Ethanol-induced hypothermia and hyperglycemia in genetically obese mice

Life Sciences, Vol. 44, pp. 1377-1385 Printed in the U.S.A.

Pergamon Press

ETHANOL-INDUCED HYPOTHERMIA AND HYPERGLYCEMIA IN GENETICALLY OBESE MICE Edwin W. Haller and Lorentz E. Wittmers, Jr. Department of Physiology, School of Medicine, University of Minnesota Duluth, Minnesota 55812-2487, USA (Received in final form llarch 7, 1989) Summary Blood glucose and rectal temperatures were monitored in two strains of genetically obese mice (C57 BL/6J ob/ob) prior to and following intragastric ethanol administration in an attempt to relate the hypothermic response to ethanol to extracellular glucose concentration. In contrast to expectation, ethanol administration was typically associated with a hyperglycemia and a hypothermic response. In the ob/ob genotype, the hypothermic response was associated with pronounced hyperglycemia which was more emphatic in older animals. The data support the conclusion that ethanol-induced hypothermia is independent of blood glucose levels. In light of the known sensitivity of ob/ob mice to insulin, it is suggested further that the observed hypothermic response was not a function of the animals' ability to transport glucose into peripheral cells. The observed hyperglycemia of the obese animals was most likely stress-related. It has been frequently shown that ethanol administration in man (i) and other species (2-7) leads to hypoglycemia and/or hypothermia. In humans the hypoglycemia is only expressed if the ethanol administration is preceded by a sufficiently long fast (I), and in the rat it has been suggested that hypo- or hyperglycemia results from ethanol administration depending upon ambient temperature as well as on the fed state (8-10). Alternatively, some investigators (ii) have suggested that increased gluconeogenesis after ethanol or during hypothermia may result from stimulation of glucagon secretion; this observation may account for the lack of a hypoglycemic response to ethanol in some experiments (11). Ambient temperature and the dose of ethanol have been suggested to determine the glycemic response to ethanol; conversely, the question remains whether the glycemic state governs the hypothermlc response to ethanol. Because of its gluconeogenic capacity (12), genetically obese (CL57 BL/6J ob/ob) mice were chosen for the present experiments as a model to test the hypothesis that post-ethanol hypothermia is related to the concentration of extracellular glucose. These genetically obese mice are hyperphagic and hypoactive (13), and they demonstrate age-dependent hyperglycemia and hyperinsulinemia (6,14). The carbohydrate metabolism of these mice differs markedly from Type II diabetics in that their plasma glucose clearance rates are equivalent to those observed in the lean littermates (12,15). In addition, the hyperglycemia observed in the ob/ob strain is likely the result of excessive endogenous glucose production and secretion and it is particularly exaggerated under conditions of stressful stimulation that involve adrenal gluconeogenlc mechanisms (15,16). 0024-3205/89 $3.00 + .00 Copyright (c) 1989 Pergamon Press plc

