Does immune challenge affect torpor duration?



1. Hibernation may alter the relationship between pathogens and their hosts; low host temperatures can prevent pathogen replication. Therefore, manipulating the timing and duration of torpor bouts could allow animals to gain an advantage over pathogens.

2. Thirty-two Turkish Hamsters (Mesocricetus brandti) were placed in short-day, cold conditions. After 10 weeks, 20 animals were challenged with an antigen to simulate a pathogen infection. Ten of these animals were returned to the cold (‘cold-challenged’). The other 10 animals were placed in warm conditions (‘warm-challenged’). Twelve animals received saline injections and were returned to the cold (‘cold-control’). Cold-challenged animals spent significantly more time in torpor than did cold-control animals.

3. After 6·5 weeks, all animals were housed in warm conditions and ceased torpor. Both cold-challenged and warm-challenged animals received a second injection of antigen. There was no correlation between time spent euthermic and level of secondary humoral response of cold-challenged animals. The secondary humoral response of the cold-challenged animals was significantly lower than that of warm-challenged animals.

4. In this study immune status influenced torpor duration, and torpor caused immunosuppression. Hibernators may manipulate body temperature in order to combat pathogens while their own immune systems are suppressed.


Many animals respond to environmental variation by entering hibernation for part of the year. The hibernation season comprises a series of torpor bouts (when body temperature and metabolism decrease greatly) interrupted by bouts of euthermy. The length and frequency of these bouts depend on the species, time of year, diet and environmental factors (Kristoffersson 1967; French 1976, 1982; Cranford 1978; Collins & Cameron 1984; Tannenbaum & Pivorum 1984; Geiser & Kenagy 1988; Thorp, Ram & Florant 1994; Thomas & Geiser 1997). Animals might also benefit by regulating torpor in response to immune status.

The alternations between torpor and euthermy create wide variations in temperature and cellular activity. These variations could potentially alter interactions between hosts and their pathogens. In fact there is evidence that torpor can confer resistance to pathogens. Beginning in the early 1900s experimenters infected hibernators with a variety of pathogens. Many studies found that torpor resulted in reduction or elimination of internal pathogens (Chute 1961; Kayser 1961, pp. 222–230; Lyman et al. 1982, pp. 196–198). Reduced food consumption may have controlled some parasites, but temperature may be much more important. Mesophilic organisms (including most internal pathogens) have optimal growth temperatures between 35 and 40 °C and cannot proliferate during torpor because of the host’s low body temperature. Mitosis at low temperatures leads to damage caused by disrupted microtubulin polymerization (Roth 1967; Nagasawa & Dewey 1972; Boltovskaya 1977; Vinogradova 1982), so neither host nor pathogen cells can undergo mitosis without risking chromosomal damage. Hibernators appear facultatively to prevent mitosis at low temperatures to avoid chromosome damage (Jaroslow et al. 1976; Kolaeva et al. 1980; Kruman et al. 1988a). DNA synthesis is also inhibited in ground squirrels throughout the hibernation season, even in the absence of torpor (Kruman et al. 1986). In contrast, mitotic activity during intertorpor euthermic bouts may be higher than normal (Vinogradova 1988). Cells may accumulate in the G2 phase during torpor, then undergo mitosis during euthermic bouts (Kruman et al. 1988b).

Pathogens are unable to proliferate while their hosts are torpid, but host functions (including immune cell proliferation) are also restricted (Jaroslow & Smith 1961; Jaroslow 1968, 1971; McKenna & Musacchia 1968; Shivatcheva 1988). Most immunological activity may be restricted to arousal bouts (Jaroslow & Serrell 1972; Shivatcheva 1988). This creates the possibility that a hibernator could manipulate the timing of its torpor bouts to maximize both the damage to pathogens and the effectiveness of its own immune response. For example, an animal infected with pathogens might spend just enough time euthermic to allow proliferation of plasma cells before returning to torpor and causing cessation of pathogen proliferation.

The main purpose of this experiment was to determine whether hibernating animals would adjust their torpor schedule in response to an immunological challenge. Other objectives were to determine whether the humoral response was related to the amount of time spent torpid and to compare the humoral responses of hibernating animals with those of euthermic animals. Actual pathogen infection could cause a variety of physiological changes that might affect time in torpor. To eliminate these confounding effects, a novel antigen was used as a challenging agent.

