We determined baseline and capture-induced glucocorticoid concentrations during two different seasons in three species of wild free-living rodents: brown lemmings (Lemmus trimucronatus), golden-mantled ground squirrels (Spermophilus saturatus), and yellow-pine chipmunks (Tamias amoenus). Initial blood samples were obtained within 3 min of capture, so that initial glucocorticoid levels reflect baseline titers of undisturbed animals. Animals were held for an additional 30 min, when a second blood sample was taken to measure stress-induced glucocorticoid titers. The primary glucocorticoid differed in each species. Lemmings secreted extremely large amounts of corticosterone (as high as 8,000 ng/ml). These high concentrations were accompanied by high corticosterone-binding globulin capacity and resistance to negative feedback. Squirrels and chipmunks secreted a mixture of cortisol and corticosterone (10–400 ng/ml). In males of all three species and female squirrels and chipmunks, glucocorticoid levels were significantly elevated 30 min after capture. Baseline and 30-min glucocorticoid levels differed seasonally in each species. Levels were higher during summer (with no snow cover) than in spring (with ∼60% snow cover) in female lemmings, higher during breeding than before hibernation in squirrels, and higher postreproductively than during breeding in chipmunks. Together, these data indicate that glucocorticoid responses to stress in these free-living species are similar to those in laboratory species, but the magnitude of the response appears to depend on life-history features specific to each species.
- corticosterone-binding globulin
all wild animals must cope with adverse environmental conditions (stressors) to survive. When encountering an adverse condition (e.g., a storm), animals initiate a suite of physiological and psychological responses designed to counteract the effects of the stressor (e.g., move to a different habitat) or adapt to the stressor (e.g., expend energy to regulate body temperature). Among the classic mediators of these physiological and psychological responses are glucocorticoids, which are released within a few minutes after the initiation of a stressor (65).
Glucocorticoid release in response to acute stressors seems crucial for survival, since even mild noxious stimuli can cause death in adrenalectomized rats (24), and small amounts of replacement glucocorticoids fail to rescue these animals (23). The importance of glucocorticoids for survival is supported by field studies. Cortisol levels are elevated in wild free-living snowshoe hares that survive a population crash (11), cortisol levels are 70 times higher in wild deer shot after a hunt (i.e., being chased by hunting dogs for hours) than in deer shot without a hunt (43), and yearly survival is poorer in cliff swallows with very low or very high corticosterone concentrations (15). High cortisol concentrations during captivity in wild European rabbits predicted survival after release (18), and survival of Galápagos marine iguanas during El Niño famine conditions is directly correlated with glucocorticoid concentrations (58). Consequently, the importance of glucocorticoids in survival, combined with their ability to regulate energy stores (22) and, in the long-term, to suppress the gonadal axis (27, 47, 62), the system important for species, rather than individual, survival, suggests that they play a fundamental role in the Darwinian fitness of an animal.
In contrast, long-term, as well as short-term, glucocorticoid exposure can lead to a multitude of deleterious effects, including immune suppression, growth inhibition, reproductive failure, and neuron death (63). These deleterious aspects of glucocorticoid action have been the focus of most of the work on stress. However, these effects are usually pathological, and the study of normal glucocorticoid physiology has often been neglected (65). Presumably, a balance must be struck between needing glucocorticoids to survive a stressor and modulating glucocorticoid secretion to prevent deleterious exposures.
Given the link between glucocorticoid secretion and survival, it is important to study naturally selected populations. Free-living mammals are still under the influence of natural selection, so that a proper physiological response to adverse environmental conditions is vital for their survival. Artificially selected populations (e.g., laboratory rats and mice) do not afford us this opportunity. Furthermore, dramatic differences in hypothalamic-pituitary-adrenal function of animals recently brought into captivity (4, 29, 59) can persist for generations (33, 40, 42).