1378

Hypothermia and llyperglycemia in Mice

Vol. 44, No. 19, 1989

Methods All experimental animals were derived in our colonies from breeding stock obtained from The Jackson Laboratories (Bar Harbor, ME). Genetically obese mice resulted from matings of C57 BL/6J heterozygotes ob/+. Lean offspring, ob/+ or +/+ genotypes, constituted the lean littermate controls. Published data suggest that homozygous (+/+) and heterozygous (ob/+) lean mice are indistinguishable with regard to food consumption, growth rates (i), blood glucose clearance (12) and glycosylated hemoglobin levels (17). Therefore, in the following discussion, the ob/+ and +/+ mice will be referred to as a single group designed as ?/+ or lean littermates. Both breeding and experimental animals were housed under conditions of controlled light (12h light/12h dark), temperature (23~I°C), and humidity (50~10%). They received food (Purina Mouse Chow ® ) and water ad libitum. All animals were fasted for 16-20 hrs (water ad libitum) prior to experimentation. This time period was required to achieve a sufficiently fasted state (glycogen depletion) in the obese animals (5,13), and to approximate those physiological conditions in humans and other species in which ethanol administration has been observed to lead to hypoglycemia (i-5). Obese mice and their lean littermates were matched as pairs with regard to age and sex. Animals were housed one pair (ob/ob and ?/+, each) to a cage. Experiments were conducted on 80 animals of both sexes ranging in age from 47 to 174 days, Group I (n=27): 52.3!i.I d (Mean_~SEM); Group II (n=28): 87.2~1.9 d; and Group III (n=25): 152.6~2.6 d. Blood glucose and rectal temperature were monitored in all animals in each age group. Blood ethanol concentration was assayed in ten mice (five of each genotype) from each age group. The hyperglycemia associated with genetic obesity has previously been reported to be age-dependent with a peak reported between 80-120 days of age (6,14). Previously we have not observed a sex difference in blood glucose concentrations of C57 BL/6J ob/ob and ?/+ mice (14). Blood glucose and rectal temperature were monitored in relation to a standard oral dose of ethanol, at zero time and at specific time intervals post-treatment (at 20, 40, 60, 90 and 120 minutes). Ethanol (12 mg/g body weight) was administered as an aqueous solution (24%) by garage. The experiments were conducted at room temperature (21.5 ~ 0.2°C). Animals were placed on a thermoneutral surface and protected from draft. Blood glucose was assayed on 25 uL blood aliquots (collected from the retro-orbital plexus) using the glucose oxidase method (YSI Glucose Analyzer). Rectal temperatures were measured with a thermocouple probe (Bailey Instruments, Inc., Model RET-3) that had been inserted into the rectum approximately 1 cm beyond the anus. Prior to experimentation, this probe had been calibrated over the expected experimental range (20-40°C) against a certified ASTM mercury thermometer. Ethanol concentration in blood was determined by the alcohol oxidase method employing a Yellow Springs Instrument analyzer according to Mason (18). All experimental data were analyzed by Fisher's exact T-test (19). Results The experimental animals, both lean and obese genotypes, appeared healthy and alert and body weights of two groups of 15 animals given in Table I were within the expected range. Pretreatment baseline values of rectal temperature and blood glucose levels following a 16-20 hour fast, for the two genotypes and the three age groups, are presented in fig. I. Compared to lean littermates the rectal temperature (fig. IA) of ob/ob mice was

Vol. 44, No. 19, 1989

Hypothermia

and Hyperglycemia

in Mice

1379

lower (p
GENOTYPE (lean)

AGE GROUP I

n Mean (g) SEM ob/ob (obese) n Mean (g) SEM

AGE GROUP II

AGE GROUP III

5 21.2 + 1.36

5 23.9 +__ 1.22

5 28.2 + 2.68

5 41.3 + 1.09

5 50.1 + 2.15

5 60.3 + 1.52

Panel B

Panel A 40

200

30

20

o

100

I-10

0

0 I

II

I11

I

Age Group

II Age Group

ob/ob

D

FIG. 1 Basal pretreatment levels of rectal temperature (Panel A) and blood glucose concentration (Panel B) of genetically obese mice and their lean littermates. The bars represent means and the standard errors of values from three separate age groups (see text). The profile of fasting blood glucose levels (fig. IB) at various ages confirms earlier findings (6, 17). Blood glucose baseline concentration was highest in the youngest ob/ob mice, and in each age group that of ob/ob mice tended to exceed (p<0.05) corresponding values in lean littermates. The statistical significance of this difference disappeared, however, in the oldest age group. Within genotype and for a given age group, administration of the standard dose of ethanol resulted in relatively constant, elevated blood ethanol concentrations (fig. 2). From 20 min after ethanol treatment onwards, for lean mice plasma alcohol levels ranged from 118 to 247 mg/dl with a mean~SEM of 198.7~3.7, and for obese mice comparable values ranged from 209

