INTRODUCTION

TOP ABSTRACT INTRODUCTION RESEARCH METHODS RESULTS DISCUSSION REFERENCES

EXPOSURE TO COLD AMBIENT temperature (T_a) produces a strong homeostatic challenge to homeothermic vertebrates. In mammals, a primary response to cold is elevation of metabolic heat production via nonshivering thermogenesis (15). This response is mediated by increased efferent sympathetic nervous system activity (27) and is very rapid; significant increases in uncoupling protein-1 mRNA in brown adipose tissue (BAT) occur in lean rats within 2.5 h of cold exposure (4). The pattern of augmented sympathetic nervous system activity in the cold has been studied by tissue norepinephrine kinetics (9, 10, 19, 39) and direct recordings in anesthetized animals (17, 22) and reveals that sympathetic activity is increased to regions other than BAT.

We speculate that secondary products of heightened resting energy expenditure include increased heart rate (HR) and mean arterial blood pressure (MAP), both of which are found in cold-exposed rats (see Ref. 31 for review). The time course of the cold-induced increase in MAP is often reported to be delayed relative to HR, which increases promptly in cold (9, 10, 14, 23, 40), but this may depend on measurement technique for MAP. Studies using tail-cuff measurement have indicated that at least 1 wk of cold exposure is required before hypertension results (31), whereas direct measurement by the catheterization technique has shown increased MAP within hours of cold exposure (12) and shows continued increases during 3-5 wk of additional exposure (33). The tail-cuff technique, but not catheterization, requires that the cold-exposed rat be warmed when MAP is measured and therefore may not be the best approach to assess the influence of cold exposure on cardiovascular function.

The primary purpose of the present study is to examine interrelationships between the cardiovascular and metabolic responses of spontaneously hypertensive rats (SHRs) and normotensive Sprague-Dawley (SD) rats when they are exposed to 4°C cold for 1 wk and when they return to the baseline T_a (23°C). The experimental situation we employed (36) offers the advantage of simultaneous, unobtrusive measurement of relevant variables during the entire daily cycle; HR and MAP were measured through telemetry and metabolic rate was measured by indirect calorimetry (O₂). Behavioral responses (food and water intake, locomotor activity) also were measured. We hypothesized that there would be a coordinated increase in HR, MAP, and O₂ in each strain, but possibly heightened responsivity in the SHR strain, which has been shown to respond to some environmental stimuli with exaggerated cardiovascular and sympathetic responses (16, 20, 21).

	RESEARCH METHODS

TOP ABSTRACT INTRODUCTION RESEARCH METHODS RESULTS DISCUSSION REFERENCES

Animals and housing. Eight male SHRs and seven normotensive SD rats (both from Harlan Sprague Dawley, Indianapolis, IN) were 12 ± 1 wk of age at the beginning of the experiment and had been housed for similar periods of time in standard laboratory conditions where T_a equals 21-22°C. The rats were anesthetized (pentobarbital sodium, 50 mg/kg) and instrumented with a catheter in the descending aorta coupled with a sensor and transmitter (TA11PA-C40; Data Sciences, St. Paul, MN) for telemetric monitoring of cardiovascular status. During recovery from surgery, the rats were housed individually in caging similar to the experimental cages, with a 12:12-h light-dark cycle and T_a 23°C. Each experimental cage (43 by 22 by 17 cm) had a polycarbonate base with an 8-cm lip containing animal bedding and sides and top constructed of wire mesh so that adequate air mixing could be achieved for the purpose of indirect calorimetry (see below). Powdered chow (Purina 5001; physiological fuel value, 3.3 kcal/g) in a food jar was reached by entering a modified stainless steel tunnel feeder that minimized spillage. Water was obtained by licking a tube located nearby. Each experimental cage was housed in a separate metabolic chamber that provided precise control of T_a (±0.1°C). The daily light-dark schedule (12:12-h) was imposed by lighting an overhead incandescent lamp (illumination ~60 lux of 2,250°K light) for 12 h each day, beginning at 2200 local time; the chamber was completely dark for the remaining 12 h. Rats remained in the metabolic chambers continuously except for a brief (15 min) daily maintenance period ~2 h before lights off.

Indirect calorimetry. The apparatus and procedures used for determination of metabolic rate have been described previously (26). Briefly, each experimental cage housed an individual rat placed in an acrylic metabolic chamber (40-liter capacity) where oxygen consumption (O₂) and carbon dioxide production (CO₂) were measured every 2.5 min by open circuit respirometry (2 l/min) using a modification of the approach described by Bartholomew et al. (3) to isolate successive samples. O₂ was adjusted for mass (ml · min¹ · kg^0.75) (5). A separate value for respiratory quotient (RQ; CO₂/O₂) was calculated for the entire 12-h dark phase and for the first 10 h of the light phase.

