- Email: [email protected]
The effect of hypothalamic temperature on the immune response in the rat
Er&r Resectrch ffullerjn, Vol. 13, pp. 247-251, 1984.e Ankho Intemational Inc.
Printed in the
U.S.A.
0361-9230/84$3.00 + .00
The Effect of Hypothalamic Temperature on the Immune Response in the Rat’ STEFANUS
BRANDT
AND MANUEL
Received
12 Apt-3 1984
BANET2
BRANDT, S. AND M. BANET. The efSect of hypothalamic temperature on the immune response in the rat. BRAIN RES BULL 13(Z)247-251, 1984.-Toclarify the role of febrile temperatures on the immune system, rats were immunized with sheep erythrocytes and their core temperature was then changed by continuously cooling or heating the preoptic area for five days. Core temperatures of up to 2°Cabove or below normal were associated with a high titre of antibodies against
sheep erythrocytes, whereas larger displacements of core temperature, as well as normal temperature, were associated with a low titre. These results are at variance with the idea that the production of antibodies is proportional to body temperature. It is suggested that the immunostimulation elicited by heating and cooling the preoptic area, and by inference that the immunostimulation associated with fever, could be due to factors other than the change in body tern~~t~re. Fever
Temperature
regulation
Antibody titre
kept at 22-25°C in a room with natural
illumination. Food and water were freely available. The animals were anaesthetized with 50 mg of sodium pentobarbitone per kg of body weight, and a thermode was implanted into the preoptic area (31. At this time, an antirotatory device which allows otherwise freedom of movement [ 14f was fixed to the calvarium. One week after the operation, oxygen consumption was determined by an open circuit method [3] for at least 45 min, and core temperature was measured by inserting a thermocouple about 60 mm beyond the anal sphincter. The experiments first began after the resting values of these parameters had been determined for several days, every day at the same hour. In four of the animals, the thermode had a the~ocou~le glued to the tip with histoacryl. In these animals, we wanted to determine the steady state relationship between the temperature of the thermode perfusate and the temperature at the tip of the thermode, and between the temperature at the tip of the thermode and oxygen consumption, and rectai temperature. For this purpose, the thermodes were perfused for four hours with water the temperature of which was randomly varied from day to day. Oxygen consumption was dete~ined during the last hour of this experimental period, whereas core and thermode temperatures were measured immediately before and after the dete~ination of oxygen consumption. Forty-one animals were used to study the relationship between core temperature and antibody titre. These experiments were done in three series, one each in summer, fall and winter, and, in each series, the animals were randomized into five groups. At the beginning of the experiments, ati animals were immunized with 1 ml of a 10% suspension of
ALTHOUGH fever is the most common symptom of infectious disease, the significance of febrile temperatures for the immune response of the host is not yet clearly understood. Since temperature influences the rate of all metabolic processes, the production of antibodies is generally thought to increase and decrease with body temperature [2,28]. The results of in vitro experiments with mammalian cells are, conflicting-temperatures of 3%40°C may however, enhance [19,27] or inhibit [is] the production of antibodie~and, in any case, it is doubtfu1 that these results can be extrapolated to complex in viva phenomena. In sivo, the febrile response appears to enhance antibody pr~uct~on [7,22]. However, and active, fever-like increase in body temperature, induced by locahy coohng the preoptic area, enhances, whereas the same increase in temperature, if passively brought about by heat exposure, depresses the immune response [6]. Thus, nonthermal factors appear to determine the immune effect of increased body temperature. The main goal of this work, therefore, was to study the quantitative relationship between body temperature and antibody titre. For this purpose, rats were immunized with sheep erythrocytes and their body temperature was either actively increased by cooling the preoptic area, a procedure which induces a fever-like response [3], or actively decreased by heating this area, a procedure which may simulate the action of a possibIe endogenuous cryogen [ZO]. The results are compatibte with the idea that the febrile stimufation of the immune system 17,223 could be mediated by a response other than the increase in body temperature. METHOD Forty
five male Wistar rats weighing 3803t7 (SE) g were
~‘This work was supported by the Deutsche Forsc~u~s~emeinssh~. *Requests for reprints shoufd be addressed to Dr. Manuel Banet, Institut fur Normale und Pathoiogische ~utschhausstraRe 2, D-3550 Marburg, Federal Republic of Germany.