A recent review indicated that many wild free-living animals can seasonally modulate glucocorticoid secretion (56). In other words, the baseline concentrations and the magnitude of glucocorticoid release in response to stressors vary depending on the time of year. These changes have not been described in standard domesticated laboratory animals. There is a robust elevation of glucocorticoid concentrations concurrent with breeding in ∼75% of the amphibian, reptile, and bird species that have been studied (56). Mammalian species, however, show more equivocal patterns: some show higher glucocorticoid concentrations during the breeding season, others show higher concentrations outside the breeding season, and still others show no seasonal changes (55). Very few studies have been carried out in mammalian species, however, and even fewer have been able to determine true baseline (unstressed) concentrations (i.e., collected plasma samples within 3 min of capture).
To determine how common seasonal modulation of glucocorticoid release is in mammals, we determined prestressed baseline and stress-induced glucocorticoid concentrations in three wild rodents at two different times of the year. A further important goal was to ensure that we had good estimates of baseline glucocorticoid concentrations by collecting the initial blood sample only within 3 min of capture.
MATERIALS AND METHODS
Species and capture protocols.
All procedures were performed according to Association for Assessment and Accreditation of Laboratory Animal Care guidelines and approved by the University of Washington Institutional Animal Care and Use Committee.
Brown lemmings (Lemmus trimucronatus) are small (adults are 30–110 g) herbivorous rodents that are easy to handle and manipulate. They range across the high Arctic and specialize in flat tundra (35, 70). We captured our animals during a high-density year near Barrow, AK (71.1° N, 156.4° W). Brown lemmings are relatively easy to catch: because of the permafrost in the Arctic, it is impossible for the animals to burrow. Consequently, lemmings rely on grass-covered runways for cover from predators. It was thus possible for us to traverse the tundra, spot a lemming, and, if we acted quickly, capture the animal by hand. We could usually catch animals in <30 s from detection, although we do not know whether they began to mount a response before being spotted. This process is labor intensive, but it ensured that we collected plasma samples within 3 min of capture. We captured brown lemmings twice during the year: in early spring when the snow cover was >50% (12–14 June 1995 and 21–26 June 1997) and during the late summer with no snow cover (26 July–1 August 1995). These two periods differed in the amount of green forage and, possibly, in the predation risks [snow cover protects the animals from avian predators (personal observations; 53)]. We chose brown lemmings, because they live in the same habitat where seasonal modulation of the glucocorticoid response had been shown in Arctic birds (56).
The Cascade golden-mantled ground squirrel (Spermophilus saturatus) is a medium-sized (∼150–300 g) rodent distributed in the Cascade Mountains of Washington State and nearby southern British Columbia. They are obligate hibernators that store large amounts of body fat to survive the winter. We used baited Tomahawk traps to capture our animals from a population of individually marked animals that was being monitored from 1984 through the date of the present study at a study site in Chelan County, WA (39). To ensure that the first blood sample was taken within 3 min of the animal being trapped, a limited number of traps (i.e., 10–15) were set within hearing distance of the trapper. A trap being sprung made a noise that informed the trapper that an animal had been captured. The trapper then rapidly collected the animal and took a blood sample. Once captured, animals are easy to handle and manipulate. We captured golden-mantled ground squirrels twice during the year: in the spring, just as animals emerge from hibernation and are starting to breed (24–26 April 1995), and in the fall, just before animals return to their hibernaculae (21–23 August 1995).
Yellow-pine chipmunks (Tamias amoenus) are small rodents (adults are 38–71 g) that also live in montane regions of the western United States (17). Similar to golden-mantled ground squirrels, yellow-pine chipmunks are obligate hibernators. However, yellow-pine chipmunks cache food for consumption during the winter and do not significantly increase fat storage before winter. We captured animals from the same field site, from the same traps using the same techniques, and at the same times as the golden-mantled ground squirrels.
Once captured, animals were subjected to the identical stressors of capture and restraint. Each animal was quickly lightly anesthetized with methyoxyflurane (Metofane) or ether. Blood (∼40 μl) was removed within 3 min of capture for baseline glucocorticoid measurements from the infraorbital sinus with use of heparinized microhematocrit capillary tubes. Since glucocorticoid levels generally do not start to rise until 3 min after stress initiation (57, 65), all samples taken within 3 min were considered to reflect baseline levels. Animals were then placed in opaque cloth bags or returned to the trap for a 30-min restraint period; then they were again lightly anesthetized, and blood (∼40 μl) was removed for stimulated glucocorticoid measurements. Because preliminary data suggested that glucocorticoid levels plateaued at 30 min, we did not restrain the animals for >30 min.
