Stimulation of drinking by bacterial endotoxins in the rat

Physiology&Behavior,Vol.54, pp. 1005-1009, 1993

0031-9384/93 $6.00 + .00 Copyright© 1993PergamonPressLtd.

Printed in the USA.

RAPID COMMUNICATION

Stimulation of Drinking by Bacterial Endotoxins in the Rat K A I D O N G W A N G , LISA W A S E L E N C H U K A N D M A R K D. E V E R E D l

Department of Physiology, College of Medicine, and Department of Oral Biology, College of Dentistry, University of Saskatchewan, Saskatoon, Saskatchewan, Canada S7N OWO

Received 23 N o v e m b e r 1992

WANG, K., L. WASELENCHUKAND M. D. EVERED. Stimulation of drinking by bacterialendotoxins in the rat. PHYSIOL BEHAV 54(5) 1005-1009, 1993.--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 ug/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 sufficientto reduce plasma osmolality and sodium concentration 2 to 3%. Drinking occurred during the period of hypothermia that frequently precedes the onset ofendotoxin-induced fever, so cannot be attributed to increased body temperature. Doses of endotoxin causing drinking inhibited both spontaneous and deprivationinduced feeding. The cause of the drinking is not known, but may involve mechanisms other than the known dehydrational signals controlling thirst. Endotoxin Antidiuresis

Lipopolysaccharide Bacterialinfection Thirst Fluid balance Fever Anorexia Rat

ENDOTOXINS are lipopolysaccharides from the cell walls of gram-negative bacteria. They stimulate many of the physiological and pathologic changes associated with bacterial infections (27). We have been investigating the effects of bacterial endotoxins on water intake and fluid balance. The importance of maintaining adequate fluid intake during illness is well recognized. Because fever, sweating, diarrhea, and some of the metabolic changes associated with bacterial infections are likely to cause dehydration, the stimulation of thirst would be expected and appropriate. Yet several investigators have reported that systemic or intracranial injections of endotoxins inhibit water intake (2,5,9,10,16). In contrast, we have found that IV injections of endotoxins from at least two bacterial sources, Escherichia coli and Salmonella minnesota, can stimulate drinking in water-replete rats. In this paper we describe the effects of several doses of these endotoxins on water intake, urine volume, and body temperature and the changes in body fluids caused by endotoxin-induced drinking. Changes in food intake were examined also, to investigate the behavioral specificity of the response.

1Requests for reprints should be addressed to Dr. Mark D. Evered.

1005

Drinking

Water intake

Urinary excretion

METHOD

Animals, Maintenance, and Surgical Preparation Long-Evans rats, weighing about 300 g at the start of the experiments, were purchased from Charles River Canada, Quebec. They were housed individually in plastic shoebox cages in a room lighted from 0700 to 1900 h. Room temperature and relative humidity were 20 to 22°C and 25 to 35%, respectively. Diet was pelleted Purina Rodent Chow #5001 and tap water. All experimental procedures were approved by the University of Saskatchewan Animal Care Committee, in accordance with the guidelines of the Canadian Council on Animal Care. At least 1 week before the experiments we implanted femoral venous cannulas, as described previously (6). Instruments were sterilized in antiseptic solution (Savlodil, Ayerst Laboratories, Montreal, Quebec) and rinsed in boiling water. Sites of incisions were shaved, cleaned, and wiped with iodine solution (1%, Proviodine, Rougier, Inc., Chambly, Quebec) and 70% alcohol. Implants were sterilized in 70% alcohol and iodine and rinsed with sterile isotonic saline just before use. Rats were anesthetized

1006 with Equithesin (12), 3 ml/kg IP and treated postoperatively with analgesic, buprenorphin-hydrochloride (Temgesic, Reckitt & Colman Pharm., Hull, England; 0.1 mg 1M). During the week of recovery, rats were handled frequently and accustomed to the test cages and procedures. All regained their preoperative body weight and full use of their operated leg.

