinical illness similar to that in humans, an illness characterized by diarrhea and significantly decreased food intake, weight gain, and survival. Furthermore, infection induces changes in the patterns of intestinal motor activity that are a specific result of infection (they are not related to decreased food intake and weight gain) and are associated with an increased rate of aboral transit (22). The present studies were performed to determine whether YE enteritis and/ or associated undernutrition affects the contractile properties of intestinal smooth muscle. MATERIALS
AND
METHODS
Model New Zealand White rabbits weighing 600-900 g were studied. Animals were initially quarantined for a 3- to &day observation period to ensure the absence of diarrhea. Because YE enteritis in the rabbit is associated with a significantly decreased food intake and weight gain, the experiments were designed to separate the potential contribution of undernutrition to any observed effect of YE enteritis on small intestinal longitudinal smooth muscle function. Experimental
Design
Rabbits were divided into three groups: infected animals (n = 9) and sham-infected control animals fed ad libitum (n = 9) or pair fed with the infected group (n = 9). At time 0 animals received an intragastric inoculation (via nasogastric tube) of lOlo organisms of YE in 10 ml NaHC03 (infected group) or 10 ml NaHC03 (sham-infected control and pair-fed groups). YE strain MCH 700s (serotype 0:3), originally isolated from a patient with diarrhea, was used. The in vivo and in vitro virulence properties of the strain, which possesses the 42-MDa plasmid, have previously been described (14, 20, 21). Daily food intake, weight gain, and YE excretion were noted. The time of study of the infected and sham-infected pair-fed groups was staggered such that the daily food intake of the infected animal was measured and an identical quantity was then fed to its sham-infected pair-fed control. Daily rectal swabs were plated on Salmonella-Shigella media for YE isolation. O’Loughlin et al. (19) documented the clinical, morphological, and biochemical alterations in acute YE infection in rabbits. They showed that mucosal injury is maximal 6 days postinfection, at which time there is extensive abscess formation, villus atrophy, and diffuse brush-border injury, with decreased disaccharidase activity. For this reason animals were killed on day 6 after inoculation and longitudinal muscle strips were obtained for contractility studies from the proximal (P), medial (M), and distal (D) segments of intestine in infected and sham-infected control and pair-fed groups. Contractility
Studies
On day 6, animals anesthetic (halothane)
0 1992 the American
Physiological
were anesthetized with an inhalational and through a midline abdominal inciSociety
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 20, 2019.
YERSINIA
ENTEROCOLITICA
sion the small intestine between ligament of Treitz and ileocecal valve was removed and cleared of feces by gently flushing with normal saline. Lengths (2 cm) of intestine were obtained from the proximal jejunum beginning at the ligament of Treitz, in the midgut, and in the distal ileum 10 cm proximal to the ileocecal valve. The gut segments were opened longitudinally at the mesenteric attachment, pinned out in oxygenated Krebs solution, and a 0.5 x 2 cm longitudinally oriented intestinal segment was obtained from the antimesenteric wall. Intestinal segments, mucosa intact, were anchored to the base of a 20-ml tissue bath, suspended from an isometric force transducer (Harvard Apparatus model 50-7905, O-50 g, Kent, UK) and bathed in Krebs solution. The components of the Krebs were (in mM) 120.3 NaCl, 5.9 CaCl, 2.5 KClz, 1.2 MgC1,, 15.4 NaHC03, 1.2 NaH2P04, and 11.5 glucose. Temperature was maintained at 37OC, and the bath was bubbled continuously with 95% Oa--5% CO,. Mechanical activity detected by the isometric force transducer was amplified by a transducer amplifier (Harvard Apparatus model 50-7970), input to a bioelectric amplifier (HewlettPackard model 8811A), and recorded on an eight-channel chart recorder (Hewlett-Packard model 7858A). Tetrodotoxin ( 10W6 M; Sigma, St. Louis, MO) was present in the buffer throughout. The basal and active length-tension relationships of the tissues were determined as follows. Segments were allowed to stabilize in the tissue bath for 1 h and were then stretched by increments of 1 mm until muscle tension began to increase with any further increase in muscle length. This length was noted as Lie From this initial length (Li), the muscle strip was then progressively stretched to 110, 120, 125, 130, 135, and 140% of Lie At each increment of stretch (100~140% of Li), the muscle was allowed to reach a level of tension that was stable (the basal tension at that level of stretch) and was then maximally contracted using 10m5 M carbachol. The carbachol-stimulated contraction was measured from the stable level of resting tension to the peak tension recorded after carbachol administration. The optimal length