The identical stressor of capture, handling, anesthesia, and restraint provided a standardized stressful stimulus that was used to compare stress responses across seasons. Inherent in these experiments was the assumption that capture, handling, anesthesia, and restraint maximally activated stress responses, so that differences resulted from physiological constraints, rather than changes in the perceived stressfulness of the stimuli (78–80).
Field test of negative feedback.
A dexamethasone (DEX) suppression test was administered to adult brown lemmings. DEX (1 mg/kg body wt im) was administered immediately following collection of the sample after 30 min of restraint, and glucocorticoid levels were assessed after 3 h, as described by Carroll et al. (19) and modified by Sapolsky and Altmann (64) for use in free-living animals. This dose has been shown to be effective in several other wild species (45, 64).
Sample processing and assays.
Samples were stored on ice in the field for up to 16 h. They were then centrifuged, and the plasma was removed and stored at −20°C. Corticosterone was assayed by RIA following the methods of Wingfield et al. (81). Briefly, all samples were equilibrated with ∼2,000 cpm of tritiated corticosterone (for subsequent recovery analysis) and extracted with dichloromethane. Dried extracts were redissolved in buffer, and corticosterone levels were assessed using a known amount of hot steroid and a corticosterone antibody (Endocrine Sciences); then dextran-coated charcoal was used to separate bound from unbound steroid. Samples were compared with a standard curve, with values adjusted by the percent recovery. Inter- and intra-assay coefficients of variation were <11% and 5%, respectively, as determined by running a plasma standard in each assay.
Cortisol was assayed with a 125I-labeled cortisol RIA kit (INCSTAR, Stillwater, MN) using the manufacturer's recommended procedure. Inter- and intra-assay coefficients of variation were <15% and 8%, respectively. Although a different glucocorticoid (e.g., cortisone) could also be present in these species, there was insufficient plasma to test for this possibility.
Corticosterone-binding globulin (CBG) capacity and affinity (Kd) for lemmings were determined by generation of a steroid-free plasma and measurement of radiolabeled corticosterone binding following the methods of Orchinik et al. (51). Briefly, samples were stripped of endogenous corticosterone with 2× volume of dextran-coated charcoal, incubated for 30 min at 55°C, and centrifuged for 10 min at 2,000 g. The final assay dilution was 1:600. For the saturation curve, extracted samples were combined with 0.25–14 nM [3H]corticosterone for 1 h at room temperature and then for 15 min at 4°C. Free and bound tritiated steroids were separated by rapid filtration over Whatman GF/B filters. Filter-bound radioactivity was then quantified by standard liquid scintillation spectroscopy. Kd was 10.79 ± 0.705 and the maximum binding was 6,809 ± 215.8 nM (Fig. 1). Six tubes were used for each individual sample: three tubes contained unlabeled corticosterone for determination of nonspecific binding, and three tubes contained assay buffer for determination of total binding. We used 44 nM [3H]corticosterone to determine CBG capacity for individual samples. This concentration should occupy 80% of binding sites on the basis of our calculated Kd estimates. Free corticosterone was estimated using the equation of Barsano and Baumann (6).
Changes in hormone levels over time (the stress response) and seasonal differences in baseline and stress-induced glucocorticoid levels were compared with repeated-measures ANOVA using Proc Mixed in SAS (version 9.1). This technique specifically allows repeated-measures ANOVA to be performed with missing values, so that the few records with single values (animals with only a baseline or only a 30-min sample when the other sample was lost) were not excluded. Analyses initially included sex and weight of the animal as cofactors, but there was no effect of either variable (P > 0.05) for chipmunks or ground squirrels. Consequently, sex and weight were removed, and the analyses were recalculated for these species. There was a significant effect of sex (F1,111 = 19.43, P < 0.0001) and weight (F1,111 = 6.22, P < 0.02) for lemmings. Consequently, separate analyses were performed for males and females, and weight was maintained as a cofactor. Nonparametric Mann-Whitney U tests were used for comparisons of lemming CBG and free corticosterone, because these data were not normally distributed according to Shapiro-Wilk W goodness-of-fit tests (all P < 0.05).