Materials and methods

The Turkish Hamsters (Mesocricetus auratus) used in this study were provided by Dr Georgiana Jagiello. They were descended from individuals in Dr Bruce Goldman’s colony, which derived from animals imported by Dr Charles Lyman. The original 14 animals were bred to obtain a total of 32 animals. During the breeding cycle, animals were maintained on a cycle of 16:8 (L:D), at ≈ 22 °C. Standard rat chow and water were provided ad libitum. Supplemental wheat germ and oatmeal were provided for reproduction and growth of juveniles. Hamsters were housed in clear plastic cages (22 cm by 42·5 cm by 20·5 cm) with wood shavings for bedding.

On 2 September 1997, the L:D schedule for all animals was switched to 10:14. On 4 October, temperature was decreased to ≈ 10 °C (8–12 °C). On 18 December, the 10 animals that had failed to begin the hibernation cycle (defined as having spent at least 2 days in torpor during the preceding week) were returned to 22 °C and 12:12 (L:D). These animals are referred to as ‘warm-challenged’. The remaining 22 animals were exposed to either the novel antigen or the saline control (randomly assigned) before return to the cold, short-day conditions, and are referred to as ‘cold-challenged’ and ‘cold-control’, respectively.


On 18 December 1997, torpid animals were handled briefly to trigger awakening from torpor, then all were anaesthetized with Metofane (methoxyflurane; Schering-Plough Corp., NJ). Opthaine (proparacaine HCl, P-4554; Sigma; MO) was used as a topical anaesthetic on the eye, and a blood sample of ≈ 75 μl was taken from the postorbital sinus. All animals were then injected subcutaneously on the dorsal thorax. Experimental animals (both warm- and cold-challenged) received 25 μg of hen eggwhite lysozyme (HEL, Sigma L-6876; MO) in 250 μl of phosphate-buffered saline (PBS). Cold-control animals received injections of 250 μl of PBS. The type of injection was unknown to the experimenter. On 3 February 1998, all animals were returned to 22 °C and 12:12 L:D to cause arousal from torpor. Within hours, all animals were active. On 4 February 1998, a second blood sample was taken from both warm- and cold-challenged animals. Animals in both groups were given a second injection of antigen. On 9 February (at the expected peak for a secondary response), a third blood sample was taken from the warm- and cold-challenged animals. Samples were refrigerated overnight, then centrifuged. Serum was stored at – 20 °C. The serum samples were analysed for anti-HEL IgG using enzyme-linked immunosorbent assay (ELISA).


The state of the animals (torpid or non-torpid) was monitored via daily observations between 0900 and 1100 hours. Animals in a hibernating posture were tested for responsiveness using a puff of air directed at the dorsal area (Hall, Bartke & Goldman 1982). Torpid animals were sprinkled with oatmeal to allow the detection of arousal bouts between examinations (Lyman & O’Brien 1977).


During the first 38 days of the hibernation period, cold-challenged animals spent an average of 30·1 days in torpor (SE = 2·5, n = 10), significantly more time than the 23·2 days spent torpid by cold-control animals (SE = 1·3, n = 12, P = 0·03, unpaired t-test). After 38 days, both cold-challenged and cold-control animals had similar patterns (Fig. 1). Cold-control animals showed a non-significant trend toward longer maximum active bouts (average 7·2 days, SE = 0·22 days vs 2·4, SE = 0·4, P = 0·07, unpaired t-test). The number of animals in torpor was not correlated with temperature or humidity in the environmental chamber during the cold, short-day cycle.

Figure 1.

. Proportion of Turkish Hamsters (Mesocricetus brandti) in torpor for animals challenged with hen eggwhite lysozyme and a control group (n = 10 animals in cold-challenged, 12 in cold-control). Both groups were maintained at 10 °C (8–12 °C) with an L:D of 10:14. Hamsters challenged with the antigen spent significantly more time in torpor (P = 0·03, unpaired t-test).