Drugs and Injection Procedure Two endotoxin preparations were tested, a trichloroacetic acid (TCA) extract of Escherichia coli, serotype 026:B6, and a TCA extract of Salmonella minnesota (Sigma, St. Louis, MO). They were dissolved in sterile isotonic NaC1 solution and stored frozen in polypropylene tubes. Solutions were thawed just before the experiment. The osmolality of the highest concentration of E. coli used, 450 ug/ml, was 299 mOsm/kg, only slightly greater than that of the saline vehicle used as control (280 mOsm/kg). Venous cannulas were flushed with 0.2 ml sterile heparinized (10 U/ml) saline about 1 h before the experiment. Endotoxin solution or the vehicle control were injected IV in a volume of 1 ml/kg, followed by 0.2 ml sterile saline to clear the cannula.

Procedures All experiments were done between 0900 and 1700 h. Rats were placed in metabolism cages 1 h before injections. Temperature and humidity of the laboratory were similar to those of the housing quarters. Tap water was provided in inverted graduated centrifuge tubes with stainless steel spouts. Unless otherwise indicated, food was not available. Urine was collected in graduated tubes below the cages. Intake and urine volume were recorded to the nearest 0.1 ml. Each of the studies described below was done on a separate group of rats. Effects on water balance and rectal temperature. In eight rats, changes in water intake, urine volume, and rectal temperature were measured hourly for 6 h after IV injection ofE. coli extract. Two doses, 150 and 450 #g/kg, and the saline vehicle were tested in each rat in mixed order and with l week to recover between injections. Food and water were freely available until the time of injections, but then food was removed to prevent interference from changes in feeding. Rectal temperature was recorded on a telethermometer (Model 46 TUC, Yellow Springs Instruments, Yellow Springs, OH) using a thermister probe (3 mm diameter, YSI probe 402) inserted 4-5 cm into the rectum and held there l rain. To accustom rats to this procedure, it was repeated several times daily during the week before experiments started. In the same way, but using a second group of nine rats, we measured the effects of IV S. minnesota extract on water intake and urine volume. Three doses, 125, 250, and 500 #g/kg, and the saline vehicle, were tested in each animal, in mixed order and with 1 week between treatments. Effects on body fluids. Blood samples were taken from venous cannulas of six rats immediately before and 2 h after injection of E. coli endotoxin, 150 #g per kg body weight. Water intake and urine volume were recorded over this period. Blood samples were taken by first attaching a heparinized syringe to the venous catheter and withdrawing 0.5 ml to clear and fill the tubing with fresh undiluted blood. Then a second heparinized syringe was connected to withdraw 0.5 ml for analysis, the first 0.5 ml was returned, and the catheter was cleared with 0.2 ml saline. Some of the sample was placed in capillary tubes for hematocrit determination. After centrifuging the remaining sample, we measured plasma osmolality by freezing point

WANG, WASELENCHUK AND EVEREI) depression (Precision Systems, Osmette model 5004), plasma sodium concentration by flame photometry (Instrumentation Laboratory, model 943), and plasma protein concentration by refractometry (Atago, model SPR-T2). Eff~'cts onJeeding. Ten rats with tbod and water freely available until the experiment were placed in metabolism cages and given E. coli endotoxin, 450 ug/kg IV or the saline vehicle. Each was given about 5 g of the same pelleted food available in their home cages and intakes over the next 4 h were measured by weighing. Water was not available. The experiment was repeated with treatments reversed 1 week later. In a second experiment another 10 rats were deprived of food (but not water) for 18 h. E. coli endotoxin, 150 ug/kg, or the vehicle was injected IV, rats were placed in the test cages with food but not water, and intakes were measured over the next h. The experiment was repeated with treatments reversed 1 week later.

Data Analysis Results are presented as means _+ 1 SE. Means were compared by analysis of variance and the Newman-Keuls test for multiple comparisons where appropriate. Differences were considered statistically significant if the probability of their occurrence by chance was less than 5%. RESULTS

Effects on Water Balance and Rectal Temperature Figure 1 illustrates the effects of IV injections of endotoxin extracted from E. coli. Both doses (125 and 450 ~tg/kg) significantly increased water intake and reduced urinary water loss, with the higher dose having the greater and longer.lasting effect (p < 0.05). The combined effect on drinking and excretion was a large dose-dependent net gain of water. Both doses also lowered rectal temperature 1 to 2°C by the first h. After the higher dose, rats remained hypothermic for about 3 h. In rats given the lower dose, rectal temperature rose above control levels by the 6th h. Endotoxin from another bacterial source, S. minnesota, also stimulated drinking, as shown in Fig. 2. All three doses tested (125, 250, and 500 #g/kg) caused similar amounts of drinking and both the pattern and volume drunk were similar to those caused by the higher dose of the E. coli extract. Only the higher dose, however, significantly reduced urine volume.

