Vagal stimulation-induced
gastric damage in rats
LYNN E. HIERLIHY, JOHN L. WALLACE, AND ALASTAIR V. FERGUSON Department of Physiology, Queen’s University, Kingston, Ontario K7L 3N6; and Department of Medical Physiology, University of Calgary, Calgary, Alberta TZN 4N1, Canada
HIERLIHY, LYNN E., JOHN L. WALLACE, AND ALASTAIR V. FERGUSON. Vagal stimulation-induced gastric damage in rats. Am. J. Physiol. 261 (Gastrointest. Liver Physiol. 24): G104GllO, 1991.-The role of the vagus nerve in the development of gastric mucosal damage was examined in urethan-anesthetized male Sprague-Dawley rats. Electrical stimulation was applied to the vagus nerves for a period of 60 min, after which macroscopic gastric damage was scored and samples of the stomach were fixed for later histological assessment. Damage scores were assigned blindly based on a 0 (normal) to 3 (severe) scale. Stimulation of vagal afferents or efferents in isolation did not result in significant damage to the gastric mucosa (P > 0.1). In contrast, stimulation of both intact vagus nerves resulted in significant gastric mucosal damage (mean damage score, 2.0 k 0.33, P < 0.01). A second series of experiments demonstrated this gastric damage to be induced within 30-60 min; extending the stimulation period to 120 min did not worsen the gastric damage scores significantly (P > 0.1). In a third study, stimulation of both intact vagus nerves after paraventricular nucleus (PVN) lesion resulted in damage scores (0.33 t 0.17) that were significantly reduced compared with intact PVN and non-PVN-lesioned animals (P < 0.01). These results indicate that the development of vagal stimulationinduced gastric damage requires the activation of both afferent and efferent vagal components and suggest further that such damage is dependent upon an intact PVN. vagus; paraventricular
nucleus
ATTENTION over the years has been directed toward understanding the relationship between the central nervous system (CNS) and gastrointestinal (GI) ulceration. In particular the common association of stress-related GI mucosal damage with CNS injury has supported the hypothesis that the CNS may play a significant role in controlling the structural integrity of the GI mucosa. It is well known that after intracranial disease, head trauma, or CNS surgery, patients are often afflicted with GI hemorrhage and acute ulceration (16). the intracrani al disease-associat!ed GI damage bei ng otherwrise known as “Cushing’s ulce r” (4,8). Cushing in 1932 (4) suggested such ulceration to be a result of vagal hYrjeractivity and gastric acid 1lypersecretion. T ‘his hyPot hesis has IDeen further suppclrted by the observation of i ncreased g:astric acid secretic3n in patients wit ;h CNS injl Iry, an effc:ct that is thought to be mediated t{hrough effects on vagal nuclei in the medulla (8, 19) initiated by centers in the hypothalamus (7). The establishment of the presence of a family of peptides in both the gut and the brain led to speculation that these peptides may act as CNS neurotransmitters involved with the control of CONSIDERABLE
G104
019% l&37/91
$1.50 Copyright
GI function. Studies examining the effects of intracerebra1 administration of such brain-gut peptides have demonstrated specific effects on gastric functions apparently due to actions of these substances in specific areas of the brain such as the hypothalamus (31). The hypothalamus exerts direct control over a wide variety of autonomic functions and has long been known to play an important role in GI function through its descending projections to’ medullary autonomic centers (28, 30). The paraventricular nucleus (PVN) of the hypothalamus in particular is in an ideal position to modify gastric functions in light of its considerable anatomic projections to preganglionic autonomic neurons in the medullary dorsal motor nucleus of the vagus (DMN) and nucleus tractus solitarius (NTS) as well as projections to the intermediolateral cell column (IML) of the spinal cord as demonstrated by anatomic tracing techniques (27, 28, 30). Electrophysiological studies have verified such connections from the PVN to both the DMN and NTS (1, 13, 18, 26), and in addition such single-unit recordings have shown that the same NTS neurons receive convergent input from both vagal afferent projections and the PVN (22). The PVN has specifically been demonstrated to be involved in a number of gastric functions including the control of acid secretion, motility, and blood flow, presumably through these connections with the vagal nuclei in the medulla. PVN stimulation elicits increases in gastric acid secretion (24), whereas PVN lesions have been found to suppress vagal afferent stimulation-induced increases in gastric acid secretion (23). Electrical stimulation of the PVN also has effects on gastric motility that have been shown to be mediated by the vagus nerves. Activation of oxytocin-secreting neurons projecting from the PVN to DMN decreases gastric motility through noncholinergic synapses on gastric smooth muscle (25), whereas activation of neurons containing thyrotropin-releasing hormone (TRH) from the PVN to DMN is thought to increase gastric motility through cholinergic synapses on gastric smooth muscle (25). Vagal nerve stimulation has also been found to increase gastric mucosal blood flow by causing vasodilatation of gastric arterioles (9, 17, 35). Recent work from this laboratory (5) has demonstrated that electrical stimulation in the PVN, but not the immediately surrounding anatomic regions, results in gastroduodenal lesions in the rat. This gastric damage in response to PVN stimulation was not observed if stimulation was preceded by bilateral vagotomy, suggesting that vagal efferents are important mediators of such gastric damage. It has previously been reported that
N:> 1991 the American
Physiological
Society
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (143.210.133.025) on November 19, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
VAGAL-INDUCED
intermittent vagal efferent stimulation results in gastric ulceration in rats (2); however, vagus nerves were ligated in this earlier study, a procedure that does not ensure complete elimination of vagal afferent innervation. Thus the present investigations were specifically designed to further examine the role of the vagus nerve in the development of gastric mucosal damage. MATERIALS
AND
METHODS
