Thermogenic capacity at subzero temperatures: how low can a hibernator go?

Abstract Hibernation in mammals is a physiological and behavioral adaptation to survive intervals of low resource availability through profound decreases in metabolic rate (MR), core body temperature (Tb), and activity. Most small mammalian hibernators thermoconform, with Tb approximating ambient temperature (Ta); arctic species are an exception, since they must actively defend what can be large thermal gradients between Tb and Ta. Here we compare the thermogenic capacity of the arctic ground squirrel (Urocitellus parryii) to that of the golden-mantled ground squirrel (Callospermophilus lateralis), a temperate-zone montane hibernator. We allowed animals to reenter torpor at sequentially lower Ta's and found that arctic ground squirrels maintained steady state torpor at Ta's as low as -26°C, through a 36-fold increase in torpid MR (TMR), compared to their minimum TMR, exhibited at a Ta of 0°C. Golden-mantled ground squirrels are able to maintain steady state torpor at Ta's at least as low as -8°C, through a 13.5-fold increase in MR, compared to their minimum TMR at a Ta of 2°C. In a second experiment, torpid animals were exposed to continuously decreasing Ta's (0.25°C/30 min); individuals of both species increased their metabolism while remaining torpid at low Ta's (as low as -30°C for arctic ground squirrels and -10°C for golden-mantled ground squirrels). Although the capacity to hibernate at subfreezing Ta's is not unique to arctic ground squirrels, their large body size, greater torpid metabolic scope, and previously ascribed capacity to supercool allow them to occupy much colder hibernacula for prolonged seasons of hibernation.

Role of the antioxidant ascorbate in hibernation and warming from hibernation.

Ground squirrels tolerate up to 90% reductions in cerebral blood flow during hibernation as well as rapid reperfusion upon periodic arousal from torpor without apparent neurological damage. Thus, hibernation is studied as a model of tolerance to cerebral ischemia and other types of brain injury. Metabolic suppression likely plays a primary adaptive role that allows hibernating species to tolerate dramatic fluctuations in blood flow. Several other aspects of hibernation physiology are also consistent with tolerance to ischemia and reperfusion suggesting that multiple neuroprotective adaptations may work in concert during hibernation. The purpose of the present work is to review evidence for enhanced antioxidant defense systems during hibernation, with a focus on ascorbate, and discuss potential roles of these antioxidants during hibernation. In concert with dramatic decreases in blood flow, nutrient and oxygen delivery, plasma concentrations of the antioxidant ascorbate [(Asc)p] increase 3-5-fold during hibernation. In contrast, during re-warming, [Asc]p declines at a relatively rapid rate that peaks at the time of maximal O(2) consumption. This peak in O(2) consumption also coincides with a brief rise in plasma urate concentration consistent with a surge in reactive oxygen species production. Overall, data suggest that elevated concentration of plasma ascorbate is poised for distribution to metabolically active tissues during the surge in oxidative metabolism that accompanies re-warming during hibernation. This pool of ascorbate, as well as increased expression of other antioxidant defense systems, may protect vulnerable tissues from oxidative stress during hibernation and re-warming from hibernation. Better understanding of the role of ascorbate in hibernation may guide use of ascorbate and other antioxidants in treatment of stroke, head trauma and neurodegenerative disease.

Thermogenic responses to body cooling during fever induced by Staphylococcus aureus cell walls in rabbits.

Hypothalamic temperature (Thypo) and metabolic heat production (M) were measured in seven conscious rabbits injected intravenously with either saline or with Staphylococcus aureus, (8.10(7) cell walls.kg-1) while being subjected to a 3-h period of ramp-like total body cooling using a chronically implanted intravascular heat exchanger. In pyrogen-injected animals cooling started (1) at the time of injection or (2) 70 min after injection. In (1) the fall in Thypo induced by heat extraction was similar (1.0 degree C) in afebrile and febrile animals. In (2) there was a transient increase in Thypo of about 0.5 degree C at a time corresponding to the start of fever resulting in a significantly smaller fall in Thypo at the end of the 3-h cooling period (0.5 degree C vs 0.9 degree C, P < 0.05, n = 5). At this time in both (1) and (2) M was lower than theoretically expected from the increase in shivering threshold during fever. However, most of this effect can be explained when available data showing a decrease in thermosensitivity during S. aureus-induced fever are taken into account. After cessation of cooling in both groups of febrile animals Thypo rose to about 1 degree C higher than the precooling level, which is comparable to the fever level in a separate series of experiments with S. aureus injection without cooling (1.2 degrees C).

Thermosensitivity is reduced during fever induced by Staphylococcus aureus cells walls in rabbits.

