Temporal relationships of blood pressure, heart rate, baroreflex function, and body temperature change over a hibernation bout in Syrian hamsters

Abstract

Hibernating mammals undergo torpor during which blood pressure (BP), heart rate (HR), metabolic rate, and core temperature (T_C) dramatically decrease, conserving energy. While the cardiovascular system remains functional, temporal changes in BP, HR, and baroreceptor-HR reflex sensitivity (BRS) over complete hibernation bouts and their relation to T_C are unknown. We implanted BP/temperature telemetry transmitters into Syrian hamsters to test three hypotheses: H-1) BP, HR, and BRS decrease concurrently during entry into hibernation and increase concurrently during arousal; H-2) these changes occur before changes in T_C; and H-3) the pattern of changes is consistent over successive bouts. We found: 1) upon hibernation entry, BP and HR declined before T_C and BRS, suggesting baroreflex control of HR continues to regulate BP as the BP set point decreases; 2) during the later phase of entry, BRS decreased rapidly whereas BP and T_C fell gradually, suggesting the importance of T_C in further BP declines; 3) during torpor, BP slowly increased (but remained relatively low) without changes in HR or BRS or increased T_C, suggesting minimal baroreflex or temperature influence; 4) during arousal, increased T_C and BRS significantly lagged increases in BP and HR, consistent with establishment of tissue perfusion before increased T_C/metabolism; and 5) the temporal pattern of these changes was similar over successive bouts in all hamsters. These results negate H-1, support H-2 with respect to BP and HR, support H-3, and indicate that the baroreflex contributes to cardiovascular regulation over a hibernation bout, albeit operating in a fundamentally different manner during entry vs. arousal.

mammals in hibernation show marked physiological changes, including decreased core body temperature (T_C), metabolic rate, blood pressure (BP), and heart rate (HR) (8, 21, 30, 32). Typically, small hibernators such as ground squirrels and hamsters go through bouts of torpor that extend over several days, during which time T_C is maintained within a few degrees of ambient temperatures that are above freezing, and metabolic rate is reduced to 2–4% of euthermic rates (8). Torpor is interrupted by periods (hours to days) of arousals during which heart rates rapidly increase from 5–10 beats per minute (bpm) to >400 bpm and T_C increases to ∼37°C (8). Three interrelated systems (cardiovascular, thermoregulatory, and metabolic) are synchronized for survival throughout the sequence of hibernation bouts, and key to this survival is a well regulated BP, which ensures that oxygen/nutrient supplies match the dramatic changes in metabolic demand. Such regulation requires a coherent temporal pattern of central and peripheral physiological responses, a relationship further delineated in this study.

In several species, changes in HR and T_C in hibernation bouts are well documented and show that changes in HR generally precede changes in T_C during both entry into and arousal from hibernation (10, 16, 45). However, we have limited information on BP changes over entire hibernation bouts. Using arterial catheters that were externalized, Lyman and O'Brien (32) recorded BP in ground squirrels during a hibernation bout, but only partial BP responses were described, possibly due to technical limitations in keeping the arterial catheter open at all times for continuous BP recordings. Nonetheless, they showed that a drop in BP preceded the decline in T_C during entry into hibernation and that during arousal, BP reached its peak value about the time that thoracic temperature reached 37°C.

These changes lead to the question of how well and how consistently BP is regulated during each of the various stages of a hibernation bout. Whereas the arterial baroreflex in euthermic mammals provides a powerful beat-to-beat regulation of BP around a set point via a defined central pathway (29, 42, 43), it is unclear whether arterial baroreflex regulation of HR continues to play a major role in regulating BP over specific phases of a hibernation bout or whether other factors (e.g., temperature) take precedence.