Corticosterone appears to be the primary glucocorticoid in lemmings. Cortisol levels were <1% of the corticosterone levels (data not shown), which is near the cross-reactivity for corticosterone of the cortisol antibody used in the assay. However, corticosterone levels in this species were extraordinarily high, with higher concentrations in females than in males (Fig. 2). The sex difference in corticosterone concentrations was paralleled by a sex difference in CBG capacity: capacity was higher in females than in males (Z1,12 = −4.48, P < 0.001; Fig. 3).
Male brown lemmings significantly elevated their corticosterone concentrations in response to 30 min of restraint and handling (Fig. 2A; F2,39 = 8.56, P < 0.001 for overall effect of sample time). This response was related to weight (F1,39 = 12.78, P < 0.001): smaller males showed a larger increase. Females, however, did not alter their corticosterone concentrations over the sampling period. Concentrations were extremely high and did not increase further in response to capture, handling, and restraint (Fig. 2B; F2,66 = 0.00, P = 0.99 for overall effect of sample time). Moreover, DEX failed to reduce corticosterone concentrations in either sex.
Female brown lemmings showed seasonal differences in corticosterone release (Fig. 2B; F1,66 = 18.70, P < 0.0001 for overall effect of season), but there was no seasonal change in their lack of response to capture, handling, restraint, and DEX injection (F2,66 = 0.49, P = 0.61 for season-sample time interaction). Throughout the sampling period, corticosterone concentrations were dramatically higher in July after the snow had melted. The seasonal change was related to weight (F1,66 = 4.97, P < 0.03): smaller females showed a greater increase in corticosterone. However, CBG capacity did not change seasonally in females (Fig. 3A; Z1,12 = 0.64, P = 0.52). In contrast, male brown lemmings showed seasonal differences in CBG capacity (Fig. 3A; Z1,12 = 2.68, P = 0.007) but not in corticosterone concentrations (Fig. 2A; F1,39 = 0.92, P = 0.34 for overall effect of season; F2,39 = 0.4, P = 0.67 for season-sample time interaction). As with corticosterone concentrations in females, CBG capacity in males was elevated in July compared with June.
These changes in CBG capacity resulted in free corticosterone patterns that differed from total corticosterone patterns. For example, in contrast to total corticosterone, baseline and stress-induced free corticosterone concentrations of male and female lemmings were similar (Fig. 3B; Z1,12 = 1.37, P = 0.17 for baseline and Z1,12 = 1.65, P = 0.10 for stress-induced free corticosterone). Furthermore, female baseline free corticosterone was higher in July than in June (Fig. 3B; Z1,22 = −2.66, P < 0.01), whereas stress-induced free corticosterone did not differ between the seasons for females (Fig. 3C; Z1,23 = −1.61, P = 0.11). For males, neither baseline (Fig. 3B; Z1,14 = −1.59, P = 0.11) nor stress-induced (Fig. 3C; Z1,13 = 1.56, P = 0.12) free corticosterone varied between the seasons.
In addition, a few juvenile lemmings were captured during both seasons. Although we could not determine sex in these young animals, there was no effect of weight; therefore, weight was removed from the analysis. Juveniles increased their corticosterone concentrations in response to capture, handling, and bleeding (Fig. 4; F1,24 = 4.28, P < 0.05 for overall effect of sampling time), but there was no seasonal difference (F1,24 = 0.36, P = 0.55) or interaction between season and sampling time (F1,24 = 0.14, P = 0.71).
