William H. Percy

Experimental colitis increases blood-brain barrier permeability in rabbits

Extraintestinal manifestations of inflammatory bowel disease are numerous. This study examined the effects of two models of acute colitis on cerebral blood flow (CBF) and permeability of the blood-brain barrier in rabbits. CBF (measured with radiolabeled microspheres), or the extraction ratio or permeability-surface area (PS) product of the blood-brain barrier to fluorescein and FITC-dextran, was measured 48 h after colitis induction with acetic acid (HAc) or trinitrobenzene sulfonic acid (TNBS). PS product for fluorescein increased (P < 0.05) in TNBS colitis (1.33 x 10(-5) +/- 0.52 x 10(-5) ml/s and 0.48 x 10(-5) +/- 0.13 x 10(-5) ml/s (mean +/- SE) for treated (n = 14) and untreated (n = 10) animals, respectively. PS product for the larger FITC-dextran was not different in TNBS colitis (0.24 x 10(-5) +/- 0.09 x 10(-5) ml/s, n = 7) compared with untreated controls (0.19 x 10(-5) +/- 0.04 x 10(-5) ml/s, n = 8). PS product for fluorescein increased (P < 0.01) in HAc colitis compared with vehicle (2.66 x 10(-5) +/- 1.46 x 10(-5) ml/s and 0.33 x 10(-5) +/- 0.05 x 10(-5) ml/s, respectively; n = 6 in each group). The extraction of fluorescein from the blood to the brain increased by 75% during TNBS colitis when compared with vehicle (P < 0.05). CBF and cerebrovascular resistance did not change from the untreated control after TNBS colitis. Our data suggest that, irrespective of induction method, acute colitis increases the permeability of the blood-brain barrier to small molecules without changing CBF.Extraintestinal manifestations of inflammatory bowel disease are numerous. This study examined the effects of two models of acute colitis on cerebral blood flow (CBF) and permeability of the blood-brain barrier in rabbits. CBF (measured with radiolabeled microspheres), or the extraction ratio or permeability-surface area (PS) product of the blood-brain barrier to fluorescein and FITC-dextran, was measured 48 h after colitis induction with acetic acid (HAc) or trinitrobenzene sulfonic acid (TNBS). PS product for fluorescein increased ( P < 0.05) in TNBS colitis (1.33 × 10-5 ± 0.52 × 10-5 ml/s and 0.48 × 10-5 ± 0.13 × 10-5ml/s (mean ± SE) for treated ( n = 14) and untreated ( n = 10) animals, respectively. PS product for the larger FITC-dextran was not different in TNBS colitis (0.24 × 10-5 ± 0.09 × 10-5ml/s, n = 7) compared with untreated controls (0.19 × 10-5 ± 0.04 × 10-5 ml/s, n = 8). PS product for fluorescein increased ( P < 0.01) in HAc colitis compared with vehicle (2.66 × 10-5 ± 1.46 × 10-5 ml/s and 0.33 × 10-5 ± 0.05 × 10-5ml/s, respectively; n = 6 in each group). The extraction of fluorescein from the blood to the brain increased by 75% during TNBS colitis when compared with vehicle ( P < 0.05). CBF and cerebrovascular resistance did not change from the untreated control after TNBS colitis. Our data suggest that, irrespective of induction method, acute colitis increases the permeability of the blood-brain barrier to small molecules without changing CBF.

Effects of purified Clostridium difficile toxin A on rabbit distal colon

BACKGROUND & AIMS Antibiotic-associated pseudomembranous colitis in humans is caused by proliferation of Clostridium difficile, which elaborates an enterotoxin toxin A that causes epithelial damage and altered motility in rabbit small intestine. The aim of this study was to assess the effects of toxin A on rabbit distal colonic motility and to relate this to histological damage and inflammatory mediator production. METHODS Two hundred micrograms per milliliter of toxin A was placed in a distal colonic loop in anesthetized rabbits, and myoelectric activity was recorded for the following 7 hours. The colon was histologically evaluated and assayed for eicosanoid production. The effects of toxin A on longitudinal and circular muscle were also assessed in vitro. RESULTS Beginning 1 hour after instillation, toxin A caused a significant increase in the number of spike bursts without altering slow wave frequency; this was associated with an increase in mucosal neutrophils and increased production of prostaglandin E2 and leukotrienes B4 and C4/D4/E4. Seven hours after administration of toxin A, mediator levels and myoelectric activity remained increased but significant mucosal damage was now also present. Toxin A did not affect longitudinal or circular muscle in vitro. CONCLUSIONS C. difficile toxin A caused a significant neutrophil infiltration and an increased myoelectric activity before producing mucosal damage. The myoelectric effect may be indirect, resulting from the production of motility-altering arachidonic acid metabolites.