247
Physiologic der Universitat,
BKANDT AND BANET
24x
washed sheep erythrocytes given IP. In the animals of the not hypothalamic temperature was control group, manipulated-an additional control group without thermodes was not necessary because the implantation does not influence antibody titre [6,7]. In the animals of the other groups, starting within one hour of immunization, an active change in core temperature was induced by cooling or heating the hypothalamus to a temperature that varied from group to group. Since we wanted to induce average core temperatures of 41, 40, 39 and between 35 and 36”C, we regularly readjusted the temperature of the thermodes’ perfusate to correct any deviation of core temperature from the desired value. This treatment was continued for five days because this is the duration of the fever induced by infection with Salmonella enteritidis in the rat [4]. During this time, core temperature was measured as frequently as it appeared necessary but at least every eight hours. Oxygen consumption was measured once a day, though, due to technical failure, this could be done in only 32 of the animals. At the end of this period, the antirotatory devices were removed from all the animals. Afterwards, core temperature was measured daily, and, under ketamine hydrochloride anaesthesia (44 mg/kg), blood samples for the determination of the titre of antibodies were regularly taken by the orbital bleeding method E251. At the end of the experiments, the titre of antibodies was determined by the methods of agglutination and haemolysis [6], and the position of the thermodes was confirmed anatomically. RESULTS
The Effect of Hypothalamic Temperature Production and Core Temperature
on Heat
The results of these experiments are summarized in Fig. 1. Oxygen consumption increased with decreasing thermode temperature, and the relationship between the parameters was linear (linear r= -0.99). Core temperature aIso increased with decreasing thermode temperature but the relationship between the parameters is best described by a second degree polynomial (non-linear r=0.99). This curve shows that body temperature reached 41°C at a thermode temperature of about 23°C and that more intense cooling, despite the additional increase in oxygen consumption it induced, had no further effect on core temperature. In this experiments, we also calculated the regression of the temperature at the tip of the thermode with respect to the temperature of the water perfusing it. Since the perfusion system was not changed throughout the experiment, this equation was used to estimate the temperature of the thermode in further experiments. It must, however, be noted that this temperature represents the maximum possible displacement of hypothakunic temperature and that the temperature of the tissue not in immediate contact with the thermode would be displaced to a lesser degree. The Effect of Hypothalamic immune Response
Temperature
on the Humorul
Except for a slight increase following immunization, the average core temperature of the control animals remained at the normal level throughout the experimental period (Fig. 2). In the three groups of animals in which the preoptic area was cooled, core temperature and oxygen consumption rose to a higher level. In two of these groups, we succeeded in keeping core temperature rather close to the desired levels of 39
16r
c.$
1L
1
c g
8
2
0
6
I
20
I
I
24
I
I
28
Thermode
1 i
!
33 temperature
I
36
I
,
-’ 36
LO
(“C)
FIG. I. Oxygen consumption (0) and core temperature ( ,) after locally changing preoptic temperature for 3-4 hours in four rats The vertical lines are standard errors.
and 40°C. In the animals in which we desired an average core temperature of 41°C. we first cooled the thermode to about 23°C. Core temperature increased approaching the desired level but then it fell slowly, though our attempts to prevent this fail by further lowering the temperature of the thermode did increase oxygen consumption further. In the animals in which the preoptic area was heated. average core temperature and oxygen consumption fell below the normal level (Fig. 2). Core temperature, however, varied slightly beyond the desired limits of 35 and 36°C because the relatively fast initial drop in temperature induced by heating the thermode was often followed by a very slow fall that lasted for up to 10 hours. Therefore, our corrective action to bring the temperature down was sometimes premature and led to a drop beyond the desired level. All groups of experimental animals, except those in which core temperature was maintained at about 40°C (thermode temperature 3 IQ, had a titre of antibodies higher than the titre of the control group (Fig. 3). This can also be seen in Table 1, which summarizes the results of these experiments. In these experiments, we controlled the average core temperature of each group of animals rather than the temperature of each animal. Due to small differences in the location of the thermode and in the thermal sensitivity of the animals, any thermode temperature induced a change in core temperature that varied somewhat from animal to animal. particularly if the thermode was heated. Consequently. the temperature of the animals in neighbouring groups (Table 1) overlapped. Therefore, the animals of the five groups were pooled and regrouped, in ranges of O.YC. according to their average core temperature during the first five postimmunization days. The average titre of antibodies of each subgroup was then calculated (Fig. 4). The curve shown in this figure is a spline function [ 161 that simply connects the average of the agglutination and haemolysis titres with a smooth line. These results show that core temperatures about 2°C above or below normal were associated with a high titre, whereas larger displacements of temperature, as web as normal temperature, were associated with a low titre of antibodies. DISCUSSION Short-term
cooling of the preoptic
area increased
both
HYPOTHALAMIC
F
TEMPERATURE
AND IMMUNE
RESPONSE
249
39-
i?! = 38: ki r 37J
6s-
o) 36Zi U 35 -
L 3 -
,I
I
0
,I 1
,li,“2
Time
3
I
c
I I
2-
1
0
5
2
L
6
8
Time (days
(days)
10
TABLE
40 Control 35 31 19
N 9 (a$) II (8s) 8 (85) 6 (4S) 7 (4S)
7.2 + 0.4 8.9 k 0.4 10.4 t 0.6 10.6 * 0.6 11.5 + 0.9 p<0.0001P
I I
11
I6
1
1
AVERAGE OXYGEN CONSUMPTION AND CORE TEMPERATURE IN THE FIRST FIVE POST-IMMUNIZATION DAYS, AND LOG, OF THE HIGHEST TITRE OF ANTIBODIES AGAINST SHEEP ERYTHROCYTES IN RATS IN WHICW HYPOTHALAMIC TEMPERATURE WAS ARTIFICIALLY CHANGED DURING THE FIRST FIVE DAYS AFTER IMMUNIZATION WITH SHEEP ERYTHROCYTES
Oxygen consumptiont (mlimin)
I I
FIG. 3. Log, of antibody titre (arbitrary units) in animals in which core temperature was manipulated by heating or cooling the area preoptica for five days, starting on day 0 one hour after immunization with sheep erythrocytes. Al1 symbols as in Fig. 2.
FIG. 2. Oxygen consumption (upper panel) and core temperature (lower panel) in animals in which the thermode was heated to an average of 40°C (0), and cooled to an average of 35 (a), 31 (A), and 19°C (A) for five days, as well as in control animals (Cl). On day 0 at noon, all animals were immunized with sheep erythrocytes. The standard errors in the lower panel are given only if larger than 0.2”C.