Advances in telemetry technology and analytic methods now make it possible to continuously record BP and HR and to calculate baroreceptor-HR reflex sensitivity (BRS) without perturbing the animal (6, 44). High negative BRS values imply that spontaneous changes in BP evoked large compensating changes in HR, bringing BP back toward its prevailing level, whereas low negative values of BRS imply that the same spontaneous change in BP evoked a small change in HR. With the availability of a method for continuously calculating BRS on the basis of spontaneous changes in BP, the relationship between cardiovascular and thermoregulatory neural controllers over each phase of a hibernation bout can be fully explored. To this end, in this study we measured BP, HR, BRS, and T_C over entire hibernation bouts to test the following three hypotheses: H-1) BP, HR, and BRS decrease concurrently during entry into hibernation and increase concurrently during arousal; H-2) the changes in BP, HR, and BRS occur before changes in T_C; and H-3) the temporal pattern of these changes is consistent over successive bouts.

METHODS

All experimental procedures and protocols in this work were reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) in compliance with the Animal Welfare Act and in accordance with the US Public Health Service Policy on the Humane Care and Use of Laboratory Animals. Syrian hamsters from our colony were bred from animals selected over generations to more readily hibernate when exposed to cold and short photoperiod. All hamsters were provided ad libitum access to food (Lab Diet 5001 Rodent Diet) and water, and the breeding colony was housed at 22 ± 2°C with a long photoperiod (14:10 h light:dark cycle).

Implantation of pressure and of pressure-temperature radiotelemetry transmitters.

Surgery was performed according to the Guidelines for Survival Surgery in Rodents provided by the IACUC. Eight hamsters (5 females; 3 males) 21–43 wk old (body wt 106–162 g) were anesthetized with isofluorane with the depth of anesthesia being maintained such that there was no response to hind-paw pinch or any other surgical manipulation. Antibiotic (Baytril, 2.5 mg/kg) and analgesic (ketoprofen, 5 mg/kg) drugs were administered immediately prior to surgery. Each hamster was placed on a servo-controlled water blanket, and body temperature was monitored via a rectal temperature probe and maintained at 37 ± 1°C. The abdominal cavity was opened through a midline incision, and the abdominal aorta was isolated below the renal artery. The catheter of a pressure transmitter was inserted into the abdominal aorta and secured with a drop of Vetbond surgical glue and a nitrocellulose patch. In the first set of experiments, six hamsters, denoted group P, were implanted with a telemetry transmitter (TA11PA-C10; Data Sciences, St. Paul MN) that encoded and transmitted only BP. Subsequently, two hamsters, denoted group PT, were implanted with a larger telemetry transmitter (TL11M2-C50-PXT; Data Sciences) that encoded and transmitted BP and T_C. The transmitter body was sutured to the abdominal musculature with 5–0 PDS II (Ethicon), the skin incision was closed with surgical clips, and clips were removed after 7 days under light isofluorane anesthesia. All animals were allowed to recover for at least 14 days postsurgery in a short photoperiod (8:16 h light:dark) room at 22 ± 2°C before being transferred to a short photoperiod (8:16 h light:dark) cold room (6 ± 2°C). The day before the animal was transferred to the cold room, its telemetry transmitter was turned on for continuous recordings until the animal was killed. The hamsters first entered hibernation within 1–10 wk after transfer to the cold room.

Data acquisition and analysis.

The telemetry transmitters were calibrated twice; once prior to implantation and again after removal from the animal. For the pressure calibrations, the transmitter was placed into a sealed chamber connected to a mercury manometer. The sealed chamber was placed in a temperature-controlled water bath at temperatures ranging from 6° to 37°C. At each temperature, chamber pressure was varied from 0 to 220 mmHg. For group PT hamsters, the transmitters were also calibrated for temperatures over the range of 6° to 37°C prior to implantation and after removal from the animal using the sealed chamber and a temperature-controlled water bath. Signals from the implanted transmitters were continuously recorded at 500–2,000 Hz. Beat-to-beat systolic blood pressure (SBP) and pulse intervals were determined offline using Dataquest analysis software (Dataquest A.R.T.).