All experiments were performed on male SpragueDawley rats (130-250 g). Food and water were provided ad libitum until the night before experimentation, at which time access to food was withdrawn. Experimental Protocol Animals were anesthetized with urethan (1.4 g/kg ip) and then fitted with an indwelling femoral arterial catheter (PE-50 Intramedic) to monitor blood pressure and heart rate. A midline incision was made to expose the neck region, and a tracheal tube was inserted to facilitate breathing. Cervical vagus nerves were then surgically isolated and laid over bipolar hook electrodes. A Grass SD9 stimulator was used to deliver supramaximal stimulation at a frequency of 5 Hz (5 V, 1-ms pulse duration) for 1 h. Such stimulation parameters were determined to be supramaximal as indicated by blood pressure changes (-30 to -70 mmHg); i.e., any further increase in stimulation intensity did not elicit any larger decreases in systemic blood pressure. Specific experimental treatments were carried out to examine the effect of vagal stimulation in the development of gastric mucosal damage. Afferent and/or efferent vagal stimulation study. Modifications of the above protocol resulted in separate experimental treatments of six groups of animals. Each group consisted of a least five animals. One group of animals had their vagus nerves sectioned proximally and stimulation applied to the distal cut ends resulting in orthodromic activation of vagal efferents and antidromic activation of vagal afferents (efferent). Conversely, a second group of animals had their vagus nerves sectioned distally and stimulation applied to the proximal cut ends resulting in orthodromic activation of vagal afferents and antidromic activation of vagal efferents (afferent). In a third group of animals in which both vagus nerves were sectioned, one nerve was stimulated at its distal cut end while the other vagus nerve of the same animal was stimulated at its proximal cut end to provide simultaneous afferent and efferent vagal stimulation (afferent/ efferent). Vagus nerves were left intact in three last groups: one group underwent intact vagal stimulation of both nerves (intact), one group underwent intact vagal stimulation of one nerve (unilateral), and a final shamstimulation control group did not undergo stimulation, although the vagi were placed over the electrodes (control). Time course study. We next examined the time course of the development of gastric damage in response to stimulation of intact vagi. In these experiments three groups of animals were utilized in which the vagi were
GASTRIC
DAMAGE
Gl05
stimulated for 30, 60, or 120 min and the effects of such stimulation on the integrity of the gastric mucosa were evaluated. PVN lesion study. One week before vagal stimulation animals were anesthetized with pentobarbital sodium (60 mg/kp ip) and placed in a stereotaxic frame. A small burr hole was made in the skull, and a parylene-coated tungsten monopolar electrode (MPI-LFOlG) was advanced toward the region of the PVN according to coordinates of Paxinos and Watson (20). PVN lesions were performed by stereotaxically positioning the stimulating electrode 7.5 mm ventral to the brain surface and passing direct current at 0.5 mA for 30 s. The animals were then allowed to recover for 1 wk, at which time intact vagal stimulation (60 min) experiments were performed as previously described for the vagi intact group of animals. The control group of PVN-lesioned animals did not undergo the vagal stimulation experiment addressing the issue of whether PVN lesion alone resulted in gastric mucosal damage. Macroscopic damage. After vagal stimulation, each animal was overdosed with anesthetic and the abdominal cavity was opened to remove the stomach and proximal duodenum. The stomach was then cut along the greater curvature and examined macroscopically. An observer unaware of the experimental treatment assigned damage scores based on an arbitrary scale from 0 to 3, where 0 indicates minimal damage and a score of 3 indicates extensive gastric damage including regions of hemorrhage. Stomachs were immersed in neutral buffered 10% formaldehyde solution in preparation for later histological assessment of gastric damage. Gastric histology. Samples of dorsal and ventral corpus regions of the stomach (2 x 10 mm) were excised. Stomachs were coded, and histological analysis was performed in random order so that the observer was unaware as to which experimental group each stomach belonged. These tissues were then immersed in fresh formaldehyde solution and processed by routine procedures. CNS histology. The rat was perfused with 0.9% saline followed by 10% formaldehyde solution administered through the left ventricle of the heart. The brain was removed and placed in 10% formaldehyde solution for at least 12 h. Coronal 100~pm sections were cut through the forebrain using a Vibratome, and such sections were stained with cresyl violet for later determination of the anatomic location of lesions. Statistical Analysis All data are expressed as means t SE. Comparisons of damage scores were made using the Mann-Whitney U test for nonparametric data. An associated probability (P value) of ~5% was considered as significant. RESULTS
Effect of Vagal Stimulation Sham experiments in which no stimulation was applied to vagus nerves did not result in any significant damage to the gastric mucosa (mean damage score, 0.39 t 0.22, n = 9). Efferent stimulation of the vagus nerves
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (143.210.133.025) on November 19, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
G106
VAGAL-INDUCED
GASTRIC
resulted in gastric damage scores (0.68 & 0.27, n = 11) that were similar to those for control animals (Fig. l), although in some animals small regions of superficial damage were observed. In contrast, stimulation of the intact vagi caused significant damage scores (2.0 + 0.33, n = 9) compared with control animals’ scores (P < 0.05) (Fig. 1). Obvious damage was indicated by lines of hemorrhagic regions and petechiae on both ventral and dorsal surfaces throughout the corpus and the pyloric region (Fig. 2). In all cases, however, the forestomach was spared from damage. No consistent damage apart from local hyperemia was observed in the duodenum. Such macroscopic gastric damage was confirmed histologically as mucosal erosions where the necrosis did not penetrate the muscularis mucosae (Fig. 3). Further experiments in which the afferent components A
3
B
0
: .: ..,., ~1: ::‘..:.,.;‘,;~~~,., :.