Characteristics of the muscularis mucosae in the acid-secreting region of the rabbit stomach

It has been suggested that muscularis mucosae excitation may augment gastric acid secretion, implying that this muscle should contract to secretagogues or stimulation of its motor innervation. The aim of this study was to characterize in vitro the responses of the muscularis mucosae in the rabbit gastric corpus to substances that modulate acid release and to intrinsic nerve stimulation. Muscularis mucosae from both fundic and antral ends of the corpus had identical mechanical properties, contracted to ACh, ADP, ATP, and histamine but relaxed to vasoactive intestinal polypeptide. Fundic but not antral muscularis mucosae contracted to bombesin and PGE2and PGF2α, whereas adenosine, AMP, CCK, gastrin, secretin, and somatostatin were without effect on any preparation. In both regions electrical field stimulation evoked TTX-sensitive responses consisting of an atropine-resistant contraction followed by an N G-nitro-l-arginine methyl ester- and indomethacin-resistant relaxation. It is concluded from the regional variability in the pharmacological properties of the gastric muscularis mucosae that if its motor activity is linked to acid secretion this would be achieved by a neurally mediated relaxation rather than a paracrine- and/or endocrine-induced alteration in tone.

Effects of antibiotics on epithelial ion transport in the rabbit distal colon in-vitro.

One side‐effect of the therapeutic use of antimicrobial agents is respiratory paralysis as a result of inhibition of skeletal neuromuscular transmission; cholinergic neuro‐effector motor transmission in the gastrointestinal tract is inhibited by the same classes of antimicrobial agent. Study of the effects of several classes of antibiotic compound on intestinal motility has suggested that antibiotic‐induced alterations of intestinal motility may be related to the onset of diarrhoea or the development of antibiotic‐associated colitis. These compounds may, however, also initiate or exacerbate diarrhoea by altering control of epithelial function, a possibility that has not previously been rigorously investigated. This series of experiments investigated the effect of six antibiotics on rabbit distal colonic epithelial ion transport.

Temporal Changes in Colonic Vascular Architecture and Inflammatory Mediator Levels in Animal Models of Colitis

This study investigated the temporal relationship between inflammatory mediator production and colonic vascular architecture in rabbit and rat trinitrobenzene sulfonic acid (TNBS) models of colitis. In both species significant colonic damage and loss of mucosal integrity occurred over time. In the rabbit there was a significant increase in TxB2 levels in the muscularis propria and mucosa at 1, 6, and 48 hr after TNBS, with a significant elevation in PGE2 production in the muscularis propria at 48 hr. However, elevated mediator levels were not associated with measurable changes in vascular architecture. In contrast, significant changes in the numbers and diameters of colonic blood vessels were observed in the rat, in the absence of significant elevations in TxB2 or PGE2 levels. These data suggest that the role of lipid mediators in acute colitis is species-dependent and, although there are corresponding gross and microscopic changes in both models, these occur through disparate mechanisms.

Temporal patterns of colonic blood flow and tissue damage in an animal model of colitis.

This study tested the hypothesis that alteredcolonic blood flow may contribute to tissue damageduring the development of colitis in the rabbit. Thiswas achieved by using radioactive microspheres tomeasure colonic blood flow at various times aftercolitis induction with trinitrobenzene sulfonic acid.Significant colonic damage occurred 6 hours post colitisinduction and persisted throughout the 5 day study. Blood flow to the muscularis propria andmucosa/muscularis mucosae compartments increasedsignificantly from 5 minutes until one hour postinduction. At 6 and 12 hours post induction colonicblood flow returned to control levels before increasingagain from 24 to 96 hours. This second increase in flowwas, however, predominantly in the mucosa/muscularismucosae compartment. Blood flow in the stomach and non-gastrointestinal tissues did not changesignificantly at any time. These data demonstrate thatincreased colonic blood flow may be disrupted in theearly stages of colitis and that this coincides with the onset of significant damage. It isconcluded that maintenance of elevated colonic bloodflow throughout the development of colitis may help toameliorate subsequent tissue injury.

Effects of Free Radicals and Leukocytes on Increases in Blood–Brain Barrier Permeability During Colitis

This study examined the role of leukocytes and free radicals on increases in permeability of the blood–brain barrier (BBB) in rabbits with acute colitis. Trinitrobenzene sulfonic acid (TNBS) was used to induce colitis in male New Zealand White rabbits. The extraction ratio of fluorescein was used as an index of the permeability of the BBB. The extraction ratio for fluorescein was 1.0 × 10−3 ± 2 × 10−4 ml/g and 1.1 × 10−3 ± 3 × 10−4 ml/g (mean ± se) for saline (N = 7) and ascorbic acid-treated (N = 8) rabbits with sham colitis. Conversely, TNBS-induced acute colitis increased the extraction ratio over 70% in saline (N = 8) and ascorbic acid-treated (N = 8) animals. Vinblastine significantly reduced the number of circulating leukocytes, whereas the permeability of the BBB was augmented by 54% in rabbits with TNBS-induced acute colitis (N = 8). Vinblastine had no effect on the permeability of the BBB in rabbits with sham colitis (N = 8). These data suggest that free radicals are not responsible for BBB disruption, and leukocytes may protect the BBB during acute colitis.