BRS was analyzed with the sequence method (6, 52) using the Nevrokard Small Animal BRS Analysis program (Nevrokard, Medistar). Briefly, this software identified spontaneous baroreflex sequences of three or more consecutive beats in which SBP progressively rose and HR progressively decreased (or SBP progressively fell and HR progressively increased) and where HR values sampled were offset (delayed) by one beat. BRS was calculated by the software as the slope of the change in HR vs. the change in SBP from regression lines fitted to each spontaneous SBP sequence and the delayed HR sequence (6, 44). The sequence method takes advantage of the ability of the baroreflex to detect within a short period of time (a few beats) a decrease (or increase) in SBP from its prevailing level (i.e., set point) to trigger a corrective increase (or decrease) in HR that brings SBP back to the prevailing level. Moreover, the sequence method isolates baroreflex sequences from all sequences; that is, sequences that fail to follow the inverse HR-SBP relationship of a negative feedback reflex are considered to be nonbaroreflex responses (34).

SBP, HR, and BRS averaged over a 24-h period starting 30 h before an animal's first hibernation bout were used as prehibernation euthermic values. One animal was terminated after nine bouts of hibernation; two animals had only one recorded bout due to technical issues; and the remainder (n = 5) had their recording terminated after 3–6 bouts of hibernation.

Fall time (for entry) and rise time (for arousal) were defined as the time in which the variable decreased or increased between 10% and 90% of its maximum change. Because there was an overshoot of SBP and HR at the end of arousal, the range over which rise time was calculated was from arousal onset to peak of the overshoot. There was no overshoot in BRS or T_C. During entry into and arousal from hibernation, changes in all variables were fit with a single exponential function to obtain time constants.

A one-way ANOVA was used to determine differences among bouts. Because there were no significant differences among bouts for any of the measured variables, subsequent analyses were performed on all available bouts combined. A one-way repeated ANOVA was used for comparing SBP and HR values at different phases of the hibernation bouts. Fall times, rise times, and time constants from the single exponential fits for SBP, HR, BRS, and T_C were compared with a one-way repeated ANOVA. A Fisher's LSD post hoc test was performed when appropriate. Data are expressed as means ± SE, with comparisons being taken as significant at P < 0.05.

RESULTS

Group P hamsters.

We recorded a total of 23 hibernation bouts from group P hamsters (those implanted with BP, but not temperature, telemetry transmitters). Thirty-minute averages of SBP and HR over 40 days from one hamster are illustrated in Fig. 1. This hamster went through nine bouts of hibernation, each lasting 1–4 days. Although this animal displayed transient undershoots in SBP when entering hibernation (Fig. 1), such undershoots were not observed in every bout or in every hamster. Over the nine bouts, this hamster spontaneously aroused from hibernation and remained euthermic for 1–2 days before reentering torpor. The average length of time it took to complete entry into hibernation for all 23 recorded bouts (245 ± 14 min) was significantly longer than that of arousal time (112 ± 5 min; paired t-test). Twenty-one of the 23 hibernation bouts began entry into hibernation during the dark cycle (with 83% between 8 P.M. and 8 A.M.; Fig. 2). In contrast, arousals were more distributed, with 12 beginning during the dark cycle and 11 during the light cycle (Fig. 2).

Fig. 1.

Thirty-minute averages of systolic blood pressure (SBP) and heart rate (HR) from one hamster. This hamster went through nine bouts of hibernation over 40 days.

Fig. 2.

A 24-h clock demonstrating the time at which each animal entered (left) and aroused (right) from hibernation. Each letter represents an animal (females in upper case, males in lower case), and the number following the letter is the hibernation bout number for that animal. The dark portion of the light cycle is designated by the gray area. Hamsters in which only BP was recorded are identified by letters A through f. Hamsters in which both blood pressure (BP) and core temperature (T_C) were recorded are identified by letters G and H.