2
.. a
.:..,..,.:..... .,.,4, 1...fP.q . (.,..., ... .. .. c. ./ .1.
‘. ..’
FIG. 1. Gastric damage scores obtained from animals after 60 min of vagal efferent stimulation (A) and vagal intact stimulation (B). Each bar indicates mean + SE of grouped data. Compared with nonstimulated group (Mann-Whitney U test): * P < 0.01.
FIG. 2. Photograph of a rat stomach cut along its greater curvature after 60 min of intact vagal stimulation. Note hemorrhagic regions and petechiae throughout gastric mucosa. Such extensive damage was assigned a score of 3 by an observer unaware of experimental treatment.
DAMAGE
of the vagus nerves were stimulated resulted in significantly reduced damage (0.17 + 0.17, n = 6) compared with the intact vagal stimulations (P < 0.05). Stimulation of either one intact vagus nerve (1.31 + 0.45, n = 8) or one afferent-one efferent within one animal (1.19 f 0.38, n = 8) was found to elicit damage that was intermediate to control scores and bilateral intact vagus nerve stimulation scores; however, such damage scores were not found to be significant (P > 0.05). Within the oneintact vagal stimulation group, no significant difference between the degree of damage as induced by stimulation of either left (1.5 f 0.65, n = 4) or right (1.25 + 0.75, n = 4) vagus nerves was observed. Therefore, in our subsequent experiments examining the time course and involvement of PVN in the development of such gastric damage, vagal stimulation-induced damage refers to gastric damage as a consequence of bilateral stimulation of intact vagi. A summary of the mean gastric damage scores as produced by no stimulation, vagal efferent, vagal afferent, or both intact vagi stimulation is presented in Fig. 4. Effect of Stimulation Period
This second series of studies was carried out to examine the time course of the development of vagally induced gastric mucosal damage. Electrical stimulation of intact vagus nerves for 0.5 h did not result in significant damage to the gastric mucosa compared with control (n = 4). As already described, similar vagal stimulation for a period of 1 h caused significant damage to the gastric mucosa (2.0 + 0.33, P < 0.05, n = 9). Significant mucosal damage in the stomach was again observed after vagal stimulation for a period of 2 h (1.46 f 0.29, P < 0.05, n = 14); however, these damage scores were not significantly different from those observed after only 1 h of stimulation (P > 0.05) (Fig. 5). Effect of PVN Lesion
Animals were assigned to experimental groups after histological verification of lesion sites. As a result animals were categorized as having a non-PVN, unilateral PVN, or bilateral PVN lesion. A schematic representation of the PVN area demonstrating the location of a bilateral PVN lesion is presented in Fig. 6. PVN lesions alone were without effect on the gastric mucosa inasmuch as no significant gastric damage was observed in any animal 1 wk after bilateral PVN lesion (0.10 & 0.10, n = 5). Electrical stimulation of intact vagus nerves in animals with a bilateral PVN lesion also did not result in significant gastric damage scores (0.3 rt 0.2, P > 0.05, n = 6) compared with earlier vagal stimulation-induced gastric damage scores (2.0 +- 0.33, P < 0.05, n = 9) observed in animals with no lesion and an intact PVN (Fig. 7). The damage scores observed after vagal stimulation in animals lesioned in non-PVN areas (1.43 f 0.20, P < 0.05, n = 7) were significantly different from those observed upon vagal stimulation in animals with a PVN lesion (0.3 + 0.2, n = 6) (Fig. 7) and similar to those induced in nonlesioned animals. Vagal stimulationinduced damage in animals with unilateral PVN lesions
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (143.210.133.025) on November 19, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
VAGAL-INDUCED
GASTRIC
DAMAGE
G107
FIG. 3. Photomicrographs of gastric mucosa of rat in which no vagal stimulation was applied (A) and intact vagal stimulation was applied for 60 min (B). Slides were stained with hematoxylin and eosin. A: normal epithelium (~75). R: destruction of epithelium is evident, with cellular debris, mucus, and blood visible in lumen. Epithelial damage involved upper one-third of mucosa and did not penetrate muscularis mucosae (~75).