At the termination of the cold-exposure period, cold-challenged animals demonstrated virtually no increase in circulating anti-HEL antibodies compared with pre-exposure levels (0, SE = 7%; Fig. 2), while warm-challenged animals had measurable levels of circulating anti-HEL antibodies (18, SE = 8%); however, the difference was not significant (P = 0·13, Mann–Whitney U-test). Five days after the second antigen exposure and return to warm conditions, the response of cold-challenged animals was still very low (10, SE = 9%) compared to warm-challenged animals (74, SE = 30%, P = 0·049, Mann–Whitney U-test). However, both warm- and cold-challenged animals had a significant humoral response to HEL on 9 February (P = 0·009 and P = 0·046, Wilcoxon matched pairs test). There was no relationship between time spent torpid and anti-HEL antibody level for the cold-challenged animals.

Figure 2.

. Turkish Hamster (Mesocricetus brandti) humoral immune responses to hen eggwhite lysozyme (HEL) as measured by ELISA. Cold-challenged animals had been kept at ≈ 10 °C, 10:14 (L:D) and had been cycling in and out of torpor. Warm-challenged animals had been kept at ≈ 22 °C, 12:12 (L:D), and had been euthermic for at least 6 weeks (n = 10 in each group). ‘Circulating’ indicates the level of anti-HEL IgG at 46 days after initial HEL exposure. ‘Secondary’ indicates the level 5 days after the second exposure. Both groups were housed at 22 °C, 12:12 beginning the day before the second antigen exposure. The secondary response of warm-challenged animals was significantly higher than that of cold-challenged animals (P = 0·049).


Hibernation may influence the balance between pathogens and host immune systems. A host benefits by manipulating the timing and duration of torpor bouts to aid in defence against pathogens. Mean torpor and euthermic bout lengths for cold-control animals in this study were similar to those observed previously for ancestors of this group of M. brandti (Jagiello et al. 1992). Cold-challenged hamsters in this study responded to an immune challenge by increasing the amount of time spent in torpor. The cold-challenged animals also tended to spend shorter time periods in euthermy.

Manipulation of euthermic bout frequency and duration could maximize the disruption of pathogen replication. A hibernator could accumulate products necessary for mounting an immune response while still torpid. After these products had reached a certain level, the animal could quickly arouse, allow its cells to undergo mitosis, then drop its temperature again. There are two benefits to this strategy. First, an animal could remain euthermic for exactly the amount of time that would create the greatest benefit to the host while minimizing proliferation time for pathogens. Second, a sudden decrease in body temperature might catch pathogen cells in the midst of their mitotic phase, thus potentially damaging those cells. The hibernator’s own cells would be far less vulnerable to cold-induced chromosome damage due to precooling mitotic inhibition. Repeated cycles of this might allow a hibernator to gain an advantage over pathogens during the course of the hibernation season.

There is some evidence that hibernators are adapted for maximizing their mitotic activity during arousal bouts. Cells in the digestive tracts of ground squirrels (Spermophilus) appear to show normal gene expression during hibernation, but little or no proliferation until immediately after arousing from torpor, when there is a dramatic increase in proliferation and migration to levels seen in active animals (Carey & Martin 1996). Analysis of cells from hibernating dormice (Muscardinus avellanarius) demonstrates the presence of nuclear bodies never observed in euthermic animals. These seem to be storage forms of newly transcribed RNA and ribonuclear proteins needed for RNA processing (Malatesta et al. 1994, 1995). The researchers who described this phenomenon speculate that this allows large amounts of pre-mRNA to be slowly synthesized during hibernation, then released quickly during early arousal. Antibody formation could also occur (slowly) during torpor as long as the proliferation of plasma cells happened during arousal bouts.

In the current experiment, animals that had recently been hibernating (cold-challenged) had much lower secondary responses than the non-hibernating animals (warm-challenged). The warm-control animals did not hibernate under the cold, short-day conditions, indicating pre-existing differences between the groups; however, this result does demonstrate that hamsters in this colony can mount an immune response to HEL. Had the warm-challenged animals failed to hibernate owing to health defects or poor nutrition, a lower immune response would have been expected. Instead their response was much more robust than that of the hibernators.