Golden-mantled ground squirrels show distinct seasonal modulation of glucocorticoid secretion (Fig. 5). In golden-mantled ground squirrels, the dominant glucocorticoid is cortisol, and they released more than twice as much cortisol in the spring as in the fall (Fig. 5A; F1,37 = 25.53, P < 0.0001). Furthermore, the ground squirrels significantly elevated their cortisol concentrations in response to 30 min of restraint and handling (F1,37 = 5.65, P < 0.03), but there was no seasonal interaction (F1,37 = 0.35, P = 0.56). Corticosterone levels also varied seasonally (Fig. 5B; F1,38 = 41.26, P < 0.0001), but at levels far below cortisol levels. Furthermore, the increase in corticosterone after capture and handling was less robust (F1,38 = 3.99, P = 0.053), but again with no seasonal interaction (F1,38 = 0.02, P = 0.88).
The dominant glucocorticoid in yellow-pine chipmunks was cortisol (Fig. 6). Yellow-pine chipmunks also showed seasonal modulation of cortisol release (Fig. 6A; F1,27 = 16.21, P < 0.0005), but, in contrast to the ground squirrels, concentrations were significantly higher in the fall than in the spring. This result was mirrored in the corticosterone concentrations (Fig. 6B; F1,29 = 23.47, P < 0.0001), even though overall corticosterone concentrations were far below cortisol concentrations. Furthermore, the chipmunks significantly elevated corticosterone and cortisol concentrations in response to 30 min of restraint and handling (F1,29 = 25.67, P < 0.0001 for corticosterone; F1,27 = 61.96, P < 0.0001 for cortisol), but with no seasonal interaction (F1,29 = 2.53, P = 0.12 for corticosterone; F1,27 = 0.01, P = 0.91 for cortisol).
All three species show distinct seasonal differences in glucocorticoid concentrations. Although seasonal rhythms of gonadal steroids have been studied for years in wild mammals, especially in mammals that have distinct breeding seasons, surprisingly few studies have examined seasonal glucocorticoid rhythms. One exception has been studies of certain Australian marsupials (e.g., Antechinus sp.), which dramatically elevate glucocorticoids at the end of the breeding season, although this is apparently a mechanism for programmed mortality (5, 13, 14, 44). Other than these Australian species, glucocorticoid samples from unstressed animals (collected within 3 min of capture) that have been collected two or more times during the year (e.g., comparing seasonal differences) have been reported for few mammalian species (for review see Ref. 56). Although some studies report no seasonal differences in glucocorticoid levels (73), most studies show a seasonal change. However, some species had the highest measured glucocorticoid levels during breeding (8, 11, 20, 61), whereas others had the lowest levels during breeding (37, 54, 71). Of special interest is that the earlier studies on yellow-pine chipmunks (37, 54) match the results reported here, and smaller adrenal mass is correlated with lower seasonal glucocorticoid concentrations in another chipmunk species (66). It appears, therefore, that wild mammals do have seasonal rhythms of baseline glucocorticoid release, although when in the year that peak occurs is not consistent from species to species.
In contrast, considerably more studies have examined wild mammals for seasonal differences in stress-induced glucocorticoid concentrations (for review see Ref. 56). The increased effort is primarily a by-product of typical trapping techniques, where traps are checked every few hours, making it impossible to ascertain how long animals have been in traps. Although glucocorticoid concentrations from these studies are clearly not baseline, because of the potentially large variation in the time individual animals have been in traps, it is difficult to interpret stress-induced concentrations. Nevertheless, most studies report seasonal differences in stress-induced glucocorticoid concentrations, but, as with baseline concentrations, some species, including male yellow-pine chipmunks (54), show a seasonal nadir during breeding (3, 29, 31, 38, 41, 48), some, including golden-mantled ground squirrels (12), show a seasonal peak during breeding (1, 11, 30, 48), and some, including female yellow-pine chipmunks (37), show no seasonal differences (2, 10, 28, 69). In addition, most (7, 16, 52), but not all (49), captive studies with nontraditional laboratory animals show seasonal glucocorticoid variation.