Effects on Body Fluids Rats given E. coli endotoxin, 150 #g/kg IV, drank and retained sufficient water in 2 h (5.0 - 0.8 ml) to dilute body fluids (Table 1). Plasma osmolality fell 3.5% and sodium concentration fell 2.3% (both p < 0.01). The small reduction in hematocrit (1.9%) was not statistically significant. However, there was a large (15%) and significant (p < 0.01) reduction in plasma protein concentration.

Effects on Feeding Doses ofendotoxin that stimulated drinking inhibited feeding. The 4-h food intake of 10 nondeprived rats given E. coil endotoxin, 450 ttg/kg IV, was 0.14 _ 0.09 g, significantly less than their intake after control treatment, 1.63 + 0.34 g (p < 0.01). E. coli endotoxin, 150 #g per kg IV, also reduced feeding after 18 h food deprivation; 1 h intakes in 10 rats after control and endotoxin treatments were 3.1 + 0.4 g and 0.5 + 0.2 g, respectively (p < 0.01).

ENDOTOXINS AND DRINKING

1007

* 8

*

*

*

Z --

intake, leading to a small net water loss. The combination of increased drinking and fluid retention caused by endotoxin reversed this trend and caused a significant, dose-dependent net gain of water. Both preparations of endotoxin we tested stimulated drinking. The higher dose of E. coli extract had a significantly greater dipsogenic effect than the lower dose. However, all three doses of the S. minnesota extract caused similar amounts of drinking; also, the pattern and intake were similar to those caused by the

"Z"

~_j 12

-

z

~ ~.~

0

16

<; N

a 4

FIG. 1. Cumulative water intakes, urine volumes, net water gains (intake--urine volume), and changes in rectal temperature in rats after IV injections at time 0 of E. coli endotoxin, (O) 150 #g/kg or (u) 450 t~g/ kg, or (O) the isotonic saline vehicle. Points indicate means _ 1 SE, n = 8 for each group. *Indicates means significantly different from control values, p < 0.05.

DISCUSSION

~t~ ~

0 --4 -- 8 I

I

I

I

I

I

I

0

1

2

3

4

5

6

TIME (h) Our findings demonstrate that bacterial endotoxins can stimulate drinking behavior in rats. Rats drank even when endotoxin decreased urinary water loss. The antidiuresis was probably caused by increased vasopressin secretion (18) and reduced renal plasma flow and glomerular filtration (4). Rats are nocturnal, and during the light phase urine volume often exceeds water

FIG. 2. Cumulative water intakes, urine volumes, and net water gains (intake--urine volume) after IV injections at time 0 of S. minnesota endotoxin, (A) 125 #g/kg, (11) 250 #g/kg, or (O) 500 #g/kg, or (O) the isotonic saline vehicle. Points indicate means _+ 1 SE, n = 9 for each group. *Indicates means significantly different from control values, p < 0.05.

1008

WANG. WASELENCHUK AND [:~VEREI) TABLE

1

WATER BALANCE AND BLOOD VARIABLES 2 h AFTER INJECTION OF E. coli ENDOTOXIN, 150 /~g/kg IV

Water intake (ml) Urine volume (ml) Net water gain (ml) Plasma osmolality (mOsm/kg) Plasma sodium (mmol/1) Plasma protein (g/dl) Hematocrit (%)

Before

After

304.3 + 1.7 144.6 +_ 1.1 6.4 _+0.1 42.5 +_0.6

5.4 _+0.8 0.4 _+0.3 5.0 _+0.8 293.7 _+ 1.6" 1 4 1 . 3 _+ I. 1 * 5.4 +_0.2* 41.7 + 0.6

Values are means + 1 SE of results from six rats. * Significantlydifferent from value before treatment, p < 0.01.