was intermediate to damage scores observed in animals with a bilateral and non-PVN lesion (0.90 + 0.33, n = 5). DISCUSSION
The present studies have shown that electrical stimulation in our vagal efferent or afferent groups results in no significant damage to the gastric mucosa of the rat. In contrast, simultaneous activation of all vagal components results in the formation of obvious gastric mucosal erosions. Such observations emphasize that both afferent and efferent vagal nerve components participate in the development of gastric damage. Destruction of the epithelium was evident upon microscopic examination of gastric tissues, with cellular debris, mucus, and blood visible in the lumen. In addition this damage was demonstrated to be confined to the upper one-third of the mucosa. The gastric damage observed in these experiments was histologically similar to acute mucosal damage observed after administration of nonsteroidal anti-inflammatory drugs or after restraint stress (33). This mucosal damage was rapidly induced (within 30-60 min), although extending the stimulation period to 120 min did not significantly increase the gastric damage scores. Observed damage did not appear to be related to changes in blood pressure induced by stimulation as a transient
decrease in blood pressure was observed upon vagal stimulation in all experimental groups, although as already described gastric damage was only elicited in the vagi intact animals. The lack of gastric injury after vagal efferent stimulation was surprising in view of the demonstration that bilateral vagotomy will abolish PVN induced gastroduodenal damage (5). Numerous neuroanatomic (27, 28,30) and electrophysiological studies (1, 13, 18,26) have demonstrated PVN efferent neural connections to the DMN of the medulla, which support a role for vagal efferents in mediating PVN-induced damage. Of the descending projections from the PVN to the dorsal medulla, however, the most intense projections have been demonstrated to the posterior part of the NTS, the area of incoming vagal information (30). In addition electrophysiological studies have reported that very few DMN neurons are activated by PVN stimulation (13). Neurons in the NTS have been shown to receive convergent input from both the PVN and the vagus nerve, providing the PVN with the ability to modify visceral function by modulation of vagal afferent information (22). The present studies clearly indicate that a vagal afferent component is required in the formation of gastric damage, since damage was only observed after concurrent activation of both afferent and efferent vagal components. In view of the observation that bilateral lesion of PVN prevented
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (143.210.133.025) on November 19, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
Gl08
VAGAL-INDUCED
GASTRIC
*
EFFERENT FIG.
efferent grouped
4. Gastric damage scores obtained (efferent), vagal afferent (afferent), data. Compared with nonstimulated
DAMAGE
p < 0.01
AFERENT
from animals after a 60-min period of no stimulation or vagal intact (intact) stimulation. Each bar indicates group (Mann-Whitney U test): * P < 0.01.
INTACT
(control), vagal mean t SE for
Bregma -1.8mm
p < **
CONTROL
p
,PVN
0.01
< 0.05
30
-MN
60
MN
120
m
5. Gastric damage scores as a result of stimulation applied to intact, vagus nerves for periods of 0, 30, 60, or 120 min. Each bar indicates mean f: SE for grouped data. Compared with nonstimulated group (Mann-Whitney U test): ** P < 0.01; * P < 0.05.
FIG. 6. Schematic representation pothalamus in region of paraventricular typical anatomic location of bilateral shaded area were considered as PVN ventricle; OT, optic tract.
of a coronal section through hynucleus (PVN) to demonstrate lesion sites. Lesions placed with lesions. Scale bar, 0.5 mm. V, 3rd
FIG.