of the muscle (L,) was the muscle length at which peak active tension developed. All subsequent experiments evaluating the tissue response to agonists were performed with the segments equilibrated at L,. The dose response to carbachol was established by exposing tissues to 0.5 log M increments of carbachol in a noncumulative fashion with two washes and a 30-min equilibration period between successive doses. The effect of graded KC1 depolarization was established by exposing tissues to 5.9, IO, 20, 30, 40, 60, and 80 mM KC1 in a noncumulative fashion with at least two washes and a 30-min equilibration between successive exposures. The increment in potassium ion was balanced by an equivalent decrement in sodium within the challenge Krebs solution. To minimize variability when data derived from a specific treatment and intestinal site were analyzed, the basal and active tension measurements were normalized to cross-sectional area (i.e., expressed as a stress with units mN/mm2). At the end of each experiment the segment was removed and blotted dry. The mucosa was scraped from the muscularis propria with a glass slide before each strip was weighed. The cross-sectional area (mm”) of each muscle strip was determined using the following equation cross-sectional = length
area (mm2) mass of wet muscle strip (mg) of muscle strip (mm) x density (mg/mm”)
where the density of smooth mg/mm? Histological studies wet muscle strip attributable smooth muscle lavers. and all
(1)
muscle was assumed to be 1.05 determined the thickness of the to the longitudinal and circular calculations of mass have been
IN THE
G279
RABBIT
corrected such that tension is normalized area of longitudinal smooth muscle alone. Statistical
to cross-sectional
Analysis
Results are expressed as means t SE, where the number of rabbits (n) was nine. Incremental stretch of a muscular tissue generates basal tension as an exponential function of the stretch. Therefore the basal tension (T, in newtons) at any length (L, expressed as percentage increment of Li) is given by T = AetmL) + C where A, m, and C are constants. For the purposes of this paper, A = 1, m defines the rate of exponential increase in tension per percentage increment in Lip and C is the y-axis intercept if tension is plotted against length on Cartesian coordinates. Because by definition tension equals zero when the tissue is at its initial resting length (percent increment of Li = 0), then by substitution C must = -I. Exposure of muscle strips to the agonist carbachol (lo-” M) at each increment of stretch generates a length-active tension curve. Tissue strips develop considerable active tension in response to the agonist, even at their initial length. If active tension is plotted against percentage increment of Li on Cartesian coordinates a hyperbolic function with a positive y- intercept (tension at 0 stretch) is generated. The tension (T, in newtons) in response to agonist at any stretched length (L, expressed as percentage increment of Li) is given by T=
T max L EDs0 + L
where Tmax represents the maximal tension developed and EDso represents the dose that produces half-maximal tension, and tensions are all normalized so that T = 0 when the percentage increment of stretch equals zero. Normalized tension in any given condition equals T - mean tension at Lie Once curve parameters were determined, the true Tmax was calculated as the normalized Tmax + mean tension at Lie The muscle length at which peak active tension developed (L,) was determined from the data for each of the nine animals in each treatment group. The difference between the means of the L, for each treatment group was compared using analysis of variance (ANOVA) with a Tukey’s test for post hoc comparison of pairs of means where multiple comparisons were possible. Dose-response experiments generate tension as a hyperbolic function of stretch or agonist dose. Therefore the active tension (T, in newtons) at any concentration of agonist (C) is given by T=
Tmax
C
EDsa + C
In this equation Tmax represents the maximal tension developed and EDs0 represents the dose that produces half-maximal tension. To determine the parameters defining the basal length tension, active length tension, and dose-response parameters of the data, nonlinear least-squares regression analysis was performed using the statistical software package Systat (Systat, Evanston, IL) and a microcomputer (Zenith 386). To define the differences that might exist between curves and to define confidence intervals for comparisons of parameter estimates (m for length-basal tension data, Tmax and EDso for lengthactive tension and dose-response data), we utilized a recently described nonlinear least-squares computer analysis (18). Comparisons were made of stress developed in response to different treatments (control, pair feeding, or infection) at a given site along the intestine and for stress developed at different sites along the intestine (proximal ieiunum. midgut. or distal ileum)
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 20, 2019.