highest E. coli dose. This suggests that even the lowest dose of S. minnesota had a maximal effect. Because the lowest S. minnesota dose caused drinking without decreasing urine volume, it appears that the renal and dipsogenic effects are separable and that the threshold is lower for the behavioral response. The drinking and antidiuretic responses were sufficient to dilute body fluids. Two hours after injection of the lower E. coli dose, 150 #g/kg IV, which caused a net gain of 5 ml water, plasma osmolality and sodium concentration fell 2 to 3%. The amounts drunk would not be expected to increase blood volume substantially, because most of the water would move osmotically into the larger intracellular and interstitial fluid compartments. Consistent with this, we observed only a small reduction in hematocrit. We did note, however, a significant fall in plasma protein concentration, a change normally associated with blood volume expansion. Here, however, it was probably caused by an endotoxin-induced shift of protein from plasma to interstitial sites (3,14). The drinking cannot be attributed to fever. Warming can stimulate drinking in animals even in the absence of dehydration (13). In these rats, however, rectal temperature fell during the first hour. Rats remained hypothermic for at least 3 h after the higher dose. Although fever is the response most frequently emphasized, hypothermia is commonly observed in rats given endotoxin systemically, and fever often takes hours to develop (8,18,33). Normally, cooling reduces drinking and increases urine flow in the rat, opposite to the changes observed here (11). Rectal temperatures eventually rose above control levels after the low dose of E. coli endotoxin, but well after drinking had ceased. It is also unlikely that a general behavioral arousal increased drinking. Endotoxin-treated rats often appeared lethargic for hours after the higher doses, rested quietly at the back of their cages, and seldom moved except to drink. Behavioral depression and reduced activity are common in infectious diseases, and may be accompanied by increased slow-wave sleep (15,21). However, in these experiments, done at a time when rats are not active, the magnitude of any behavioral depression is difficult to judge. To investigate the behavioral specificity more directly, we examined the effects of dipsogenic doses of endotoxin on feeding. Doses of endotoxin that stimulated drinking inhibited both spontaneous and deprivation-induced feeding. These findings are consistent with previous reports of endotoxin's anorexic effect (1,22,28). Thus, we conclude that endotoxin-induced water intake is not associated with fever or a general increase in ingestive activity.

Also, the body fluid dilution caused by endotoxin-induced water intake suggests that the rats were not drinking simply to repair a fluid deficit. This is consistent with our findings (Wang and Evered, manuscript under review) that dipsogenic doses of endotoxin do not increase plasma osmolality or Na + concentration, decrease blood volume, or decrease arterial or venous pressure. Our findings are consistent also with the report that endotoxin, at doses similar to those used here, stimulates vasopressin secretion by a mechanism other than the known dehydrational controls (18). The mechanism may involve interleukins, cytokine mediators of many other actions of endotoxin. It has been reported that IV interleukin-l~/stimulates vasopressin secretion (29) and that continuous 1P infusions of interleukin-1~ increase daytime water intake in rats (32). Our findings are contrary to previous reports that bacterial endotoxins inhibit thirst. For example, two early studies (5,16) demonstrated a reduction in daily water intake in rats and mice after endotoxin treatment. Because endotoxins also inhibit feeding, and because food intake is a major determinant of daily water intake (20), the hypodipsia noted in these studies may have been, at least partially, secondary to endotoxin-induced anorexia. Consistent with this, it has been reported that endotoxin does not reduce daily water intake when food is withheld (23). Also, our findings may help explain the observation (25) that endotoxin has a weaker anorexic effect when food-deprived rats are tested with liquid rather than solid food: an endotoxininduced increase in water intake might have partially offset the inhibitory effect on food intake. It is much harder to reconcile our findings with reports that endotoxins given systemically or intracranially inhibit drinking induced by water deprivation and other physiological and pharmacological thirst stimuli (2,9,10,16). Some caution is necessary when comparing endotoxin studies because of the variation in activity of different endotoxin preparations (17). However, we have recently confirmed that endotoxin inhibits deprivation-induced drinking (Wang and Evered, manuscript under review), using rats of the same strain, age, and sex as those used here, tested at the same time of day, and using doses of the same endotoxin preparation (E. coli endotoxin, 450 ~g/kg IV) shown here to be dipsogenic in water-replete rats. Thus, our findings suggest that the action of endotoxin is dependent on the state of hydration of the rat. It may be that a general malaise caused by endotoxin is exacerbated by the additional stress of dehydration or other experimental treatments used to stimulate thirst. Alternatively, dehydration may interact in some way with the metabolism of endotoxin or its mediators, including the interleukins. For example, water deprivation has been reported to increase the febrile response to endotoxin by increasing the production of endogenous pyrogen (probably interleukin 1) (26). If interleukin 1c~ and 1~ have opposite effects on thirst, as recently suggested (31), differences in the ratio of these in blood or brain may be an important factor. In summary, we have found that endotoxins extracted from at least two bacterial genera stimulate drinking in rats. Water intake increased in spite of fluid retention, hypothermia, inhibition of feeding, and, perhaps, a general malaise. Rats drank and retained enough water to dilute body fluids. Whatever the mechanism, we can add increased drinking to the behavioral changes caused by bacterial endotoxins. At [east some of these appear to be beneficial to the infected animal (7,24). Changes in thermoregulatory behaviors contribute to the development of fever, a response that may facilitate the host's attack on the bacteria (19). The inhibition of feeding