In summary, most studies that have measured glucocorticoid levels in wild free-living mammals have found seasonal differences. The three species studied here add to this work. However, in contrast to birds, reptiles, and amphibians (56), there is no consistent period during the annual cycle when glucocorticoids are at a peak. The peaks and nadirs appear to vary by species. There are no hypotheses that successfully predict when a species should have higher glucocorticoid concentrations. Two of these, proposing that seasonal glucocorticoid differences relate to seasonal metabolic demands or seasonal modulation of glucocorticoid-induced behavioral changes, are likely only relevant in some species under limited conditions (56), and it is not clear how they would be relevant to the species in the present study. One recent hypothesis focused on hibernators and predicted, on the basis of the role of glucocorticoids in glucose and lipid trafficking, that hibernators that undergo prehibernation hyperphagia should show elevated glucocorticoid concentrations in the fall during the hyperphagia period (22). Although one study in marmots supported this hypothesis (3), our data do not. The golden-mantled ground squirrels, which undergo prehibernation hyperphagia, had higher cortisol concentrations in the spring, whereas the yellow-pine chipmunks, which do not undergo prehibernation hyperphagia had higher corticosterone concentrations in the fall. These findings contradict the prediction.
One other hypothesis has proposed that glucocorticoid concentrations should be elevated during those periods in the annual cycle when there is a greater risk of exposure to stressors (56). For example, glucocorticoid concentrations may track greater predation risk. Although there is little empirical evidence for this hypothesis, there is some support. For instance, low glucocorticoid concentrations are linked to population cycles driven by predation (9). Furthermore, anecdotal evidence suggests that differences in predation pressure might explain the seasonal difference in corticosterone concentrations in female, but not male, brown lemmings. Preliminary data from counting the number of male and female lemming carcasses in snowy owl nests (a major lemming predator) suggest that the sex ratio is roughly equivalent in June, when there is extensive snow cover, but becomes heavily female biased in July, when the snow cover is gone (unpublished data).
Clearly, much more work is needed to understand seasonal modulation of glucocorticoid release in wild mammals. It appears to be a widespread phenomenon, however; therefore, understanding how and why glucocorticoid release is modulated is likely to be fundamental to our understanding of the stress response.
Corticosterone concentrations in brown lemmings.
Brown lemmings have the highest glucocorticoid concentrations ever reported for a mammal. Several bat species have baseline cortisol levels of ∼400 ng/ml, with stress-induced levels peaking at ∼1,000 ng/ml (76, 77); prairie voles have baseline corticosterone levels of ∼800 ng/ml, with stress-induced peaks at 1,300–2,500 ng/ml (25, 72). These levels are similar to those in male brown lemmings but are far below the 3,000–4,000 ng/ml in female brown lemmings (Fig. 1). The higher concentrations in females than in males parallel laboratory data from domesticated rats (74) and Townsend's vole (Microtus townsendii) (46). In several female brown lemmings, corticosterone levels exceeded 8,000 ng/ml. Glucocorticoid levels this high would be lethal in many species, so either these high corticosterone levels are necessary for survival in brown lemmings or brown lemmings have adaptations making target tissues less sensitive to the corticosterone signal. The data presented here begin to address this issue.
Once released, the availability of glucocorticoids to their target tissues is primarily regulated in two ways. The first is through CBG binding (75). Only free glucocorticoids are believed to be available to tissues (60), although this is under some debate (56). Under normal physiological conditions, ∼90% of circulating glucocorticoids are bound to CBG (67). Therefore, glucocorticoid availability can be regulated by altering CBG concentrations, even when plasma glucocorticoid concentrations do not change. Altering glucocorticoid and CBG levels can be synergistic, since increasing glucocorticoid levels can reduce CBG levels (26, 68), thereby greatly augmenting glucocorticoid availability to target tissues. Furthermore, CBG binding is saturated at near-basal glucocorticoid levels (75); therefore, small rises in glucocorticoid concentrations result in large increases in free hormone available to receptors (21). Consequently, the increase in CBG capacity during July in the lemmings could act as a buffer to limit corticosterone availability to receptors.
