Use of Animal Models in Peptic Ulcer Disease

Transcription

1 Use of Animal Models in Peptic Ulcer Disease HERBERT WEINER, MD, DR MED (HON) Considerable progress has been made in the understanding of the formation of gastric erosions by the use of animals. The role of gastric acid secretion in their pathogenesis has been clarified. Gastric erosions are associated with the presence of acid in the stomach and slow gastric contractions. With several different experimental procedures, the animal's body temperature falls; preventing the fall averts erosions. A fall in body temperature or exposure to cold are associated with the secretion of thyrotropin-releasing hormone (TRH), and both increased and decreased concentration of corticotropin-releasing factor (CRH) in discrete regions of rat brains. Thyrotropin-releasing hormone when injected into specific sites in the brain produces gastric erosions and increases acid secretion and slow contractions, whereas CRH has the opposite effects. One of the major sites of interaction of the two peptides is in the dorsal motor complex of the vagus nerve. Thyrotropin-releasing hormone increases serotonin (5-HT) secretion into the stomach. Serotonin counterregulates acid secretion and slow contractions. Many other peptides injected into discrete brain sites stimulate or inhibit gastric acid secretion. Key words: peptic gastric ulceration, body temperature, role of peptides. INTRODUCTION Pavlov pioneered the use of dogs to describe the mediating role of the vagus nerve in gastric acid secretion, the anticipatory ("psychic") phase of secretion, and its conditioning. And, Selye produced gastric erosions in rats by a very large number of injurious and noninjurious (such as restraint) agents. Ever since then, animals have been used to explore two related themes: a) the central (brain) regulation of gastroduodenal function and b) the pathogenesis of gastroduodenal erosions. These two topics can, of course, be related when experimental procedures have been devised to alter the appropriate regulatory brain mechanisms either by making electrolytic lesions in, by stimulating discrete areas of the brain, or injecting areas with active agonistic or antagonistic substances. A long tradition also exists of producing erosions, not by a direct approach to the brain, but by a large number of experimental procedures (conceived of as "stressful"), many of which, however, do not occur in the natural life of the animal. Reliable animal models of disease have revised our understanding of the multifactorial nature and From the From the Department of Psychiatry and Biobehavioral Sciences, University of California Los Angeles, School of Medicine, Los Angeles, California. Address reprint requests to: Herbert Weiner, MD, Dr Med (Hon), University of California Los Angeles School of Medicine, Department of Psychiatry and Biobehavioral Sciences, 760 Westwood Plaza, Room C8-225/NPI, Los Angeles, CA Received for publication August 1, 1995; revision received December 4, complexity of human gastroduodenal erosions (1), and have lead to a variety of improvements in treatment to reduce gastric acid secretion, or enhance gastric mucosal resistance. However, two caveats must be made: a) the validity of erosion formation produced in rats for our understanding of the pathogenesis of human gastroduodenal ulcers has continued to be questioned; and b) the discovery of the key (ie, local) role of Helicobacter pylori infection in human ulcer disease, has seemingly trivialized the part played by stressful experience mediated by the brain in its pathogenesis. (I shall return to this topic later in this report.) Experimental Methods to Produce Gastric Erosions Restraining rats by wrapping them in wire mesh, placing them in a tight-fitting tube, or tying them to a board has been a traditional method of producing gastric erosions (2 9). Usually, animals were not fed 24 hours before or during the procedure. Initially, the validity of the model for human disease was questioned, because the erosions produced by some forms of restraint, especially when associated with food deprivation, tended to be in the rumen of the stomach (10), whereas those in the glandular portion did not usually penetrate to the muscularis mucosa (11). However, this limitation was overcome by the careful control over the experimental design (eg, production of complete immobilization, reduction of the time of food deprivation before restraint), to produce true ulceration (12). Therefore, the aim of inciting gastric erosions that 524 Psychosomatic Medicine 58: (1996) /96/ $03.00/0 Copyright 1996 by the American Psychosomatic Socie

2 PEPTIC ULCER ULCERATION/ANIMAL MODELS were the analogue of the human lesion was attained by the total immobilization of the animal. The restraint technique (combined or not with exposing the animals to cold or water), or electrical shocks applied to the tail or feet, and other such painful procedures, are experimental artifacts. The animal could not predict, prevent, avoid, or escape intermittent or continuous electric shocks, regularly or randomly administered. But the effects on an animal that could predict the onset of signaled shocks differed from those that could not. The extent of the gastric erosions produced was six times greater in rats that received shocks to the tail that were not regularly paired with a warning tone, than in those exposed to shocks that were signaled. When a rat could learn to avoid receiving shocks by turning a wheel that switched off the current, the number and extent of gastric erosions were reduced when compared with experimental animals that were not provided with one (13). The turning-off of the current could also be made to coincide with another brief electric shock the animal was punished for its attempts to avoid the initial shock. When this procedure was followed, rats developed even more extensive gastric erosions when compared to those of a yoked, unpunished partner. Conversely, animals that could not prevent receiving shocks, but were allowed to maul a neighboring rat, developed fewer gastric erosions (14). If the rat was allowed a choice between predictable and unpredictable shock, it "elected" the former and had fewer gastric erosions (15). These experiments illustrate important principles not originally predicted. Preventing or avoiding a painful or stressful experience reduced its damage; punishing attempts to do so maximized its effects. During the 1950s and 1960s (16) some of the physiological correlates of restraint and other experimental procedures were clarified. For example, vagotomy reduced and adrenalectomy increased the incidence of erosions in rats (6, 11). The complex interactions that influenced erosion formation were attested to by the fact that genetically determined factors variations in levels of serum pepsinogen in rats or the time of day restraint was applied, also played a role in the incidence of lesions. These variables could, however, be made irrelevant if restraint was maintained for a prolonged period when it became the prepotent variable. Ader (17-19) showed that the distribution of values for serum pepsinogen levels in populations of rats and humans were similar. Considerable differences in mean values occurred between and within various strains of rats. Male rats of the Osborne-Mendel strain had the highest, and those of the Wistar strain the lowest, values. Female rats of each of four different strains tended to have higher values of serum pepsinogen than male rats of the same strain, but differences between strains were less apparent in females. The role of variations in pepsinogen levels was revealed only when rats were restrained. The reason for this was that the location in the stomach of the erosions depended in part on the nature of the stressful experimental procedure. When rats were subjected to restraint with minimal food deprivation, the lesions were largely confined to the body of the stomach (20, 21). No relationship could be found between serum pepsinogen levels and erosion formation in the rumen of the rat's stomach. The body of the rat's stomach secretes acid and pepsin. About 29% of restrained animals, which fell into the highest 15% of the distribution of serum pepsinogen levels, were more susceptible to erosion formation in the body of the stomach, than those whose levels were in the lower 15% (19). The values for serum pepsinogen levels in rats were not fixed. Restraint lowered high, and increased low, basal pepsinogen levels. Circadian variations in such levels also occurred, which corresponded with the activity cycle of the animal, being higher at the beginning of the night and lower during the day when rats were asleep. Restraining rats for 6 hours at the peak of the activity cycle was associated with gastric erosions, which did not occur when restraint was applied at its nadir (17). But the circadian variation in pepsinogen levels did not seem to account for the results of this second procedure (21). Corticosterone levels rose before nightfall and then fell. Despite the fact that restraint increased corticosterone levels in the rat, their circadian variations did not seem to play a role in the formation of erosions (22). This second group of studies reopened the question of the part played by the corticosteroids in the formation of gastric erosions. In fact, the role of the corticosteroids in gastric erosion formation has still remained unsettled. The fact that procedures, such as prolonged restraint (with or without exposure to cold) or unavoidable electrical shocks, were both followed by their secretion and by the formation of erosions, does not mean that their positive correlation was a causal one. Murphy et al. (23), and Weiss (24-26) suggested that it was likely to be so. But erosions occurred even when adrenalectomy was followed by restraint (7); in fact, their incidence was increased! No correlation could be found in less prolonged experiments using restraint; in fact, an Psychosomatic Medicine 58: (1996) 525

3 H. WEINER inverse relationship between corticosteroid levels and erosion formation existed when a "backward" form of Pavlovian conditioning was employed (27) (eg, the physiological stimulus was delivered just before the neutral cue). At present, it is likely that the correlation between erosions and raised corticosteroids levels is a temporal one. Therefore, one must conclude that Selye's proposal that corticosterone played a primary pathogenic role in erosion formation was not likely to be correct. In part, it was predicated on the fact that in large doses the corticosteroids may be ulcerogenic in the rat (28) and in humans. The corticosteroids are known to inhibit prostaglandin (PG) secretion, which protects the gastric mucosa against erosions (29). The Roles of Prior Experience, Gastric Acid Secretion, and Body Temperature in Gastric Erosion Formation A third group of experiments (30-37) led to revised concepts about the roles of prior experience as a risk factor for, and of increased gastric acid secretion in, the formation of erosions (32). These experiments also highlighted the critical role of a fall in body temperature before the development of gastric erosions, during many experimental procedures designed to produce them. Male rats when prematurely weaned at 15 days, and when placed in an experimental conflict situation for variable periods of up to 4 weeks, at 120, 200, and 260 days of age, all developed rumenal ulcers; the females did not (21). Undetermined nutritional factors predisposed male rats to erosions and were postulated to mediate the effects of premature weaning (33, 34). Erdos6va and her colleagues (35) reported that male rats, before sexual maturity, when weaned at 16 days of age and then later immobilized at various ages, were more susceptible to gastric erosions in the body of the stomach. Sexually immature rats had a significantly higher incidence of lesions than rats who weaned themselves at 21 days. However, no further investigations were carried out to determine the reason why premature weaning should create a "window of susceptibility" to erosion formation during restraint. Prematurely separated rats, when restrained or not fed (or both), at 22, 30, and 40 days of age developed erosions in the glandular portion of the stomach with an incidence of 80-95% (36). (The incidence in normally weaned animals of the same age is 10%.) This effect was mediated by a fall in body temperature during restraint or food deprivation (30). Maintaining the body temperature of restrained rats averted the lesion. The animals did not ulcerate when restrained at 17 days of age. After 40 days of age, the incidence of gastric erosions progressively declined in separated animals when restrained; by 200 days of age it was 20%. The effects of a challenge in prematurely separated animals was, therefore, age-dependent. The initial separation did not produce erosions but placed them at profound risk for the lesions on subsequent challenge at a later age (of days) but not at all ages. An incapacity to maintain body temperature after separation was manifested only on challenge, and was not observed when the animal was left to its own devices in an environment whose temperature was about 20 C. The disturbance in body temperature regulation was originally brought about by depriving the animal of its mother's milk. If young separated rats were enabled to lap up milk, their body temperature was maintained when restrained (32). Prematurely separated rats were permanently underweight. They also stored less brown fat; when restrained, the stores were more rapidly depleted than those of their normal peers. As a consequence their body temperature was not maintained as in normally weaned animals (37). Cooling normal rats artificially in order to produce a fall in body temperature linearly induces increased gastric acid secretion. But when 30-day old, prematurely separated rats were cooled so that their body temperature fell to 29 C, they actually secreted less hydrochloric acid than did nonseparated rats (32, 37). When mature male rats of the same age, representing seven different rat strains, were food-deprived and then subjected to restraint in water maintained at 18.5 C, all rapidly showed a fall of body temperature of 5 to 8 C. Members of certain strains (Long- Evans, Wistar Kyoto, and Fisher-344) had a higher incidence of erosions. They also all showed a much slower recovery to normal body temperature than did members of other strains (eg, Lewis, Wistar) (38). One may conclude that a fall in body temperature and recovery from it were important variables in mediating the effects of restraint and/or cold. In fact, it is now possible to classify the many different procedures used to produce gastric erosions into two main groups: a) those associated with a fall in core body temperature: ie, restraint, immobilization in water (whether cold or not), a low ambient temperature, premature separation followed by restraint, increased motor activity produced by feeding the animal once daily (39), and so forth and b) those that promote an increase in body temperature: ie, pro- 526 Psychosomatic Medicine 58: (1996)