Thermosensitivity (TS) and threshold core temperature for metabolic cold defence were determined in six conscious rabbits before, and at seven different times after i.v. injection of killed Staphylococcus aureus (8 x 10(7) or 2 x 10(7) cell walls x kg(-1)) by exposure to short periods (5-10 min) of body cooling. Heat was extracted with a chronically implanted intravascular heat exchanger. TS was calculated by regression of metabolic heat production (M) and core temperature, as indicated by hypothalamic temperature. Threshold for cold defence (shivering threshold) was calculated as the core temperature at which the thermosensitivity line crossed preinjection resting M. The shivering thresholds followed the shape of the fever response. TS was significantly reduced (up to 49%) during the time course of fever induced by the highest dose of pyrogen only. At both high and low doses of pyrogen TS correlated negatively with shivering threshold (r = 0.66 and 0.79 respectively) with similar slopes. The reduction in TS during fever was thus associated with the increase in shivering threshold resulting from the pyrogen injection and not by the dose of pyrogen. Model considerations indicate, however, that changes in sensitivity of the thermosensory input to the hypothalamic controller may affect threshold changes but cause negligible TS changes. It is more likely that the reduction in TS is effected in the specific hypothalamic effector pathways.

Effect of total body core cooling during poly I:C-induced fever in rabbits.

At ambient temperature (Ta) 20 and 10 degrees C, metabolic heat production and hypothalamic temperature (Thypo) were measured to determine the fever response in six rabbits injected with polyriboinosinic-polyribocytidylic acid (poly I:C; 5 mg/kg iv). Similar measurements were made in afebrile and febrile animals subjected to 3 h of body cooling, in which heat was extracted with a chronically implanted intravascular heat exchanger in a ramplike manner. The fever time course showed a biphasic pattern. During cooling in the febrile experiments, Thypo remained constant or even slightly increased during the time corresponding to the first phase of fever but rapidly fell during the second phase because of a depressed shivering response. The net effect at the end of the cooling period was that Thypo decreased by 0.4 and 0.6 degree C more than in the afebrile cooling experiments at Ta 20 and 10 degrees C, respectively. The results indicate normal shivering responses during phase I of poly I:C-induced fever and depressed shivering in phase I, possibly because of a reduced thermosensitivity.

Poly I:C-induced fever elevates threshold for shivering but reduces thermosensitivity in rabbits.

Shivering threshold and thermosensitivity were determined in six conscious rabbits at ambient temperature (Ta) 20 and 10 degrees C before and at six different times after saline injection (0.15 ml iv) and polyriboinosinic-polyribocytidylic acid (poly I:C)-induced fever (5 micrograms/kg iv). Thermosensitivity was calculated by regression of metabolic heat production (M) and hypothalamic temperature (Thypo) during short periods (5-10 min) of square-wave cooling. Heat was extracted with a chronically implanted intravascular heat exchanger. Shivering threshold was calculated as the Thypo at which the thermosensitivity line crossed resting M as measured in afebrile animals at Ta 20 degrees C. There were negligible changes in shivering threshold and thermosensitivity in saline-injected rabbits. In the febrile animals, shivering threshold generally followed the shape of the biphasic fever response. At Ta 20 degrees C, shivering threshold was higher than regulated Thypo during the initial rising phase of fever and was lower during recovery. At Ta 10 degrees C the shivering thresholds were always higher than regulated Thypo except during recovery. Thermosensitivity was reduced by 30-41% during fever.

Synchrony between breathing and shivering in three muscles of bantam hens exposed to cold eggs.

Unrestrained, incubating bantam hens (Gallus domesticus) were exposed to cold, water-perfused eggs to induce shivering. Mean rectified electromyogram (EMG activity) was recorded from the iliotibialis, gastrocnemius, and pectoralis muscles, together with breathing and O2 consumption. EMG activity in all three muscles varied in synchrony with breathing, as confirmed by spectral analysis. The highest shivering intensity occurred during end expiration. This synchrony was also present when respiratory frequency was increased by hypercapnic/hypoxic stimuli and tended to improve with increasing respiratory frequency and tidal volume. The absolute modulation of shivering intensity remained relatively constant with increasing average intensity, indicating that breathing-modulated motor units are activated first. In the less active pectoralis muscle, all activity was modulated. Modulation is probably not restricted to one type of muscle fiber because it occurs in both predominantly red hindquarter muscles and the white pectoralis muscle. It is tentatively suggested that respiratory drive-related influence from the brain stem has a synchronizing effect on spinal cord activity. A bursting activity pattern may be more favorable for muscle circulation than a tonic discharge pattern.

Control of shivering and heart rate in incubating bantam hens upon sudden exposure to cold eggs.