1380

Hypothermia and Hyperglycemia

in Mice

Vol. 44, No. 19, 1989

to 404 mg/dl with a mean of 314.8+4.7. This difference was statistically significant at the p<0.05 level. --During the latter portion of the experimental period (90-120 min), the blood glucose concentration tended to fall in the ob/ob mice. Since plasma ethanol concentrations were stable from 20 min post-treatment onward, steady state volumes of distribution (V D) were calculated for ethanol based on body weight. The following formula was applied: VD = [ V I ( C I / C 2 ) - V l ] / ( b w ) , where: VD = c a l c u l a t e d volume o f d i s t r i b u t i o n , ml V1 = i n j e c t i o n volume, ml

C 1 = ethanol concentration in the injectate, bw = body weight, g C 2 = plasma ethanol concentration, mg/dl The were 3A). was and

mg/dl

values for V D were equal in the two genotypes in Age Group I, but they higher in obese than in lean animals in the other two age groups (fig. Relative blood ethanol (fig. 3B) concentration (mg/dl/g body weight) significantly (p<0.05) higher in lean animals in age groups II (13.7%) III (33.7%).

Panel

Panel A 400

400 '

300 "

300 "

B

-o 200"

100'

0

20

40

60

i

9O

2O0 "

100 "

0 120

0

Time (rnm)

I

Age Group I

20

40

60

90

120

Time (rnin)

~]

Age Group II

FIG.

['~

Age Group III

2

Blood ethanol concentration (mean + SEM) in three age groups of mice. Values of lean littermates are shown in Panel A and those from obese homozygotes in Panel B. Ethanol was administered by gavage at zero time of the experiment. See text for discussion. Whereas the glycemic response of lean mice to ethanol treatment was minimal, obese mice at all ages demonstrated varying degrees of hyperglycemia with ethanol treatment (fig. 4). Their response was considerably greater than that of the lean littermates, at any time. Basal blood glucose levels in all animals regardless of age or genotype were in the same range except for the youngest group of ob/ob mice. The most pronounced hyperglycemic response occurred in the youngest age group of obese mice.

Vol. 44, No. 19, 1989

Hypothermia and Hyperglycemia in Mice

1381

Ethanol treatment produced a hypothermia in all genotypes regardless of initial blood glucose concentration or age (fig. 5). Within genotype and independent of age, baseline temperatures were the same (fig. IA). However, rectal temperatures of obese mice were significantly (p<0.05) lower than those of their lean littermates by about 1.5°C. Only the youngest age group of obese mice had a final rectal temperature at 120 min post-ethanol administration that was lower than that of their lean littermates. For all other animals, regardless of their starting basal temperatures, the corresponding rectal temperatures reached at the end of the experimental period were not statistically distinguishable.

Panel A

B

Panel lO

081 8

u.I co

6 4

._~ 2

I

II Age

I

ill

~l

III

AgeGroup

Group

I

ob/ob

J~J

FIG. 3

The means and SEM of values of distribution (Panel A) and blood level of ethanol (Panel B) as calculated on the basis of body weight are shown for the three Age Groups of obese mice and their lean littermates as indicated in the figure. See text for further discussion. At the dose administrated, the hypothermic response to ethanol occurred immediately and appeared to have reached its maximum within the first 20 minutes after treatment. No statistically significant changes in rectal temperature were detected subsequently in any of the experimental groups. Although not statistically significant, the oldest group of ob/ob mice maintained uniformly lower rectal temperatures than any other genotype. Discussion In the present experiments, the typical hypothermic responses to ethanol were demonstrated in genetically obese (C57 BL/6J ob/ob) mice and their lean littermates. This hypothermia was equivalent in both genotypes. However, although all mice became hyperglycemic after ethanol, the hypothermic response was dissociated from any level of hyperglycemia. Furthermore, the hypothermic response to ethanol was not age-dependent. All animals appeared to be healthy and in good physical condition. Their body weights were within previously reported (5) ranges appropriate to genotype and age.