Thermode temperature* (“0
I 12
Core temperature? (“C) 35.6 2 0.3 37.3 k 0.1 38.8 c 0.2 39.9 2 0.1 39.5 t 0.1 p
Log, antibody titret (arbitrary units) Agglutination 7.3 2 0.6 5.7 k 0.4 8.5 t 0.4 4.9 2 0.4 8.2 i 0.7 p
Haemolysis
7.7 + 0.8 5.8 4 0.3 8.5 + 0.4 4.5 t 0.9 8.6 + 0.9 pi0.0001$, *These are the estimated temperatures at the surface of the thermode and represent the highest possible displacement of hy~th~amic temperature. tThe values are the means and the standard errors. SNumber of animals in which oxygen consumption was measured. PAnalysis of variance comparing all the values in the column.
HKANDT AND BANET
1.510
N 6 :: II 5 I
33
.L_l..
Average
1.
^... _. ..._l
31
35 core
temperature
39
-ILl
J°C)
FIG. 4. Log, of antibody titre as determined by the methods of agglutination (0) and haemolysis (0) versus core temperature in animals in which core temperature was manipulated by heating and cooling the preoptic area, as well as in the corresponding control animals. The digits show the number of animals in each subgroup, and the vetiical lines the standard errors.
heat production and core temperature. However, despite a linear increase in the production of heat, temperature did not rise above 41°C. At this core temperature, heat production and heat loss may have passively balanced each other under our experimental conditions. On the other hand. 41°C is a temperature that is rarely exceeded under various conditions 191, including fever in rats j4]. Thus, it is possible that the animals actively increased heat loss to prevent core temperature from rising above this level. When the preoptic area was intensely cooled for more than a few hours, both heat production and core temperature fell to levels below those elicited by short-term cooling. These results (see also [3]) tend to reinforce the idea that some emergency mechanisms may reduce excessively high core temperatures. What these mechanisms may be is not known. ACTH and corticosteroids. however, are thought to limit the febrile rise in temperature [21], and the secretion rate of these hormones increases when the preoptic area is cooled [ fO,lS]. The mechanism of the antipyretic action of corticosteroids is not known. ACTH, however, appears to inhibit activity in heat conservation and heat production effector pathways 1211. Thus, it seems possible that ACTH. and perhaps corticosteroids, may also limit the rise in temperature induced by intense and prolonged cooling of the preoptic area. Since the rate of metabolic processes is temperature dependent, it is often assumed that the production of antibodies in mammals changes in parallel to body temperature [28]. Experiments in lower vertebrates tend to support this view [2]. Our experiments, however, show a complex relationship between body temperature and antibody titre (Table 1, Fig. 4) and thus challenge this view. The animals in which the thermode was cooled to an average of 19°C had the highest metabolic response but core temperature rose only to 39.S”C (Table I), probably because much of the heat produced was lost to the thermode. Since these animals had a high titre of anitbodies, it seems reason-
able to suggest that the nonthermnl responses elicited by cooling the preoptic area are not immunosuppressive. Therefore, the low titre of the animals in which core temperature was increased to 39.9”C can he attributed to a depressive effect of high temperature. This agrees with experiments itr t+ro, which show that temperatures above 39-40°C inhibit the production of antibodies Jl8,2?]. Hypothermia normally decreases the production ot antibodies in mammals [26.28]. and the mononucle~i! phagocytic system, a system which appears to be necessary for the development of the primary immune response [23]. is also inhibited by low tempetatures ;,I t*iiw [ 171. Resides. the in vitro response of lymphocytes to polyclonal mitogens is decreased at 35°C’ [I]. Thus. the decreasing immune rcsponse associated with large falls in core temperature can also be attributed to a direct effect of temperature. Changes in temperature of up to 3”r, however. \VCIC associated with a high titre of antibodies. This immune stimulation could have been the consequence of a direct temperature effect--small changes in temperature inducing stimulation, whereas large changes inducing depression of the immune response. This, however, implies that there are two optimal temperatures, perhaps reflecting the optimal range of distinct lymphoctye subs&. On the other hand, caution should be exercised before this stimulating immune effect of heating or cooling the anterior hypothalamus ih at-tributed to the change in body temperature, as some 111r?rl-tt experiments appear to suggest 119.271. The viability of lymphocytes in t*itro is often very low 1131 and temperature de pendent [13,18], and the results of these experiments may thus be biassed. Besides. the thermal sensitivity of an intact organism is higher than that of its isolated components 1301. Thus. the possibility that any change in body temperaturr may even depress the function of lymphocytes i/l I+,Y~cannot be excluded. If it were so. what could then be the uusc of the immune stimuIation induced by heating and cooling the hypoth~amus~~ Previous experiments have shown that nonthermal factors determine the immune effect of elevated body temperature [6], whereas the present ones show that thermal stimulation of the hypothalamus can induce the same degree of immunostimulation independently of the direction in which body temperature changes. Therefore. ws would like to suggest an alternative hypothesis for further consideration, and that is that the immune stimulation elicited by cooling and heating the preoptic area is mediated hy factors other than the change in body temperature. The nature of these nonthermaf factors is not known at present. However. the neuroendocrine system appears to play a pivotal role in the control of immune function (8.121. and hy~th~amic temperature influences the secretion rate of most hormones. In this context, it may be of interest to note that the only common endocrine responses to hypothalamic cold and warm stimuli appear to be an increase in the becretion rates of ACTH and corticosteroids (J IO. I I. I.‘] but SW [24]). These hormones, which at physiological levels appeal to stimulate the immune system [ 131, are thought to be the main mediators of the known effects of the hypothalamus on immunity [29]. Since the local cooling of the preoptic area inducea 11 fever-like response [3]. the results of the present cvxriments are compatible with the idea that certain febrile nonthermal responses may enhance the defences of the host animal [4,5]. However, whether this enhancement of the host’s defences has any reievant effect on the course of infectious disease is yet to he shown.