For all 23 recorded bouts, SBP dropped from ∼100 (at euthermia) to ∼30 mmHg after entry into hibernation and then gradually and significantly rose to ∼60 mmHg over several days, at which time arousal began (Table 1). Consistently upon arousal, there was a significant transient overshoot in SBP that lasted for approximately 1 h (Table 1). As was the case with SBP, there were no significant bout-to-bout differences in HR at each hibernation phase, and there was a significant transient overshoot in HR (Table 1). However, unlike SBP, HR remained low throughout the torpor period (at 10–14 bpm) after the initial drop during entry (Table 1).

Table 1.

Systolic blood pressure and heart rate over a hibernation bout

Hibernation Phase	SBP, mmHg	HR, bpm
Before hibernation	100 ± 2a	325 ± 5a
Hibernation lowest	30 ± 2b	11.4 ± 0.2b
Arousal onset	60 ± 2c	13.0 ± 0.3b
Arousal overshoot	137 ± 2d	484 ± 5c

Values are means ± SE; n = 23 bouts.

SBP, systolic blood pressure; HR, heart rate; bpm, beats per minute.

For each variable (column), values that share the same superscript letter do not differ significantly (one-way repeated ANOVA followed by Fisher's LSD post hoc test).

Regular and skipped heart beats over a single bout.

Figure 3 shows example traces of recorded blood pressure waveforms at different stages of hibernation (Fig. 3A) and beat-to-beat SBP and HR over one complete hibernation bout from one animal (Fig. 3B). Prior to entry into hibernation (i.e., when the animals were housed at cold ambient temperature and short photoperiod but were euthermic), the hamsters displayed periodic episodes of bradycardia that appeared to be skipped beats (Fig. 3, A1). This also occurred, albeit less frequently, when the hamsters were housed at room temperature (data not shown). During entry into hibernation, the interbeat interval became longer for both the regular and the skipped beats (Fig. 3, A2) until HR reached a minimum (10–14 bpm) in torpor (Fig. 3, A3). During most of the arousal period, there were only a few skipped beats (Fig. 3, A4). However, they were more frequent after the transient HR overshoot near the end of arousal.

Fig. 3.

A: example traces of recorded BP during various stages of a hibernation bout: euthermia (A1), entrance (A2), deep hibernation (A3), and arousal (A4). B: beat-to-beat SBP and HR over a complete bout of hibernation from one hamster. Black/white bar at bottom indicates dark/light cycle.

Temporal relationship of SBP, HR, and BRS over a hibernation bout in group P hamsters.

Data from 23 recorded hibernation bouts for the six group P hamsters (Fig. 4) show a temporal relationship among SBP, HR, and BRS. That is, at the onset of entry into hibernation both SBP and HR decreased, followed significantly later by the onset of the BRS decrease (Fig. 4, A and B). This lag period, when BRS remained high while both SBP and HR were exponentially decreasing, occurred during the first 2 h (of the ∼4 h) it took to enter torpor (Fig. 4A). Times for the 10–90% decrease were significantly different among all variables, with SBP being the longest and BRS being the shortest (Fig. 4C). These differences in the rate at which variables decreased were also observed in the time constants obtained by fitting data with a single exponential decay curve (Fig. 4C). Similarly, arousal began with an initial slow increase in HR (from 13.0 ± 0.3 bpm to 35 ± 1 bpm), followed by increased SBP, and then increased BRS (Fig. 4, A and B). At the onset of the BRS increase, SBP and HR were 124 ± 4 mmHg and 248 ± 21 bpm, respectively. Times for the 10–90% increase were significantly longer for BRS, and time constants from the fitted exponential curves confirmed that BRS increased the slowest (Fig. 4C).

Fig. 4.