Regional Variations in Neurokinin Receptor Subtype Contributions to Muscularis Mucosae and Epithelial Function in Rat Colon

It is known that the muscularis mucosae and mucosa are not pharmacologically homogeneous throughout the rat colon. The aim of this study was to simultaneously characterize all three neurokinin (NK) receptors in the muscularis mucosae and mucosa along the length of the rat colon. Strips of proximal, mid, and distal colonic muscularis mucosae were prepared for isometric recording or sheets of muscle-free mucosa were mounted in Ussing chambers for measurement of short-circuit current. In both muscularis mucosae and mucosa the greatest responses to substance P were found in the proximal region. Use of selective agonists revealed the presence of all three NK receptors in both structures, however, selective antagonism suggests that only NK2 receptors in the muscularis mucosae and NK1 receptors in the mucosa are physiologically relevant. In conclusion, substance P–induced responses in the rat colon are region-specific and not mediated by a single NK receptor subtype common to both structures.

Types of Adrenoreceptors Mediating Responses of Rabbit Gastric Muscularis Mucosae

This study investigated adrenoreceptor-mediated responses of muscularis mucosae from the fundic and antral ends of the rabbit gastric corpus. Norepinephrine-induced fundic muscularis mucosae contractions were enhanced by propranolol and converted to relaxations by phentolamine. Methoxamine, but not clonidine, elicited large fundic contractions. Fundic muscle responded to low isoproterenol concentrations with atenolol- and butoxamine-resistant relaxations, and to high concentrations with atenolol-sensitive contractions. Norepinephrine evoked propranolol-resistant relaxations of antral muscularis mucosae that were enhanced by phentolamine. Methoxamine and clonidine elicited small antral contractions. Lower concentrations of isoproterenol caused atenolol-resistant antral relaxations that were enhanced by butoxamine; higher concentrations produced weak excitation. Fundic and antral relaxations to isoproterenol were abolished by cyanopindolol. Fundic muscularis mucosae possesses excitatory α1-, β1- and inhibitory β3-adrenoreceptors. Excitatory β2- and inhibitory β3-adrenoreceptors predominate in the antral region. The heterogeneous adrenoreceptor-mediated responses of the gastric muscularis mucosae suggest that adrenergic modulation of its motor activity is unlikely to be linked to acid secretion.

In vitro evidence that rabbit distal colonic muscularis mucosae has a Clostridium difficile toxin A receptor

In the rabbit ileum Clostridium difficile toxin A causes inflammation and mucosal damage via a specific glycoprotein receptor that contains α-d-galactose. In rabbit colon toxin A also causes inflammation, and this is associated with increased myoelectric activity and eicosanoid production. The present in vitro study was undertaken to determine if a toxin A receptor on one or more layers of colonic smooth muscle could mediate the motor effects of this agent. Toxin A (20-100 μg/ml) was without effect on longitudinal and circular muscle but had two different effects on the muscularis mucosae. Initial exposure to the toxin caused increased numbers of spontaneous contractions and a small, atropine-, tetrodotoxin-, and indomethacin-resistant increase in resting tone. More importantly, however, 30-min exposure to toxin A resulted in attenuated muscularis mucosae responses to acetylcholine and K+. Both the small excitatory and the larger inhibitory effects of toxin A were abolished by pretreatment with the lectin BS-1, which binds to toxin A receptors, but not by the nonreceptor-binding lectin DBA. These data strongly suggest that toxin A causes significant motor effects on the distal colonic muscularis mucosae via a receptor-mediated mechanism. These mechanical data were supported by the presence of histologically demonstrable toxin A and BS-1 binding sites on the muscularis mucosae but not on either the longitudinal or circular muscle layers, both of which were unresponsive to the toxin. By depressing muscularis mucosae function and, ultimately, mucosal movement as a result of toxin A production, C. difficile may promote its own proliferation, thus further contributing to the development of antibiotic-associated colitis.In the rabbit ileum Clostridium difficile toxin A causes inflammation and mucosal damage via a specific glycoprotein receptor that contains alpha-D-galactose. In rabbit colon toxin A also causes inflammation, and this is associated with increased myoelectric activity and eicosanoid production. The present in vitro study was undertaken to determine if a toxin A receptor on one or more layers of colonic smooth muscle could mediate the motor effects of this agent. Toxin A (20-100 microg/ml) was without effect on longitudinal and circular muscle but had two different effects on the muscularis mucosae. Initial exposure to the toxin caused increased numbers of spontaneous contractions and a small, atropine-, tetrodotoxin-, and indomethacin-resistant increase in resting tone. More importantly, however, 30-min exposure to toxin A resulted in attenuated muscularis mucosae responses to acetylcholine and K+. Both the small excitatory and the larger inhibitory effects of toxin A were abolished by pretreatment with the lectin BS-1, which binds to toxin A receptors, but not by the nonreceptor-binding lectin DBA. These data strongly suggest that toxin A causes significant motor effects on the distal colonic muscularis mucosae via a receptor-mediated mechanism. These mechanical data were supported by the presence of histologically demonstrable toxin A and BS-1 binding sites on the muscularis mucosae but not on either the longitudinal or circular muscle layers, both of which were unresponsive to the toxin. By depressing muscularis mucosae function and, ultimately, mucosal movement as a result of toxin A production, C. difficile may promote its own proliferation, thus further contributing to the development of antibiotic-associated colitis.