4 PEPTIC ULCER ULCERATION/ANIMAL MODELS longed, unavoidable electric shocks (40) and lateral hypothalamic lesions (41). The recognition that a fall in body temperature was the consequence of a number of experimental manipulations, culminating in gastric erosions, was a direct indication of the participation of the brain in the pathogenetic process. Before that time it was not certain that the effects of restraint required the brain's involvement, especially as increased corticosteroid levels, when they occur, were temporal, not causal, correlates of the procedure. (They were believed to rise as a result of the release of adrenocorticotropin (ACTH).) The Activity Stress Model Placing rats in a cold environment for 2 to 3 hours understandably lowered their body temperature, but it has not been so clear why merely immobilizing them also did so. When restrained they did not struggle; in fact, they initially fell asleep (31). Of course, immobilization interfered with motor activity but only when applied at the peak of the circadian activity cycle were gastric erosions produced (22). Increasing the motor activity of rats by the simple maneuver of feeding them for 1 hour once a day and giving them free access to run in a wheel markedly increased their motor activity (39). As 4-5 days passed, some rats progressively ran more and more, lost body weight, and died. At autopsy, large gastric erosions were found. In addition, a number of other physiological changes occurred before death: corticosteroid levels were high, testosterone levels fell, and in those animals which died the body temperature fluctuated wildly by 6-9 C in any 24-hour period (42). When no such fluctuations were observed, rats survived. But neither food restriction nor enhanced activity by themselves produced gastric erosions (39). Nor, could it be concluded that gastric erosions preceded the rats' increased motor activity (43). The critical variable preceding gastric erosion, seemed to have been a disregulation of body temperature. Furthermore, a combination of increased gastric contractions (44) and acid secretion, mediated in part by histamine and its H-2 receptor, seemed to initiate gastric erosions (43). The Proximate Mechanisms of Gastroduodenal Erosions: The Role of Gastric Hydrochloric Acid Secretion The recognition that a fall in body temperature was a critical variable associated with several experimental procedures, which produced gastric erosions, explained the effects of several different procedures and unified a number of disparate observations (30, 32, 45, 46). Cold restraint usually increased gastric acid secretion (46, 47) mediated by the release of histamine in the stomach and by H-2 receptors under vagal influence (48, 49). Increased gastric acid secretion, until recently, has been held to be mainly responsible for the formation of gastric erosions, despite the fact that it had long been known that lesions did not necessarily occur with exposure to cold, or to cold restraint (9, 50, 51). In the 30-day-old, prematurely weaned rat, cooling produced an average fall in body temperature of 8 C, but such a rat actually secreted less hydrochloric acid into the lumen of the stomach than did its nonseparated peers. In addition the response of prematurely separated rats to various secretogogues was virtually the same as that of normal rats of the same age (32). Exposure of normal rats to unavoidable electric shock did not increase acid secretion (52); gastric erosions still occurred in them with cold restraint, during which acid secretion was partially suppressed (53, 54). However, when acid secretion was totally inhibited by cimetidine or ranitidine, cold restraint was not associated with erosion formation (54). Thus acid in the stomach played a permissive role in erosion formation; it needed to be present in some amounts, but by itself it did not suffice to incite The Role of Gastric Contractions in the Production of Gastric Erosions About the time that the role of hydrochloric acid secretion in the pathogenesis of experimental erosions was clarified, changes in gastric contractions during stressful experience were beginning to be given serious consideration. A technique for measuring contractions by means of miniature strain-gauge force transducers attached to the external wall of the stomach was devised (40, 53, 55, 56). By this method, stressful experiences of several kinds were shown to produce a pattern of slow (0.5-2/minute or less), relatively regular rhythms in all parts of the stomach (53). The wave form and frequency (6-7/minute), but not always the amplitude, of the regular feeding pattern of contractions differed from the slow ones. The slow contractions were incited by lowering the body temperature (56), during cold restraint (53, 56), with predictable and unpredictable electric Psychosomatic Medicine 58: (1996) 527

5 H. WEINER shocks (40), in the restrained, prematurely separated, 30-day old rat, and after the placing of lateral hypothalamic: lesions (58). The slow contractions outlasted the procedure for as long as 2 hours (40). They were invariably associated with erosion formation. Only by their complete suppression (without altering gastric acid secretion) during cold restraint by pretreating rats with papaverine ( mg/kg) were erosions averted (53, 59). A transition in the rhythm of gastric contractions, usually associated with feeding, occurred when preceded by a variety of stressful procedures (including a fall in body temperature), and seemed to be another necessary condition in the pathogenesis of gastric erosions in rats, especially, and if combined with the presence of a certain amount of hydrochloric acid in the stomach lumen. However, the manner in which persistent, slow contractions might contribute to the production of erosions has remained controversial. They may predispose to reduced blood flow and patchy hypoxemia in the gastric mucosa (60-62). As more accurate methods have been developed for the measurement of mucosal blood flow while recording gastric contractions during cold restraint, this hypothesis has not been substantiated (63). In fact, other investigators have suggested that the slow contractions directly and mechanically injure the mucosa (64). The two variables gastric acid secretion and contractions normally covary (55, 65) in rats and in humans. What remains to be seen is whether their covariance is altered in human beings as in rats, during stressful experiences. Changes in motility and acid secretion are not the only ones to play a role in erosion formation for example, diminished HCOJ secretion in the stomach and/or duodenum have been described. The combination of four factors increased acid secretion, slow contractions, rapid gastric emptying, and diminished duodenal HCOJ secretion produced duodenal ulcers in rats, when injected three times with histamine (40 mg/kg) during a period of cold restraint (66). The most likely explanation of the formation of gastroduodenal ulceration was that, in the presence of increased amount of hydrochloric acid, slow, persistent gastric contractions accelerated gastric emptying. Acid entered the duodenum, backdiffused into die duodenal mucosa, whose first line of defense (HCOJ) had been reduced. This "local" end process was not linear but interactive, requiring a minimum of four variables; it could be averted by anticholinergic agents and bilateral vagotomy (51, 56, 57). Therefore, in the past 15 years, one of the key questions being investigated was: How does a stressinduced fall in body temperature mediated by the vagus nerve produce slow contractions, changes in acid secretion, and gastric erosions? Regulation of Gastric Acid Secretion Clinicians focus on the proximal mechanisms of gastroduodenal ulceration in particular on excessive gastrin and, therefore, gastric acid and pepsin secretion with gastrinoma or H. pylori infection, and on decreased mucosal "defense" due to the inhibition of PG synthesis by nonsteroidal anti-inflammatory drugs. Physiologists are, however, interested in the regulation of acid secretion and its disturbances. Some but not all the interacting, regulatory factors are displayed in Figure 1. A simple function acid secretion by the parietal cells of the stomach depends on neural (vagal-cholinergic), endocrine (gastrin), and paracrine (histamine) stimulation. However, pepsinogen secretion from the chief cells is regulated by gastrin and histamine. As the figure also shows each is stimulated or inhibited by the other, as well as by additional local (gastric) and neural and endocrine factors emanating from medullary vagal and raphe nuclei, sympathetic neurons, pituitary hormones, and specific hypothalamic and (not shown) limbic areas. The number of variables displayed in Figure 1 would alone generate a minimum of 36 possible interactions. Yet the figure does not display the regulation of gastric motility, mucus production, or mucosal cell turnover. Although it may very well be that, in some forms of human ulcer disease, increased pepsin and gastric acid secretion together, and diminished HCOJ production play the most important pathogenetic roles, the fact is that they may each come about by means of diverse distal regulatory disturbances. THE ROLES OF BRAIN PEPTIDES AND THE VAGUS NERVE IN THE FORMATION OF GASTRIC EROSIONS Beginning in 1980, several lines of evidence began to converge that pointed to the role of the tripeptide, pyro-glu-his-pronh 2 (TRH), in mediating the stressinduced fall in body temperature to alter the functions of the stomach, and to produce gastric erosions. They were as follows: 1. Intracerebroventricular (ICV) TRH in picomolar doses stimulated gastric acid secretion (67, 68), and 528 Psychosomatic Medicine 58: (1996)

6 PEPTIC ULCER ULCERATION/ANIMAL MODELS FALL IN PVN B T? N. Raphe' I* TRH J* PVN COLD 0-END Positive Influence Negative Influence Fig. 1. Some of the known factors involved in the control of gastric acid secretion and the formation of gastric erosions. Note in particular the regulatory influences at the level of the DMV and NA, and of the stomach. Cold and a fall in body temperature (BT) release TRH and CRH, TRH mainly increases gastric acid (H + ) secretion by acetylcholine [ACh) and histamine acting respectively through muscarinic (M a ) and histamine receptors in the stomach. TRH promotes the release of serotonin [5-HT] and bicarbonate ions [HCO^). CRH also stimulates HCO3 secretion by the action of/3-endorphin [fi-end). However, gastrin plays a major role in regulating H + secretion but is counter-regulated by antral acidification, and its own release is inhibited by somatostatin (STS) which is under vagal and CRH control. Gastrin, ACh, and histamine potentiate each other's activities in promoting H + secretion. Abbreviations: + = excitatory: - = inhibitory; AP, anterior pituitary gland; avp, arginine vasopressin; CRH, corticotropin-releasing factor; DMV, dorsal motor nucleus of the vagus nerve; H +, hydrogen ion; NA, nucleus ambiguus: PP, posterior pituitary gland; PVN, paraventricular nucleus of the hypothalamus; TRH, thyrotropin-releasing hormone. intracisternal (IC) TRH promoted gastric emptying (69), pepsin secretion (70-72), and also produced gastric erosions (69, 71, 72, 75). But injections at these sites did not reveal where in the brain TRH acted to produce these effects. 2. Exposing rats to an ambient temperature of 4 C increased the content of TRH 5- to 10-fold in the median eminence and in third ventricular fluid (76, 77). The release of TRH by cold, and presumably by a fall in body temperature, was mediated by a a - adrenergic receptors (78). In addition to maintaining and regulating the secretion of thyrotropin (TSH) tonically, TRH also regulated short-term, phasic responses (79, 80) in TSH to cold. 3. TRH is widely distributed in the brain. Twelve percent of the brain's total content of this peptide was present in the brain stem, especially in the dorsomotor nucleus of the vagus nerve (DMV), the nucleus ambiguous (NA), and the nucleus tractus solitarius (NTS) (81), where nerve terminals containing TRH and receptors for it were found (82-84). The medullary sites of the cell bodies of the TRH containing neurons and terminals in the DMV and NA were the nuclei raphe pallidus, magnus, and obscurus (81, 83-86). The DMV contained the cell bodies of preganglionic neurons of the vagus nerve to the stomach (87, 88), in particular to the antrum and pylorus in the cat (85). Cutting the vagus nerve abolished all the effects of ICV or IC TRH (72, 89, 90), which were also partially blocked by atropine (69, 74, 75, 91). Site of Action of TRH in the Brain Stem: Its Role in Gastric Acid Secretion Therefore, a number of clues pointed to the DMV and NA as the likely brain stem sites of action of Psychosomatic Medicine 58: (1996) 529