Cardiovascular telemetry. A telemetry receiver (RPC-1; Data Sciences) was positioned under the rat cage within the metabolic chamber. The pulses from the receiver were relayed to a calibrated pressure output adapter (R11CPA; Data Sciences), where they were converted to analog voltages representing blood pressure waveforms. The signals were amplified, filtered with a time constant of 1 ms, and relayed to a 12-bit analog-to-digital converter board in a computer that sampled the signals at 500 Hz with a resolution of 0.1 mmHg. Custom-written online software processed the blood pressure waveform, detected the systolic and diastolic blood pressures, and calculated MAP and HR. The average systolic and diastolic blood pressures, MAP, and HR were calculated for each 30-s period of the light-dark cycle and were stored on a floppy disk as 16-bit integers for off-line analysis. Extremely low or high blood pressures resulting from transient signal interference occurred rarely and were automatically detected and excluded from the 30-s averages.

Food and water intake. Daily food and water intakes were obtained by measuring on/off weights of the feeder (corrected for spillage) and water bottle during the daily maintenance period. The feeder and water bottle assemblies were equipped with a photobeam arrangement that indicated when the rats were ingesting food or water (see Ref. 36 for details).

Locomotor activity. One-half of the rats were run in metabolic chambers instrumented to record locomotor activity. These chambers rested on a fulcrum positioned across the narrow axis of the cage in which the rat lived. One end of the chamber was attached to a stiff strain gauge transducer (alpha-load beam; BLH electronics) that prevented rocking of the chamber on the fulcrum and allowed determination of the position of the rat within the chamber along the long axis. The electronic signal from the strain gauge was amplified, low-pass filtered with a 3-dB point of 10 Hz, and routed to an analog-to-digital converter where it was sampled at 20 Hz, thus allowing determination of the position of the animal's center of gravity along the long axis. To detect locomotor activity as the animal moved in the chamber, the rate of change in the chamber's center of gravity was low-pass filtered with an average time of 500 ms. Locomotor activity, measured in meters traveled, was accumulated in 30-s periods and stored with a 1-mm resolution.

Protocol. After recovery from surgery and the initial adaptation to the housing conditions, rats were transferred to the metabolic chambers (T_a 23°C) for an adaptation period of at least 2 days before data collection. Food and deionized water were available ad libitum throughout the experiment. When cardiovascular and metabolic data and food and water intakes were judged to be stable for three successive baseline days, T_a was decreased from 23 to 4°C at a rate of 10°C/h, beginning at the onset of the dark phase. The cardiovascular, metabolic, ingestive behavior, locomotor-activity, and body weight responses to cold exposure were observed for 1 wk. Each rat was removed from the cold for only 5-10 min/day during the daily maintenance period when the food, water, and rat were weighed, and the food, water, and chamber-bedding were refreshed. After the 7-day exposure to 4°C, T_a was returned to 23°C at the onset of the dark phase at a rate of 12°C/h for a 5-day recovery period. Because of a computer malfunction, SD rats remained in cold for 1 additional day.

Data analysis and statistics. Cardiovascular and locomotor measures were collected and stored in 30-s bins; metabolic measures were collected and stored in 2.5-min bins. All measures were transformed into 10-min bins that were used in computing separate values for the light and dark phases. Because the final 2 h of the light phase (which began with daily chamber maintenance) were excluded in these computations, values reported for dark and light phases are based on 12 and 10 h of data, respectively. The responses of the two strains were compared by repeated-measures ANOVA. Two main sets of analyses were carried out. "Cold" analyses were concerned with changes in body mass, food and water intake, MAP, HR, O₂, RQ, and locomotor activity during 7 days of cold exposure (data for the eighth day of cold experienced by the SD rats are plotted in the figures but are not included in the statistical analyses comparing strains. The responses of SD rats on this extra day of cold were similar to day 7 in each measure.) "Recovery" analyses were concerned with the extent to which the measured variables recovered to baseline levels during the 5 days when T_a was 23°C after cold exposure ended. In both the cold and recovery analyses, the baseline level for each measure was represented by the mean of the 3 baseline days immediately preceding cold exposure. In measures where separate dark- and light-phase data were obtained, separate ANOVAs were carried out for each phase. Comparable analyses were also carried out on "basal data" calculated for the MAP, HR, and O₂ measures. Basal data were obtained by sorting light- and dark-phase data each day to identify the twelve 10-min bins with the lowest O₂ values in each phase. Cumulatively, these data identified 2 h in each phase when metabolic rate was lowest. Mean values of MAP, HR, and O₂ calculated for each rat during these basal periods were used in the basal analyses. Tukey tests determined significant differences between means in post hoc pairwise comparisons. Statistical significance levels were set at P < 0.05. In all analyses, n = 8 for SHRs and n = 7 for SD rats, except for the following measures where n = 4: 1) locomotor activity (only 4 chambers were instrumented to quantify locomotor data), 2) SHR caloric intake (data from 4 SHRs implementing an improved feeder design that minimized food spillage were utilized), and 3) SHR RQ (due to temporary measurement difficulty, CO₂ data for 4 SHRs were unreliable during cold exposure).