ENDOTOXINS AND DRINKING

1009

may be beneficial also, by limiting the availability of nutrients essential for growth of pathogenic organisms (28,34). A general reduction in activity a n d increased sleeping m a y help in terms of safety and tissue repair (15). Increased d r i n k i n g with fluid retention also would benefit an infected a n i m a l by correcting or perhaps even preventing the dehydration likely to be associated with infection. Clinically, the importance of m a i n t a i n i n g adequate hydration in the t r e a t m e n t of sepsis is well recognized (30). We speculate that rats may have evolved

m e c h a n i s m s for early detection of bacterial infection that cause them to drink and retain water even before debilitating deficits develop. ACKNOWLEDGEMENTS Supported by the Medical Research Council of Canada and the Heart and Stroke Foundation of Saskatchewan. We thank Ewa Michalska for excellent technical assistance and Robert Hutchinson for the artwork and photography.

REFERENCES 1. Baile, C. A.; Naylor, J.; McLaughlin, C. L; Catanzano, C. A. Endotoxin-induced fever and anorexia and elfazepam-stimulated feeding in sheep. Physiol. Behav. 27:271-277; 1981. 2. Calapai, G.; Squadrito, F.; Massi, M.; Caputi, A. P.; De Caro, G. Endotoxin inhibition of drinking behavior in the rat. Pharmacol. Res. 22:161-170; 1990. 3. Chien, S.; Sinclair, D. G.; Dellenback, R. J.; et al. Effect of endotoxin on capillary permeability to macromolecules. Am. J. Physiol. 207: 518-522; 1964. 4. Churchill, P. C.; Bidani, A. K.; Schwartz, M. J. Renal effects of endotoxin in the male rat. Am. J. Physiol. Renal Fluid Electrol. Physiol. 253(22):F244-F250; 1987. 5. Dubos, R. J.; Schaedler, R. W. The effect of bacterial endotoxins on the water intake and body weight of mice. J. Exp. Med. 113: 921-924; 1961. 6. Evered, M. D. Relationship between thirst and diazoxide-induced hypotension in rats. Am. J. Physiol. Regul. Int. Comp. Physiol. 259(28):R362-R370; 1990. 7. Ewald, P. W. Evolutionary biology and the treatment of signs and symptoms of infections disease. J. Theoret. Biol. 86:169-176; 1980. 8. Feldberg, W.; Saxena, P. N. Prostaglandins, endotoxin and lipid A on body temperature in rats. J. Physiol. 249:602-615; 1975. 9. Foea, A.; Matera, G.; Mastroeni, P.; Caputi, A. P. Evidence that prostaglandins within preoptic area may mediate the antidipsogenic effect of Escherichia coli endotoxin in the rat. Circ. Shock 17:137145; 1985. 10. Foca, A; Nicoletta, P.; Matera, G.; Mastroeni, P.; Caputi, A. P. Antidipsogenic effect of endotoxin in the rat. Circ. Shock 11:341-350; 1983. 11. Fregly, M. J. Water and electrolyte exchange in rats exposed to cold. Can. J. Physiol. Pharmacoi. 46:873-881; 1968. 12. Gandal, C. P. Avian anaesthesia. Fed. Proc. 28:1533-1534; 1969. 13. Grace, J. E.; Stevenson, J. A. F. Thermogenic drinking in the rat. Am. J. Physiol. 220:1009-1015; 1971. 14. Green, T. P.; Johnson, D. E.; Marchessault, R. P. Gatto, C. W. Transvascular flux and tissue accrual of Evans blue: Effects of endotoxin and histamine. J. Lab. Clin. Med. 111:173-183; 1988. 15. Hart, B. L. Animal behavior and the fever response: Theoretical considerations. J. Am. Vet. Med. Assoc. 187:998-1001; 1985. 16. Holmes, J. E.; Miller, N. E. Effects of bacterial endotoxin on water intake, food intake and body temperature in the albino rat. J. Exp. Med. 118:649-658; 1963. 17. Horan, M. A.; Little, R. A.; Rothwell, N. J.; Strijbos, P. J. L. M. Comparison of the effects of several endotoxin preparations on body temperature and metabolic rate in the rat. Can. J. Physiol. Pharmacol. 67:1011-1014; 1989. 18. Kasting, N. W.; Mazurek, M. F.; Martin, J. B. Endotoxin increases vasopressin release independently of known physiological stimuli. Am. J. Physiol. Endocrinol. Metab. 248(11):E420-E424; 1985.

