Comparison of the effects of bacterial lipopolysaccharide and muramyl dipeptide on food intake

Tolerance to appetite suppression induced by peptidoglycan

Physiologically realistic peptidoglycan (PG) fragments, derived from Neisseria gonorrhoeae, were shown previously to dose-dependently suppress food consumption and body weight gain in rats following single intraperitoneal injections. The present study, examining the effects of repeated daily injection of PG, provides additional support to our underlying hypothesis, i.e., that soluble PG fragments contribute to the loss of appetite commonly associated with bacterial infections. An initial intraperitoneal injection of purified, soluble, macromolecular, extensively O-acetylated PG fragments (S-O-PG) (240 micrograms/kg of body weight) decreased overnight food consumption in male Lewis rats (150 g) by approximately 35% relative to animals receiving diluent alone (P < 0.05). However, subsequent daily injections of S-O-PG resulted in progressively smaller effects on food consumption until, by the fourth day, rats were completely nonresponsive (tolerant) to S-O-PG-induced hypophagia. Rats that developed tolerance to the effects of S-O-PG on appetite were also tolerant to three other known hypophagic agents, lipopolysaccharide (LPS), muramyl dipeptide, and interleukin-1, when challenged one day after establishment of S-O-PG tolerance. Similarly, rats developed tolerance to repeated injections of muramyl dipeptide or LPS and were cross-tolerant to S-O-PG when challenged 1 day later. However, 30 days after establishment of S-O-PG tolerance, rats remained nonresponsive to S-O-PG but regained full responsiveness to LPS-mediated hypophagia. Thus, at least two mechanisms of tolerance to S-O-PG hypophagia exist: an early tolerance which is nonspecific and a late tolerance which is specific for S-O-PG. Late, but not early, tolerance to S-O-PG-mediated suppression of appetite was associated with an increase in specific anti-PG antibody activity as measured in an enzyme-linked immunosorbent assay.

Pavlovian conditioning of LPS-induced responses: effects on corticosterone, splenic NE, and IL-2 production

The present study used a taste aversion paradigm to condition lipopolysaccharide (LPS)-induced suppression of splenic lymphocyte interleukin-2 (IL-2) production, with concurrent measurement of corticosterone production and splenic norepinephrine (NE) content). In training, two groups of rats received saccharin and IP LPS in a paired (P) manner and a third group in a specifically unpaired (U) manner. In the test, the unpaired group (group U) and one of the paired (group P) groups were re-exposed (R) to the cue and the other not (NR). An additional group controlled for the effects of cues (conditional stimulus) and fluid deprivation (negative control; NC). A robust taste aversion in the P-R group was accompanied by suppression of IL-2 production, reduced splenic NE content, and elevated corticosterone production, relative to combined controls (i.e., groups U-R, P-NR, and NC). The conditioned modulation of IL-2 secretion, along with the concomitant alteration of adrenocortical and sympathetic mediators, supports the involvement of bidirectional central nervous-immune system pathways in this paradigm.

Bacterial products and the control of ingestive behavior: clinical implications

Bacterial products such as lipopolysaccharides (LPS) and muramyl peptides are delivered in the course of infections. They trigger the host's acute phase responses to bacterial infections and are probably involved in the accompanying hypophagia because LPS and muramyl dipeptide (MDP, the minimal immunologically active muramyl peptide) reduce food intake after parenteral administration in animals. LPS and MDP inhibit feeding synergistically through separate but interacting mechanisms. The hypophagic effects of LPS and MDP are presumably mediated by the combined actions of interleukin-1, tumor necrosis factor, and other cytokines. More work is required to understand the interactions between these cytokines, and between bacterial products and cytokines, before cytokine antagonists can be used for treatment of the hypophagia during bacterial infections. As the hypophagia seems to be an early mechanism of host defense, a treatment should be carefully considered. If an intervention is indicated because of a patient's poor condition, inhibitors of eicosanoid synthesis and glucocorticoids may hold more promise for therapy because such substances block LPS and MDP hypophagia. Although LPS can reduce food intake by direct action on the brain, presently available evidence indicates that systemic LPS acts primarily in the periphery to generate a neural signal that is transmitted to the brain and inhibits feeding through the vagus. The exact site where LPS acts on peripheral nerves remains to be identified. LPS hypophagia is conditionable, but conditioning cannot solely account for LPS hypophagia under most test conditions. Whether MDP hypophagia is also conditionable and mediated by vagal afferents is not yet known. All in all, the putative mediators and mechanisms of LPS and MDP hypophagia suggest some options for a treatment of the hypophagia during bacterial infection, but present knowledge about the mechanisms and interactions of the involved substances is still fragmentary and requires further investigation.