G280
YERSINIA
ENTEROCOLITICA
within a single treatment. When multiple comparisons were made between more than two curves, a Bonferroni adjustment was used so that the adjusted P value was equal to the unadjusted P value divided by the number of multiple comparisons. Comparisons were considered significantly different whenever the adjusted P < 0.05. To test for the presence of a potential interaction between the effects of treatment and location, we modeled (using multivariate least-squares regression analysis and the statistical package Systat) the stress developed in response to the classification variables level of applied stimulus, treatment (control, pair feeding, or infection), the location (distal ileum, midgut, and proximal jejunum), and the interaction term (treatment x location) stress = constant
+ level of stimulus + location
+ treatment + (treatment
X
location)
Separate analyses were made for the data defining the active length-stress response (7 levels of stimulus between 100 and 140% of Li), the dose response to carbachol(9 levels of stimulus between 10v8 and 10m4 M carbachol), and the dose-response to KC1 (7 levels of stimulus for concentrations of KC1 between 0 and 80 mM). A squared correlation coefficient ( r2, which defines the proportion of the variance of the data that is predicted by the model) and the significance level for the contribution of each of the parameters to the predictive capability of the model are reported. Histological
Technique
Muscle strips from control, infected, and pair-fed animals were fixed in formaldehyde solution, embedded in paraffin, sectioned, and stained with hematoxylin and eosin. By use of a Leitz Wetzlar microscope (FRG) at 250 power, the thickness of the longitudinal and circular muscle layers was calculated with an ocular micrometer and standard image-splitting micrometry.
IN
THE
RABBIT
of the muscularis propria attributable to the longitudinal smooth muscle layer in tissues from control, infected, and pair-fed groups at the level of the proximal jejunum, midgut, or distal ileum (Table 2). Basal length-stress response. Figure 1 shows the lengthbasal stress curve for longitudinally oriented tissue strips from the proximal jejunum of control, pair-fed, and infected groups of animals. Similar data were obtained but are not shown for midgut and distal ileal tissue strips from the same animals. In each treatment group, at each site, there was an exponential increase in basal stress with incremental stretch. Estimates of the curve parameter wl for the different treatments at each location were not significantly different from one another. Active length-stress response. The length vs. active stress in response to lo-” M carbachol for longitudinally oriented tissue strips from the midgut of control, pairfed, and infected animals are plotted in Fig. 2. Visual inspection of these length-active stress curves shows that the y-intercept (active tension at Li), slope (rate of increase in stress with incremental stretch), and the T,,, (maximal stress generated at L,) are diminished for infected compared with control or pair-fed tissues and increased for pair-fed compared with control or infected tissues. The muscle length at which peak active stress developed (L,) was similar for all three treatments at all three sites. The T,,, and L, for each treatment at all three sites are summarized in Table 3. There was no significant difference in L, (which averaged 125% of Li) among control, pair-fed, or infected tissues at any site. However, there was a consistent and significant (P < 0.05) decrease in the maximum active stress generated Table 2. Proportion of thickness of muscularis propria attributable to longitudinal smooth muscle layer
RESULTS
Control
Clinical Nine animals were inoculated with YE on day 0. All excreted YE in the stool in follow-up compared with none in the control or pair-fed groups-Infected animals demonstrated a significantly (P < 0.001) decreased food intake and weight gain compared with controls but not pair-fed animals (Table 1).
Proximal jejunum Midgut Distal ileum
_T1
Values
are means
Pair
Fed
Infected
P
0.22t0.07 0.17t0.03
0.13t0.02 0.13t0.01
0.14t0.02 0.12t0.01
0.19
0.17&0.03
0.15t0.02
0.13t0.02
0.57
c(-
0.21
t SE.
lo.0
. CONTROL A PAIR-FED q INFECTED
8.0 --
Contractility
of Intestinal Longitudinal Smooth Muscle
Longitudinal smooth muscle thickness. There was no significant difference in the proportion of the thickness Table 1. Effect of Y. enterocolitica enteritis on food intake and weight gain Control
Protocol Avg daily food intake, g Body wt, g Day 0 before inoculation Day 6 Values are means P < 0.01.
Pair
Fed
Infected
ys
4*o-
1
41.1t10.6*
41.4t10.6*
818*54
835225
831t22
1,060+54
A
m
70.0t3.1
+: SE. * Significantly
EL-+ w cr OJE 6.0-&E \ az
870t30* different
100 LENGTH
789t41* from
control
group,
120 140 (% of Li)
Fig. 1. Length-basal stress curves for strips from proximal jejunum of control, (Protocol 1). Li, initial length.
longitudinally oriented tissue pair-fed, and infected animals
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 20, 2019.