7 H. WEINER TRH in promoting efferent vagal activity, in regulating gastric function, and in the formation of gastric erosions. Injection of TRH in doses of 1 to 10 pmol ( ng) increased gastric acid secretion (92, 93) when injected into the DMV of rats, but it had no effect if injected into the area postrema or a variety of other medullary nuclei. The action of TRH in these experiments was very brief. The tripeptide is known to be rapidly cleaved by endopeptidases into several metabolites pyro- Glu-His-Pro, pyro-glu-his-pro-oh, pyro-glu-his, and cyclo(his-pro) none of which were active in promoting acid secretion (94). The effective dose and action of a stable, synthetic analogue, pyro-glu- His(3,3' dimethyl)-pronh 2 (Rx 77368), was the same as TRH but lasted much longer when injected IC (94, 95). RX in doses of ng injected into the DMV, NTS, and NA produced a dose-dependent increase of gastric acid output which slowly built up and attained a plateau level 40 minutes after injection. The secretory response was abolished by vagotomy (94, 96, 97). Co-secretion of TRH and Serotonin Thyrotropin-releasing hormone is co-localized with serotonin (5-HT) and substance P in the nucleus raphe' obscurus (98). Pretreatment of rats with a 5-HT reuptake inhibitor given IV or IC potentiated gastric acid secretion induced by IC RX (99). When TRH or analogue and 5-HT were injected in that order into the DMV, both gastric motility (100) and acid secretion were enhanced over the levels attained by injecting RX alone (101). Serotonin was released at the DMV by stimulating the nucleus raphe' obscurus with kainic acid but only in fed and not in fasted rats. Because bilateral vagotomy abolished the release of 5-HT in fed rats, it was believed that reflex vagal afferent activity stimulated by food in the stomach, interacted with efferent discharge from raphe" neurons (102). The Release of Gastric Serotonin and Hydrochloric Acid by TRH Vagal stimulation ( ) increased the gastric acid and the 5-HT content of the stomach in rats and cats. But the exact reasons for this correlated phenomenon did not become clear until later (97, 105, 106), although it was known that systemic administration of 5-HT inhibited stimulated gastric acid secretion (107, 108) and diminished the motility of the stomach (109, 110). But the action of endogenously released 5-HT (presumably from enterochromaffin, enterochromaffin-like and mast cells, and enteric neurons in the stomach) was unknown. The IC injection of RX (100 ng) produced a dose-dependent tripling of gastric acid secretion, but also a 9- to 10-fold increase of 5-HT content of the lumen of the stomach. Gastric acid secretion after 3 to 30 ng of RX IC could be increased by an additional 43% after the stomach's content of 5-HT was depleted by 66%, and 5-HT secretion into the lumen fell by 57%, after pretreatment of the animal with p-chlorphenylalanine (p-cpa) (300 mg/kg) an inhibitor of tryptophan hydroxylase. However, p- CPA also depleted the whole brain of 5-HT by 95%. Therefore, increased gastric secretion subsequent to IC RX injection could not initially be ascribed to a reduction of the gastric secretion of 5-HT. However, when central but not gastric stores of 5-HT were depleted by the neurotoxin, 5',7'-dihydroxytryptamine (5',7'-DHT), while the brain's noradrenergic neurons were protected against its action by desipramine, no enhancement of gastric acid secretion after IC RX occurred. 5',7'-DHT reduced whole brain 5-HT by 57% without changing the stomach's content. Even when 5',7'-DHT was combined with another neurotoxin, p-chloroamphetamine, to reduce brain 5-HT content by another 7%, the gastric secretory response to RX remained unchanged (97, 106). The exact sites of action in the brain stem of RX in stimulating 5-HT secretion in the stomach were the same ones that enhanced gastric acid secretion the DVM and NTS, and the NA but not other sites in the medulla. Thirty nanograms of the peptide analogue injected into the DVM increased the secretion rate of 5-HT 13 times (ill). Therefore, one may conclude that 5-HT secretion into the stomach was stimulated by vagal efferent discharge (it was blocked by pretreatment with atropine), and counter-regulated gastric acid secretion in the stomach. Curiously enough, gastric acid secretion when provoked by peripherally administered pentagastrin, histamine, or a cholinergic agonist (bethanechol) was not further enhanced when the stomach was depleted of 5-HT by p-cpa (97, 106). Thus the counter-regulatory effect of 5-HT on gastric acid secretion depended on their both being centrally and simultaneously stimulated. Not only did p-cpa pretreatment enhance acid secretion but it also doubled the gastric contractile response produced by IC RX (100 ng). In contrast, administration of. 5',7'-DHT did not (97, 106). 530 Psychosomatic Medicine 58: (1996)

8 PEPTIC ULCER ULCERATION/ANIMAL MODELS The Vagal Control of Gastric Acid Secretion and the Role of TRH Even such a simple function as gastric acid secretion is under complex regulatory control. Increased vagal efferent activity stimulated electrically produced the release of gastrin (counter-regulated by somatostatin (STS)), of acetylcholine acting via a M 1 muscarinic receptor, and of histamine mediated by H-2 receptors, and opposed by STS. The fact that injection of TRH and its analogue increased gastric acid secretion in rats was supported by the observation that the firing rates of vagal efferent neurons were increased by RX microinfused into the DMV (112, 114, 115). But, curiously, RX (15 ng) increased gastric acid secretion only through the release of acetylcholine and histamine, acting through their respective receptors; no gastrin was secreted (115). Differential Effects of TRH and STS on Gastric Acid and Serotonin Secretion: Not Only Vagal Control There seems to be no limit to the complexity of the system being described! Although 5-HT potentiated the effects of TRH in exciting vagal efferent neurons arising in the DMV, it counter-regulated the effects of vagal stimulation on gastric acid secretion. Somatostatin inhibited the effects on local gastrin- and histamine-stimulated acid secretion, but its analogue promoted acid and 5-HT secretion by the stomach when injected IC. Surprisingly, however (and obviously in contrast to TRH analogue), it inhibited the release of 5-HT by the stomach (116). Furthermore, it seemed that gastric 5-HT secretion produced by a TRH analogue was not only mediated by increased vagal discharge but by adrenal catecholamines. In fact, STS suppressed such catecholamine secretion incited by TRH (117). TRH and the Stimulation of Slow Gastric Contractions To recapitulate, restraining rats in the cold (and other standard stressful experiences) was invariably associated with slow, irregular (high amplitude), long-lasting gastric contractions and gastric erosions, which were also associated with a fall in body temperature of 8-12 C (118). Intracisternal TRH (1 jug) or RX ( ng) stimulated these slow contractions in awake rats, which could be abolished by bilateral vagotomy (90), and were associated with gastric erosions (71, 73, 74, 91). The site of action of TRH and its analogue were exactly the same as those that stimulated increased acid secretion. Once stimulated, the contractions lasted for 1 hour (119, 120). Injection of control sites in the medulla the hypoglossal nucleus and the lateral reticular nucleus had no effect (119). In summary, TRH and RX injected into specific medullary sites known to contain TRH nerve terminals and receptors (83, 84) in the rat and cat (85, 121) simulated the effects of cold restraint. The stimulatory effects of TRH on gastric acid secretion and contractions were independent of TSH secretion (70, 85) or changes in blood pressure and heart rate (85). The fact that TRH and its analogue promoted both acid secretion and slow contractions, and produced gastric erosions in rats, has not proven that cold restraint or other procedures released the tripeptide from known sites (eg, the DMV) in the medulla oblongata. However, suggestive evidence has been obtained that this line of thought was valid. TRH was released in the rat medulla oblongata under basal conditions or when stimulated by K + (122). The ICV injection of a specific antibody to TRH decreased gastric acid secretion (123, 124) and TSH release in rats exposed to cold (125, 126), and averted gastric erosion produced by cold restraint (126). TRH Stimulates Acid Secretion From Additional Brain Sites Therefore, considerable evidence has been accumulated that TRH and its analogue increased gastric acid, acetylcholine, histamine, and 5-HT secretion, and incited slow gastric contractions and gastric erosions when injected IC, ICV, or into the DMV and NA. Many of the same effects have been produced by electrically stimulating, or by the injection of RX near the nerve cell bodies of the medullary raphe nuclei whose axons terminate in the DMV (84, ). Thyrotropin-releasing hormone also increased gastric acid secretion and contractions when injected into the ventromedial (VMH), lateral (LH), and paraventricular (PVN) nuclei of the hypothalamus (131, 132) and the central nucleus of the amygdala (133). However, much larger doses of TRH were needed to produce the same effect than were required when the DMV was microinjected. Psychosomatic Medicine 58: (1996) 531

9 H. WEINER Corticotrop in -Releasing Hormone (CRH), Stressful Experience, Gastric Function, and Erosions Corticotrop in-releasing hormone is a 41-amino acid peptide (134, 135). It is chemically identical in rats and humans (136). The peptide is widely distributed throughout the brain and the body. The largest concentration of CRH in the brain is contained in the perikarya of neurons in the medial parvocellular portion of the PVN, whose axons pass to the median eminence where it is released into the portal circulation of the pituitary gland ( ). Corticotropin-containing neurons, terminals, and receptors are present throughout the brain in the forebrain, amygdala, central gray matter, the locus ceruleus, the dorsal tegmental and parabrachial nuclei, NTS, DVM, and NA, the A a and A 2 nuclear complexes of the medulla as well as in preganglionic sympathetic neurons of the spinal cord ( , 146). Of particular note is the direct neuronal pathway from the PVN to the locus ceruleus, to the NTS and the DVM whose transmitter is oxytocin (145, 146). The peptide has also been identified in the adrenal medulla, liver, lung, stomach, duodenum, and pancreas ( ). Release of CRH During Stressful Experiences Immunoreactive CRH was not usually detectable in the blood serum of rats and humans under normal and some life-threatening or stressful experiences ( ). However, its concentrations increased from baseline levels in the hypophysial portal circulation, and in the adrenal venous blood of rats and dogs during and after hemorrhage, and during exposure to cold (151, 153, 154). Acute and chronically stressful experiences (for instance, cold restraint, restraint alone, unpredictable shock, or injury) altered the CRH content of the median eminejnce, the PVN, and other nuclei known to contain it (155, 156). Unpredictable shock plus cold restraint at 4 C reduced CRH levels in the median eminence and DMV, but increased it in the periventricular nucleus and the locus ceruleus (155). Physiological Effects of CRH The CRH induces the secretion of the proopiomelanocortin gene products ACTH, /3-endorphin, and so forth from the anterior pituitary gland (134, 157,158). In addition, it has wide-ranging behavioral and bodily effects which cannot alone be ascribed to the secretion of the pituitary hormones. Of particular importance are its autonomic and metabolic effects: It stimulated sympathetic efferent, but reduced parasympathetic activity; glucagon, epinephrine, norepinephrine were secreted under its influence. But the secretion of the growth and luteinizing hormones were inhibited. It raised the heart rate and blood pressure. When injected into specific brain sites or ICV (in hypophysectomized animals), CRH diminished gastric acid secretion and emptying and small bowel transit and motility, but increased gastric and duodenal HCO^" secretion. It speeded up colonic transit ( ). Virtually all of these changes in physiology can be reversed by the subsequent ICV injection of a specific CRH antagonist, a-helical CRH (158, ). CRH and Gastric Erosions It is fairly certain that TRH was released when rats were cold restrained and their body temperature fell; TRH-injected IC or ICV incited gastric erosions. The CRH was also selectively secreted under these conditions, but it protected against the formation of erosions. The injection of CRH into the lateral hypothalamus or ICV failed to incite gastric erosions (75,167). In fact, CRH (5 jag), injected ICV or into the ventromedial nucleus of the hypothalamus, reduced the extent of gastric erosions associated with cold restraint (137, 168, 169). If CRH played a role in inciting such erosions (rather than attenuating them), its inhibition by the administration of a-helical CRH would be expected to reduce their severity; but this did not occur (168). In fact, CRH seemed to have some of the exact opposite effects to TRH on gastroduodenal function! But the matter may not be quite as simple as just stated. The protective effect of ICV CRH did not seem to manifest itself in food-deprived animals (169), in fact it may have enhanced erosions (75). (As already mentioned, the gastric erosions of food deprivation differed in location from those associated with restraint.) These discrepancies may have been the result of not taking into account the ages of animals. Older rats tended to respond to CRH by mounting a diminished, acute corticosterone response compared to younger ones. Yet older animals had a more sustained corticosteroid elevation to stressful experiences (170). The protective effect on the stomach of IV CRH against stress was seen only in younger animals (171). 532 Psychosomatic Medicine 58: (1996)

10 PEPTIC ULCER ULCERATION/ANIMAL MODELS CRH and Gastric Acid Secretion In rats and dogs, especially when challenged IC and ICV, CRH inhibited gastric acid secretion (142, 167, ), which was reversed by the administration of ICV (but not IV) a-helical CRH ( ). Intracisternal CRH also diminished increased gastric acid secretion in rats, stimulated by IV pentagastrin, or IC TRH (167). In dogs, ICV CRH decreased only pentagastrin- not histamine-stimulated acid secretion (175). Adrenalectomy and ganglionic and a.,- and a 2 - noradrenergic blocking agents prevented the decrease in gastric acid secretion, suggesting that the central nervous system action of CRH was mediated by sympathetic, noradrenergic (but not adrenergic) means, probably by stimulating STS secretion. But the effects of CRH were not altered by vagotomy (see, however, Ref. 167) or hypophysectomy (175, 176, 180, 181). To complicate the matter further, vasopressin may have been another mediator of stress-promoted gastric acid secretion, and of that induced by CRH, because a specific vasopressin inhibitor reversed the suppression of acid secretion brought about by CRH (175, 176, ). Vasopressin secretion into the blood stream was promoted by CRH (176) and interacted with CRH to potentiate ACTH secretion (183). Although the specific sites of action in the brain of CRH to inhibit gastric secretion are not completely known, one likely place was the DMV. However, CRH also had the same effect when applied to the VMH (184, 185), LH (167, 184), and the PVN (184); the lowest dose of CRH was required when the latter was injected (184). CRH was unique among peptides in diminishing gastric acid concentration, and at the same time stimulating the volume of gastric fluid, and gastric HCOJ production and concentration when injected into the three hypothalamic sites (but not when the IC or ICV routes were used) (173, 184, 185). CRH and Gastroduodenal Bicarbonate Secretion The presence of HCO^~ in cells and its secretion into lumen of the stomach and duodenum, respectively, constitute a "first line of defense" against protons and their back-diffusion into mucosal cells (186). Both in human duodenal ulcer diseases (187), and in the cold-restrained rat given histamine (66), duodenal HCOJ secretion was reduced. Duodenal HCOJ secretion is regulated both locally and by vagal and sympathetic efferent activity. Increased vagal activity (eg, sham feeding, hypoglycemia, ICV TRH, and STS) stimulated HCO^ secretion ( ) in rats and dogs. Noradrenergic (eg, a 2 -sympathetic activity) inhibited it ( ). Both IV and ICV CRH stimulated gastroduodenal HCO^ secretion an effect that was counteracted by a-helical CRH (194, 195). The secretory responses to CRH were similar to that produced by stressful experience (195). However, the effect of CRH on HCOJ secretion was counterintuitive. If the peptide increased noradrenergic sympathetic discharge it should have reduced it, unless it acted through some other mediator. In fact, this is the case. No HCOJ secretory response in hypophysectomized rats occurred after CRH injection or in unoperated rats after the administration of naloxone. Conversely, the response was unaltered after sympathetic ganglionic blockade or adrenalectomy. By implication, IV or ICV CRH released j3-endorphin which stimulated duodenal HCOjj" secretion an inference supported by the fact that the administration of this opioid peptide did produce such an effect (195). CRH and Gastric Motor Function Immobilization of rats affected the complex function of gastric emptying in opposite ways. In restrained rats, emptying and passage through the small bowel were both delayed. Intracerebroventricular (but not IV) CRH reproduced these effects which were blocked by ICV (not IV) injection of a-helical CRH. The antagonist also abolished the delay in gastric emptying after stressful experiences (178, ). The effects of ICV CRH on emptying were mediated by increased sympathetic activity because they could be abolished by a ganglionic blocking agent (chlorisondamine) and by an a n -norepinephrine antagonist (198, 200). ]However, other data suggest that the vagus nerve also participated in delaying emptying, or, at least, in abolishing slow contractions (201), or gastric cyclic migratory motor activity (202), after ICV CRH infusion. Slow Gastric Contractions and CRH: Interaction with TRH Centrally administered CRH protected rats against gastric erosions, but the manner in which it does so is still not known. However, the peptide diminished the motility of the stomach and some of its normal rhythms. The slow gastric contractions in rats, produced by IC RX 77368, were reduced by IC CRH in a dose-dependent manner. The contractions, however, Psychosomatic Medicine 58: (1996) 533