ICV administration of anti-NPY antisense oligonucleotide: effects on feeding behavior, body weight, peptide content and peptide release

Neuropeptide-Y (NPY) is a potent stimulator of feeding, and chronic administration of the peptide has been shown to increase body weight. This study determined the chronic effects of repeated daily injections of an antisense phosphorothioate oligonucleotide complementary to the rat mRNA for NPY (aNPY) on food intake, feeding behavior and body weight change in rats. Five micrograms of the aNPY oligonucleotide in ten microliters of vehicle or a missense control oligonucleotide were administered intracerebroventricularly (ICV) for seven consecutive days. Cumulative food intake, meal size and meal duration were significantly lowered in aNPY-treated animals. Body weight change of aNPY-injected animals was significantly lower than controls, and the effect was reversed after treatments ceased. A two-bottle taste aversion paradigm was employed to determine the behavioral specificity of the anorectic effect, and the phosphorothioate oligonucleotide was found not to be aversive at the dosage used. Following an additional five day injection period, animals were killed and paraventricular nuclei (PVN) were dissected. In vitro release and tissue content of NPY from this brain area were evaluated by heterologous radioimmunoassay. Content of NPY was unchanged in this brain area. Paradoxically, in vitro release of NPY was increased in aNPY-treated animals.

Peptidoglycan fragments decrease food intake and body weight gain in rats

We hypothesized that peptidoglycan (PG) fragments decrease appetite in rats. Male Lewis rats (150 g; n > or = 7) received intraperitoneal (i.p.) injections of purified soluble PG fragments that had been treated with polymyxin B-agarose to remove residual endotoxin. Food consumption and body weight gain were determined at intervals after injection. Single i.p. injections of macromolecular extensively O-acetylated PG (S-O-PG) and non-O-acetylated PG fragments (24 to 240 micrograms/kg) reduced food intake and body weight gain in a dose-dependent fashion during the first 12 h after injection. Low-molecular-weight disaccharide peptide monomers with nonreducing 1,6-anhydro-N-acetylmuramic acid ends and muramyl dipeptide (MDP; 1.6 mg/kg) were also appetite and weight gain suppressants, albeit at least 10-fold less potent than S-O-PG; however, muramidase-derived monomers and peptide cross-linked dimers with reducing muramic acid ends were inactive. Appetite suppression was not limited to the Lewis rat strain since another strain, F344, exhibited similar decreases in food intake after injection of S-O-PG or MDP. Oral administration of MDP or S-O-PG, at concentrations 3 and 20 times higher, respectively, than those that were active i.p., failed to elicit a hypophagic response. We conclude that soluble PG fragments are potent suppressants of food consumption and body weight gain in rats and that, although macromolecular PG is more potent than low-molecular-weight fragments, neither O-acetylation nor glycosidic linkage of PG fragments is required for activity. We speculate that PG fragments may contribute to loss of appetite during bacterial illness.

Stimulation of drinking by bacterial endotoxins in the rat

We investigated the effects of endotoxins on water balance, rectal temperature, and food intake in male Long-Evans rats with femoral venous catheters. Extracts of Escherichia coli or Salmonella minnesota, in doses ranging from 125 to 500 micrograms/kg IV, stimulated drinking and reduced urinary water loss for several hours. The net gain of 5 ml water 2 h after the lowest dose of E. coli endotoxin was sufficient to reduce plasma osmolality and sodium concentration 2 to 3%. Drinking occurred during the period of hypothermia that frequently precedes the onset of endotoxin-induced fever, so cannot be attributed to increased body temperature. Doses of endotoxin causing drinking inhibited both spontaneous and deprivation-induced feeding. The cause of the drinking is not known, but may involve mechanisms other than the known dehydrational signals controlling thirst.

Does a learned taste aversion contribute to the anorectic effect of bacterial lipopolysaccharide?

The present study addressed the possible role of a conditioned taste aversion in the anorectic effect of bacterial lipopolysaccharide (LPS) in the rat. Pairing an intraperitoneal (IP) injection of LPS (100 micrograms/kg b.wt.) with the subsequent presentation of a familiar diet (FD) or of a novel-tasting saccharin diet (SD) for several hours did not affect FD or SD intake when the same diet was offered several days later after 12 h of food deprivation. However, food intake during the second presentation of SD was reduced when food was not withheld prior to the test. In a similarly designed experiment, the antipyretic and antiinflammatory drug indomethacin (5 mg/kg b.wt., IP) attenuated the anorectic effect of LPS during the initial pairing, but did not affect the inhibition of SD intake in LPS-pretreated rats during the second feeding test. The antiemetic trimethobenzamide (5 mg/kg b.wt., IP) failed to influence the anorectic effect of LPS. Lesion of the area postrema (AP) and the adjacent nucleus of the solitary tract (NST) was found to enhance the anorectic effect of LPS, but the development of tolerance to this effect remained unchanged in AP/NST-lesioned animals. In spite of the ability of LPS to induce a taste aversion that inhibits feeding under certain conditions (novel-tasting diet, no food deprivation prior to the feeding test), the findings indicate that a learned taste aversion is not the only contributor to the anorectic effect of LPS.