YERSINIA 20.0
. CONTROL A PAIR-FED w INFECTED T 1
ENTEROCOLITICA
IN
THE
25.0
-8
120 140 LENGTH (% of Li)
Control
fed
LO Infected
Proximal 8.0tl.O
14.4k1.7"
Control
Pair
fed
131k1.9
126t3.8
15.1*1.2*
5.7+0.2*"r
Distal 10.3tl.O
16.4tl.l*
7.8+0.5*?
124t3.7
120t3.8
126t3.5
fed
118t5.0
tension generated; L,, optimal Values are means t SE. T,,,, maximal length. * Significantly different from control, P < 0.05. 7 Significantly different from pair fed, P < 0.05.
by infected compared with control tissues and a significant increase in the maximum active stress generated by tissues from the pair-fed compared with control groups at all three sites in the intestine. Dose-response to carbachol. Figure 3 is a plot of the active stress generated in response to carbachol over the range 10m8 to 10B4 M b y 1ongitudinally oriented tissue strips from the midgut of control, pair-fed, and infected animals. All tissues were studied at L, (125% of Li). Similar plots were obtained for the proximal jejunum and distal ileum from the same animals but are not shown. The T,,, and EDs0 for each treatment at all three sites have been summarized in Table 4. At each site there was a consistent and significant (P < 0.01) decrease in the maximum active stress generated in infected compared with control tissues and a significant (P < 0.01) increase in the maximum active stress generated by tissues from the pair-fed compared with control group. There were no significant differences in EDs0 among control, pair-fed, or infected tissues. Dose response to KCI. Figure 4 graphs the active stress generated in response to graded depolarization with KC1 over the range 5.9-80 mM. All tissues were studied at L, (125% of Li). Data for tissue strips from the distal ileum
Control
Control
Pair
Infected
8.1t0.4
11.7rtO.5"
3.8+0.2*j-
13.2k0.7
17.6t0.5*
4.5+0.2*-F
13.4t0.5
15.5tl.O*
Proximal
E&o,
nM
Pair
fed
Infected
jejunum 8t3
8&3
6k2
13t4
8t2
lOk4
3t2
6t4
6t2
Midgut
Distal
123t3.3
ileum 127t4.3
mN/mm2
122t4.3
Midgut 9.9t0.8
max9
Infected
jejunum
3.8&0.2*-f-
-4 ’ M
Fig. 3. Active stress developed in response to carbachol by longitudinally oriented tissue strips from midgut of control, pair-fed, and infected animals. Where no error bars are visible, they are obscured by symbol showing mean.
T
mN/mm2
Pair
-6 LOG[CARBACHOL],
Table 4. Dose response to carbachol
Table 3. Active length-stress response to carbachol (low5 M) max9
CONTROL A PAIR-FED
0.0’
Fig. 2. Length vs. active stress in response to 10m5 M carbachol for longitudinally oriented tissue strips from midgut of control, pair-fed, and infected animals. Where no error bars are visible, they are obscured by symbol showing mean.
T
-0
T
0.0 I. 100
G281
RABBIT
ileum
11.3~0.5""f
Values are means t SE. T,,,,, maximal tension; EDSO, dose producing half-maximal response. * Significantly different from control, P < 0.01. T Significantly different from pair fed, P < 0.01.
1 0.0
. CONTROL A PAIRFED n INFECTED
T
q
60
80
8.0 6.0
0
20
40
[KCI],
mM
Fig. 4. Active stress generated in response to graded depolarization with KC1 by longitudinally oriented tissue strips from distal ileum of control, pair-fed, and infected animals. Where no error bars are visible, they are obscured by symbol showing mean.
of control, pair-fed, and infected animals are shown. Similar plots were obtained for the proximal jejunum and midgut. The T,,, and EDso for each treatment at all three sites are summarized in Table 5. With increasing
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 20, 2019.