Group data (n = 23 bouts) of hemodynamic changes over a hibernation bout from group P. A: 5-min averages of SBP, HR, and baroreceptor-HR reflex sensitivity (BRS) during entry into and arousal from hibernation. Horizontal dotted lines represent baseline values before entering hibernation. Standard error bars are omitted for clarity. Vertical lines indicate onset of entry into and arousal from hibernation. B: lag time of onset of hemodynamic changes for entry into and arousal from hibernation. C: duration (10–90%) and time constant of fall (entry) and rise (arousal) time. Bars sharing the same letter in each graph do not differ significantly (one-way repeated ANOVA; Fisher's LSD tests).

The changes over time described above suggest that entry and arousal patterns for BP, HR, and BRS differ, and this is confirmed by the hysteresis plots for these three variables (Fig. 5). Although the initial concurrent decreases in SBP and HR were somewhat linear (Fig. 5A; early phase of entry), BRS remained unchanged. At the later phase of entry into hibernation HR reached near minimum, whereas SBP continued to decrease and BRS began to fall (Fig. 5; late phase entry). During torpor, SBP gradually increased while HR and BRS remained low (Fig. 5; torpor). And during the initial phase of arousal, both SBP and HR increased while BRS remained low (Fig. 5; early phase arousal). HR continued to increase when SBP reached the peak of its overshoot (Fig. 5; late phase arousal), at which time BRS began to increase and continued to do so until all variables returned to prehibernation level (Fig. 5; recovery phase).

Fig. 5.

Hysteresis plots of HR vs. SBP (A); BRS vs. SBP (B); and BRS vs. HR (C). Data are from nine bouts of hibernation in one hamster.

Group PT hamsters.

We recorded a total of nine hibernation bouts from group PT hamsters (those implanted with BP and temperature telemetry transmitters). As was the case for group P hamsters, all hibernation bouts began entry into hibernation during the dark cycle and arousals were more distributed (Fig. 2). SBP and HR measurements for the nine bouts recorded in group PT hamsters (Table 2) were also similar to those of group P hamsters (Table 1) as was the temporal relationship of cardiovascular variables over hibernation bouts. Thus the larger size of telemetry transmitters in group PT did not appreciably affect these measurements.

Table 2.

Systolic blood pressure, heart rate, and body temperature over a hibernation bout

Hibernation Phase	SBP, mmHg	HR, bpm	TC, °C
Before hibernation	96 ± 6a	336 ± 7a	35.6 ± 0.1a
Hibernation lowest	29 ± 3b	11.0 ± 0.3b	7.8 ± 0.1b
Arousal onset	47 ± 3c	12.7 ± 0.5b	7.8 ± 0.1b
Arousal overshoot	140 ± 5d	492 ± 5c

Values are means ± SE; n = 9 bouts.

SBP, systolic blood pressure; HR, heart rate; bpm, beats per minute; T_C, body temperature.

For each variable (column), values that share the same superscript letter do not differ significantly (one-way repeated ANOVA followed by Fisher's LSD post hoc test).

Temporal relationship of cardiovascular variables to T_C over a hibernation bout in group PT hamsters.

Data from the nine recorded hibernation bouts for group PT hamsters showed that entry into hibernation started with concurrent decreases in SBP and HR, followed by declining T_C, and then declining BRS (Fig. 6, A and B). The rate of decrease in T_C during entry into hibernation was significantly less than that of the other measured variables (Fig. 6C). In contrast, T_C recovery time during arousal (as indicated by duration and time constant) was similar to that of SBP and HR, whereas BRS recovery was significantly slower (Fig. 6C).

Fig. 6.

Group data (n = 9 bouts) of hemodynamic and T_C changes over a hibernation bout from group PT. A: 5-min averages of SBP, HR, BRS, and T_C during entry into and arousal from hibernation. Horizontal dotted lines represent baseline values before entering hibernation. Standard error bars are omitted for clarity. Vertical lines indicate onset of entry into and arousal from hibernation. B: lag time of onset of hemodynamic and T_C changes for entry into and arousal from hibernation. C: duration (10–90%) and time constant of fall (entry) and rise (arousal) time. Bars sharing the same letter in each graph do not significantly differ (one-way repeated ANOVA; Fisher's LSD tests).