HYPOTHALAMIC
TEMPERATURE
AND IMMUNE
RESPONSE
251
REFERENCES 1. Ashman, R. B. and A. J. Nahmias. Enhancement of human lymphocyte responses to phytomitogens in vitro by incubation at elevated temperatures. Clin Exp Irnmunol 29: 464-467, 1977. 2. Avtalion, R. R., A. Wojdani, Z. Malik, R. Shahrabani and M. Duczyminer. Influence of environmental temperature on the immune response in fish. Curr Top Microbial Immunol61: l-35, 1973. 3. Banet, M. Fever and survival in the rat. The effect of enhancing fever. Pflugers Arch 381: 35-38, 1979. 4. Banet, M. Fever and survival in the rat. Metabolic versus temperature response. Experientia 37: 1302-1304, 1981. 5. Banet, M. The biological function of fever: an alternative view. Funkt Biol Med 2: 21 l-218, 1983. 6. Banet, M., D. Fischer, K. U. Hartmann, H. Hensel, and U. Hilling. The effect of whole body heat exposure and of cooling the hypothalamus on antibody titre in the rat. PjTugers Arch 391: 25-27, 1981. 7. Banet, M., S. Brandt and H. Hensel. The effect of continuously cooling the hypothalamic preoptic area on antibody titre in the rat. Experienfia 38: 965-966, 1982. 8. Besedovsky, H. and E. Sorkin. Network of immuneneuroendocrine interactions. Clin Exp Immunol27: l-12, 1977. 9. Bligh, J. Temperature Regulution in Mammals and Other Vertebrates. Amsterdam: North-Holland, 1973. 10. Chowers, I., H. T. Hammel, S. B. Strommeand S. M. McCann. Comparison of effect of environmental and preoptic cooling on plasma cortisol levels. Am J Physiol 207: 577-582, 1964. Il. Chowers, I., H. T. Hammel, J. Eisenman, R. M. Abrams and S. M. McCann. Comparison of effect of environmental and preoptic heating and pyrogen on plasma cortisol. Am J Physiol 210: 606-610, i966. _. 12. Comsa, J., H. Leonhardt and H. Werkele. Hormonal coordination of the immune response. Rev Physiol Biochem Pharmacol 92: 115-191, 1982. _ 13. Duff, G. W. and S. K. Durum. Fever and immunoregulation: Hyperthermia, interleukins 1 and 2, and T-cell proliferation. Yale J Biol Med 55: 437-442, 1982. 14. Fantino, M. F. Antirotatory device permitting free movement in animals with several catheters. J Appl Physiol 42: 275-280, 1977. 15. Gale, C. C., M. Jobin, D. W. Proppe, D. Notter and H. Fox.
Endocrine thermoregulatory responses to local hypothalamic cooling in unanesthetized baboons. Am J Physiol 219: 193-201, 1970.
16. Greville, T. N. E. Theory and Applications of Spline Functions. New York: Academic, 1969. 17. Huggins, C. and W. J. Noonan. An increase in reticuloendothelial cells in outlying bone marrow consequent upon a local increase in temperature. J Exr, Med 64: 275-280, 1936 18. Izumi, A., S. Koga-and M. Maeta. Effects of in vitro hyperthen-ma on murine and human lymphocytes. Cancer 51: 20612065, 1983. 19. Jampel, H. D., G. W. Duff, R. K. Gershon, E. Atkins and S. K. Durum. Fever and immunoregulation. III. Hyperthermia augments the primary in vitro humoral immune response. J Exp Med 157: 1229-1238, 1983. 20. Kluger, M. J., A. J. Thurnbull, W. I. Cranston, A. J. Wing, M. P. Cross and B. A. Rothenburg. Endogenous cryogen excreted by the kidneys. Am J Physiol 241: R271-R276, 1981. 21. Lipton, J. M., J. R. Glyn and J. A. Zimmer. ACTH and a-melanotropin in central temperature control. Fed Proc 40: 2760-2764, 1981. 22. Marat, B. A. Effect of the body temperature change on proceeding of the primary immune response in rabbits. Fiziol Zh SSSR 11: 1707-1709. 1981. 23. Pierce, C. W. and J. A. Kapp. The role of macrophages in antibody responses in vitro. In: Immunobiology of the Macrophage, edited by D. S. Nelson. New York: Academic, 1976, pp. l-31. 24. Proppe, D. W. and C. C. Gale. Endocrine thermoregulatory responses to local hypothalamic warming in unanesthetized baboons. Am J Physiol 219: 202-207, 1970. 25. Riley, V. Adaptation of orbital bleeding technic to rapid serial blood studies. Proc Sot Exp Biol Med 104: 751-754, 1960. 26. Roberts, N. J. Temperature and host defence. Microbial Rev 43: 241-259,
1979.
27. Saririan, K. and D. A. Nickerson. Enhancement of murine in vitro antibody formation by hyperthermia. Cell Immunol 74: 306312,
1982.
28. Schmidt, K. L. and V. R. Ott. Immunreaktionen bei hoher Korpertemperatur. Med Welt 25: 1963-1%8, 1974. 29. Stein, M., S. J. Schleifer and S. E. Keller. Hypothalamic influences on immune responses. In: Psychoneuroimmunology, edited by R. Ader. New York: Academic Press, 1981, pp. 429447. B. Thermostability of cells and proteins of 30. Ushakov, poikilotherms and its significance in speciation. Physiol Rev 44: Sl8-560,
1964.