G282 Table
YERSINIA
5. Dose response
to KC1 at L,
T Ill&IX, mN/mm2 Control
Pair
0.9kO.3
3.2tl.l"
fed
ENTEROCOLITICA
Infected
Control
Proximal 2.OkO.6
E&0,
nM
Pair
fed
Infected
jejunum 20.4t10.5
36t18.8
10.9k10.5
Midgut l.lIfrO.4
7.4t3.6"
2.2+0.6-j-
2.OkO.5
9.7*1.1*
2.5+0.5-j-
22.4t18.1
Distal
19.8t14.6
20.1t20.9
20*9
26S3
ileum 25tlO
Values are means t SE. L,, optimal length; TmBX, maximal EDso, dose producing half-maximal response. * Significantly from control, P < 0.001. t Significantly different from pair
tension; different fed, P
midgut > proximal jejunum) (23), we next tested the validity of this relationship on the data reported in these experiments. It is apparent that T,,,, in distal ileal tissues was consistently greater than in midgut, which was greater than in proximal jejunum for any treatment (control, pairfeeding, or infection) in response to stimulation with carbachol (Tables 3 and 4) or KC1 (Table 5). Although the EDSOs were all similar, the differences in T,,, were statistically significant (P < 0.05) for comparisons made between the most widely separated segments: distal ileum and proximal jejunum. Thus stress development is dependent on both treatment (control, pair feeding, or infection) and intestinal location (distal ileum, midgut, or proximal jejunum). If the T,,, responses of infected animals are expressed as a percentage of the pair-fed response, there appears to be a diminution of the effect of infection along an aboral gradient (26 vs. 76% of pair-fed controls for the proximal jejunum and distal ileum, respectively, for the data in Table 3; 32 vs. 72% of pair-fed controls for the proximal jejunum and distal ileum, respectively, for the data in Table 4). This trend suggested an interaction of aboral location with the effect of treatment. To test for any significant interaction or a local influence of intestinal location on the effect of treatment we utilized the data summarized in Tables 3-5 and for each set of data modelled the stress developed in response to stimulation against the level of stimulus, treatment, location and the interactive component (treatment X location). The model predicted 61% of the variance in the stress
IN
THE
RABBIT
developed (r2 = 0.61) for the length-active stress (Table 3) and the dose-response to carbachol (Table 4) data but only 43% of the variance in the dose-response to KC1 (Table 5) data. In each case the separate effects of level of stimulus, treatment, location, and the interactive component (treatment x location) were highly significant (P < 0.001). Thus stress development depends on the level of stimulation but is significantly affected by treatment and intestinal location. In addition to the separate effects of treatment and location, these results suggest that local factors that depend on aboral location along the gut modify the effect of treatment. DISCUSSION
Our findings demonstrate that in the rabbit animal model YE enteritis and undernutrition induce separate and contrasting alterations of the contractile properties of longitudinal smooth muscle in the small intestine. Tissues from infected animals, which manifested a significantly decreased food intake and weight gain compared with controls, had a significantly reduced responsiveness to carbachol compared with controls. In contrast, tissues from animals pair fed with the infected group (an undernourished but uninfected group of animals) had a significantly increased responsiveness to carbachol compared with ad libitum-fed controls. The decreased contractile response in infected animals and the increased responsiveness of undernourished animals were not neurally mediated inasmuch as they occurred in the presence of tetrodotoxin. There was no change in the EDSO of carbachol in tissues of infected or pair-fed compared with control animals, and the increased responsiveness of undernourished compared with control tissues and the decreased responsiveness of infected compared with undernourished tissues were reproduced with KC1 depolarization, suggesting that the mechanism involves an alteration of smooth muscle function that is not receptor mediated. The effect of inflammation on the function of intestinal smooth muscle has been studied in the human and several animal models (3). The decreased active contractile response to carbachol observed in response to YE enteritis is similar to that reported in colonic smooth muscle in human (25) and rabbit (2, 28) colitis or to the inhibition of rabbit ileal smooth muscle contractility that occurs in response to direct in vitro exposure to Clostridium dif’icile toxin B (7). However, it is in contrast to the response observed in animal models of intestinal parasitism. In rat animal models the nematodes Nippostrongylus brasiliensis and Trichinella spiralis induce timedependent reversible inflammatory changes in the jejunal mucosa and submucosa and marked hypertrophy of the muscularis propria. Parasitic infection is associated with a significantly increased development of tension by jejunal longitudinal smooth muscle in response to carbachol and 5’-hydroxytryptamine (4, 5, 26). Because the increased tension was expressed per unit cross-sectional area of muscle, the findings were independent of the trophic changes in smooth muscle observed in nematodeinfected rat intestine. The altered contractility of intestinal smooth muscle in the parasitized models was, as in
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 20, 2019.