11 H. WEINER were suppressed only by IV CRH in 10 times the dose effective by the IC route. Large high amplitude contractions provoked by IC 2-deoxy-D-glucose were also abolished by IC but not IV CRH. But the contractions produced by continuous IV carbachol infusion remained unaltered either by IC or IV CRH (100 ng) (203). These results strongly suggested that CRH inhibited only centrally stimulated contractions. Simultaneous microinjection of RX and CRH into the DMV failed to produce slow contractions (204). Therefore, one site of inhibition of the action of RX by CRH was the DMV. The IC doses of CRH required to suppress slow contractions were less by a factor of 10 than those necessary to suppress gastric acid secretion. The inference that could be drawn was that stimulated slow gastric contractions and acid secretion were reduced by CRH through an inhibition of vagal efferent activity in an differentiated manner (167, 201, 205). Additional Inhibitors of TRH and Its Analogue at the DMV: Bombesin and Interleukin-I/3 A number of peptides additional to CRH seemed to interact with TRH or its analogue at the level of the DMV. Bombesin. This 14-amino acid peptide, originally isolated from the skin of amphibians, has a mammalian counterpart in gastrin-releasing peptide. Immunoreactivity and receptors for bombesin-like peptides occur in the hypothalamus, area postrema, dorsal vagal complex (206, 207), stomach, and gut. When injected IC, bombesin inhibited food intake and gastric acid secretion and emptying; prevented gastric erosions; lowered the body temperature; and stimulated grooming, motor activity, and epinephrine secretion (89, ). The PVN (211), and DMV (212) seemed to be two of the sites of action of bombesin because, when injected there, it suppressed stimulated (by IC TRH analogue) gastric acid secretion (212) and slow contractions in a dosedependent manner, but not when they were elicited by IV carbachol (213). Yet, IC bombesin inhibited IV pentagastrin-induced gastric acid secretion an effect which was reversed by a specific bombesin antagonist iv-acetyl-grp (20-26)-0-CH 3 (214). The concentration (in pmol) of bombesin used to inhibit gastric secretion made it a more potent inhibitor weight for weight than CRH. Interleukin-Ifi (IL-Ifi). IL-I/3 is a generic name for two cytokines, each with a molecular mass of 17 kda, which share the same two receptors. (The receptors do not have intrinsic tyrosine kinase activity.) The IL-I/3 play a key role in inflammation but also in noninflammatory processes. They modulate pain, stimulate the release of CRH, promote slow wave sleep, and decrease food intake. Many different cell types both within (microglia, astrocytes, selected neurons, and microvascular endothelium) and outside (macrophages, monocytes, fibroblasts, and osteoblasts) the brain secrete IL-I/3, especially on stimulation by bacterial cell products (215, 216). At the cellular level, IL-I secretion in the brain has been induced by substance P, and in return is regulated the secretion of transforming growth factor-/3, nerve and fibroblast growth factors. The mrna for IL-^ has been identified in the VM, PVN, lateral and arcuate hypothalamic nuclei, olfactory bulb, cerebellar granule cells, hippocampus, and in glia. Interleukin-I receptors have been identified in the basolateral amygdala, hippocampus, anterior and medial dorsal thalamus, cerebellum, mesencephalic nucleus of cranial nerve V, and on 5-HT-containing neurons in the dorsal raphe nuclei. Hypothalamic neurons that increased their firing rates in rats placed in a cold environment were also excited by IL-I/3 infusion. Blocking the IL-I receptor (II), inhibited a rise in body temperature produced by IL-I/3 (215, 217, 218); and the mrna for IL-I/3 was induced in the brains of immobilized rats (219). Interleukin-I/3 inhibited all gastric functions when injected ICV, IC, IV, or into the anterior preoptic nucleus and PVN in fentomolar quantities (220). Human recombinant IL-I/3 inhibited basal and stimulated gastric acid and pepsin secretion, gastric emptying, postprandial intestinal activity, and prevented gastric erosion formation induced by restraining rats, or IC injection of TRH ( ). Additionally, murine recombinant IL-I/3 inhibited slow gastric contractions, promoted by microinjection of the TRH analogue into the DMV of rats. The suppressive effects of picogram amounts of IL-I/3 on contractions was reversed by IC injections of the IL-I receptor antagonist (223). Although we are far from understanding when, where, or under what conditions IL-I/3, its receptor, and receptor antagonist are expressed, a provisional synthesis may be possible. Substance P was colocalized with TRH and 5-HT in nuclei raphe obscurus and profundus. Cold and restraint seemed to induce the expression of TRH a potent activator of DMV efferent neurons. Serotonin and TRH potentiated each others' actions at the DMV. Substance P also induced IL-I/3 which stimulated CRH secretion (221), and all three counter-regulate the excitatory 534 Psychosomatic Medicine 58: (1996)

12 PEPTIC ULCER ULCERATION/ANIMAL MODELS effects of TRH on vagal efferent discharge to the stomach. The Role of Other Peptides in Gastric Function and Possibly in Erosion Formation A number of additional peptides inserted into the brain, ICV, or IC are known to affect gastric function, in particular gastric acid secretion (Table 1]. Specifically, STS, vasoactive intestinal peptide (VIP), gastrin, cholecystokinin, and oxytocin stimulated secretion, but have not, with the exception of VIP been shown to produce erosions. Conversely, calcitonin and its gene-related peptide, neuropeptide Y, neurotensin, the opioid peptides, insulin growth factor II, and platelet-activating factor inhibit gastric secretion (for review, see Refs. 224 and 225). At this time, the realistic role of this multitude of peptides in erosion formation, except for TRH and CRH, has not been determined. For example, we still do not know whether the other peptides are released when body temperature falls, or whether they respond to stressful experiences. The effects of many of the peptides have been studied in rats, cats, ferrets, rabbits, and dogs (225). But their relevance to human peptic ulcers remains to be established. THE TRANSDUCTION OF STRESSFUL PROCEDURES BY THE BRAIN Despite a great deal of progress, we do not understand how the traditional experimental procedures are transduced by the brain to produce gastric erosions. Most of our information comes from neurophysiological studies of increased electrical activity of neurons in the medial, central, and lateral nuclei of the amygdala, and the bed nuclei of the stria TABLE 1. Peptides Known to Stimulate or Inhibit Gastric Acid Secretion When Centrally Injected into Mammals Stimulate Inhibit Cholecystokinin Bombesin (animals) Castrin Gastrin-releasing peptide Oxytocin Calcitonin and gene-related peptides Somatostatin and analogues CRH TRH and analogue Insulin growth factor II (RX 77368) Vasoactive intestinal lnterleukin-l/3 peptide Neuropeptide Y Opioid peptides Platelet-activating factor terminalis of rats during the initial phase of restraint. However, in those rats that later developed erosions, a rebound inhibition of multiple neuronal units in the amygdala followed. Whereas in rats without erosions, the initial increase in neuronal discharge rates was followed by a return to the baseline rate. Reproducible, distinct patterns of neuronal activity were, therefore, produced by restraint, each correlated with erosion formation, or not (224, 225). Similar results have been obtained by sampling evoked neuronal activity in granule cells of the dentate gyrus of the hippocampus: it was suppressed by restraint, but only in erosion-susceptible rats. Restraint also activated neurons in the anterior cingulate gyrus of rats (225). Exposure of animals to cold increased the activity of neurons in the VMN of the hypothalamus. And indirect evidence was obtained, implicating the central nucleus of the amygdala during activity stress (eg, see Ref. 226). Neurochemical changes in the brain have been described during various forms of stressful procedures. The highest levels of the principal metabolite of norepinephrine, 3-methoxy-4-hydroxyphenylethyleneglycol sulfate (MHPG-SOJ, occurred in all the major regions of the brain (from cerebral cortex to the lower brain stem) after rats were exposed to the activity stress procedure for 5 days. The levels of MHPG-SO 4 in these brain regions progressively increased day by day. The levels of MHPG-SO 4 attained (especially in the cerebral cortex, thalamus, hippocampus, and midbrain) were considerably higher than in rats exposed either to 16-hour restraint for 6 successive days, or 5 days of 1 hour each of electric shocks (227). One hour of restraining rats produced about the same increases in levels of MHPG-SO4 in all brain regions as did 1 hour of electrical shocks to the feet. But when fear was conditioned in rats, the levels were increased over those found in control animals but only in the amygdala, hypothalamus, and locus ceruleus (228). Therefore, regional changes in MHPG-S0 4 depended on the experimental procedure used. The central nucleus of the amygdala sends axons directly to the dorsal vagal complex, and via the stria terminalis to the hypothalamus. The hypothalamus also receives direct input from the hippocampus through fimbria-fornix neurons which may, however, play an insignificant role in gastric erosion formation. The ventral subiculum of the hippocampus directly connects to the central nucleus of the amygdala and may be the more relevant pathway in the production of gastric erosions (225). By electrical stimulation, placing lesions, microinjection of peptides and of biogenic amines, or the Psychosomatic Medicine 58: (1996) 535

13 H. WEINER use of their receptor antagonists, it has been possible to trace out the neuronal pathways involved in the regulation of gastric acid, pepsin, bicarbonate, and serotonin secretion, the stimulation of gastric motility, and in the production of gastric erosions. That circuitry is displayed in Figure 2 (229). The validity of these neuronal arrangements (Table 2) has partly been established by the injections of TRH and CRH ( ), but not for most of the other peptides previously mentioned. The final common pathway in the process of gastric erosion formation is the DMV and NA and their preganglionic vagal efferent fibers. But the sympathetic nervous system, and the hormones of the pituitary and adrenal glands also participate in To ana from (hi Penpnenj VII. IX. X Cranlol Nerves Fig. 2. Projections of the dorsal vagal complex (nucleus of the solitary tract and dorsal motor nucleus). This is a highly schematized view of the nuclei that have direct or monosynaptic connections with the dorsal vagal complex (DVQ. The broader lines represent the more prominent pathways, but not necessarily the more important ones (see text). For the sake of simplicity, the efferents from the DVC are not indicated. ACE, central nucleus of the amygdala; DMH, dorsomedial hypothalamus; PVN, para ventricular nucleus; BNST, bed nucleus of the stria terminalis; VMH, ventromedial hypothalamus; NTS, nucleus of the solitary tact; DMN, dorsal motor nucleus. TABLE 2. Sites of Action in the Brain and Effect of Gastric Function of TRH and CRH" Nucleus TRH CRH Amygdala: Central nucleus Hyphothalamus: Paraventricular n. Lateral n. Ventromedial n. Medulla: Dorsal motor n. of vagus nerve N. ambiguus N. raphe (magnus, obscurus, pallidus) t (CE) t (CE) t (GAS) t (GAS) t (CAS) t (GAS) T (GC) T (CE) T (5-HT) t (Blood flow) T (CAS) t (CC) T (CE) T (5-HT) t (Blood flow) I (CAS) T (5-HT) " Key: CRH, corticotropin-releasing factor; CAS, gastric acid secretion; GC, gastric contractions; CE, gastric erosions; HCOJ, bicarbonate secretion; 5-HT, serotonin secretion; and TRH, thyrotropin-releasing hormone.? J, (CAS) t (Volume) i (GAS) t (Volume) t (HCOJ) 1 (CAS) T (HCO3-) t (Gastrin) I (CE) I (CAS) 4 (GC) the regulation of gastric acid secretion and, possibly in erosion formation. In fact, the matter is even more complex that Figure 1 portrays. For example, TRH injected into the DMV inhibited vagal afferent activity mediated through the overlying NTS (112, 113). Furthermore, the pituitary gland stimulated by CRH also regulates H + secretion by the stomach by releasing j8-endorphin and arginine vasopressin. Further credence to this belief was given by the fact that CRH injected into the PVN in nanogram doses inhibited gastric acid secretion and stimulated HCOJ secretion effects that were reversed by the CRH antagonist. Injection of CRH into the VMH had the same effects, and also averted gastric erosions produced by 4 hours of cold restraint (184). Lateral hypothalamic lesions were followed by increased basal acid secretion, slow gastric contractions and gastric erosions (58). Microinjection of LH with CRH also inhibited gastric acid and promoted HCOJ secretion (167, 184). Injection of these two hypothalamic nuclei with TRH-stimulated gastric acid secretion (131).? 536 Psychosomatic Medicine 58: (1996)