Comparison of the feeding responses to bacterial lipopolysaccharide and interleukin-1 beta

To further investigate the idea that endogenous interleukin-1 plays a major role in the anorectic effect of bacterial lipopolysaccharide (LPS), feeding responses to recombinant human interleukin-1 beta (rhIL-1 beta) and LPS, as well as crosstolerance or -sensitization between both compounds, were investigated in rats. Intraperitoneally (IP) injected paracetamol (50 mg/kg body weight = b.wt.) markedly attenuated the anorectic effect of rhIL-1 beta (50,000 LAF units/kg b.wt., IP), but was clearly less effective in attenuating the anorectic effect of LPS (100 micrograms/kg b.wt., IP). As in previous experiments of ours, repeated IP injections of rhIL-1 beta (three injections of 50,000 LAF units/kg b.wt. on experimental days 1, 4, and 10) resulted in sensitization to the anorectic effect rhIL-1 beta, whereas repeated IP injections of LPS (three injections of 100 micrograms/kg b.wt. every second day) resulted in LPS-tolerance. Sensitization to the anorectic effect of rhIL-1 beta did not affect the anorectic response to LPS. Likewise, LPS-tolerance did not alter the anorectic response to rhIL-1 beta. RhIL-1 beta suppressed feeding by reducing meal frequency and meal size. In contrast, the anorectic effect of LPS was due entirely to a reduction of meal frequency. The results indicate that rhIL-1 beta and LPS do not affect feeding through exactly the same mechanism.

Enhancement or loss of the hypophagic effect of interleukin-1 upon chronic administration

To further characterize the effect of interleukin-1 on food intake, we tested whether a tolerance to the hypophagic effect of recombinant human interleukin-1 beta (rhIL-1 beta) develops with repeated injections or continuous infusion in rats. Daily intraperitoneal (IP) injections of rhIL-1 beta (25,000 LAF units/kg b.wt.) for 4 days did not result in tolerance to rhIL-1 beta's hypophagic effect. The hypophagic effect of the same dose of rhIL-1 beta actually increased if injections were given every second day, when the hypophagic effect of the preceding injection had subsided. A dose of rhIL-1 beta that usually did not affect food intake (5000 LAF units/kg b.wt.) reduced food intake if injected repeatedly. Continuous infusion of rhIL-1 beta (25,000 LAF units/kg b.wt/day) via IP-implanted osmotic minipumps caused a strong initial suppression of feeding followed by the development of tolerance to the hypophagic effect of the infused rhIL-1 beta. Nevertheless, hypophagia caused by a subsequent IP injection of rhIL-1 beta (25,000 LAF units/kg b.wt.) was enhanced. As specific antibodies to rhIL-1 beta could be detected in sera of only three of 11 rhIL-1 beta-infused rats, the observed tolerance was probably not due to a humoral immune response. The results demonstrate that, dependent on test conditions, chronic administration of rhIL-1 beta in the rat can lead to an enhancement or to a loss of its hypophagic effect. The reasons for this difference remain unclear.

Mechanisms of vasopressin's anorectic effect

The present experiments investigated the effect of vasopressin (VP) on food intake in rats under various conditions. VP (1.25-10 micrograms/kg body weight = b.wt.) injected intraperitoneally (IP) at the onset of the dark phase of the lighting cycle inhibited feeding in a dose-dependent manner. The suppression of feeding induced by VP was primarily due to a delayed onset of the first meal after injection and was reversed by a V1-receptor antagonist (7 micrograms/kg b.wt., IP), by the Ca(++)-channel blocker verapamil (5 mg/kg b.wt., IP) and by the alpha-adrenergic receptor antagonist phentolamine (500 micrograms/kg b.wt.), but not by dissection of the hepatic branch of the vagus. In further experiments VP inhibited gastric emptying. This effect was not reversed by phentolamine. VP had also an aversive effect, but this effect was weaker than that of LiCl and probably not involved in VP-induced hypophagia. The results suggest that VP reduces feeding through a V1-receptor-mediated activation of an alpha-adrenergic mechanism. The inhibition of gastric emptying or a possible stimulation of hepatic oxidative metabolism by VP seems to be not essential for VP's effect on feeding. The results are consistent with a role of VP in stress-induced anorexia in rats.