1382

Hypothermia

and H y p e r g l y c e m i a

in M i c e

Vol. 44, No. 19, 1989

Panel B

Panel A 400

400

300

300

E

200

200

o

(.9 100

~oo

0

,

,

,

,

1

0

'

'

'

'

1

'

'

50

'

'

1

100

Time

'

0

I

'

1

0

150

I

'

"

(min)

'

I

'

'

'

'

100

Time

FIG.

'

50

I 150

(min)

4

Means and SEM of blood glucose concentration in lean (Panel A) and genetically obese (Panel B) mice after ethanol treatment at time of the experiment, in different age groups (open squares - Group I, diamonds - Group II, closed squares - Group III). Panel

Panel A

B

37-

36"

35"

35"

~

F-

34"

33"

32

34"

33"

'

'

'

'

1'

'

'

'

'

50

I

100 Time

'

'

'

'

32

I

0

150

50

1 O0

150

Time (min)

(min)

FIG.

5

Means and SEM of rectal temperature in lean (Panel A) and genetically obese (Panel B) mice after ethanol treatment at time zero of the experiment, in different age groups (open squares - Group I, diamonds Group II, closed squares - Group III). H y p o t h e r m i a is the typical response to ethanol variety of conditions in a number of different

administration under a species (9, I0, 21-23).

-

A

Vol. 44, No. 19, 1989

Hypothermia

and Hyperglycemia

in Mice

1383

number of studies have suggested nevertheless that ethanol induces a hypothermia. On close examination, these studies indicate that the observed glycemic response can be linked to ~nbient temperature and the state of feeding or fasting of the animal (21). This was most clearly demonstrated by Souza et al. (9, I0). The implication from their studies is that ambient temperature and satiation determine the glycemic response to ethanol; however, in our studies the glycemic response was the result of ethanol treatment and hypothermia was independent of blood glucose levels. Since our experiments were carried out only at one ambient temperature, 21.5°C, we cannot address the question of the role of ambient temperature in this response. The measured blood ethanol concentrations reasonably reflect the administered dose of the drug. In a detailed study of the pharmacokinetics of ethanol after oral administration in the fasted state, Wilkinson et al. (24) showed that blood ethanol concentrations are primarily controlled by gastric emptying which is followed by complete absorption and systemic availability of the entire oral dose of ethanol, The concentration range for which this model has been validated falls within the limits of the present experiment and agrees with our observations. Within each genotype, blood ethanol concentrations attained by oral administration remained relatively constant during the experiment. Only in the oldest ob/ob mice was there a tendency for blood ethanol to fall during the latter part (90-120 min) of the experimental period. No ready explanation for this phenomenon can be offered, and no effect was detected on the conclusions drawn from the present study. It has been suggested that the metabolic clearance rate of ethanol is enhanced with age (25). The striking difference in blood ethanol concentration between genotypes requires further comment. Even though ethanol was given on the basis of body weight, subsequent calculation of the apparent volume of distribution of ethanol reflects a steady state condition that appears to obtain at 20 min post-treatment, i.e., plasma ethanol remains stable during the remainder of the experimental period. For this reason, the calculated volumes of distribution, in this case, can serve as a basis for comparison. Thus, the data indicate that in each genotype that volume is less in lean littermates, and it increases with body weight and age in obese mice. The calculated values parallel the changes observed in body weight with age. When volume of distribution and blood ethanol concentration are calculated relative to body weight, these observed differences disappear. Such considerations would indicate that the volume of distribution in both genotypes is a function of body weight, which does not necessarily reflect blood volume or extracellular fluid volume. With the considerably greater proportion of body fat a greater v o l ~ e of distribution of ethanol might be expected in obese mice as suggested by the present data. The general, the mammalian response to ethanol ingestion is characterized by simultaneous lowering of body temperature and of blood glucose concentration (1-7, 25). Therefore, the ?/+ genotype, i.e., the lean littermate without known abnormalities in carbohydrate and fat metabolism, might have been expected to present with these symptoms. Instead, these animals remained essentially normoglycemic with perhaps a slight upward trend (see fig. 4A). Since ethanol administration may also induce and enhance other stressful responses (27), it may be appropriate to regard the glycemic response observed in both lean and obese animals in this light. Certainly, both the extent and the dynamics of the hyperglycemia appear identical to that induced by another stressor, i.e., the administration of an anesthetic dose of ether (15). It is interesting to note that the