14 PEPTIC ULCER ULCERATION/ANIMAL MODELS Injection of TRH into the central nucleus of the amygdala when rats were cold restrained (4 C) enhanced the number and severity of erosions. (Their severity was also markedly reduced by dopamine and neurotensin (230).) The effect of TRH was abolished by methyl atropine, suggesting that it was mediated by the vagus nerve (224, 230). Left out of the foregoing account is the possibility that, once injured, the repair of the gastroduodenal mucosa was impaired when a rat was subjected to "stressful" procedures. Gastrin enhanced the regeneration and/or repair of mucosal cells by stimulating DNA synthesis (233, 234). Mediated by gastrin secretion, the vagus nerve seemingly played a trophic role in mucosal growth and regeneration. Support for this notion came from the demonstration that in dogs when subjected to chronic sham feeding, which stimulated maximal gastric acid output by vagal means, an increased density of the gastric parietal and chief cells was produced (235). In addition to gastrin, epidermal growth factor (EGF) and transforming growth factor-/3 (TGF-/3) also stimulated the growth of the gastroduodenal mucosa (236). When the bodily source of EGF was removed, rats were more susceptible to stress-induced erosions, which were prevented in other experiments by EGF administration (237, 238). This growth factor also accelerated the healing of erosions (237, 239), as did basic fibroblast growth factor when duodenal ulcers were incited in rats by cysteamine (240). A promising line of animal research might well be to determine whether stressful experiences influence the induction or secretion of growth factors. VALIDITY OF ANIMAL MODELS FOR HUMAN GASTRODUODENAL ULCERATION For many years gastroenterologists have questioned the validity of animal models for human peptic ulceration (241) and have focused their attention on proximal pathogenetic mechanisms (242), rather than on the distal ones the roles of stressful experimental procedures, their transduction by the brain, and the brain's capacity to alter gastric function to produce erosions. The doubts of gastroenterologists have been heightened recently by the discovery of the association of gastroduodenal H. pylori infection with gastric and duodenal ulcers (243), while neglecting several conceptual problems that this approach has posed. In about 80% of all older patients with peptic ulcer this bacterium has been isolated from the stomach (244). In 93%, specific antibodies against it can be detected in the serum of older hospitalized patients with gastric ulcer, and in 92% with duodenal ulcer. But, the authors of these studies neglected the meaning of the fact that in 78% of the control subjects without any disease, the same serum antibodies could be detected. They stated that the odds ratio of developing the two forms of peptic ulcer were increased when the specific antibodies against H. pylori were present in the serum of patients (245). Another issue has been understated. The presence of H. pylori in the stomach has been associated with a number of other symptoms and diseases: dyspepsia (246), active gastritis and duodenitis (247), gastric carcinoma (248), and non-hodgkin's gastric lymphoma (249). Therefore, the bacterium lacks pathogenetic specificity. The argument for its pathogenic role in peptic ulcer has been strengthened by the observation that treatment with a combination of an antibiotic (eg, tetracycline), bismuth subsalicylate, and metronidazole, designed to eradicate H. pylori from the stomach and duodenum, healed peptic ulcer disease in 74% and 95% of patients, respectively (250). However, 75% of patients with duodenal ulcer in one series responded favorably to placebo without the disappearance of the infection a statistically significant lower rate than in those given combined treatment. And an occasional patient, in whom the bacterium had been eradicated, later had a recurrence of a peptic ulcer (251). Therefore, one may conclude that H. pylori plays some role in the pathogenesis of peptic ulcer and gastritis. Intentional ingestion of the bacterium did produce an acute gastritis (252), but not ulceration. The question has remained how the infection contributed to the pathogenesis of gastritis and ulceration. Patients with this infection had excessive gastrin responses to the eating of a meal, as a result of which they had higher gastric acid secretion rates than noninfected subjects (253). Even when no ulcer was present, infected subjects responded with higher serum gastrin levels to meals than did noninfected ones (254). In addition, infection has been shown to raise pepsinogen-i levels, which fell with antibiotic treatment (255). Infection was also associated with a lowering of hydrophobicity of the gastric mucus layer due probably to an endopeptidase secreted by H. pylori. Treatment restored the composition of the mucus (255). The bacterium has been found to secrete a protein cytotoxin (256) and a chemotactic factor (257) which may incite an inflammatory response and gastric mucosal metaplasia (258). The fact that serum antibodies against H. pylori were produced suggested Psychosomatic Medicine 58: (1996) 537

15 H. WEINER that immunocompetent cells in the lamina propria of the stomach wall mount an immune response. The role of these cells has been much neglected in the study of gastroduodenal disease. They expressed mrnas for histamine (H-2), gastrin, muscarinic (M.,_ 5 ) acetylcholine, and dopamine (D^) receptors, which parietal cells did not seem to (259). Therefore, immunocytes (eg, macrophages, monocytes, CD4 + cells, and so forth) may be activated by infection and secrete gastrin, histamine, and so forth. Presumably, H. pylori by raising gastrin secretion increased gastric acid and pepsin production, impaired mucus protection, and incited metaplasia and inflammation, leading to gastritis, duodenitis, and peptic ulceration. However, the pathogenetic mechanisms responsible for gastric non-hodgkins lymphoma and adenocarcinoma have remained unexplained. Most (80%) normal peoples' stomachs are colonized by H. pylori, yet only 1-2% ever developed peptic ulcers ( ). Therefore, the bacterium may be necessary but not sufficient for one form of the disease, leaving room for many other contributing factors including psychosocial ones. Thus peptic ulcers remain a multifactorial and heterogenous group of diseases (1, 255). SUMMARY This review does not fulfill the promise that an animal model would account for the pathogenesis of human peptic ulceration. The main impediment to providing such an explanation is that the processes which culminate in ulceration are so complex. Even the usual regulatory processes underlying the physiology of the stomach for example, a function as simple as hydrogen ion production and secretion result from multiple feedback processes in interaction with each other (Figure l). Because of the number of interacting variables, many different combinations of antecedent pathogenetic processes are possible, or even likely. The immediate antecedents of erosion formation seem to be back-diffusion of protons into mucosal cells, occurring when a variety of protective factors (eg, 5-HT, HCOJ, and mucus secretion) were diminished. However, in rats another essential condition must be met slow gastric contractions which seem to damage the.mucosa by unknown means. The discovery that in many experimental models of erosion formation in rats, the body temperature fell, and when prevented, no lesions occurred, has accelerated progress. Because TRH was secreted in the brain of rats exposed to cold, much effort has been expended in verifying that TRH did indeed incite acid secretion and erosions when injected into DMV, PVN, and amygdala. Slow gastric contractions were generated with DMV injections of TRH or its analogue. But TRH also incited gastric 5-HT and HCOJj" secretion to counteract acid secretion. Thus it is likely that TRH secretion participated in the cold restraint model, in which gastric acid secretion was usually increased. Yet it was not always so (9, 50, 51). Thus the pathogenesis of erosions on cold restraint has not been fully explained. Nonetheless, exposure to cold also promoted CRH secretion. When CRH was injected into the VMH, it attenuated damage to the gastric mucosa after 4 hours of cold restraint. When injected IC and into the DMV, CRH also abolished stimulated slow gastric contractions, inhibited gastric emptying, and increased HCO^ secretion into the stomach lumen, thereby raising its ph (185). Therefore, during cold restraint, a spectrum of acid secretory responses may have occurred, brought about by the relative predominance either of TRH or CRH. To add further to the problem, TRH incited proton secretion only through the release of acetylcholine and histamine. But CRH also increased serum gastrin levels in rats (185) and incited gastric STS release, known to inhibit both gastrin and histamine secretion. Much progress has been made in the past years in providing some support to the idea that gastric erosions can be brought about by stressful experiences mediated by the brain. Some of the brain circuitry translating such experiences has been mapped for example, the central nucleus of the amygdala, PVN, VMH, and LH nuclei, and the DMV and NA. Many of these nuclei are monosynaptically connected. And each of them are responsive in opposite directions to TRH and CRH to alter the principal gastric functions. REFERENCES 1. Weiner H: From simplicity to complexity ( ): The case of peptic ulceration I. Human studies. Psychosom Med 53: , Selye H: A syndrome produced by diverse nocuous agents. Nature (Lond) 148:84-85, Selye H: The alarm reaction. Can Med Assoc J 34:706, Bonfils S, Lambling A: Psychological factors and psychopharmacological actions in the restraint-induced gastric ulcer. In Skoryna SC (ed), Pathophysiology of Peptic Ulcer. Montreal, McGill University Press, 1963, Psychosomatic Medicine 58: (1996)