the formation of gastric damage after electrical stimulation of the intact vagi, it appears likely that this vagal afferent component involves a connection from the NTS to the PVN. The observation that similar stimulation of the vagus nerves was still able to produce significant damage in animals lesioned in non-PVN regions further supports a specific role for the PVN in the formation of such gastric erosions. The presence of gastric acid is required for the formation of gastric damage in most experimental models, although even high concentrations of acid alone are not sufficient to induce damage in the normal rat stomach (32). This suggests that, upon intact vagal stimulation,
other factors predisposing the stomach to damage are present or protective factors are absent. Stimulation of the vagus nerve has been shown to induce changes in gastric motility that could themselves affect the ability of the gastric mucosa to defend itself. Both PVN and DMN stimulation have been demonstrated to cause an increase in phasic contractions with an underlying decrease in gastric “tone,” these effects being mediated by the vagus nerve (25,35). Such mechanical movements of the stomach wall could cause abrasive rubbing of the gastric mucosa predisposing it to injury as suggested by Mersereau and Hinchey (15). Other studies have implicated gastric motility patterns in gastric damage involving vagal stimulation (2, %I), and vagal afferent stimulation has been shown to cause an increase in peripheral levels of oxytocin (29), which could induce changes in gastric emptying. Vagal nerve stimulation has also been
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (143.210.133.025) on November 19, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
VAGAL-INDIJCED
*
INTACT
p
BLAT
gastric damage in rats
LYNN E. HIERLIHY, JOHN L. WALLACE, AND ALASTAIR V. FERGUSON Department of Physiology, Queen’s University, Kingston, Ontario K7L 3N6; and Department of Medical Physiology, University of Calgary, Calgary, Alberta TZN 4N1, Canada
HIERLIHY, LYNN E., JOHN L. WALLACE, AND ALASTAIR V. FERGUSON. Vagal stimulation-induced gastric damage in rats. Am. J. Physiol. 261 (Gastrointest. Liver Physiol. 24): G104GllO, 1991.-The role of the vagus nerve in the development of gastric mucosal damage was examined in urethan-anesthetized male Sprague-Dawley rats. Electrical stimulation was applied to the vagus nerves for a period of 60 min, after which macroscopic gastric damage was scored and samples of the stomach were fixed for later histological assessment. Damage scores were assigned blindly based on a 0 (normal) to 3 (severe) scale. Stimulation of vagal afferents or efferents in isolation did not result in significant damage to the gastric mucosa (P > 0.1). In contrast, stimulation of both intact vagus nerves resulted in significant gastric mucosal damage (mean damage score, 2.0 k 0.33, P < 0.01). A second series of experiments demonstrated this gastric damage to be induced within 30-60 min; extending the stimulation period to 120 min did not worsen the gastric damage scores significantly (P > 0.1). In a third study, stimulation of both intact vagus nerves after paraventricular nucleus (PVN) lesion resulted in damage scores (0.33 t 0.17) that were significantly reduced compared with intact PVN and non-PVN-lesioned animals (P < 0.01). These results indicate that the development of vagal stimulationinduced gastric damage requires the activation of both afferent and efferent vagal components and suggest further that such damage is dependent upon an intact PVN. vagus; paraventricular
nucleus
ATTENTION over the years has been directed toward understanding the relationship between the central nervous system (CNS) and gastrointestinal (GI) ulceration. In particular the common association of stress-related GI mucosal damage with CNS injury has supported the hypothesis that the CNS may play a significant role in controlling the structural integrity of the GI mucosa. It is well known that after intracranial disease, head trauma, or CNS surgery, patients are often afflicted with GI hemorrhage and acute ulceration (16). the intracrani al disease-associat!ed GI damage bei ng otherwrise known as “Cushing’s ulce r” (4,8). Cushing in 1932 (4) suggested such ulceration to be a result of vagal hYrjeractivity and gastric acid 1lypersecretion. T ‘his hyPot hesis has IDeen further suppclrted by the observation of i ncreased g:astric acid secretic3n in patients wit ;h CNS injl Iry, an effc:ct that is thought to be mediated t{hrough effects on vagal nuclei in the medulla (8, 19) initiated by centers in the hypothalamus (7). The establishment of the presence of a family of peptides in both the gut and the brain led to speculation that these peptides may act as CNS neurotransmitters involved with the control of CONSIDERABLE
G104
019% l&37/91
$1.50 Copyright
GI function. Studies examining the effects of intracerebra1 administration of such brain-gut peptides have demonstrated specific effects on gastric functions apparently due to actions of these substances in specific areas of the brain such as the hypothalamus (31). The hypothalamus exerts direct control over a wide variety of autonomic functions and has long been known to play an important role in GI function through its descending projections to’ medullary autonomic centers (28, 30). The paraventricular nucleus (PVN) of the hypothalamus in particular is in an ideal position to modify gastric functions in light of its considerable anatomic projections to preganglionic autonomic neurons in the medullary dorsal motor nucleus of the vagus (DMN) and nucleus tractus solitarius (NTS) as well as projections to the intermediolateral cell column (IML) of the spinal cord as demonstrated by anatomic tracing techniques (27, 28, 30). Electrophysiological studies have verified such connections from the PVN to both the DMN and NTS (1, 13, 18, 26), and in addition such single-unit recordings have shown that the same NTS neurons receive convergent input from both vagal afferent projections and the PVN (22). The PVN has specifically been demonstrated to be involved in a number of gastric functions including the control of acid secretion, motility, and blood flow, presumably through these connections with the vagal nuclei in the medulla. PVN stimulation elicits increases in gastric acid secretion (24), whereas PVN lesions have been found to suppress vagal afferent stimulation-induced increases in gastric acid secretion (23). Electrical stimulation of the PVN also has effects on gastric motility that have been shown to be mediated by the vagus nerves. Activation of oxytocin-secreting neurons projecting from the PVN to DMN decreases gastric motility through noncholinergic synapses on gastric smooth muscle (25), whereas activation of neurons containing thyrotropin-releasing hormone (TRH) from the PVN to DMN is thought to increase gastric motility through cholinergic synapses on gastric smooth muscle (25). Vagal nerve stimulation has also been found to increase gastric mucosal blood flow by causing vasodilatation of gastric arterioles (9, 17, 35). Recent work from this laboratory (5) has demonstrated that electrical stimulation in the PVN, but not the immediately surrounding anatomic regions, results in gastroduodenal lesions in the rat. This gastric damage in response to PVN stimulation was not observed if stimulation was preceded by bilateral vagotomy, suggesting that vagal efferents are important mediators of such gastric damage. It has previously been reported that