1384

Hypothermia

and Hyperglycemia

in Mice

Vol. 44, No. 19, 1989

glycemic state in ob/ob mice exists even in the presence of documented hyperinsulinemia (5). Potter and Morris (28) have reported that oral ethanol administration caused a dose-dependent reduction of plasma insulin and an elevation of plasma glucagon concentrations in both fed and fasted rats. In fasted rats, blood glucose remained within normal limits, but in fed rats a hyperglycemia could be observed. Unfortunately, ambient temperature was not reported in this study. Furthermore, it is possible that in these rats glucagon secretion represented a stress response. As we have shown in previous experiments (29), the sensitivity of ob/ob mice to insulin is unaltered compared to that in lean littermates although the dose-response curve has been shifted upwards. Therefore, application of the ob/ob mouse model to the present question is appropriate. Thus, we suggest that the observed hypothermia in ob/ob mice after ethanol is not likely the result of intracellular hypoglycemia. Several experiments (7,11,28) have suggested that ethanol may induce glucagon secretion independent of a coexisting hypoglycemia, and that this response could be potentiated by ethanol-induced adrenal medullary catecholamine secretion. Even in normal rats, stressful stimuli have been reported to enhance the hypothermic effect of ethanol (27, 30). Although these experiments do not address this point directly, the possibility of a significant role of stress in the observed hyperglycemic response in the present experiments cannot be excluded entirely. Previously, we have found that the genetically obese mouse is highly sensitive to adrenal-mediated stressful stimuli (12). Ethanol administration can be considered as such a stimulus. Ethanol treatment was likely to have increased both glucagon and adrenal catecholamine secretion (7, ii), which would have resulted in enhanced gluconeogenesis and glycogenolysis. Involvement of adrenal mechanisms in this hyperglycemic response is supported by early work by Jauhonen and Hassinen (26), who reported that adrenalectomy in rats markedly reduced the hyperlipidemic and hyperglycemic responses to metabolites of ethanol. The present experiments as well as previous data from our laboratory (16) indicate that in the ob/ob genotype, diethyl ether produces an increase of blood glucose (presumed to be due to gluconeogenesis) by involving adrenal mechanisms. The extent of hypothermia observed in our experiments was approximately equal in the two genotypes and at all ages. The most pronounced expression of stress sensitivity, i.e., hyperglycemia, occurred in the youngest age group of the ob/oh mice. It is interesting to note the equal magnitude of hypothermia after ethanol in the two genotypes regardless of their basal blood glucose concentration or rectal temperature. This observation therefore supports reports on other species (22,31), which also suggest that the magnitude of the hypothermia was independent of extracellular glucose concentration. The information provided by our experiments raises the possibility that the observed ethanol-induced hypothermia may also be independent of the animals' ability to transport glucose into peripheral cells. Thus, the present studies suggest that, in the genotypes of mice tested here, the hypothermic response to ethanol is independent of the glycemic state. Nevertheless, ethanol must also be regarded as a powerful stress stimulus which likely operates by similar mechanisms in the two genotypes studied here and which may produce responses evoked by other stressors.

Vol. 44, No. 19, 1989

Hypothermia and Hyperglycemia in Mice

1385

Acknowledgments The authors express their thanks to E. Tahtinen and J. Gearns for their expert assistance. This work was supported in part by grants from the NSF (PCM-8003740) and the NIH (S06RR08212-04AI).