	RESULTS

TOP ABSTRACT INTRODUCTION RESEARCH METHODS RESULTS DISCUSSION REFERENCES

Body weight, caloric intake, and water intake. Figure 1A indicates that on baseline and during exposure to cold, SD rats weighed about 65 g more on average than SHRs (F_Strain = 17.04, df = 1,13; P = 0.001), and there was some indication of weight loss (about 6 g on average) in response to cold (F_Days = 3.47, df = 7,91; P < 0.01). In the recovery period, both strains increased body weight (F_Day = 25.32, df = 5,65; P < 0.001), averaging about 4% above baseline level by the last day (Fig. 1A). Daily food intake (Fig. 1B), which did not differ between the strains in any part of the experiment, promptly increased in response to cold exposure (F_Days = 24.22, df = 7,63; P < 0.001) and returned to baseline levels 1 day after T_a returned to 23°C (F_Days = 4.13, df = 5,65; P < 0.01; post hoc tests). During cold, food intake reached a plateau about 50% higher than baseline by about the fifth day. SD rats drank about 9 ml more water each day than SHRs (Fig. 1C), but this strain difference reached significance only in the recovery period (F_Strain = 8.10, df = 1,13; P < 0.05). Cold exposure resulted in an increase in water intake in both strains (F_Days = 14.89, df = 7,90; P < 0.001), but this increase was sluggish relative to the rapid change found in food intake; post hoc tests indicated that water intake first became reliably greater than baseline on day 3 of cold. Water intake leveled off ~25-30% above baseline by about the fifth day of cold exposure. In the recovery period, water intake decreased to baseline by the third day (F_Days = 9.26, df = 5,64; P < 0.001; post hoc tests).

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Means ± SE body weight (A), caloric intake (1 g = 3.3 kcal; B), and water intake (C) on each day of the experiment for spontaneously hypertensive rats (SHRs) and Sprague-Dawley (SD) rats. SHRs, n = 8 except in B where n = 4; SD rats, n = 7. X-axis labels B1-B3, baseline days; R1-R5, recovery days; ambient temperature (Ta), 23°C in both cases. The days of cold exposure (4°C) are labeled 1-8.

View larger version (41K): [in this window] [in a new window]   Fig. 2.   Response of SHRs (n = 8) and SD rats (n = 7) in the light and dark phases on each day of the experiment. Means ± SE values are plotted for mean arterial blood pressure (MAP; A and B), heart rate (HR; C and D), and weight-specific metabolic rate (O2; E and F). Full-phase averages were computed from all 10-min bin data in the 12-h dark phase and in the first 10 h of the light phase. Basal averages were computed from the 12 10-min bins identified in each phase that had the lowest O2 values. The y-axis is scaled the same in the full-phase and basal panels for each measure. X-axis labels as in Fig. 1.

View larger version (44K): [in this window] [in a new window]   Fig. 3.   Means ± SE difference scores calculated from the full-phase data in Fig. 2. Daily scores calculated for each rat relative to the mean of its 3 baseline days were averaged to obtain means for each strain. Data from the eighth day of cold exposure in the SD rats are not shown. Means are plotted separately for the dark (A, C, and E) and light (B, D, and F) phases. X-axis labels as in Fig. 1.