N:> 1991 the American
Physiological
Society
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (143.210.133.025) on November 19, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
VAGAL-INDUCED
intermittent vagal efferent stimulation results in gastric ulceration in rats (2); however, vagus nerves were ligated in this earlier study, a procedure that does not ensure complete elimination of vagal afferent innervation. Thus the present investigations were specifically designed to further examine the role of the vagus nerve in the development of gastric mucosal damage. MATERIALS
AND
METHODS
All experiments were performed on male SpragueDawley rats (130-250 g). Food and water were provided ad libitum until the night before experimentation, at which time access to food was withdrawn. Experimental Protocol Animals were anesthetized with urethan (1.4 g/kg ip) and then fitted with an indwelling femoral arterial catheter (PE-50 Intramedic) to monitor blood pressure and heart rate. A midline incision was made to expose the neck region, and a tracheal tube was inserted to facilitate breathing. Cervical vagus nerves were then surgically isolated and laid over bipolar hook electrodes. A Grass SD9 stimulator was used to deliver supramaximal stimulation at a frequency of 5 Hz (5 V, 1-ms pulse duration) for 1 h. Such stimulation parameters were determined to be supramaximal as indicated by blood pressure changes (-30 to -70 mmHg); i.e., any further increase in stimulation intensity did not elicit any larger decreases in systemic blood pressure. Specific experimental treatments were carried out to examine the effect of vagal stimulation in the development of gastric mucosal damage. Afferent and/or efferent vagal stimulation study. Modifications of the above protocol resulted in separate experimental treatments of six groups of animals. Each group consisted of a least five animals. One group of animals had their vagus nerves sectioned proximally and stimulation applied to the distal cut ends resulting in orthodromic activation of vagal efferents and antidromic activation of vagal afferents (efferent). Conversely, a second group of animals had their vagus nerves sectioned distally and stimulation applied to the proximal cut ends resulting in orthodromic activation of vagal afferents and antidromic activation of vagal efferents (afferent). In a third group of animals in which both vagus nerves were sectioned, one nerve was stimulated at its distal cut end while the other vagus nerve of the same animal was stimulated at its proximal cut end to provide simultaneous afferent and efferent vagal stimulation (afferent/ efferent). Vagus nerves were left intact in three last groups: one group underwent intact vagal stimulation of both nerves (intact), one group underwent intact vagal stimulation of one nerve (unilateral), and a final shamstimulation control group did not undergo stimulation, although the vagi were placed over the electrodes (control). Time course study. We next examined the time course of the development of gastric damage in response to stimulation of intact vagi. In these experiments three groups of animals were utilized in which the vagi were
GASTRIC
DAMAGE
Gl05
stimulated for 30, 60, or 120 min and the effects of such stimulation on the integrity of the gastric mucosa were evaluated. PVN lesion study. One week before vagal stimulation animals were anesthetized with pentobarbital sodium (60 mg/kp ip) and placed in a stereotaxic frame. A small burr hole was made in the skull, and a parylene-coated tungsten monopolar electrode (MPI-LFOlG) was advanced toward the region of the PVN according to coordinates of Paxinos and Watson (20). PVN lesions were performed by stereotaxically positioning the stimulating electrode 7.5 mm ventral to the brain surface and passing direct current at 0.5 mA for 30 s. The animals were then allowed to recover for 1 wk, at which time intact vagal stimulation (60 min) experiments were performed as previously described for the vagi intact group of animals. The control group of PVN-lesioned animals did not undergo the vagal stimulation experiment addressing the issue of whether PVN lesion alone resulted in gastric mucosal damage. Macroscopic damage. After vagal stimulation, each animal was overdosed with anesthetic and the abdominal cavity was opened to remove the stomach and proximal duodenum. The stomach was then cut along the greater curvature and examined macroscopically. An observer unaware of the experimental treatment assigned damage scores based on an arbitrary scale from 0 to 3, where 0 indicates minimal damage and a score of 3 indicates extensive gastric damage including regions of hemorrhage. Stomachs were immersed in neutral buffered 10% formaldehyde solution in preparation for later histological assessment of gastric damage. Gastric histology. Samples of dorsal and ventral corpus regions of the stomach (2 x 10 mm) were excised. Stomachs were coded, and histological analysis was performed in random order so that the observer was unaware as to which experimental group each stomach belonged. These tissues were then immersed in fresh formaldehyde solution and processed by routine procedures. CNS histology. The rat was perfused with 0.9% saline followed by 10% formaldehyde solution administered through the left ventricle of the heart. The brain was removed and placed in 10% formaldehyde solution for at least 12 h. Coronal 100~pm sections were cut through the forebrain using a Vibratome, and such sections were stained with cresyl violet for later determination of the anatomic location of lesions. Statistical Analysis All data are expressed as means t SE. Comparisons of damage scores were made using the Mann-Whitney U test for nonparametric data. An associated probability (P value) of ~5% was considered as significant. RESULTS
Effect of Vagal Stimulation Sham experiments in which no stimulation was applied to vagus nerves did not result in any significant damage to the gastric mucosa (mean damage score, 0.39 t 0.22, n = 9). Efferent stimulation of the vagus nerves
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (143.210.133.025) on November 19, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
G106
VAGAL-INDUCED
GASTRIC
resulted in gastric damage scores (0.68 & 0.27, n = 11) that were similar to those for control animals (Fig. l), although in some animals small regions of superficial damage were observed. In contrast, stimulation of the intact vagi caused significant damage scores (2.0 + 0.33, n = 9) compared with control animals’ scores (P < 0.05) (Fig. 1). Obvious damage was indicated by lines of hemorrhagic regions and petechiae on both ventral and dorsal surfaces throughout the corpus and the pyloric region (Fig. 2). In all cases, however, the forestomach was spared from damage. No consistent damage apart from local hyperemia was observed in the duodenum. Such macroscopic gastric damage was confirmed histologically as mucosal erosions where the necrosis did not penetrate the muscularis mucosae (Fig. 3). Further experiments in which the afferent components A
3
B
0
: .: ..,., ~1: ::‘..:.,.;‘,;~~~,., :.