Printed in the
U.S.A.
0361-9230/84$3.00 + .00
The Effect of Hypothalamic Temperature on the Immune Response in the Rat’ STEFANUS
BRANDT
AND MANUEL
Received
12 Apt-3 1984
BANET2
BRANDT, S. AND M. BANET. The efSect of hypothalamic temperature on the immune response in the rat. BRAIN RES BULL 13(Z)247-251, 1984.-Toclarify the role of febrile temperatures on the immune system, rats were immunized with sheep erythrocytes and their core temperature was then changed by continuously cooling or heating the preoptic area for five days. Core temperatures of up to 2°Cabove or below normal were associated with a high titre of antibodies against
sheep erythrocytes, whereas larger displacements of core temperature, as well as normal temperature, were associated with a low titre. These results are at variance with the idea that the production of antibodies is proportional to body temperature. It is suggested that the immunostimulation elicited by heating and cooling the preoptic area, and by inference that the immunostimulation associated with fever, could be due to factors other than the change in body tern~~t~re. Fever
Temperature
regulation
Antibody titre
kept at 22-25°C in a room with natural
illumination. Food and water were freely available. The animals were anaesthetized with 50 mg of sodium pentobarbitone per kg of body weight, and a thermode was implanted into the preoptic area (31. At this time, an antirotatory device which allows otherwise freedom of movement [ 14f was fixed to the calvarium. One week after the operation, oxygen consumption was determined by an open circuit method [3] for at least 45 min, and core temperature was measured by inserting a thermocouple about 60 mm beyond the anal sphincter. The experiments first began after the resting values of these parameters had been determined for several days, every day at the same hour. In four of the animals, the thermode had a the~ocou~le glued to the tip with histoacryl. In these animals, we wanted to determine the steady state relationship between the temperature of the thermode perfusate and the temperature at the tip of the thermode, and between the temperature at the tip of the thermode and oxygen consumption, and rectai temperature. For this purpose, the thermodes were perfused for four hours with water the temperature of which was randomly varied from day to day. Oxygen consumption was dete~ined during the last hour of this experimental period, whereas core and thermode temperatures were measured immediately before and after the dete~ination of oxygen consumption. Forty-one animals were used to study the relationship between core temperature and antibody titre. These experiments were done in three series, one each in summer, fall and winter, and, in each series, the animals were randomized into five groups. At the beginning of the experiments, ati animals were immunized with 1 ml of a 10% suspension of
ALTHOUGH fever is the most common symptom of infectious disease, the significance of febrile temperatures for the immune response of the host is not yet clearly understood. Since temperature influences the rate of all metabolic processes, the production of antibodies is generally thought to increase and decrease with body temperature [2,28]. The results of in vitro experiments with mammalian cells are, conflicting-temperatures of 3%40°C may however, enhance [19,27] or inhibit [is] the production of antibodie~and, in any case, it is doubtfu1 that these results can be extrapolated to complex in viva phenomena. In sivo, the febrile response appears to enhance antibody pr~uct~on [7,22]. However, and active, fever-like increase in body temperature, induced by locahy coohng the preoptic area, enhances, whereas the same increase in temperature, if passively brought about by heat exposure, depresses the immune response [6]. Thus, nonthermal factors appear to determine the immune effect of increased body temperature. The main goal of this work, therefore, was to study the quantitative relationship between body temperature and antibody titre. For this purpose, rats were immunized with sheep erythrocytes and their body temperature was either actively increased by cooling the preoptic area, a procedure which induces a fever-like response [3], or actively decreased by heating this area, a procedure which may simulate the action of a possibIe endogenuous cryogen [ZO]. The results are compatibte with the idea that the febrile stimufation of the immune system 17,223 could be mediated by a response other than the increase in body temperature. METHOD Forty
five male Wistar rats weighing 3803t7 (SE) g were
~‘This work was supported by the Deutsche Forsc~u~s~emeinssh~. *Requests for reprints shoufd be addressed to Dr. Manuel Banet, Institut fur Normale und Pathoiogische ~utschhausstraRe 2, D-3550 Marburg, Federal Republic of Germany.