As can be observed in the hysteresis plots of cardiovascular variables vs. T_C for group PT hamsters (Fig. 7), curves for SBP vs. T_C and HR vs. T_C during arousal from hibernation did not retrace those recorded for entry into hibernation. Rather, the open loop indicates that the relationship for both SBP vs. T_C and HR vs. T_C were fundamentally different during arousal vs. entry. Specifically, upon entry into torpor, there was an initial rapid decline in both SBP and HR with a more gradual and delayed decline in T_C (Fig. 7, A and B; early phase entry), followed by a much slower decline in SBP and HR when T_C fell rapidly (Fig. 7, A and B; late phase entry). In contrast, BRS remained relatively high during the early phase of entry (i.e., until T_C reached ∼30°C), after which it rapidly declined, reaching a low when T_C was ∼20°C (Fig. 7C). During arousal, rapid increases in SBP and HR (early phase arousal) preceded the rapid increase in T_C (late phase arousal), and BRS remained low until T_C was almost back to euthermic temperature (Fig. 7C; recovery phase).

Fig. 7.

Hysteresis plots of SBP, HR, and BRS vs. T_C. Data are from six bouts of hibernation in one hamster.

DISCUSSION

The data from this study, which is the first to continuously measure BP, HR, T_C, and BRS over successive hibernation bouts, show that the temporal pattern of changes in these variables during a hibernation bout is consistent across bouts (summary in Fig. 8). This consistency provides insight to how these variables are related and to the role that baroreflex may play in regulating blood pressure as body temperature changes during the various stages of a mammalian hibernation bout. In particular, our novel finding that BRS (designated by the slope of each line in Fig. 8) remained high during the early stages of entry into hibernation implies that the baroreflex is dynamically and effectively altering HR, thereby ensuring that BP closely tracks a steadily declining set point during (indicated by Entry arrow in Fig. 8). To our knowledge, this is the first report demonstrating that the baroreflex has a clearly defined and important role during initial entry into a hibernation bout.

Fig. 8.

A simplified hysteresis plot of changes in all variables over a hibernation bout (from entry into hibernation through arousal) on the basis of all recorded bouts. The slope of each line indicates BRS at each point. The color in the background indicates changes in core body temperature. The loop is open because the interbout period is not included in the diagram.

A second novel finding in our study is that during arousal, BRS increased very slowly, reaching its euthermic value after BP, HR, and T_C had already done so (Fig. 8, Arousal arrow). These results imply that full dynamic baroreflex control is not attained until late in the arousal period. Moreover, the fact that hysteresis plots of pairs of variables display open loops (Figs. 5, 7, and 8) indicates that the cardiovascular control system operates in a fundamentally different manner during entry into and arousal from hibernation.

We hypothesized that the changes in BP, HR, and BRS occurred in unison as the hamsters entered and aroused from hibernation and prior to changes in T_C (H-1 and H-2). However, BP, HR, and BRS did not change in unison (thereby negating H-1), and although the changes in BP and HR preceded the changes in T_C (consistent with H-2), those in BRS did not. Moreover, the unexpected finding that BRS remained high during the initial entry into hibernation, then fell and remained low until late in arousal (Fig. 8) is consistent with the baroreflex regulation of HR being more important during the initial phase of entry into hibernation and the late phase of arousal than during the intervening periods. The overall temporal pattern itself was similar from bout to bout, supporting H-3.