YERSINIA
ENTEROCOLITICA
our experiments, thought to be related to changes in excitation-contraction coupling at steps distal to, or separate from, receptor occupancy. The pathological injury associated with YE enteritis in the rabbit animal model has been documented (19). By day 6 after infection, gross and microscopic morphological abnormalities are identified. On gross inspection the terminal ileum, cecum, and proximal colon are distended, hyperemic, and fluid filled. There is a patchy distribution of microabscesses surrounding bacteria in the lamina propria of the crypt region, sometimes causing the formation of small ulcers. These inflammatory infiltrates extend into the muscularis mucosa but not into the muscularis propria. Although these morphological changes are evident throughout the intestinal tract, they are more extensive in the ileocecal region. In addition there is a significant reduction in villus height and mucosal enzyme activity and a significant increase in cell proliferation in the ileum, but not in the proximal intestine or midgut, of infected compared with pair-fed and control animals. In contrast to all the findings showing that YE enteritis localizes and has its greatest impact on the mucosa of the terminal ileum, we found that the changes in contractility associated with infection and undernutrition were evident in the proximal jejunum, midgut, and terminal ileum. Although the response was generalized and suggested the involvement of a mediator(s) with a systemic response, we also demonstrated a significant interaction between the effects of treatment and location. The interaction supports the presence of additional local factors that depend on aboral location and modify the ability of treatment to produce an effect. It is very likely that the gross and microscopic localization of YE infection and its associated inflammatory response in the distal small intestine is one such local factor. However, the mechanisms of effect and the identity of the mediators responsible for the generalized and local effects of undernutrition and infection are unknown. The potential role of inflammatory mediators and cytokines needs further investigation. Support for this hypothesis is provided by the results of recent experiments in another animal model. In the nematodeinfected rat, jejunal muscle from worm-free segments excluded from the rest of the gut before infection manifested increased tension generation. In this model daily steroid therapy (betamethasone 3.0 mg/kg SC daily) abolished the increase in myeloperoxidase activity associated with infection and attenuated the increased stress generated in worm-free jejunal muscle from infected animals. These results suggest that alterations in intestinal smooth muscle function in the nematode-infected rat do not require the presence of the parasite in the lumen and may be mediated by systemic mechanisms that are part of the host inflammatory response (17). The finding that undernutrition resulted in a significant increase in intestinal smooth muscle contractility compared with controls was unexpected. During prolonged undernutrition protein synthesis and energy utilization in the viscera are maintained at the expense of peripheral protein degradation (18). However, the effect of undernutrition on striated muscle structure and function varies depending on the muscle’s contractile activity
IN
THE
RABBIT
G283
and oxidative capacity (6, 9, 10, 12). Effects on intestinal smooth muscle have not been well studied. Malnutrition of growing animals causes a significant reduction in total body weight and intestinal wet weight (11). Thus we had projected that undernutrition of the infected and pairfed groups of animals in our study would result in decreased muscle mass and a proportionate reduction in the total active tension developed by the tissue. Not only is this the opposite of what was observed, but the data are expressed as newtons per millimeter squared of intestinal longitudinal smooth muscle, and the differences in contractility that we describe cannot be attributed to differences in the mass of intestinal longitudinal smooth muscle between treatment groups. There is little literature on the effect of undernutrition on intestinal smooth muscle contractility. Interestingly, Weisbrodt et al. (27) examined a related topic, the functional and structural effects of nutrient exclusion in a rat model of jejunoileal bypass. In their study circular muscle strips from the proximal, mid, and distal portions of the bypassed intestine and from the distal portion of the in-continuity intestine generated greater maximal active stress than tissues from unoperated controls. However, increases in stress development were also observed in strips taken from equivalent areas of sham-operated animals compared with unoperated controls. These data suggest that systemic factors, perhaps related to laparotomy, intestinal transection, or the transient nutritional and metabolic alterations induced by surgical bypass of -70% of the pylorus-to-cecum length can alter intestinal smooth muscle contractility. Perhaps the undernutrition of the growing rabbits utilized in our study provoked a similar stress response that mediated the increase in contractility. The increased responsiveness of undernourished compared with control tissues and the decreased responsiveness of infected compared with undernourished tissues occurred in the presence of tetrodotoxin and probably represents an alteration of smooth muscle function. Carbachol induces a receptor-mediated contraction of intestinal smooth muscle, whereas KC1 depolarizes the membrane and induces contraction independently of membrane receptors. The increase in carbachol-stimulated maximum stresses observed in undernourished compared with control tissues was reproduced by KC1 depolarization. Although infection caused a decrease in carbacholstimulated maximum stress to a level significantly less than that seen in tissues from ad libitum-fed control animals, KC1 depolarization caused a reduction that was significantly less than that observed in pair-fed undernourished tissues but not different from ad libitum-fed control tissues. In tissues from infected animals, the portion of the decrease in carbachol-stimulated maximal stress that could not be reproduced by KC1 depolarization may be receptor dependent and perhaps reflect an alteration in receptor number. However, the changes in contractility produced by undernutrition and the major proportion of that induced by infection were reproduced with KC1 depolarization. In addition, the EDSOs of the control, pair-fed, and infected tissue responses to carbachol were not significantly different. This suggests that undernutrition and infection induce alterations in me-
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 20, 2019.