Other mammalian species alter CBG capacity. The very high corticosterone levels in prairie voles compared with rats are partially buffered by a twofold increase in CBG (72), and seasonal changes in glucocorticoids in marmots are paralleled by seasonal changes in bound glucocorticoids, presumably CBG (3). However, in none of the earlier studies, including the present study, did changing CBG capacity fully compensate for altered corticosterone levels. Interestingly, the Kd of 10.79 for lemming CBG is similar to Kd of 14.72 for prairie voles (50), even though the binding capacity of 6,809 nM for lemmings is far higher than 716 nM (50) or 1,553 nM (72) estimated for prairie voles.
The second way to regulate glucocorticoid availability is through negative feedback. Feedback occurs primarily at the pituitary, paraventricular nucleus of the hypothalamus, and hippocampus, with the pituitary and paraventricular nucleus mediating fast feedback and the hippocampus mediating slow feedback (32, 34, 36). Each of these feedback loops serves to inhibit continued glucocorticoid release after initiation of a stressor. Although numerous studies have demonstrated that the efficacy of negative feedback shows a daily rhythm (for review see Ref. 36), few studies have examined whether negative-feedback efficacy varies seasonally as well. There may also be species differences in how effective negative feedback might be in inhibiting glucocorticoid release. For example, prairie voles (25, 72) and several bat species (76, 77) have extremely high corticosterone levels, but corticosterone levels are resistant to negative-feedback effects in prairie voles (72). Similarly, Antechinus species are resistant to negative feedback (45). This suggests that high glucocorticoid concentrations can be maintained by an alteration of the feedback mechanism that prevents such high levels in rats and other vertebrates. A similar mechanism appears to be working in brown lemmings. Although it is possible that glucocorticoid receptors in lemmings are altered so that they no longer bind DEX or that a higher dose of DEX might have been effective, this has not been demonstrated in any mammalian species. Instead, it is likely that the lemmings were resistant to DEX stimulation of negative feedback. The implications of feedback resistance for a species are still not clear.
One important caveat for the corticosterone concentrations in lemmings is that the present study was conducted during years of relatively high density (required to successfully capture animals), and high density can be associated with poor health and elevated glucocorticoid concentrations (11). However, there were no overt signs of negative health, disease, or parasite infections, although admittedly we did not rigorously examine animals for these traits. In addition, we were able to capture a substantial number of juveniles, which indicates reproductive success and suggests that animals were healthy. Consequently, although we think it is unlikely, these results may only apply to lemmings during a high-density year.
Perspectives and Significance
There are three main contributions from this study. 1) These data provide one of the few examples where baseline corticosteroid concentrations are known with a reasonable degree of confidence for free-living mammalian species. Sampling within 3 min of capture provides the opportunity for a deeper understanding of glucocorticoid physiology of natural populations. 2) The corticosterone concentrations in lemmings are, to the best of our knowledge, the highest ever described for a mammal. These concentrations demonstrate that the variation between species is even greater than is known so far and that our understanding of the physiological functions of glucocorticoids within and between species is far from complete. Furthermore, this species potentially would be an excellent new model for understanding glucocorticoid resistance (especially given that they failed the DEX suppression test). 3) The data clearly indicate that glucocorticoid responses show seasonal differences in these free-living species and highlight the importance of species-specific life-history features in the magnitude of the response. Seasonal glucocorticoid variations are poorly understood in free-ranging mammals relative to other vertebrate groups. Given the biomedical importance of understanding glucocorticoid physiology in mammals and the wealth of data regarding stress physiology in captive mammals, the addition of data from free-ranging mammals should provide a broader prospective on the nature and evolutionary role of these responses.
This work was supported by National Institute of Neurological Disorders and Stroke Grant 1RO1 NS-30240-01 and National Science Foundation Grant OPP-9300771 to J. C. Wingfield, Environmental Protection Agency Science to Achieve Results Fellowship FP-91649101-0 to N. E. Cyr, and a grant from the American Psychological Association and National Science Foundation Grant IOB-0542099 to L. M. Romero.
We are deeply indebted to Robert Suydam and the North Slope Borough of Alaska for logistical support. We also thank K. K. Soma for help catching lemmings, C. Breuner for advice on the CBG assay, and L. Erkmann for technical support.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 the American Physiological Society