View larger version (50K): [in this window] [in a new window]   Fig. 4.   Detail of MAP, HR, and O2 responses during the last hour of the light phase and the first 4 h of the dark phase on days when Ta began to change at lights out. Data are in 10-min bins. Difference scores for each rat were calculated between each bin, and the mean of its data in the final hour of the light phase. Average difference scores ± SE are shown for each strain. In cold onset (A, C, and E): Ta, 23°C in the final hour of the light phase; in cold offset (B, D, and F): Ta, 4°C during that hour. The time course of change in Ta is shown in each panel by the solid line whose high value is 23°C and low value is 4°C (no axis scale shown). In each panel, the data for both strains are plotted to overlap so that the SD rat data are plotted in front of SHR data. This format helps visualize bins in which there is greater responsiveness by SHRs. A and B: MAP; C and D: HR; E and F: O2.

MAP. MAP was higher in SHRs than in SD rats in both the cold and recovery analyses of light- and dark-phase data (F_Strain  35.26, df = 1,13; P < 0.001; Fig. 2A). Exposure to cold resulted in elevated MAP in both strains in the light and dark phases (F_Days  74.53, df = 7,91; P < 0.001) but not in an equivalent fashion (F_{Strain × Day}  13.51, df = 7,91; P < 0.001); pairwise comparisons indicated that cold initially induced a greater elevation above baseline in SHRs. This effect is shown most clearly in Fig. 3, A and B. By the fourth (light phase) or fifth (dark phase) day in cold, the greater elevation in MAP shown by SHRs disappeared; in the remainder of the cold exposure, the strains remained nondifferential (dark phase) or actually reversed such that SHRs had a smaller elevation above baseline than SD rats (light phase). On days 5 and 6 of cold, light-phase MAP was statistically indistinguishable from baseline in SHRs but remained statistically above baseline on all cold days in SD rats (Fig. 3B); in the dark phase, however, MAP in cold remained significantly elevated above baseline in both strains (Fig. 3A).

View larger version (30K): [in this window] [in a new window]   Fig. 5.   Means ± SE distance traveled by each strain in the light and dark phases on each day of the experiment. Because all of the chambers did not have activity sensors, the data in this figure are based on n = 4 for each strain. X-axis labels as in Fig. 1.

HR. In analyses of both cold and recovery data in the light and dark phases, HR was significantly lower in SHRs than in SD rats (F_strain  20.72, df = 1,13; P < 0.001), which reflected the baseline difference as well (Fig. 2C). HR became elevated in both phases on exposure to cold (F_Days  112.72, df = 7,91; P < 0.001), and in the light phase, the strains showed comparable changes during cold exposure (no strain-by-day interaction); after an initial increase of about 160 beats/min above baseline values, HR decreased to about 120 beats/min above baseline values by the final 3 days in cold (Fig. 3D). In the dark phase, however, a significant interaction occurred (F_{Strain × Day} = 3.29, df = 7,91; P < 0.01) and post hoc tests indicated that the rate of increase in HR across the first 3 days in cold was greater in the SHRs (Figs. 2C and 3C). During the recovery period, the changes in HR in both the light and dark phases (F_Days  21.52, df = 5,65; P < 0.001) showed no significant strain-by-day interactions. Post hoc tests indicated that after the first day of recovery, when HR was still above baseline, there was a gradual decrease to below baseline values that became statistically significant in the dark phase by day 4 and in the light phase by day 3.

O₂. There was no significant main effect of strain in analyses of cold or recovery O₂ data (Figs. 2, E and F, and 3, E and F). Exposure to cold resulted in a large increase in O₂ in both phases (F_Days  285.74, df = 7,91; P < 0.001) that differed between strains only in the dark phase (F_{Strain × Day} = 2.86, df = 7,91; P = 0.01). Pairwise comparisons indicated that dark-phase O₂ was significantly greater in SD rats on days 3 and 6 of cold (P < 0.05; Figs. 2E and 3E). In the recovery analyses, the only significant effect was a change across days in both phases (F_Days  5.20, df = 5,65; P < 0.05), which pairwise comparisons indicated was attributable to above-baseline O₂ on the first day, followed by the recovery to baseline on subsequent days in the recovery period.

RQ. Analysis of the RQ data in the dark and light phases indicated some differences between the strains. In the baseline period, dark-phase RQ was significantly higher in SHRs than in SD rats (0.98 ± 0.005 vs. 0.94 ± 0.005, respectively); the light-phase values for the two strains did not differ (SHRs, 0.95 ± 0.005; SD rats, 0.92 ± 0.005). The principal effect of the cold exposure was a decrease in RQ that averaged about 0.06 in the dark phase and 0.045 in the light phase on the early days in cold. On the later days in cold, there was some recovery in RQ and it promptly returned to baseline when T_a increased in the recovery period.