247
Physiologic der Universitat,
BKANDT AND BANET
24x
washed sheep erythrocytes given IP. In the animals of the not hypothalamic temperature was control group, manipulated-an additional control group without thermodes was not necessary because the implantation does not influence antibody titre [6,7]. In the animals of the other groups, starting within one hour of immunization, an active change in core temperature was induced by cooling or heating the hypothalamus to a temperature that varied from group to group. Since we wanted to induce average core temperatures of 41, 40, 39 and between 35 and 36”C, we regularly readjusted the temperature of the thermodes’ perfusate to correct any deviation of core temperature from the desired value. This treatment was continued for five days because this is the duration of the fever induced by infection with Salmonella enteritidis in the rat [4]. During this time, core temperature was measured as frequently as it appeared necessary but at least every eight hours. Oxygen consumption was measured once a day, though, due to technical failure, this could be done in only 32 of the animals. At the end of this period, the antirotatory devices were removed from all the animals. Afterwards, core temperature was measured daily, and, under ketamine hydrochloride anaesthesia (44 mg/kg), blood samples for the determination of the titre of antibodies were regularly taken by the orbital bleeding method E251. At the end of the experiments, the titre of antibodies was determined by the methods of agglutination and haemolysis [6], and the position of the thermodes was confirmed anatomically. RESULTS
The Effect of Hypothalamic Temperature Production and Core Temperature
on Heat
The results of these experiments are summarized in Fig. 1. Oxygen consumption increased with decreasing thermode temperature, and the relationship between the parameters was linear (linear r= -0.99). Core temperature aIso increased with decreasing thermode temperature but the relationship between the parameters is best described by a second degree polynomial (non-linear r=0.99). This curve shows that body temperature reached 41°C at a thermode temperature of about 23°C and that more intense cooling, despite the additional increase in oxygen consumption it induced, had no further effect on core temperature. In this experiments, we also calculated the regression of the temperature at the tip of the thermode with respect to the temperature of the water perfusing it. Since the perfusion system was not changed throughout the experiment, this equation was used to estimate the temperature of the thermode in further experiments. It must, however, be noted that this temperature represents the maximum possible displacement of hypothakunic temperature and that the temperature of the tissue not in immediate contact with the thermode would be displaced to a lesser degree. The Effect of Hypothalamic immune Response
Temperature
on the Humorul
Except for a slight increase following immunization, the average core temperature of the control animals remained at the normal level throughout the experimental period (Fig. 2). In the three groups of animals in which the preoptic area was cooled, core temperature and oxygen consumption rose to a higher level. In two of these groups, we succeeded in keeping core temperature rather close to the desired levels of 39
16r
c.$
1L
1
c g
8
2
0
6
I
20
I
I
24
I
I
28
Thermode
1 i
!
33 temperature
I
36
I
,
-’ 36
LO
(“C)
FIG. I. Oxygen consumption (0) and core temperature ( ,) after locally changing preoptic temperature for 3-4 hours in four rats The vertical lines are standard errors.
and 40°C. In the animals in which we desired an average core temperature of 41°C. we first cooled the thermode to about 23°C. Core temperature increased approaching the desired level but then it fell slowly, though our attempts to prevent this fail by further lowering the temperature of the thermode did increase oxygen consumption further. In the animals in which the preoptic area was heated. average core temperature and oxygen consumption fell below the normal level (Fig. 2). Core temperature, however, varied slightly beyond the desired limits of 35 and 36°C because the relatively fast initial drop in temperature induced by heating the thermode was often followed by a very slow fall that lasted for up to 10 hours. Therefore, our corrective action to bring the temperature down was sometimes premature and led to a drop beyond the desired level. All groups of experimental animals, except those in which core temperature was maintained at about 40°C (thermode temperature 3 IQ, had a titre of antibodies higher than the titre of the control group (Fig. 3). This can also be seen in Table 1, which summarizes the results of these experiments. In these experiments, we controlled the average core temperature of each group of animals rather than the temperature of each animal. Due to small differences in the location of the thermode and in the thermal sensitivity of the animals, any thermode temperature induced a change in core temperature that varied somewhat from animal to animal. particularly if the thermode was heated. Consequently. the temperature of the animals in neighbouring groups (Table 1) overlapped. Therefore, the animals of the five groups were pooled and regrouped, in ranges of O.YC. according to their average core temperature during the first five postimmunization days. The average titre of antibodies of each subgroup was then calculated (Fig. 4). The curve shown in this figure is a spline function [ 161 that simply connects the average of the agglutination and haemolysis titres with a smooth line. These results show that core temperatures about 2°C above or below normal were associated with a high titre, whereas larger displacements of temperature, as web as normal temperature, were associated with a low titre of antibodies. DISCUSSION Short-term
cooling of the preoptic
area increased
both
HYPOTHALAMIC
F
TEMPERATURE
AND IMMUNE
RESPONSE
249
39-
i?! = 38: ki r 37J
6s-
o) 36Zi U 35 -
L 3 -
,I
I
0
,I 1
,li,“2
Time
3
I
c
I I
2-
1
0
5
2
L
6
8
Time (days
(days)
10
TABLE
40 Control 35 31 19
N 9 (a$) II (8s) 8 (85) 6 (4S) 7 (4S)
7.2 + 0.4 8.9 k 0.4 10.4 t 0.6 10.6 * 0.6 11.5 + 0.9 p<0.0001P
I I
11
I6
1
1
AVERAGE OXYGEN CONSUMPTION AND CORE TEMPERATURE IN THE FIRST FIVE POST-IMMUNIZATION DAYS, AND LOG, OF THE HIGHEST TITRE OF ANTIBODIES AGAINST SHEEP ERYTHROCYTES IN RATS IN WHICW HYPOTHALAMIC TEMPERATURE WAS ARTIFICIALLY CHANGED DURING THE FIRST FIVE DAYS AFTER IMMUNIZATION WITH SHEEP ERYTHROCYTES
Oxygen consumptiont (mlimin)
I I
FIG. 3. Log, of antibody titre (arbitrary units) in animals in which core temperature was manipulated by heating or cooling the area preoptica for five days, starting on day 0 one hour after immunization with sheep erythrocytes. Al1 symbols as in Fig. 2.
FIG. 2. Oxygen consumption (upper panel) and core temperature (lower panel) in animals in which the thermode was heated to an average of 40°C (0), and cooled to an average of 35 (a), 31 (A), and 19°C (A) for five days, as well as in control animals (Cl). On day 0 at noon, all animals were immunized with sheep erythrocytes. The standard errors in the lower panel are given only if larger than 0.2”C.
