Some metabolic aspects of tolerance to bacterial endotoxin. J. Bacteriol. 90:970-977. 1965.-The tolerance to bacterial endotoxins which is produced in mice given a series of daily injections of heat-killed Salmonella typhimurium failed to occur when actinomycin D was administered with the heat-killed cells. Neither ethionine nor 2-thiouracil, when given with endotoxin, altered the development of tolerance. An injection of endotoxin, actinomycin D, or ethionine lowered the activity of the liver enzyme tryptophan pyrrolase more significantly at either 4 or 17 hr postinjection in normal mice than in tolerant mice. Similarly, an injection of either saccharated iron oxide or Thorotrast lowered liver tryptophan pyrrolase activity more extensively in normal than in tolerant animals. Activation of the reticuloendothelial system (RES) of tolerant mice, as determined by an accelerated rate of carbon clearance from the blood, was observed, but this was prevented by the appropriate dose of actinomycin D. Similar results were obtained when saccharated iron oxide, rather than endotoxin, was used to activate the RES, but these animals were not resistant to endotoxin and their tryptophan pyrrolase was normally diminished after an injection of endotoxin. Thus, RES activation may occur without tolerance developing. A more nearly normal level of enzyme activity appears to be characteristic of the tolerant state.

Bacterial endotoxins in mice reduced the induction by cortisone of two hepatic enzymes, tryptophan oxygenase, and phosphoenolpyruvate carboxykinase, they prevented the glyconeogenesis in liver induced by the same hormone, and they induced in intact animals the liver enzyme tyrosine-a-ketoglutarate transaminase, all in proportion to their LD50. When cortisone was given in the least amount (100 ,ug), it resulted in near maximal induction of tryptophan oxygenase; a smaller amount of endotoxin reduced significantly the level of enzyme than that required when 5 mg of hormone was injected. The smallest amount of endotoxin that prevented tryptophan oxygenase induction was given intravenously to adrenalectomized mice in which 25 ,ug of cortisone was administered. The amount (0.01 ,g) is 1/40,000th of the LD50. The other metabolic processes subject to alteration by endotoxin required at least 100 to 400 times as much. This property of endotoxin can serve as a sensitive bioassay, although the dose-response curve is steep.

Inhibition of inducible liver enzymes by endotoxin and actinomycin D. J. Bacteriol. 92:107-115. 1966.-Bacterial endotoxin at the LD50 level lowers liver tryptophan pyrrolase in mice, it prevents for 16 to 20 hr the induction of the enzyme by a concurrent injection of cortisone, it lowers significantly but does not prevent substrate induction, and it reduces the enzymatic activity promptly and significantly when administered during the course of hormonal induction. The LD50 amount of actinomycin D has a similar effect on tryptophan pyrrolase, except that its inhibition of induction by cortisone persists for a longer period of time. Endotoxin in the intact mouse induces tyrosine-a-ketoglutarate transaminase almost as well as cortisone, but not in the adrenalectomized animal, a fact that suggests induction of this enzyme is due to release of endogenous adrenal hormones. Actinomycin D, on the other hand, has an effect on this transaminase similar to that on tryptophan pyrrolase. The site of action of endotoxin and actinomycin D would appear to be similar for one of the two enzymes studied and different for the other, a relationship that requires a specificity difficult to imagine for a material as complex as endotoxin.

Mice of different strains were protected against the lethal effect of bacterial endotoxin by concurrent injections of cortisone. Either inadequate amounts of cortisone or excessive quantities of endotoxin voided the protection. Analyses of blood sugar, liver glycogen, muscle glycogen, and total body carbohydrate in the skinned eviscerated carcass were carried out on different strains of mice given endotoxin and/or cortisone. Poisoned animals were virtually depleted of all carbohydrate while mice given cortisone alone had concentrations of carbohydrate from three to four times that of normal mice. Mice given a lethal amount of endotoxin and a protective dose of cortisone had two to three times as much carbohydrate as animals injected with the same amount of endotoxin alone but significantly less than that found in normal mice. Dibenzyline failed to alter the lethal effect of endotoxin and to reduce the carbohydrate loss that accompanied endotoxin administration. Endotoxin, at the dosage level employed, lowered the temperature of mice 2°–3°C. during the first hour or two postinjection and the temperature remained essentially unaltered during the next 4 to 5 hours. Loss in body carbohydrate in endotoxin-poisoned mice cannot be explained, therefore, as the result of an elevated metabolic rate accompanying hyperthermia. Endotoxin prevented the conversion of injected glucose into liver glycogen but not into muscle glycogen. Mouse liver mitochondria, in the presence of endotoxin, released from ATP approximately the same amount of inorganic phosphate as that released in the presence of dinitrophenol.

The role of the adrenal cortex in an animal's response to bacterial infections and to bacterial endotoxins has been the subject of much research and speculation.Just recently, Brooke, Kass, and Hechter (1) reported that adrenalectomized rats are killed by one one-thousandth the dose of endotoxin required to kill normal animals. This dramatic observation is far superior to any other in the literature in emphasizing the protection afforded by adrenal hormones against the toxic manifestations of an endotoxin. It conveys no information, however, as to why this should be true. In the preceding paper (2), it was indicated that the therapeutic action of cortisone in endotoxin-poisoned mice was linked, perhaps, to the higher levels of carbohydrate it produced. Since Long et al. (3) found that the glyconeogenesis initiated by adrenal cortical hormones was accompanied by protein degradation and a negative nitrogen balance, it is to be assumed that the carbohydrate formed in mice given both endotoxin and cortisone is synthesized, at least in part, from body protein. This premise is subject to experimental verification.The classical relationship which permits urinary nitrogen excretion to be used in calculating protein catabolized can now be combined with a simple procedure (4) for measuring total body carbohydrate to establish a balance between the two quantities. Results of such experiments comprise part of this report.Other facets of the problem have also been evaluated. It must be assumed that the adrenal cortical hormones produced endogenously in response to adrenoeorticotropic hormone (ACTH) adtively protect experimental animals against injections of endotoxins. The importance of studies with ACTH instead of with exogenous cortisone is, therefore, apparent. Under selected conditions, the inability of ACTH to substitute, metabolically, for cortisone in animals poisoned with endotoxin is described.Comparison is made also between the effect of an endotoxin derived from