To distinguish shivering released by cooling of the brood patch from shivering released by low core temperature, incubating bantam hens were exposed to water perfused eggs. Responses to a period of egg cooling were compared to the recovery period after egg temperature had been returned to 40 degrees C, but cloacal temperature (Tb) still was low. At an ambient temperature (Ta) of 23 degrees C and exposure of the hens to between two and eight eggs cooled to 10-35 degrees C (series 1), electromyographic (EMG) activity of musculus iliotibialis increased rapidly with an occasional overshoot, and was higher during egg cooling than during recovery. This hysteresis in EMG activity and Tb was weakly correlated to egg temperature and clutch size. Heart rate (HR) showed an almost parallel increased to shivering except that a maximum HR was reached at high shivering intensities. These responses were also present at a Ta of 37 degrees C when the hens were slightly hyperthermic before exposure to eight eggs at 20 degrees C (Series 2). At the highest starting Tbs EMG activity increased linearly after a drop in Tb. Shivering in m. pectoralis showed a lower threshold Tb and lower activity than m. iliotibialis during egg cooling, and immediately ceased at the end of egg cooling. Total body thermosensitivity estimated from the recovery periods at low and high Ta was -9.7 and -6.4 W kg-1 degrees C-1, respectively. It is concluded that shivering in incubating birds warming cold eggs probably is stimulated both by peripheral and central thermoreceptors. The peripheral component shows phasic properties typical for skin receptors.

Diffusive resistance of avian eggshell pores.

Resistance to gas diffusion through the avian eggshell resides in the microscopic pores which penetrate the shell. We calculated the resistance to water vapor diffusion of individual pores in the shells of 23 species of avian eggs, based on measurements of pore dimensions taken from drawings of 321 pore casts published by Tyler (1962, 1964, 1965) and Tyler and Simkiss (1959). Diffusive resistances were calculated from Fick's first law, using a 100-segment model of each pore. In addition, we added 2 series resistances, calculated from Stefan's law, to account for boundary layer resistances at the inner and outer pore apertures. Convective resistances for the same 100-segment model were computed from Poiseuille's law. A special, symmetrically branching model is presented for the diffusive resistance of the branched pores of ostrich eggshells, based on the drawings of Tyler and Simkiss (1959). The total aperture resistance was less than 6.2% of total pore resistance, while the outside aperture effect was on average only 1.5%. The calculated average pore conductance for all species was 5.4 micrograms (day Torr)-1, about three times higher than the average value of 1.6 micrograms (day Torr)-1 obtained by dividing measured shell conductance by the number of pores (Ar and Rahn, 1985). A possible explanation for this discrepancy is advanced. However, it is to be noted that in spite of the discrepancy, both calculated and functional values of pore conductance appear to be independent of egg mass.

Threshold for shivering in aerobic and anaerobic muscles in bantam cocks and incubating hens.

Electromyographic activity (EMG) from the musculus pectoralis (breast muscle), m. iliotibialis (thigh muscle) and m. gastrocnemius (leg muscle), cloacal temperature (Tb) and O2 consumption were measured in bantam cocks (Gallus domesticus) exposed to different ambient temperatures (Ta). The same parameters were measured in bantam hens incubating eggs artificially thermoregulated to 40 degrees and 25 degrees C (Te). EMG activity appeared in thigh and leg muscles at Ta below 32 degrees C (Tsh). This temperature probably represents the thermoneutral temperature (TNT) of the cock. EMG activity in breast muscles appeared at Ta below 20 degrees C, or 4 degrees C below the lower critical temperature (Tc). All muscles were quiet when the hen incubated 40 degrees C egg at Ta = Tsh. When Te was abruptly changed to 25 degrees C, EMG activity in the iliotibialis muscle appeared 3 min before the activity in the pectoralis muscle. Tb dropped from 41.2 degrees to 40.6 degrees C in 14 min. When Te was returned to 40 degrees C, the EMG activity in the pectoralis muscle disappeared almost at once, while the iliotibialis muscle was active until Tb returned to normal. Aerobic muscles seem to be responsible for shivering thermogenesis between Tc and Tsh, while anaerobic muscles are recruited at lower Ta or when the heat loss during incubation becomes severe.

Influence of eggshell pore shape on gas diffusion.

We describe the influence of pore shape in the hen's eggshell on calculated resistance to diffusion. These calculations are based on measurements of 70 pore casts, pictures of which were published by Tyler (In: Recent Advances in Food Sciences, Butterworth, London, Vol. 1, 1962). Single-pore resistance was estimated from Fick's first law as the summed resistances of 100 cylindrical segments in series and the calculated boundary layer resistance at the pore apertures. In addition, we constructed an average pore profile from all measurements. The calculated average single pore water vapor conductance (3.2 micrograms X day-1 X torr-1) is about twice the measured literature value. The presence of organic material in the pore openings is discussed as one of several possible reasons for this discrepancy. Average diameter should not be used to calculate conductance of trumpet-shaped pores. The narrow part of the pores plays a dominant role in total pore conductance, and pore shape must be taken into account when estimating conductance from pore dimensions.