2
.. a
.:..,..,.:..... .,.,4, 1...fP.q . (.,..., ... .. .. c. ./ .1.
‘. ..’
FIG. 1. Gastric damage scores obtained from animals after 60 min of vagal efferent stimulation (A) and vagal intact stimulation (B). Each bar indicates mean + SE of grouped data. Compared with nonstimulated group (Mann-Whitney U test): * P < 0.01.
FIG. 2. Photograph of a rat stomach cut along its greater curvature after 60 min of intact vagal stimulation. Note hemorrhagic regions and petechiae throughout gastric mucosa. Such extensive damage was assigned a score of 3 by an observer unaware of experimental treatment.
DAMAGE
of the vagus nerves were stimulated resulted in significantly reduced damage (0.17 + 0.17, n = 6) compared with the intact vagal stimulations (P < 0.05). Stimulation of either one intact vagus nerve (1.31 + 0.45, n = 8) or one afferent-one efferent within one animal (1.19 f 0.38, n = 8) was found to elicit damage that was intermediate to control scores and bilateral intact vagus nerve stimulation scores; however, such damage scores were not found to be significant (P > 0.05). Within the oneintact vagal stimulation group, no significant difference between the degree of damage as induced by stimulation of either left (1.5 f 0.65, n = 4) or right (1.25 + 0.75, n = 4) vagus nerves was observed. Therefore, in our subsequent experiments examining the time course and involvement of PVN in the development of such gastric damage, vagal stimulation-induced damage refers to gastric damage as a consequence of bilateral stimulation of intact vagi. A summary of the mean gastric damage scores as produced by no stimulation, vagal efferent, vagal afferent, or both intact vagi stimulation is presented in Fig. 4. Effect of Stimulation Period
This second series of studies was carried out to examine the time course of the development of vagally induced gastric mucosal damage. Electrical stimulation of intact vagus nerves for 0.5 h did not result in significant damage to the gastric mucosa compared with control (n = 4). As already described, similar vagal stimulation for a period of 1 h caused significant damage to the gastric mucosa (2.0 + 0.33, P < 0.05, n = 9). Significant mucosal damage in the stomach was again observed after vagal stimulation for a period of 2 h (1.46 f 0.29, P < 0.05, n = 14); however, these damage scores were not significantly different from those observed after only 1 h of stimulation (P > 0.05) (Fig. 5). Effect of PVN Lesion
Animals were assigned to experimental groups after histological verification of lesion sites. As a result animals were categorized as having a non-PVN, unilateral PVN, or bilateral PVN lesion. A schematic representation of the PVN area demonstrating the location of a bilateral PVN lesion is presented in Fig. 6. PVN lesions alone were without effect on the gastric mucosa inasmuch as no significant gastric damage was observed in any animal 1 wk after bilateral PVN lesion (0.10 & 0.10, n = 5). Electrical stimulation of intact vagus nerves in animals with a bilateral PVN lesion also did not result in significant gastric damage scores (0.3 rt 0.2, P > 0.05, n = 6) compared with earlier vagal stimulation-induced gastric damage scores (2.0 +- 0.33, P < 0.05, n = 9) observed in animals with no lesion and an intact PVN (Fig. 7). The damage scores observed after vagal stimulation in animals lesioned in non-PVN areas (1.43 f 0.20, P < 0.05, n = 7) were significantly different from those observed upon vagal stimulation in animals with a PVN lesion (0.3 + 0.2, n = 6) (Fig. 7) and similar to those induced in nonlesioned animals. Vagal stimulationinduced damage in animals with unilateral PVN lesions
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (143.210.133.025) on November 19, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
VAGAL-INDUCED
GASTRIC
DAMAGE
G107
FIG. 3. Photomicrographs of gastric mucosa of rat in which no vagal stimulation was applied (A) and intact vagal stimulation was applied for 60 min (B). Slides were stained with hematoxylin and eosin. A: normal epithelium (~75). R: destruction of epithelium is evident, with cellular debris, mucus, and blood visible in lumen. Epithelial damage involved upper one-third of mucosa and did not penetrate muscularis mucosae (~75).