G284
YERSINIA
ENTEROCOLITICA
chanical efficiency of smooth muscle function at a point distal to or separate from receptor occupancy. Mechanisms might potentially include changes in 1) the intrinsic elasticity or resistance of the tissues to stretch due to co1 lagen deposition, edema, and so forth, 2) the smooth muscle cell content of contractile proteins, or 3) calciummediated excitation-contraction coupling. The first alternative seems less likely because the basal lengthtension curves of control pair-fed and infected tissues were similar. Further experiments need to be performed to test the latter two potential mechanisms. In summary, these studies have demonstrated that in a rabbit animal model of YE enteritis intestinal longitudinal smooth muscle shows contrasting alterations in contractility in response to the infectious enteritis (which is associated with undernutrition) and to undernutrition alone. Although the data suggest th .at there is a change in muscle function at a step distal to or separate from receptor occupancy, additional work will need to be done to identify the precise mechanism and the mediators involved. The authors are grateful to Wendy Weinstein for typing the manuscript. The work was supported through funding by the Canadian Foundation for Ileitis and Colitis and by Medical Research Council of Canada Grant MT-10014. Address for reprint requests: R. B. Scott, Dept. of Pediatrics, Faculty of Medicine, University of Calgary, 3330 Hospital Dr. NW, Calgary, Alberta T2N lN4, Canada. Received
5 November
1990; accepted
in final
form
9 September
1991.
REFERENCES 1. Brown, B. P., S. Anura, and D. D. Heistad. Responsiveness of longitudinal and circular muscle layers of the portal vein. Am. J. Physiol. 242 (Gastrointest. Liver Physiol. 5): G498-G503, 1982. 2. Cohen, J. D., H. W. Kao, S. T. Lechago, and W. J. Snape, Jr. Effect of acute experimental colitis on rabbit colonic smooth muscle. Am. J. Physiol. 251 (Gastrointest. Liver Physiol. 14): G538G545, 1986. 3. Collins, S. M., J. D. Huizinga, and D. L. Vermillion. Smooth muscle function in the inflamed bowel. In: Inflammatory Bowel Disease, edited by H. Freeman. Boca Raton, FL: CRC, 1989, p. 5976. 4. Farmer, S. G., J. M. Brown, and D. Pollock. Increased responsiveness of intestinal and vascular smooth muscle to agonists in rats infected with Nippostrongylus brasiliensis. Arch. Int. Pharmacodyn. Ther. 263: 217-227,1983. 5. Fox-Robichaud, A. E., and S. M. Collins. Altered calcium handling properties of jejunal smooth muscle from the nematode infected rat. Gastroenterology 91: 1462-1469, 1986. 6. Gardner, P. F., G. Montanaro, D. R. Simpson, and V. R. Edgerton. Effects of glucocorticoid treatment and food restriction on rat hindlimb muscles. Am. J. Physiol. 238: (Endocrinol. Metab. 1): El24E130, 1980. 7. Gilbert, R. J., C. Pothoulakis, and J. T. LaMont. Effect of purified Clostridium difficile toxins on intestinal smooth muscle. II. Toxin B. Am. J. Physiol. 256 (Gastrointest Liver Physiol. 19): G767-G772,1989.