In contrast to the post-hibernation immunosuppression observed in this study, and the seasonal immunosuppression described in non-torpid hibernators (Sidky, Hayward & Ruth 1972), there is some evidence that the immune system of hibernators may be adapted to rapid activation and enhanced activity on arousal from torpor. For example, circulating lymphocytes in ground squirrels and rats are reduced by 50% during hypothermia. In the non-hibernator, it takes 12 h for levels to return to normal. In the hibernator, levels return to normal almost immediately on rewarming and are twice normal values 12 h later (Andjus et al. 1971). In hedgehogs (Erinaceus europaeus), leucocytes attain normal levels within 3 h of arousal, then fall sharply on return to hibernation (Suomalainen & Rosokivi 1973). Leucocytes may migrate from blood into intestine and lung tissue when an animals is torpid, then move back into blood during arousal periods (Inkovaara & Suomalainen 1973).

Despite immune inhibition, hibernators seem to rid themselves of pathogens during hibernation (Kayser 1961; pp. 222–230). The ability of hibernators to defeat pathogens during a time of immunosuppression may be explained by the low body temperatures that characterize torpor. The inhibitory effect of cold may well augment the host defences so that a lower humoral immune response is necessary. Short arousal bouts may be more beneficial than long ones because pathogens are given minimal time for proliferation. If a ground squirrel has a few days of euthermy between antigen presentation and torpor, it can produce peak numbers of plasma cells that are only slightly below normal, while animals exposed to antigen during hibernation have a very weak response (McKenna & Musacchia 1968; Jaroslow 1971). Not only is the response weaker than normal, it takes about seven times longer to develop in a hibernating animal. However, pathogen reproduction may be slowed too. Secondary responses, which are less dependent on cell proliferation, appear to develop (albeit slowly) during torpor (Jaroslow 1968). These findings demonstrate that there are advantages to brief periods of euthermia during which lymphocytes can process antigens, differentiate and begin proliferation.

Observed differences between cold-challenged and cold-control animals in this experiment disappeared after 38 days. The exact time required for mounting an immune response in hibernation may vary with an individual’s changing metabolic rate between the torpid and euthermic states. This means that individuals challenged at the same time with the same antigen may not produce a peak immune response at the same time. Continuous monitoring of circulating antibodies via a method such as aortic cannulation (Pengelley, Asmundson & Uhlman 1971) would help to determine the peak of humoral response in hibernators.

There was no relationship between serum levels of anti-HEL antibody and time in torpor. A similar lack of correlation between antibody production and time in euthermy was observed in ground squirrels (Spermophilus tridecemlineatus; McKenna & Musacchia 1968). One possible explanation for this phenomenon is that there is a threshold of required time in euthermy. If this is the case, an assay performed earlier or later in the postchallenge period may have revealed differences between individuals related to time in torpor. Again, continuous monitoring would answer this question.

Immunological status (challenged vs unchallenged) is certainly not the only factor that influences timing and duration of torpor. Diet, season and environmental factors clearly influence the timing and duration of torpor bouts as well. However, challenge to the immune system apparently can change an animal’s torpor schedule. Therefore, the influence of immune status on hibernation must be considered if we are to get a clearer picture of the costs and benefits of torpor vs euthermy during the hibernation season.

Changes in torpor pattern are likely to influence other facets of hibernation ecology. Although there are clear benefits to torpor, food manipulation experiments demonstrate that hibernators of some species will remain euthermic if the food supply allows it (Brown & Bartholomew 1969; French 1976; Reichman & Brown 1979). This may be because hibernation increases vulnerability to predation (Young 1990). If an animal stores food in its burrow, it cannot protect its cache while torpid. Therefore, increasing time in torpor may save some energy that would be spent maintaining euthermic body temperature, but there are apparently costs to this strategy.

Different species show very different torpor patterns and might differ in response to pathogens. Species that have very short, infrequent euthermic bouts might actually benefit by increasing time in euthermy in response to pathogen infection. This would allow hibernators to have brief periods of B-cell proliferation so that antibodies could be produced during torpor. Further comparative studies of this phenomenon are needed in order to get a clearer picture of how pathogens affect the ecology of hibernation.


We thank Dr Georgiana Jagiello of Columbia University for the generous donation of the hamsters used in this study. M. Archibeq and the rest of the UCSB Central Vivarium staff provided excellent logistical support. Dr R. Jacobs shared laboratory equipment. This work was conducted while R.S.B. was a Postdoctoral Associate at the National Center for Ecological Analysis and Synthesis, a Center funded by NSF (Grant #DEB-94–21535), the University of California – Santa Barbara, and the State of California.