Experiments in which BRS was continuously determined were feasible because of advances in the availability of small radiotransmitters for monitoring BP and in analytical methods for determining BRS. Two techniques for characterizing baroreflex function are in common use. One involves short-term infusions of phenylephrine or sodium nitroprusside to change BP while HR or sympathetic nerve activity is measured. However, data can be obtained at only a few time points over a hibernation bout using this procedure. We therefore chose the sequence method in which BRS is calculated from the slope of spontaneous baroreflex sequences (6, 44). Direct evidence supporting the validity of classifying responses as baroreflex vs. nonbaroreflex sequences derives from sinoaortic denervation experiments in which denervation significantly reduced (∼90%) the baroreflex sequence number and slope while having no effect on nonbaroreflex sequences (5, 6, 13, 28, 36). Additionally, slopes obtained from the sequence method are highly correlated with those generated from traditional drug-induced methods (7, 37). Thus, this method is a reliable technique for estimating BRS (44), and it opened the way for continuous calculation of BRS in a hibernating species during all stages of a hibernation bout.

Entry into hibernation.

Seasonal mammalian hibernators have developed adaptations to minimize their energy expenditure during winter months when food and resources are limited, and facultative hibernators such as the Syrian hamster exhibit similar adaptations when exposed to cold and short photoperiod. Early clues that the decrease in body temperature and metabolic rate remained regulated during entry into hibernation included observations that the fall in T_C during entry was significantly slower than the passive and unregulated drop in temperature occurring in dead animals (30, 31), and intermittent shivering and periodic transient rewarming occurred during entry (31, 45). Subsequent thermode experiments involving experimentally increasing hypothalamic temperature during entry into hibernation resulted in decreased thermogenesis (metabolic rate), whereas thermode-induced decreased hypothalamic temperature resulted in increased thermogenesis. These results clearly showed that a central negative feedback controller of T_C remained functional throughout entry into hibernation even as the temperature set point declined (17).

Cardiovascular regulatory mechanisms in awake and behaving Syrian hamsters are similar to those of nonhibernating mammalian species. Studies on the latter showed that the arterial baroreflex has two important features (14). One is its ability to provide powerful beat-to-beat negative feedback regulation of BP when an animal is resting. The second is to establish the new prevailing BP when the BP set point is reset to a different level (14). An example of the latter is the exercise pressor reflex: it was initially believed that the baroreflex was switched off during exercise, thereby enabling a parallel increase in BP and HR. However, it is now accepted that the parallel increase in BP and HR is due to baroreflex resetting (9, 14, 38–40). In such studies on active animals, including humans, the high temporal resolution (seconds) of the sequence method for calculating BRS enables its use to assess baroreflex function not only when the set point is stable, but also when it is changing.

Our measurements of BRS in Syrian hamsters clearly show that the baroreflex is not switched off during entry into a hibernation bout (i.e., BRS remained high during initial entry into hibernation; Figs. 4 and 6). Moreover, changes in HR and T_C do not follow the pattern observed during acute cooling in anesthetized animals (12). Thus the baroreflex can be included, along with thermocontrollers, in the neural systems that control early entry into hibernation. That is, just as a central thermal neurocontroller dynamically controls effectors that ensure T_C closely follows a declining temperature set point, the baroreflex dynamically controls HR such that BP closely follows a declining BP set point in the early entry stages of hibernation.

During the later phase of entry into hibernation, hamster BRS rapidly declined to a low level and remained low. The exponential decline in BRS suggests that baroreflex control of HR was rapidly withdrawn as T_C fell from 30°C to 20°C and remained minimal as T_C further declined to 12°C (Fig. 7). Milsom et al. (35) cite studies showing that the HR changes elicited by drugs or stimulation of autonomic nerves innervating the heart became minimal at low T_C.

Torpor.

Throughout torpor (1–4 days), BRS remained at a low level, implying that baroreflex control of HR was greatly attenuated. In some hamsters in some trials, a transient undershoot in SBP was observed (Figs. 1 and 3) before SBP settled at a low level. Regardless of the presence of an initial transient undershoot, BP slowly increased during torpor, whereas T_C declined to its lowest level (∼8°C). Because HR remained low and stable during these periods, these changes in BP are likely due to changes in peripheral resistance. The T_C decrease may contribute to increased resistance directly by reducing vascular reactivity or indirectly by changes in blood viscosity, and/or blood flow redistribution (1, 19, 23, 46, 47). There may also have been a gradual accumulation of circulating hormones and/or vasoactive substances (such as renin, angiotensin, and aldosterone) that contributed to increased peripheral resistance (2, 24).