was intermediate to damage scores observed in animals with a bilateral and non-PVN lesion (0.90 + 0.33, n = 5). DISCUSSION
The present studies have shown that electrical stimulation in our vagal efferent or afferent groups results in no significant damage to the gastric mucosa of the rat. In contrast, simultaneous activation of all vagal components results in the formation of obvious gastric mucosal erosions. Such observations emphasize that both afferent and efferent vagal nerve components participate in the development of gastric damage. Destruction of the epithelium was evident upon microscopic examination of gastric tissues, with cellular debris, mucus, and blood visible in the lumen. In addition this damage was demonstrated to be confined to the upper one-third of the mucosa. The gastric damage observed in these experiments was histologically similar to acute mucosal damage observed after administration of nonsteroidal anti-inflammatory drugs or after restraint stress (33). This mucosal damage was rapidly induced (within 30-60 min), although extending the stimulation period to 120 min did not significantly increase the gastric damage scores. Observed damage did not appear to be related to changes in blood pressure induced by stimulation as a transient
decrease in blood pressure was observed upon vagal stimulation in all experimental groups, although as already described gastric damage was only elicited in the vagi intact animals. The lack of gastric injury after vagal efferent stimulation was surprising in view of the demonstration that bilateral vagotomy will abolish PVN induced gastroduodenal damage (5). Numerous neuroanatomic (27, 28,30) and electrophysiological studies (1, 13, 18,26) have demonstrated PVN efferent neural connections to the DMN of the medulla, which support a role for vagal efferents in mediating PVN-induced damage. Of the descending projections from the PVN to the dorsal medulla, however, the most intense projections have been demonstrated to the posterior part of the NTS, the area of incoming vagal information (30). In addition electrophysiological studies have reported that very few DMN neurons are activated by PVN stimulation (13). Neurons in the NTS have been shown to receive convergent input from both the PVN and the vagus nerve, providing the PVN with the ability to modify visceral function by modulation of vagal afferent information (22). The present studies clearly indicate that a vagal afferent component is required in the formation of gastric damage, since damage was only observed after concurrent activation of both afferent and efferent vagal components. In view of the observation that bilateral lesion of PVN prevented
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (143.210.133.025) on November 19, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
Gl08
VAGAL-INDUCED
GASTRIC
*
EFFERENT FIG.
efferent grouped
4. Gastric damage scores obtained (efferent), vagal afferent (afferent), data. Compared with nonstimulated
DAMAGE
p < 0.01
AFERENT
from animals after a 60-min period of no stimulation or vagal intact (intact) stimulation. Each bar indicates group (Mann-Whitney U test): * P < 0.01.
INTACT
(control), vagal mean t SE for
Bregma -1.8mm
p < **
CONTROL
p
,PVN
0.01
< 0.05
30
-MN
60
MN
120
m
5. Gastric damage scores as a result of stimulation applied to intact, vagus nerves for periods of 0, 30, 60, or 120 min. Each bar indicates mean f: SE for grouped data. Compared with nonstimulated group (Mann-Whitney U test): ** P < 0.01; * P < 0.05.
FIG. 6. Schematic representation pothalamus in region of paraventricular typical anatomic location of bilateral shaded area were considered as PVN ventricle; OT, optic tract.
of a coronal section through hynucleus (PVN) to demonstrate lesion sites. Lesions placed with lesions. Scale bar, 0.5 mm. V, 3rd
FIG.
the formation of gastric damage after electrical stimulation of the intact vagi, it appears likely that this vagal afferent component involves a connection from the NTS to the PVN. The observation that similar stimulation of the vagus nerves was still able to produce significant damage in animals lesioned in non-PVN regions further supports a specific role for the PVN in the formation of such gastric erosions. The presence of gastric acid is required for the formation of gastric damage in most experimental models, although even high concentrations of acid alone are not sufficient to induce damage in the normal rat stomach (32). This suggests that, upon intact vagal stimulation,
other factors predisposing the stomach to damage are present or protective factors are absent. Stimulation of the vagus nerve has been shown to induce changes in gastric motility that could themselves affect the ability of the gastric mucosa to defend itself. Both PVN and DMN stimulation have been demonstrated to cause an increase in phasic contractions with an underlying decrease in gastric “tone,” these effects being mediated by the vagus nerve (25,35). Such mechanical movements of the stomach wall could cause abrasive rubbing of the gastric mucosa predisposing it to injury as suggested by Mersereau and Hinchey (15). Other studies have implicated gastric motility patterns in gastric damage involving vagal stimulation (2, %I), and vagal afferent stimulation has been shown to cause an increase in peripheral levels of oxytocin (29), which could induce changes in gastric emptying. Vagal nerve stimulation has also been
Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (143.210.133.025) on November 19, 2018. Copyright © 1991 the American Physiological Society. All rights reserved.
VAGAL-INDIJCED
*
INTACT
p
BLAT