16 PEPTIC ULCER ULCERAT1ON/ANIMAL MODELS 5. Brodie DA: Ulceration of the stomach produced by restraint in rats. Gastroenterology 43: , Brodie DA: Experimental peptic ulcer. Gastroenterology 55: , Brodie DA, Hanson HM: A study of the factors involved in the production of gastric ulcers by the restraint technique. Gastroenterology 38: , Brodie DA, Marshall RW, Moreno OM: The effect of restraint stress and gastric secretion in the chronic gastric fistula rat. Physiologist 4:14, Brodie DA, Marshall RW, Moreno OM: The effect of restraint on gastric acidity in the rat. Am J Physiol 202: , Par6 WP, Temple LJ: Food deprivation, shock stress and stomach lesions in the rat. Physiol Behav 11: , Brodie DA: Stress ulcer as an experimental model of peptic ulcer disease. In Pfeiffer CJ (ed), Peptic Ulcer. Philadelphia, Lippincott, 1971, Lee YH, Bianchi RG: Use of experimental peptic ulcer models for drug screening. In Pfeiffer CJ (ed), Peptic Ulcer. Philadelphia, Lippincott, 1971, Weiss JM: Psychological factors in stress and disease. Sci Am 226: , Weiss JM, Pohorecky LA, Salman S, et al: Attenuation of gastric lesions by psychological aspects of aggression in rats. J Comp Physiol Psychol 90: , Gliner JA: Predictable vs. unpredictable shock: Preference behavior and stomach ulceration. Physiol Behav 9: , Par6 WP, Glavin GB: Restraint stress in biomedical research. A review. Neurosci Biobehav Rev 10: , Ader R, Beels CC, Tatum R: Blood pepsinogen and gastric erosions in the rat. Psychosom Med 11:1-12, Ader R: Plasma pepsinogen level in rat and man. Psychosom Med 25: , Ader R: Plasma pepsinogen level as a predictor of susceptibility to gastric erosions in the rat. Psychosom Med 25: , Conger JJ, Sawrey WL, Turrell ES: The role of social experience in the production of gastric ulcers in hooded rats placed in a conflict situation. J Abnorm Soc Psychol 57: , Sawrey WL, Weisz JD: An experimental method of producing gastric ulcers. J Comp Physiol Psychol 49: , Ader R: Behavioral and physiological rhythms and the development of gastric erosions in rats. Psychosom Med 29: , Murphy HM, Widcman CH, Brown TS: Plasma corticosterone levels and ulcer formation in rats with hippocampal lesions. Neuroendocrinology 28: , Weiss JM: Effects of coping responses on stress. J Comp Physiol Psychol 65: , Weiss JM: Somatic effects of predictable and unpredictable shock. Psychosom Med 32: , Weiss JM: Effects of coping behavior with and without a feedback signal on stress pathology in rats. J Comp Physiol Psychol 77:22-30, Murison R, Isaksen E: Gastric ulceration and adrenocortical activity after inescapable and escapable preshock in rats. Scand J Psychol Suppl 1: , Robert A, Nezamis JE: Histopathology of steroid-induced ulcers: An experimental study in the rat. Arch Pathol 77: , Allen A, Garner A: Mucous and bicarbonate secretion in the stomach and their possible role in mucosal protection. Gut 21: , Ackerman SH, Hofer MA, Weiner H: Early maternal separation increases gastric ulcer risk in rats by producing a latent thermoregulatory disturbance. Science 201: , Ackerman SH, Hofer MA, Weiner H: Sleep and temperature regulation during restraint stress in rats is affected by prior maternal separation. Psychosom Med 41: , Ackerman SH: Premature weaning, thermoregulation and the occurrence of gastric pathology. In Weiner H, Hofer MA, Stunkard AJ (eds), Brain, Behavior and Bodily Disease. New York, Raven Press, 1981, Ader R, Tatum R, Beels CC: Social factors affecting emotionality and resistance to disease in animals. 1. Age of separation from the mother and susceptibility to gastric ulcers in the rat. J Comp Physiol Psychol 53: , Ader R: Social factors affecting emotionality and resistance to disease in animals. III. Early weaning and susceptibility to gastric ulcers in the rat. A control for nutritional factors. J Comp Physiol Psychol 55: , Erdos6va R, Flandera V, Kresek J, et al: The effect of premature weaning on the sensitivity of rats to experimental erosions of the gastric mucosa. Physiol Bohemoslov 14: , Ackerman SH, Hofer MA, Weiner H: Age at maternal separation and gastric erosion susceptibility in the rat. Psychosom Med 37: , Ackerman SH: Disease consequences of early maternal separation. In Weiner H, Hellhammer D, Florin I, et al. (eds), Neuronal Control of Bodily Function: Frontiers of Stress Research. Toronto, Hans Huber, 1989, Pare 1 WP: Technique and strain comparisons in stress ulcer. Ann NY Acad Sci 597: , Pare WP, Houser VP: Activity and food restriction effects on gastric glandular lesions in the rat: The activity-stress ulcer. Bull Psychon Soc 2: , Garrick T, Minor RT, Buack S, et al: Predictable and unpredictable shock stimulates gastric contractility and causes mucosal injury in rats. Behav Neurosci 103: , Grijalva CV, Roland B: The involvement of the hypothalamus in the production of stomach ulceration. In Weiner H, Hellhammer D, Florin I, et al. (eds), Neuronal Control of Bodily Function: Frontiers of Stress Research. Toronto, Hans Huber, 1989, Morrow NS, Geiselman PJ, Ohning G, et al: Thermoregulatory dysfunction in rats exposed to activity-stress. Soc Neurosci Abstr 18:491, Morrow NS, Garrick T: Effects of preadaptation to restricted feeding and cimetidine treatment on gastric mucosal injury and wheel running during activity-stress. J. Psysiol. (Paris) 87: , Morrow NS, Keifer SW: Changes in the electrogastrogram and the onset of excessive wheel running in rats during activity stress. Soc Neurosci Abstr 15:273, Tran TA, Gregg RV: Hypothermia in restraint-induced gastric ulcer in parabiotic rats. Gastroenterology 67: , Levine RJ, Senay EC: Studies on the role of acid in the pathogenesis of experimental stress ulcers. Psychosom Med 32:61-65, Witty RT, Long JF: Effect of ambient temperature on gastric secretion and food intake in the rat. Am J Physiol 219: ,1970 Psychosomatic Medicine 58: (1996) 539

17 H. WEINER 48. Levine RJ, Senay EC: Histamine in the pathogenesis of stress ulcers in the rat. Am J Physiol 24: , Muramalsu M, Arai I, Tamaki J, et al: Central regulation of gastric acetylcholine metabolism and acid output. Analysis using stress and 2-deoxy-D-glucose administration in rats. Neurochem Int 8: , Menguy R: Effects of restraint stress on gastric secretion in the rat. Dig Dis Sci 5: , Watanabe K: Some pharmacological factors involved in formation and prevention of stress ulcer in rats. Chem Pharm Bull 14: , Pare 1 WP, Livingston A: Shock predictability and gastric secretion in the chronic gastric fistula rat. Physiol Behav 11: , Garrick T, Buack S, Bass P: Gastric motility is a major factor in cold restraint-induced lesion formation in rats. Am J Physiol 25O:G191-G199, Garrick T, Goto Y, Buack S, et al: Cimetidine and ranitidine protect against cold restraint induced ulceration in the rat by suppressing gastric acid secretion. Dig Dis Sci 32: , Mersereau WA, Hinchey EJ: Synergism between acid and gastric contractile activity in the genesis of ulceration and hemorrhage in the phenylbutazone-treated rat. Surgery 901: , Mersereau WA, Hinchey EJ: Hypothermia-induced gastric hyper- contractility in the genesis of the restraint ulcer. Can J Surg 24: , Yano S, Matsukura H, Shibata M, et al: Stress procedures lowering body temperature augment gastric motility by increasing the sensitivity to acetylcholine in rats. J Pharmacobiodyn 5: , Grijalva CV, Garrick TR, Jacobs K: Lateral hypothalamic lesions are a.ssociated with stimulation of gastric contractility in the rat. Soc Neurosci Abtr 15:632, Yano S, Akahane M, Harada M: Role of gastric motility in development of stress-induced gastric lesions of rats. Jpn J Pharmacol 28: , Hase T, Moss B: Microvascular changes of gastric mucosa in the development of stress ulcer in rats. Gastroenterology 65: , Sakaguchi Y, Nakamura N, Nagasu T, et al: Gastric mucosal blood flow in the onset of stress-induced ulcers gastric mucosal blood flow, gastric wall microvascular structure and norepinephrine in the gastric wall. In Kasuya Y, Tsuchiya M, Nagao F, et al. (eds), Gastrointestinal Function Regulation and Disturbances. New York, Elsevier, 1983, Livingstone EH, Garrick TR, Scremin U, et al: Heterogenous distribution of gastric mucosal bloodflow with restraint stress in the rat. Dig Dis Sci 38: , Gnrrick T, Leung FW, Buack S, et al: Gastric motility is stimulated but overall blood flow is unaffected during cold restraint in the rat. Gastroenterology 91: , Mersereau WA, Hinchey EJ: Relationship between the gastric myoelectric and mechanical activity in the genesis of ulcers in indomethacin-insulin treated rats. Dig Dis Sci 33: , Wingate DL: Backwards and forwards with the migrating complex. Dig Dis Sci 26: , Takeuchi K, Furukawa O, Okabe S: Induction of duodenal ulcers in rats under water-immersion stress conditions. Influence of stress on gastric acid and duodenal alkaline secretion. Gastroenterology 91: , Tache Y, Vale W, Brown M: Thyrotropin-releasing hormone- CNS action to stimulate gastric acid secretion. Nature (Lond) 287: , TacW Y, Vale W, Rivier J, et al: Brain regulation of gastric secretion: Influence of neuropeptides. Proc Natl Acad Sci USA 77: , Maeda-Hagiwara M, Watanabe H, Watanabe K: Central nervous system action of TRH to stimulate gastric emptying in rats. Regul Pept 17: , Tache Y, Lesiege D, Vale W, et al: Gastric hypersccretion by intracisternal TRH: Dissociation from hypophysiotropic activity and role of central catecholamine. Eur J Pharmacol 107: , Tache Y, Maeda-Hagiwara M, Goto Y, et al: Central nervous system action of TRH to stimulate gastric function and ulceration. Peptides 9(Suppl):9-13, Tache Y, Goto Y, Hamel D, et al: Mechanisms underlying intracisternal TRH-induced stimulation of gastric acid secretion in rats. Regul Pept 13:31-30, Bagarani M, Goto Y, Debas H, et al: Effect of TRH and TRH antiserum on ulcer formation [Abstract]. Dig Dis Sci 29:7S, Maeda-Hagiwara M, Watanabe H: Intracerebroventricular injection of a TRH analogue, y-butyrolactone-y-carbonyl-i.- histidyl-prolinamide. induces gastric lesions and gastric acid stimulation in rats. Naunyn-Schmiedeberg's Arch Pharmacol 330: , Nakane T, Kanie N, Audhya T, et al: The effects of centrally administered neuropeptides on the development of gastric lesions in the rat. Life Sci 36: , Arancibia S, Tapia-Arancibia L, Assenmacher I, et al: Direct evidence of short-term cold-induced TRH release in the median eminence of unanesthetized rats. Neuroendocrinology 37: , Arancibia S, Assenmacher I: Secretion de TRH dans le troisieme ventricle cerebral lors de l'exposition aigue froid chez le rat on anesthesie. Effet des drogues adrenergiques. CR Soc Biol (Paris) 181: , Arancibia S, Tapia-Arancibia L, Astier H, et al: Physiological evidence for a,-adrenergic facilitatory control of the coldinduced TRH release in the rat, obtained by push-pull cannulation of the median eminence. Neurosci Lett 100: , Bowers CY, Serially AV, Enzmann F, et al: Porcine thyrotropin releasing hormone is (pyro)glu-his-pro(nh 2 ). Endocrinology 86: , Hefco E, Krulich L, Illinew P, et al: Effect of acute exposure to cold on the activity of the hypothalamic-pituitary-thyroid system. Endocrinology 97: , Kubek MJ, Rea MA, Hodes ZI, et al: Quantitation and characterization of thyrotropin-releasing hormone in vagal nuclei and other regions of the medulla oblongata of the rat. J Neurochem 40: , Mantyh PW, Hunt SP: Thyrotropin-releasing hormone (TRH) receptors. Localization by light microscopic autoradiography in rat brain using [ 3 H][3-Me-His 2 ]TRH as the radioligand. J Neurosci 5: , Manaker S, Winokur A, Rostene WH, et al: Autoradiographic localization of thyrotropin-releasing hormone receptors in the rat central nervous system. J Neurosci 5: , Palkovits M, Mezey E, Eskay RL, et al: Innervation of the nucleus of the solitary tract and the dorsal vagal nucleus by thyrotropin-releasing hormone-containing raphe neurons. Brain Res 373: , Hornby PJ, Rossiter CD, Pineo SV, et al: Thyrotropin-roleas- 540 Psychosomatic Medicine 58: (1996)