I. 2. 3. 4. 5. 6. 7. 8. 9. I0. II. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31.

References N. FREINKEL and R.A. ARKAY, Psychosom. Med. 28 551-563 (1966). L. ALARI, B. SJOQUIST, and T. LEWANDER, Drug Alcohol Depend. 19 369-373 (1987). S. FIDECKA, E. TAMBORSKA, R. LANGWINSKI, Acta Physiol. Pol. 37 168-175 (1986). R.A.L. BLATT and M. HAMBI, Int. J. Obesity. 6 391-397 (1982). P.U. DUBUC, R.S. RISTIMAKI, P.J. CAHN, P.L. WILLIS, Horm. Metab. Res. 15 120-123 (1983). L. HERBERG, L. and D.L. COLEMAN, Metabolism 36 59-99 (1977). V.P. JAUHONEN, Horm. Metab. Res. I0 214-219 (1978). M.L. OLIVEIRA-DE-SOUZA and J. MASUR, Pharmacol. Biochem. Behav. 15 551-554 (1981). M.L. SOUZA and J. MASUR, Pharmacol. Biochem. Behav. 16 903-908 (1982). M.L. SOUZA and J. MASUR, Pharmacol. Biochem. Behav. 20 649-652 (1984). H.J. SEITZ, W. KRONE, H. WILKE, W, TARNOWSKI, Pflugers Arch. 389 115-120 (1981). L.E. WITTMERS, Jr., G. CIZADLO, E. HALLER, G. BOWN, Physiologist 21 129 (1978). L. HERBERG, L. MAJOR, U. HENNINGS, D. GRUNKLEE, D. FREYTAG, IF.A. GRIES, Diabetologia 6 292-299 (1970). L.E. WITTMERS, Jr. and E.W. HALLER Metabolism 32 1093-1100 (1983). E. HALLER, L.E. WITTMERS, Jr., G. CIZADLO, G. BOWN, Physiologist 21 129 (1978). L.E. WITTMERS and E.W. HALLER, J. Comp. Path. 92 5199-525 (1982). P.U. DUBUC, Metabolism 35:1567-1574 (1976). M. MASON, Am. Soc. Brewing Chem. J. 40 78-79 (1982). G.W. SNEDECOR and W.G. COCHRAN Statistical Methods, p. 83. The Iowa State University Press, Ames, IA. (1980). A.D. LE, H. KALANT, J.M. KHANNA, Pharmacol. Biochem. Behav. 25 667-672 (1986). H. KALANT and A.D. LE, Pharmacol. Ther. 23 313-364 (1984). N.M. WILSON, P.M. BROWN, S.M. JUUL, S.A. PRESTWITCH, P.H. SONKSEN, Brit. Med. J. 282 849-853 (1981). P. LOMAX, J.G. BAJOREK, W.A. CHESAREK, R.R.J. CHAFFEE, Pharmacology 21 288-294 (1980). P.K. WILKINSON, A.J. SEDMAN, E. SAKMAR, D.R. KAY, J.G. WAGNER, J. Pharmacokin. Biopharm. 5 207-224 (1977). J.F. OTT, B.E. HUNTER and D.W. WALKER, Alcoholism. Clin. Exp. Res. 9 59 -65 (1985). V.P. JAUHONEN and I.E. HASSINEN, Arch. Biochem. Biophys. 191 358-366 (1978). J. PERIS and C.L. CUNNINGHAM, Life Sci. 38 273-279 (1986). D.E. POTTER and J.W. MORRIS, Experientia 36 1003-1004 (1980). E.W. HALLER and L.E. WITTMERS, Endocrinology 104 (Suppl.) 170 (1979). C.L. CUNNINGHAM and L.L. BISCHOF, Physiol. Behav. 40 377-382 (1987). M. MAJOR and R. FRANCESCONI, The Nature and Treatment of H)Tpothermia, R.S. POZOS and L.E. WITTMERS (eds.), p. I00, University of Minnesota Press, Minneapolis, MN (1983).