Thermode temperature* (“0
I 12
Core temperature? (“C) 35.6 2 0.3 37.3 k 0.1 38.8 c 0.2 39.9 2 0.1 39.5 t 0.1 p
Log, antibody titret (arbitrary units) Agglutination 7.3 2 0.6 5.7 k 0.4 8.5 t 0.4 4.9 2 0.4 8.2 i 0.7 p
Haemolysis
7.7 + 0.8 5.8 4 0.3 8.5 + 0.4 4.5 t 0.9 8.6 + 0.9 pi0.0001$, *These are the estimated temperatures at the surface of the thermode and represent the highest possible displacement of hy~th~amic temperature. tThe values are the means and the standard errors. SNumber of animals in which oxygen consumption was measured. PAnalysis of variance comparing all the values in the column.
HKANDT AND BANET
1.510
N 6 :: II 5 I
33
.L_l..
Average
1.
^... _. ..._l
31
35 core
temperature
39
-ILl
J°C)
FIG. 4. Log, of antibody titre as determined by the methods of agglutination (0) and haemolysis (0) versus core temperature in animals in which core temperature was manipulated by heating and cooling the preoptic area, as well as in the corresponding control animals. The digits show the number of animals in each subgroup, and the vetiical lines the standard errors.
heat production and core temperature. However, despite a linear increase in the production of heat, temperature did not rise above 41°C. At this core temperature, heat production and heat loss may have passively balanced each other under our experimental conditions. On the other hand. 41°C is a temperature that is rarely exceeded under various conditions 191, including fever in rats j4]. Thus, it is possible that the animals actively increased heat loss to prevent core temperature from rising above this level. When the preoptic area was intensely cooled for more than a few hours, both heat production and core temperature fell to levels below those elicited by short-term cooling. These results (see also [3]) tend to reinforce the idea that some emergency mechanisms may reduce excessively high core temperatures. What these mechanisms may be is not known. ACTH and corticosteroids. however, are thought to limit the febrile rise in temperature [21], and the secretion rate of these hormones increases when the preoptic area is cooled [ fO,lS]. The mechanism of the antipyretic action of corticosteroids is not known. ACTH, however, appears to inhibit activity in heat conservation and heat production effector pathways 1211. Thus, it seems possible that ACTH. and perhaps corticosteroids, may also limit the rise in temperature induced by intense and prolonged cooling of the preoptic area. Since the rate of metabolic processes is temperature dependent, it is often assumed that the production of antibodies in mammals changes in parallel to body temperature [28]. Experiments in lower vertebrates tend to support this view [2]. Our experiments, however, show a complex relationship between body temperature and antibody titre (Table 1, Fig. 4) and thus challenge this view. The animals in which the thermode was cooled to an average of 19°C had the highest metabolic response but core temperature rose only to 39.S”C (Table I), probably because much of the heat produced was lost to the thermode. Since these animals had a high titre of anitbodies, it seems reason-
able to suggest that the nonthermnl responses elicited by cooling the preoptic area are not immunosuppressive. Therefore, the low titre of the animals in which core temperature was increased to 39.9”C can he attributed to a depressive effect of high temperature. This agrees with experiments itr t+ro, which show that temperatures above 39-40°C inhibit the production of antibodies Jl8,2?]. Hypothermia normally decreases the production ot antibodies in mammals [26.28]. and the mononucle~i! phagocytic system, a system which appears to be necessary for the development of the primary immune response [23]. is also inhibited by low tempetatures ;,I t*iiw [ 171. Resides. the in vitro response of lymphocytes to polyclonal mitogens is decreased at 35°C’ [I]. Thus. the decreasing immune rcsponse associated with large falls in core temperature can also be attributed to a direct effect of temperature. Changes in temperature of up to 3”r, however. \VCIC associated with a high titre of antibodies. This immune stimulation could have been the consequence of a direct temperature effect--small changes in temperature inducing stimulation, whereas large changes inducing depression of the immune response. This, however, implies that there are two optimal temperatures, perhaps reflecting the optimal range of distinct lymphoctye subs&. On the other hand, caution should be exercised before this stimulating immune effect of heating or cooling the anterior hypothalamus ih at-tributed to the change in body temperature, as some 111r?rl-tt experiments appear to suggest 119.271. The viability of lymphocytes in t*itro is often very low 1131 and temperature de pendent [13,18], and the results of these experiments may thus be biassed. Besides. the thermal sensitivity of an intact organism is higher than that of its isolated components 1301. Thus. the possibility that any change in body temperaturr may even depress the function of lymphocytes i/l I+,Y~cannot be excluded. If it were so. what could then be the uusc of the immune stimuIation induced by heating and cooling the hypoth~amus~~ Previous experiments have shown that nonthermal factors determine the immune effect of elevated body temperature [6], whereas the present ones show that thermal stimulation of the hypothalamus can induce the same degree of immunostimulation independently of the direction in which body temperature changes. Therefore. ws would like to suggest an alternative hypothesis for further consideration, and that is that the immune stimulation elicited by cooling and heating the preoptic area is mediated hy factors other than the change in body temperature. The nature of these nonthermaf factors is not known at present. However. the neuroendocrine system appears to play a pivotal role in the control of immune function (8.121. and hy~th~amic temperature influences the secretion rate of most hormones. In this context, it may be of interest to note that the only common endocrine responses to hypothalamic cold and warm stimuli appear to be an increase in the becretion rates of ACTH and corticosteroids (J IO. I I. I.‘] but SW [24]). These hormones, which at physiological levels appeal to stimulate the immune system [ 131, are thought to be the main mediators of the known effects of the hypothalamus on immunity [29]. Since the local cooling of the preoptic area inducea 11 fever-like response [3]. the results of the present cvxriments are compatible with the idea that certain febrile nonthermal responses may enhance the defences of the host animal [4,5]. However, whether this enhancement of the host’s defences has any reievant effect on the course of infectious disease is yet to he shown.