18 PEPTIC ULCER ULCERATION/ANIMAL MODELS ing hormone: Distribution in the vagal nuclei of the cat and physiological effect upon microinjection. Am J Physiol 257: G454-G462, 1989 Rinamin L, Miselis RR: Ultrastructural localization of thyrotropin-releasing hormone-like immunoreactivity (TRH-LI) in the dorsal vagal complex in the rat. Dig Dis Sci 34:985, 1989 Scharoun SL, Barone FC, Wayner MJ, et al: Vagal and gastric connections to the central nervous system determined by the transport of horseradish peroxidase. Brain Res Bull 13: ,1984 Shapiro RE, Miselis RR: The central organization of the vagus nerve innervating the stomach of the rat. J Comp Neurol 238: , 1985 Somiya H, Tonoue T: Neuropeptides as central integrators of autonomic nerve activity: Effects of TRH, SRIF, VIP and bombesin on gastric and adrenal nerves. Regul Pept 9:47-52, 1984 Garrick T, Buack S, Veiseh A, et al: Thyrotropin-releasing hormone (TRH) acts centrally to stimulate gastric contractility in rats. Life Sci 40: , 1987 Goto Y, Tach6 Y: Gastric erosions induced by intracisternal thyrotropin-releasing hormone (TRH) in rats. Peptides 6: , 1985 Okuma Y, Osumi Y, Ishigawa T, et al: Enhancement of gastric acid output and mucosal blood flow by tripeptide thyrotropin releasing hormone microinjected into the dorsal motor nucleus of the vagus in rats. Jpn J Pharmacol 43: , 1987 Rogers RC, Hermann GE: Dorsal medullary oxytocin, vasopressin.oxytocin antagonist, and TRH effects on gastric acid secretion and heart rate. Peptides 6: , 1985 Stephens RL, Ishikawa T, Weiner H, et al: TRH analogue, RX 77368, injected into dorsal vagal complex stimulates gastric secretion in rats. Am J Physiol 254:G639-G643, 1988 Tache' Y, Goto Y, Lauffenburger M, et al: Potent central nervous system action of p-glu-his-(3,3'-dimethyl)pro NH 2, a stabilized analog of TRH, to stimulate gastric secretion in rats. Regul Pept 8:71-78, 1984 Ishikawa T, Yang H, Tache Y: Medullary sites of action of the TRH analogue, RX 77368, to stimulate gastric acid secretion in the rat. Gastroenterology 95: , 1988 Stephens RL, Garrick T, Weiner H, et al: Endogenous serotonin produces an inhibitory tone on vagally stimulated gastric function. Ann NY Acad Sci 597: , 1990 Johansson O, HSkfelt T, Pernow B, et al: Immuno-histochemical support for three putative transmitters in one neuron: coexistence of 5-hydroxy-tryptamine, substance P, and thyrotropin releasing hormone-like inmunoreactivity in medullary neurons projecting to the spinal cord. Neuroscience 6: , 1981 Shockley RA, LePard KJ, Stephens RL Jr: Fluoxetine pretreatment potentiates intracisternal TRH analogue-stimulated gastric acid secretion in rats. Regul Pept 38: , McTigue DM, Rogers RC, Stephens RL Jr: Thyrotropinreleasing hormone analogue and serotonin interact within the dorsal vagal complex to augment gastric acid secretion. Neurosci Lett 144:61-64, Mohammed JR, Saska TA, Chi J, et al: Stimulation of the nucleus raphe' obscurus produces marked serotonin release into the dorsal medulla of fed but not fasted rats glutamatergic dependence. Brain Res 695: , 1995 Ahlman H, Demagistris L, Zinner M, et al: Release of immunoreactive serotonin into the feline gut lumen in response to vagal nerve stimulation. Science 213: , Ahlman H, Dahlstrom A: Vagal mechanisms controlling serotonin release from the gastrointestinal tract and pyloric motor function. J Auton Nerv Syst 9: , Cho CH, Hung KM, Ogle CW: The aetiology of gastric ulceration induced by electrical vagal stimulation in rats. Eur J Pharmacol 110: , Stephens RL, Tach6 Y: Intracisternal injection of TRH analogue, RX 77368, stimulates serotonin release into the gastric lumen in the rat. Am J Physiol 256:G377-G383, Stephens RL, Garrick T, Weiner H, et al: Serotonin depletion potentiates gastric secretory and motor responses to vagal and not peripheral gastric stimulants. J Pharmacol Exp Ther 251: , Canfield SP, Spencer JE: Inhibition of acid secretion in the rat isolated stomach preparation by 5-hydroxytryptamine. Int J Tissue React 5: , Cho CH, Ogle CW: The inhibitory action of 5-hydroxytryptamine on gastric secretory function in rats. Br J Pharmacol 87: , Bech K, Andersen D: Effect of serotonin on pentagastrinstimulated gastric acid secretion and gastric antral motility in dogs with gastric fistula. Scand J Gastroenterol 20: , Bech K: Effect of serotonin on bethanechol-stimulated gastric acid secretion and gastric antxal motility in dogs. Scand J Gastroenterol 21: , Yang H, Stephens RL, Tach<5 Y: Medullary sites of action for TRH to stimulate gastric luminal release of serotonin (5-HT) in the rat [Abstract]. Gastroenterology 96:A557, McCann MJ, Hermann GE, Rogers RC: Thyrotropin-releasing hormone: Effects on identified neurons of the dorsal vagal complex. J Autonom Nerv Syst 26: , Rogers RC, McCann MJ: Effects of TRH on the activity of gastric inflation-related neurons in the solitary nucleus in the rat. Neurosci Lett 104:71-76, Wei JY, Tachg Y: Alterations of efferent discharges of the gastric branch of the vagus nerve by intracisternal injection of peptides influencing gastric function in rats [Abstract]. Gastroenterology 98:A531, Yanagisawa K, Yang H, Walsh JH, et al: Role of acetylcholine, histamine and gastrin in the acid response to intracisternal injection of TRH analog, RX 77368, in the rat. Regul Pept 27: , Stephens Jr. RL: Disparate effects of intracisternal RX and ODT8-SS on gastric acid and serotonin release: Role of adrenal catecholamines. Regul Pept 36:21-28, Fisher DA, Brown MR: Somatostatin analog: Plasma catecholamine suppression mediated by the central nervous system. Endocrinology 107: , Garrick T: The role of gastric contractility and brain thyrotropin-releasing hormone in cold restraint-induced gastric mucosal injury. Ann NY Acad Sci 597:51-70, Garrick T. Stephens R, Ishikawa T, et al: Medullary sites for TRH analogue stimulation of gastric contractility in the rat. Am J Physiol 256:GlOll-GlO15, Yang H, Trauner M, Livingston E, et al: TRH analogue, RX 77368, microinjected into the raphfi pallidus and obscurus stimulates gastric contractility in rats [Abstractl. Gastroenterology 98:A553, Lynn RB, Han J, Feng H.-S., et al: TRH injection into the dorsal motor nucleus of the vagus (DMV) increases gastric Psychosomatic Medicine 58: (1996) 541

19 H. WEINER acid output imd the force of gastric contractions in anesthetized cats [Abstract]. Gastroenterology 94:A Mendez M, Joseph-Bravo P, Cisneros M, et al: Regional distribution of in vitro release of thyrotropin releasing hormone in rat brain. Peptides 8: , Koch Y, Goldhaber G, Fireman I, et al: Suppression of prolactin and thyrotropin secretion in the rat by antiserum to thyrotropin-releasing hormone. Endocrinology 100: , Hernandez DE, Jennes L, Emerick SG: Inhibition of gastric acid secretion by immunoneutralization of endogenous brain thyrotropin-releasing hormone. Brain Res 401: , Mori M, Kobayashi I, Wakabayashi K: Suppression of serum thyrotropin (TSH) concentrations following thyroidectomy and cold exposure by passive immunization with antiserum to thyrotropin-releasing hormone (TRH) in rats. Metabolism 27: , Szabo M, Frohman LA: Suppression of cold-stimulated hyrotropin secretion by antiserum to thyrotropin-releasing hormone. Endocrinology 101: , Basso N, Bagarani M, Pekary E, et al: Role of thyrotropinreleasing hormone in stress ulcer formation in the rat. Dig Dis Sci 33: , White RL, Rossiter CD, Hornby PJ, et al: Excitation of neurons in the medullary raphe results in increases in gastric acid and pepsin production. Am J Physiol, 260:G91- G96, White RL, Rossiter CD, Hornby PJ, et al: Raphe obscurus stimulation produces increases in gastric secretion and motility. Soc Neurosci Abstr 15:661, Yang H, Ishikawa T, Tache Y: Microinjection of TRH into the raph6 pallidus (RP) stimulates gastric acid and serotonin (5-HT) secretion in the rat [Abstract], Gastroenterology 96: A557, Maeda-Hagiwara M, Watanabe H: Inhibitory effects of intrahypothalamic injection of calcitonin on TRH-stimulated gastric acid secretion in rats. Jpn J Pharmacol 39: , Morrow NS, Novin D, Garrick T: Microinjection of thyrotropin-realeasing hormone in the paraventricular nucleus of the hypothalmus stimulates gastric contractivity. Brain Res 644: , Morrow NS, Hodgson D, Garrick T: Microinjection of thyrotropin-releasing hormone in the paraventricular nucleus of the hypothalamus stimulates gastric contractility. Soc Neurosci Abstr 644: , Vale W, Spiess J, Rivier C, et al: Characterization of a 41-residue ovine hypothalamic peptide that stimulates secretion of corticotropin and beta-endorphin. Science 213: , Spiess J, Rivier J, Rivier C, et al: Primary structure of corticotropin-releasing factor from ovine hypothalamus. Proc Natl Acad Sci USA 78: , Shibahara S, Morimoto Y, Furutani Y, et al: Isolation and sequence-analysis of the human corticotropin-releasing factor precursor gene. EMBO J 2: , Bugnon C, Fellman D, Gouget A, et al: Corticoliberin in rat brain: Immunocytochemical identification of a novel neuroglandular system. Neurosci Lett 30:25-30, Merchenthaler I, Vigh S, Petrusz P, et al: Immunocytochemical localization of corticotropin-releasing factor (CRF) in the rat brain. Am J Anat 165: , Swanson LW, Sawchenko PE, Rivier J, et al: Organization of ovine corticotropin-releasing factor immunoreactive cells and fibers in the rat brain: An immunohistochemical study. Neuroendocrinology 36: , De Souza EB, Perrin MH, Insel TR, et al: Corti- cotropinreleasing factor receptors in rat forebrain: Auto-radiographic identification. Science 224: , Merchenthaler I: Corticotropin-releasing factor (CRF)-like immunoreactivity in the rat central nervous system: Extrahypothalamic distribution. Peptides 5:53-69, Udelsman R, Harwood JP, Millan MA, et al: Functional corticotropin- releasing factor receptors in the primate peripheral sympathetic nervous system. Nature (Lond) 319: , Skofitsch F, Hamill GS, Jacobowitz DM: Capsaicin depletes corticotropin-releasing factor-like immunoreactive neurons in the rat spinal cord and medulla oblongata. Neuroendocrinology 38: , Krukoff TL: Segmental distribution of corticotropin-releasing factor-like and vasoactive intestinal peptide-like immunoreactivity in presumptive sympathetic preganglionic neurons of the cat. Brain Res 382: , Sawchenko PE, Swanson LW: Immunohistochemical identification of neurons in the paraventricular nucleus of the hypothalamus that project to the medulla or to the spinal cord in the rat. J Comp Neurol 205: , Schipper J, Steinbusch HWM, Vermes I, et al: Mapping of CRF-immunoreactive nerve fibers in the medulla oblongata and spinal cord of the rat. Brain Res 267: , Kruseman ACN, Linton EA, Rees LH, et al: Corticotropinreleasing factor immunoreactivity in human gastrointestinal tract. Lancet 2: , Hashimoto K, Murakami K, Hattori, T, et al: Corticotropinreleasing factor (CRF)-like immunoreactivity in the adrenal medulla. Peptides 5: , Petrusz P, Merchenthaler I, Ordronneau P, et al: Corticotropin-releasing factor (CRF)-like immunoreactivity in the gastro-entero-pancreatic endocrine system. Peptides 5:71-78, Wolter HJ: Corticotropin-releasing factor is contained within perikarya and nerve fibres of rat duodenum. Biochem Biophys Res Commun 122: , Gibbs DM, Vale W: Presence of corticotropin releasing factor-like immunoreactivity in hypophysial portal blood. Endocrinology 11: , Sasaki A, Liotta AS. Luckey MM, et al: Immunoreactive corticotropin-releasing factor is present in human maternal plasma during the third trimester of pregnancy. J Clin Endocrinol Metab 59: , Plotsky PM, Vale W: Hemorrhage-induced secretion of corticotropin-releasing factor-like immunoreactivity in the rat hypophysial portal circulation and its inhibition by glucocorticoids. Endocrinology 114: , Williams CL, Peterson JM, Villar RG, et al: Corticotropinreleasing factor directly mediates colonic responses to stress. Am J Physiol 253:G582-G586: Chappell PB, Smith MA, Kilts CD, et al: Alterations in corticotropin-releasing factor-like immunoreactivity in discrete rat brain regions after acute and chronic stress. J Neurosci 6: , Nakane T, Audhya T, Kanie N, et al: Evidence for a role of endogenous corticotropin-releasing factor in cold, ether, immobilization and traumatic stress. Proc Natl Acad Sci USA 82: , Rivier CL, Plotsky PM: Mediation by corticotropin releasing 542 Psychosomatic Medicine 58: (1996)