19. Kluger, M. J. Is fever beneficial? Yale J. Biol. Med. 59:89-95; 1986. 20. Kraly, F. S. Physiology of drinking elicited by eating. Psychol. Rev. 91:478-490; 1984. 21. Krueger, J. M.; Walter, J.; Dinarello, C. A.; Wolff, S. M.; Chedid, L. Sleep-promoting effects of endogenous pyrogen (interleukin-l). Am. J. Physiol. Regul. Int. Comp. Physiol. 246(15):R994-R999; 1984. 22. Langhans, W.; Balkowski, G.; Savoldelli, D. Differential feeding responses to bacterial lipopolysaccharide and muramyl dipeptide. Am. J. Physiol. Regul. Int. Comp. Physiol. 261(30):R659-R664; 1991. 23. Langhans, W.; Harlacher, R.; Balkowski, G.; Scharrer, E. Comparison of the effects of bacterial lipopolysaccharide and muramyl dipeptide on food intake. Physiol. Behav. 47:805-813; 1990. 24. Legrand, E. K. An evolutionary perspective of endotoxin: A signal for a well-adapted defense system. Med. Hypotheses 33: 49-56; 1990. 25. McCarthy, D. O.; Kluger, M. J.; Vander, A. J. Suppression of food intake during infection: Is interleukin involved? Am. J. Clin. Nutr. 42:1179-1182; 1985. 26. Morimoto, A.; Murakami, N.; Ono, T.; Watanabe, T. Dehydration enhances endotoxin fever by increased production of endogenous pyrogen. Am. J. Physiol. Regul. Int. Comp. Physiol. 25 I(20):R41R47; 1986. 27. Morrison, D. C.; Ryan, J. L. Endotoxins and disease mechanisms. Annu. Rev. Med. 38:417-432; 1987. 28. Murray, M. J.; Murray, A. B. Anorexia ofinfeetion as a mechanism of host defense. Am. J. Clin. Nutr. 32:593-596; 1979. 29. Naito, Y.; Fukata, J.; Shindo, K.; et al. Effects of interleukins on plasma arginine vasopressin and oxytocin levels in conscious, freely moving rats. Biochem. Biophys. Res. Commun. 174:1189-1195; 1991. 30. Natanson, C.; Danner, R. L.; Reilly, J. M.; et al. Antibiotics vs. cardiovascular support in a canine model of human septic shock. Am. J. Physiol. Heart Circ. Physiol. 259(28):H1440-HI447; 1990. 31. Osaka, T.; Kannan, H.; Kawano, S.; Ueta, Y.; Yamashita, H. Intraperitoneal administration of recombinant human interleukin-1~ inhibits osmotic thirst in the rat. Physiol. Behav. 51:1267-1270; 1992. 32. Otterness, I. G.; Seymour, P. A.; Golden, H. W.; Reynolds, J. A.; Daumy, G. O. The effects of continuous administration of murine interleukin-la in the rat. Physiol. Behav. 43:797-804; 1988. 33. Stitt, J. T.; Shimada, S. G. A comparison of the febrile responses of the Brattleboro and Sprague-Dawley strains of rats to endotoxin and endogenous pyrogens. Can. J. Physiol. Pharmacol. 65:13771381; 1977. 34. Wing, E. J.; Young, J. B. Acute starvation protects mice against Listeria monocytogenesis. Infect. Immun. 28:771-776; 1980.