Arousal from hibernation.

The central nervous system returned the hamsters to a waking state in less than half the time it took to enter hibernation. Indeed, sudden thermal, auditory, and tactile stimuli that arouse the animal are termed alarm signals, and rapid arousal allows the animal to behaviorally respond to environmental challenges (20, 21). In the absence of alarm signals, hamsters routinely arouse spontaneously after a few days in torpor. Although the chemical and neural signals involved in these arousals are yet to be fully identified, it is likely that an early step is activation of the ascending arousal system in the brainstem (3, 4, 18, 20). A novel finding in this study is that BP slowly rose throughout torpor, suggesting that baroreceptor signals encoding this increase may be one such signal, because neurons from the nucleus tractus solitarius project to the parabrachial nucleus (34), a key brainstem nucleus of the ascending arousal system (18).

During arousal, increased BP and HR support the thermogenic demands for oxygen. Initial rewarming is due primarily to sympathetic activation of brown adipose tissue metabolic activity (nonshivering heat production) (22), whereas later in arousal, motorneuron activation of shivering contributes to the rewarming. Interestingly, BRS was not fully reactivated until BP and HR increased above prehibernation levels and T_C was near euthermia, indicating the importance of the baroreflex in returning overshoots in BP and HR to prehibernation levels.

Cardiac arrhythmias.

One striking phenomenon consistently observed in all mammalian hibernators, including the Syrian hamster (Fig. 3), is the regular occurrence of skipped beats during various stages of hibernation or torpor (15, 27, 32, 33). These skipped beats are likely to be of sinus bradycardia nature (26, 33, 41) because previous work has shown that every P wave of electrocardiographic recordings during bradyarrhythmia was followed by a QRS complex in the absence of episodes of conduction block, atrioventricular nodal block, or other forms of ectopy (33, 41). Studies in black bears suggested that these bradycardias may be part of the exaggerated respiratory sinus arrhythmias (27). Although it is well documented that cardiac parasympathetic nerve activity is involved in the generation of skipped beats (32, 35), their significance is unclear.

Perspectives and Significance

Several cardiovascular responses to cold temperature in humans have been observed, including accidental hypothermia (11) and therapeutic cooling (25). Organ damage after rewarming is one major issue with human hypothermia conditions. Hibernators have solved this problem; that is, they have natural adaptations that enable them to hibernate at near zero temperatures without ill effects after rewarming. The dramatic and rapid increases in BP and HR that occur before changes in T_C during arousal (Fig. 8) suggest that establishing adequate perfusion before raising T_C may be particularly important in preventing organ damage during and after rewarming. This specific temporal relationship could be used for developing new strategies/guidelines for effective rewarming procedures for accidental hypothermia and therapeutic cooling.

GRANTS

This study was supported by National Heart, Lung, and Blood Institute Grant R01-HL-091763.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: B.A.H., J.M.H., and C.-Y.C. conception and design of research; S.M.C. and J.S.H. performed experiments; S.M.C., C.S., J.G., M.S., and C.-Y.C. analyzed data; B.A.H., S.M.C., J.S.H., C.S., J.G., M.S., J.M.H., and C.-Y.C. interpreted results of experiments; C.-Y.C. prepared figures; C.-Y.C. drafted manuscript; B.A.H., S.M.C., J.S.H., J.M.H., and C.-Y.C. edited and revised manuscript; B.A.H., S.M.C., J.S.H., C.S., J.G., M.S., J.M.H., and C.-Y.C. approved final version of manuscript.