20 PEPTIC ULCER ULCERATION/AN1MAL MODELS factor (CRF) of adenohypophysial hormone secretion. Annu Rev Physiol 48: , Rivier CL, Rivier J, Vale W: Inhibition of adrenocorticotropic hormone secretion in the rat by immunoneutralization of corticotropin-releasing factor. Science 218: , Brown MR, Fisher LA, Spiess J, et al: Corticotropin-releasing factor: Actions on the sympathetic nervous system and metabolism. Endocrinology 111: , Brown MR, Fisher LA: Central nervous system effects of corticotropin-releasing factor in the dog. Brain Res 280:75-79, Brown MR, Fisher LA: Brain peptide regulation of adrenal epinephrine secretion. Am J Physiol 247:E41-E46, Rivier J, Rivier C, Vale W: Synthetic competitive antagonists of corticotropin-releasing factor: Effect on ACTH secretion in the rat. Science 224: , Krahn DD, Gosnell BA, Grace M, et al: CRF antagonist partially reverses CRF- and stress-induced effects on feeding. Brain Res Bull 17: , Rivier C, Rivier J, Vale W: Stress-induced inhibition of reproductive functions: Role of endogenous corticotropinreleasing factor. Science 231: , Lenz HJ: Extrapituitary effects of corticotropin-releasing factor. Horm Metab Res 16(Suppl):17-23, Brown MR, Gray TS, Fisher LA: Corticotropin-releasing factor receptor antagonist: Effects on the autonomic nervous systems and cardiovascular function. Regul Pept 16: , TachiS Y, Goto Y, Gunion MW, et al: Inhibition of gastric acid secretion in rats by intracerebral injection of corticotropinreleasing factor. Science 222: , Gunion MW, TachS Y: Gastric mucosal damage is inhibited by intraventromedial hypothalamic corticotropin-releasing factor. Soc Neurosci Abstr 12:644, Krahn DD, Wright B, Billington CJ, et al: Exogenous corticotropin-releasing factor inhibits stress-induced gastric ulceration. Soc Neurosci Abstr 12:1063, Sapolsky RM, Krey LC, McEwen BS: The adrenocortical stress-response in the aged male rat: Impairment of recovery from stress. Exp Gerontol 18:55-64, Bakke HK, Murison R: Plasma corticosterone and restraint induced gastric pathology: Age-related differences after corticotropin releasing factor. Life Sci 45: , TachS Y, Hamel D, Gunion M: Inhibition of gastric acid secretion in rats by intracisternal or intrathecal injection of rat corticotropin-releasing factor (rcrf). Dig Dis Sci 29:86S, Tach6 Y, Goto, Y, Gunion MW, et al: Inhibition of gastric acid secretion in rats and dogs by corticotropin-releasing factor. Gastroenterology 86: , Tach6 Y, Gunion M: Corticotropin-releasing factor: Central action to influence gastric secretion. Fed Proc 44: , Lenz HJ, Hester SE, Brown MR: Corticotropin-releasing factor: Mechanisms to inhibit gastric acid secretion in conscious dogs. J Clin Invest 75: , Lenz HJ, Raedler A, Greten H, et al: CRF initiates biological actions within the brain that are observed in response to stress. Am J Physiol 252:R34-R39: Stephens RL, Yang H, Rivier J, et al: Intracisternal injection of CRF antagonist blocks surgical stress-induced inhibition of gastric secretion in the rat. Peptides 9: , Lenz HJ, Raedler A, Greten H, et al: Stress-induced gastrointestinal secretory and motor responses in rats are mediated by endogenous corticotropin-releasing factor. Gastroenterology 95: , Lenz HJ: Neuroendocrine regulation of canine gastric secretion, emptying and blood flow. J Neuroendocrinol 1: , Driige G, Raedler A, Greten H, et al: Pathways mediated CRF-induced inhibition of gastric acid secretion in rats. Am J Physiol 256:G214-G219, Lenz HJ, Driige G: Neurohumoral pathways mediating stressinduced inhibition of gastric acid secretion in rats. Gastroenterology, 98: Lenz HJ, Fisher LA, Vale WW, et al: Corticotropin-releasing factor, sauvagine, and urotensin. I. Effects on blood flow. Am J Physiol 249:R85-R90, Rivier C, Vale W: Interaction of corticotropin-releasing factor and arginine vasopressin on adrenocorticotropin secretion in vivo. Endocrinology 113: , Gunion MW, Tache Y: Intrahypothalamic microinfusion of corticotropin-releasing factor inhibits gastric acid secretion but increases secretion volume in rats. Brain Res 411: , Gunion MW, Kauffman GL, Tach6 Y: Intrahypothalamic micro-infusion of corticotropin-releasing factor elevates gastric bicarbonate secretion and protects against cold-stress ulceration in rats. Am J Physiol 258:G152-G157, Flemstrom G: Gastric and duodenal mucosal bicarbonate secretion. In Johnson LR (ed), Physiology of the Gastrointestinal Tract, Vol 2, New York, Raven Press, 1987, Isenberg JI, Selling JA, Hogan DL, et al: Impaired proximal duodenal mucosal bicarbonate secretion in duodenal ulcer patients. N Engl J Med 316: , Nylander O, Flemstrom G, Delbro D, et al: Vagal influence on gastroduodenal HCOJ secretion in the cat in vivo. Am J Physiol 252:G522-G528: Konturek SJ, Thor P, Bilski J, et al: Cephalic phase of gastroduodenal alkaline secretion. Am J Physiol 252:G742- G747, Lenz HJ, Vale WW, Rivier JE: TRH-induced vagal stimulation of duodenal HCO3 mediated by VIP and muscarinic pathways. Am J Physiol 257:G677-G682, Lenz HJ, Forquignon. I. Stimulation of duodenal bicarbonate secretion in conscious rats by cerebral somatostatin-28: Role of neurohumoral pathways. Gastroenterology 99: , Fandriks L, Jonson C, Nylander O: Effects of splanchnic nerve stimulation and of clonidine on gastric and duodenal HCO^ secretion in the anesthetized cat. Acta Physiol Scand 130: , Jonson C, Fandriks L: Splanchnic nerve stimulation inhibits duodenal HCOJ secretion in the rat. Am J Physiol 255: G709-G712, Flemstrom G, Jedstedt G 1 Stimulation of duodenal mucosal bicarbonate secretion in the rat by brain peptides. Gastroenterology 97: , Lenz HJ: Regulation of duodenal bicarbonate secretion during stress by corticotropin-releasing factor and fi-endorphin. Proc Natl Acad Sci USA 86: , Gue M, Fioramonti J, Bueno L: Comparative influences of acoustic and cold stress on gastrointestinal transit in mice. Am J Physiol 253:G124-G128: Bueno L, Gue M: Evidence for the involvement of corticotropin-releasing factor in the gastrointestinal disturbances induced by acoustic and cold stress in mice. Brain Res 441:1-4, 1988 Psychosomatic Medicine 58: (1996) 543

21 H. WEINER 198. Lenz HJ, Burlage M, Raedler A, etal: Central nervous system 217. Shobitz B, De Kloet ER, Holsboer F: Gene expression and effects of corticotropin-releasing factor on gastrointestinal function of interleukin I, interleukin 6 and tumor necrosis transit in the rat. Gastroenterology 94: , 1988 factor in the brain. Progr Neurobiol 44: , Lenz HJ: Neurohumoral pathways mediating stress-induced 218. Rothwell NJ, Dantzer R (eds): Interleukin-I in the Brain, changes in rat gastrointestinal transit. Gastroenterology 97: Oxford, Pergamon Press, , Minami M, Kuraishi Y, Yamaguchi T, et al: Immobilization 200. Williams CL, Villar RG, Peterson JM, et al: Stress-induced stress induces interleukin-1/3 mrna in the rat hypothalamus. Neurosci Lett 123: , 1991 changes in intestinal transit in the rat: A model for the irritable bowel syndrome. Gastroenterology 94: , 220. Tache' Y, Saperas ES: Potent inhibition of gastric acid secretion and ulcer formation by centrally and peripherally ad Tache Y, Maeda-Hagiwara M, Turkelson CM: Central nervous system action of corticotropin-releasing factor to in ministered interleukin-i. Ann NY Acad Sci 659: , hibit gastric emptying in rats. Am J Physiol 253:G241-G245, 221. Suto G, Kiraly A, Tach6 Y: Interleukin-I/3 acts in the brain to 1987 inhibit gastric emptying in rats: Mediation through prostaglandin and corticotropin releasing factor. Gastroenterology 202. Bueno L, Fargeas MJ, Gue M, et al: Effects of corticotropinreleasing factor on plasma motilin and somatostatin levels 106: , 1994 and gastrointestinal motility in dogs. Gastroenterology 91: 222. Robert A, Saperas ES, Zhang W, et al: Gastric cytoprotection , 1986 by intracisternal interleukin-i/3 in the rat. Biochem Biophys 203. Garrick T, Veiseh A, Sierra A, et al: Corticotropin-releasing Res Commun 174: , 1991 factor acts centrally to suppress stimulated gastric contrac Morrow NS, Quinonez G, Weiner H, et al: Interleukin-I/3 in tility in the rat. Regul Pept 21: , 1988 the dorsal vagal complex inhibits TRH analogue-induced 204. Heymann-Monnikes I, Tache Y, Trauner M, et al: CRF stimulation of gastric contractility. Am J Physiol 269:G196- microinjected into the dorsal vagal complex inhibits TRH G202, 1995 analog- and kainic acid-stimulated gastric contractility in 224. Henke PG: Recent studies of the central nucleus of the rats. Brain Res 554: , 1991 amygdala and stress ulcers. Neurosci Biobehav Rev 12: Fisher LA, Brown MR: Corticotropin-releasing factor and 150,1988 angiotensin II: Comparison of CNS actions to influence 225. Henke PG: Limbic system modulation of stress ulcer development Ann NY Acad Sci 597: , 1990 neuroendocrine and cardiovascular function. Brain Res 296: 41-47, Morrow NS, Grijalva CV, Geiselman PJ, et al: Effects of 206. Moody TW, O'Donohue TL, Jacobowitz DM: Biochemical amygdaloid lesions on gastric erosion formation during localization and characterization of bombesin-like peptides exposure to activity stress. Physiol Behav 53: , in discrete regions of rat brain. Peptides 2:75-79, Zarbin MA, Kuhar MJ, O'Donohue TL, et al: Autoradio Tsuda A, Tanaka M: Neurochemical characteristics of rats graphic localization of ( 125 I-Tyr4) bombesin-binding sites in exposed to activity stress. Ann NY Acad Sci 597: , rat brain. J Neurosci 5: , Cowan A, KhunawarP, ZhuXZ, etal: Effects of bombesin on 228. Tanaka M, Tsuda A, Yokoo H, et al: Involvement of the brain behavior. Life Sci 37: , 1985 noradrenaline system in emotional changes caused by stress 209. Tach6 Y, Brown M, Collu R: Central nervous system actions in rats. Ann NY Acad Sci 597: , 1990 of bombesin-like peptides. In Collu R, Ducharme C Jr, 229. Grijalva CV, Novin D: The role of the hypothalamus and Barbeau A, et al. (eds), Brain Neurotransmitters and Hormones. New York, Raven Press, 1982, physiology. Ann NY Acad Sci 59: , 1990 dorsal vagal complex in gastrointestinal function and patho Tache Y, Ishikawa T, Gunion M, et al: Central nervous 230. Henke PG, Sullivan RM, Ray A: Interactions of thyrotropinreleasing hormone with neurotensin and dopamine in the system action of bombesin to influence gastric secretion and ulceration. Ann NY Acad Sci 547: , 1988 central nucleus of the amygdala during stress ulcer formation in rats. Neurosci Lett 91:95-100, Gunion MW, TachS Y: Bombesin microinfusion into the paraventriculai nucleus suppresses gastric acid secretion in 231. Hernandez DE: The role of brain peptides in the pathogenesis of experimental stress gastric ulcers. Ann NY Acad Sci the rat. Brain Res 422: , Ishikawa, T, Tache Y: Bombesin microinjected into the 597:28-35, 1990 dorsal vagal complex inhibits vagally stimulated gastric acid 232. Tach6 Y, Yang H: Brain regulation of gastric acid secretion secretion in the rat. Regul Pept 24: , 1989 by peptides. Ann NY Acad Sci 597: , Heymann-Monnikes I, Livingston EH, Tache Y, et al: Bomb Johnson LR, Guthrie PD: Mucosal DNA synthesis: A short esin microinjected into the dorsal vagal complex inhibits term index of the trophic action of gastrin. Gastroenterology TRH-stimulated gastrin contractility in rats. Brain Res 533: 67: , , Solomon TE: Trophic effects of pentagastrin on gastrointestinal tract in fed and fasted rats. Gastroenterology 91: Stephens RL Jr, Horner P, Drapeau G: N-Acetyl-GRP (20-26J-0-CH3 reverses intracisternal bombesin-induced inhibition of gastric acid secretion in rats. Peptides 12: , 235. Thirlby RC, Feldman M: Effect of chronic sham feeding on 116, maximal gastric acid secretion in the dog. J Clin Invest 215. Hopkins SJ, Rothwell NJ: Cytokines and the nervous system. 73: , 1984 I. Expression and recognition. Trends Neurosci 18:83-88, 236. Todaro GJ, Fryling C, De Laro JE: Transforming growth 1995 factors produced by certain human tumor cells: Polypeptides that interact with epidermal growth factor receptors Rothwell NJ, Hopkins SJ: Cytokines and the nervous system. II. Actions and mechanisms of action. Trands Neurosci Proc Natl Acad Sci USA 77: , : , Olsen PS, Poulsen SS, Therkelsen K, et al: Oral administra- 544 Psychosomatic Medicine 58: (1996)