IN
THE
RABBIT
8. Goldberg, A. L., and T. W. Cheng. Regulation and significance of amino acid metabolism in skeletal muscle. Federation Proc. 37: 2301-2307, 1978. 9. Goldspink, G., and P. S. Ward. Changes in rodent muscle fibre type during post-natal growth, undernutrition and exercise. J. Physiol. Lond. 296: 453-469, 1979. 10. Goodman, M. N., M. A. McElaney, and N. B. Ruderman. Adaptation to prolonged starvation in the rat: curtailment of skeletal muscle proteolysis. Am. J. Physiol. 242 (EndocrinoZ. Metab. 4): E321-E327,1981. 11. Hatch, T. F., E. Lebenthal, D. Branski, and J. Krasner. The effect of early postnatal acquired malnutrition on intestinal growth, disaccharidases and enterokinase. J. Nutr. 109: 1874-1879, 1979. 12. Kelsen, S. G., M. Ference, and S. Kapoor. Effects of undernutrition on the diaphragm. J. AppZ. Physiol. 58: 1354-1359, 1985. 13. Kohl, S. Yersinia enterocolitica infections in children. Pediatr. CZin. N. Am. 26: 433-443, 1979. 14. Lian, C.-J., and C. H. Pai. Inhibition of human neutrophil chemiluminescence by plasmid-mediated outer membrane proteins of Yersinia enterocolitica. Infect. Immun. 49: 145-151, 1985. 15. Maki, M., J. Vesikari, I. Rantala, and P. Gronroos. Yersiniosis in children. Arch. Dis. Child. 55: 861-865, 1980. 16. Marks, M., C. Pai, L. Lafleur, L. Lackman, and 0. Hammerberg. Yersinia enterocolitica gastroenteritis: a prospective study of clinical, bacteriologic and epidemologic features. J. Pediatr. 96: 26-31, 1980. 17. Marzio, L., P. Blennerhassett, S. Chiverton, D. L. Vermillion, J. Langer, and S. M. Collins. Altered smooth muscle function in worm-free gut regions of TrichineZZa-infected rats. Am. J. Physiol. 259 (Gastrointest. Liver Physiol. 22): G306-G313, 1990. 18. Meddings, J. B., R. B. Scott, and G. H. Fick. Analysis and comparison of sigmoidal curves: application to dose-response data. Am. J. Physiol. 257 (Gastrointest. Liver Physiol. 20): G982-G989, 1989. 19. O’Loughlin, E. V., G. Humphreys, I. Dunn, J. Kelly, C. J. Lian, C. Pai, and D. G. Gall. Clinical, morphological and biochemical alterations in acute intestinal yersiniosis. Pediatr. Res. 20: 602-608, 1986. 20. Pai, C. H., and L. DeStephano. Serum resistance associated with virulence in Yersinia enterocolitica. Infect. Immun. 35: 605611, 1982. 21. Pai, C. H., V. Mors, and T. Seemayer. Experimental Yersinia enterocoZitica enteritis in rabbits. Infect. Immun. 28: 238-244, 1980. 22. Scott, R. B., D. G. Gall, and S. C. Diamant. Intestinal motility during acute Yersinia enterocolitica enteritis in rabbits. Can. J. Physiol. PharmacoZ. 67: 553-560, 1988. 23. Scott, R. B., and D. T. M. Tan. Differential smooth muscle contractility along the rabbit small intestine (Abstract). Gastroenterozogy 98: A389, 1990. 24. Simmonds, S. D., M. A. Noble, and H. J. Freeman. Gastrointestinal features of culture-positive Yersinia enterocolitica infection. Gastroenterology 92: 112-117, 1987. 25. Snape, W. J., Jr., R. Williams, E. A. Mayer, and D. Root. Disturbances in colonic contraction in colonic muscle removed from patients with ulcerative colitis (Abstract). Gastroenterology 92: Al648,1987. 26. Vermillion, D. L., and S. M. Collins. Increased responsiveness of jejunal longitudinal muscle in TrichineZZa-infected rats. Am. J. Physiol. 254 (Gastrointest. Liver Physiol. 17): Gl24-G129, 1988. 27. Weisbrodt, N. W., P. R. Nemeth, R. L. Bowers, and W. A. Weems. Functional and structural changes in intestinal smooth muscle after jejunoileal bypass in rats. Gustroenterology 88: 958963,1985. 28. Xie, Y. N., S. N. Reddy, W. Gerthoffer, and W. J. Snape. Effect of mucosal inflammation on colonic smooth muscle contraction (Abstract). GastroenteroZogY 98: A402, 1990.
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 20, 2019.