HYPOTHALAMIC
TEMPERATURE
AND IMMUNE
RESPONSE
251
REFERENCES 1. Ashman, R. B. and A. J. Nahmias. Enhancement of human lymphocyte responses to phytomitogens in vitro by incubation at elevated temperatures. Clin Exp Irnmunol 29: 464-467, 1977. 2. Avtalion, R. R., A. Wojdani, Z. Malik, R. Shahrabani and M. Duczyminer. Influence of environmental temperature on the immune response in fish. Curr Top Microbial Immunol61: l-35, 1973. 3. Banet, M. Fever and survival in the rat. The effect of enhancing fever. Pflugers Arch 381: 35-38, 1979. 4. Banet, M. Fever and survival in the rat. Metabolic versus temperature response. Experientia 37: 1302-1304, 1981. 5. Banet, M. The biological function of fever: an alternative view. Funkt Biol Med 2: 21 l-218, 1983. 6. Banet, M., D. Fischer, K. U. Hartmann, H. Hensel, and U. Hilling. The effect of whole body heat exposure and of cooling the hypothalamus on antibody titre in the rat. PjTugers Arch 391: 25-27, 1981. 7. Banet, M., S. Brandt and H. Hensel. The effect of continuously cooling the hypothalamic preoptic area on antibody titre in the rat. Experienfia 38: 965-966, 1982. 8. Besedovsky, H. and E. Sorkin. Network of immuneneuroendocrine interactions. Clin Exp Immunol27: l-12, 1977. 9. Bligh, J. Temperature Regulution in Mammals and Other Vertebrates. Amsterdam: North-Holland, 1973. 10. Chowers, I., H. T. Hammel, S. B. Strommeand S. M. McCann. Comparison of effect of environmental and preoptic cooling on plasma cortisol levels. Am J Physiol 207: 577-582, 1964. Il. Chowers, I., H. T. Hammel, J. Eisenman, R. M. Abrams and S. M. McCann. Comparison of effect of environmental and preoptic heating and pyrogen on plasma cortisol. Am J Physiol 210: 606-610, i966. _. 12. Comsa, J., H. Leonhardt and H. Werkele. Hormonal coordination of the immune response. Rev Physiol Biochem Pharmacol 92: 115-191, 1982. _ 13. Duff, G. W. and S. K. Durum. Fever and immunoregulation: Hyperthermia, interleukins 1 and 2, and T-cell proliferation. Yale J Biol Med 55: 437-442, 1982. 14. Fantino, M. F. Antirotatory device permitting free movement in animals with several catheters. J Appl Physiol 42: 275-280, 1977. 15. Gale, C. C., M. Jobin, D. W. Proppe, D. Notter and H. Fox.
Endocrine thermoregulatory responses to local hypothalamic cooling in unanesthetized baboons. Am J Physiol 219: 193-201, 1970.
16. Greville, T. N. E. Theory and Applications of Spline Functions. New York: Academic, 1969. 17. Huggins, C. and W. J. Noonan. An increase in reticuloendothelial cells in outlying bone marrow consequent upon a local increase in temperature. J Exr, Med 64: 275-280, 1936 18. Izumi, A., S. Koga-and M. Maeta. Effects of in vitro hyperthen-ma on murine and human lymphocytes. Cancer 51: 20612065, 1983. 19. Jampel, H. D., G. W. Duff, R. K. Gershon, E. Atkins and S. K. Durum. Fever and immunoregulation. III. Hyperthermia augments the primary in vitro humoral immune response. J Exp Med 157: 1229-1238, 1983. 20. Kluger, M. J., A. J. Thurnbull, W. I. Cranston, A. J. Wing, M. P. Cross and B. A. Rothenburg. Endogenous cryogen excreted by the kidneys. Am J Physiol 241: R271-R276, 1981. 21. Lipton, J. M., J. R. Glyn and J. A. Zimmer. ACTH and a-melanotropin in central temperature control. Fed Proc 40: 2760-2764, 1981. 22. Marat, B. A. Effect of the body temperature change on proceeding of the primary immune response in rabbits. Fiziol Zh SSSR 11: 1707-1709. 1981. 23. Pierce, C. W. and J. A. Kapp. The role of macrophages in antibody responses in vitro. In: Immunobiology of the Macrophage, edited by D. S. Nelson. New York: Academic, 1976, pp. l-31. 24. Proppe, D. W. and C. C. Gale. Endocrine thermoregulatory responses to local hypothalamic warming in unanesthetized baboons. Am J Physiol 219: 202-207, 1970. 25. Riley, V. Adaptation of orbital bleeding technic to rapid serial blood studies. Proc Sot Exp Biol Med 104: 751-754, 1960. 26. Roberts, N. J. Temperature and host defence. Microbial Rev 43: 241-259,
1979.
27. Saririan, K. and D. A. Nickerson. Enhancement of murine in vitro antibody formation by hyperthermia. Cell Immunol 74: 306312,
1982.
28. Schmidt, K. L. and V. R. Ott. Immunreaktionen bei hoher Korpertemperatur. Med Welt 25: 1963-1%8, 1974. 29. Stein, M., S. J. Schleifer and S. E. Keller. Hypothalamic influences on immune responses. In: Psychoneuroimmunology, edited by R. Ader. New York: Academic Press, 1981, pp. 429447. B. Thermostability of cells and proteins of 30. Ushakov, poikilotherms and its significance in speciation. Physiol Rev 44: Sl8-560,
1964.