Locomotor activity. Figure 5 shows that exposure to cold resulted in an increase in locomotor activity by the two strains in both the dark and the light phases (F_Days  5.80, df = 7,42; P < 0.001). In these analyses, there was no main effect of strain nor was there a strain-by-day interaction. Pairwise comparisons of data in the dark phase indicated that activity was elevated above baseline beginning on the second day in cold. In the light phase, activity was significantly elevated above baseline on the first and second days in cold but not on the remaining days. Analyses of locomotor data in the recovery period yielded no significant effects; when T_a returned to 23°C, locomotor activity returned to baseline levels in both strains in the dark and light phases.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESEARCH METHODS RESULTS DISCUSSION REFERENCES

The present experiment shows that MAP, HR, and O₂ are promptly elevated by cold onset and decreased by cold offset in hypertensive and normotensive rats. With two main exceptions, the pattern of results found in the two strains was similar. One difference was that in the MAP and HR measures, the initial response by SHRs to change in T_a was relatively large compared with the normotensive SD rats. At cold onset, the SHRs had greater elevations in these measures and at cold offset the SHRs had greater depression in MAP, which actually fell significantly below baseline levels for several days in both strains, indicating a postcold period of hypotension. The other difference was that during several days of cold exposure, the SHRs showed a greater tendency to normalize MAP, especially in the light phase when locomotor activity and feeding were less likely to occur. In fact, after several days of cold exposure, basal levels of MAP decreased to values that were statistically indistinguishable from baseline in the SHRs. Although MAP in SD rats also decreased over time during cold exposure, it remained significantly elevated above baseline throughout the cold period. The overall pattern of our results is consistent with the hypothesis that the initial metabolic and cardiovascular responses to cold exposure reflect elevated sympathetic activity that is required to defend core body temperature. It is suggested that as sympathetic activation is sustained in cold, compensatory mechanisms become engaged that help normalize MAP, and these mechanisms are expressed more fully in SHRs than in SD rats. When cold is abruptly terminated after 1 wk, these compensatory mechanisms are expressed as hypotension for a few days.

Fregly and colleagues (11, 12, 23, 24, 29, 31-33) have established that cold exposure elevates MAP through both neural and renal mechanisms. The reported time course of activation of these mechanisms likely depends on the measurement technique for MAP. Many of Fregly's experiments employed the tail-cuff method for measuring MAP and that work indicated rather sluggish changes in MAP in response to cold (e.g., several days or weeks). However, one study from his laboratory reported a rapid response to cold onset when MAP was measured (in SD rats) by the chronic catheterization method (12). That finding is more in line with our present results that were obtained by the telemetry method. Furthermore, our telemetry data and Fregly's data from the catheterization method (12) also indicate that an increase in T_a following cold exposure inhibits the signals that elevate HR and O₂ during cold. This is perhaps mediated by withdrawal of sympathetic activation following cold offset. Fregly's data showed that a decrease in MAP occurred within hours of rewarming, but that MAP rebounded over a period of several days and then remained elevated compared with precold levels even 23 days after cold exposure ends. However, these data on the kinetics of change in MAP are limited by infrequent measurements; after the initial 24-h postcold period, MAP was measured at only three time points over 3 wk. The more complete picture provided by our data shows that after 1 wk in 4°C, an increase in T_a to 23°C is accompanied by 1) a decrease in MAP that is evident even during the first few hours after T_a change begins, 2) a period of postcold hypotension for several days, which is more severe in SHRs than in SD rats, and 3) the development of bradycardia between 3 and 4 days after removal from cold. Additional studies of the kinetics of changes in cardiovascular functioning during long-term cold exposure and during postcold recovery periods would provide important information about the mechanisms responsible for cold-induced changes in hypertension. Such studies would be best carried out using telemetry measurement of cardiovascular responses.

What mechanisms might account for the compensatory decrease in MAP toward baseline during sustained cold exposure in SHRs? Because we did not measure cardiac output in these studies, we do not know the respective contributions of cardiac output and systemic vascular resistance to the arterial blood pressure response during sustained cold and rewarming. It has been shown previously that cardiac output increases during cold exposure (1, 13, 35). Atterhog et al. (1) found increased MAP and cardiac output and lowered systemic vascular resistance in humans. Hales et al. (13) also reported increased cardiac output, as well as lowered total peripheral resistance (TPR) in cold sheep, although the reduction in TPR was not statistically significant. Although we did not measure cardiac output or TPR, the sustained significant increases in HR and O₂ throughout cold exposure suggest that cardiac output remains elevated during cold exposure. However, compensatory reductions in MAP during sustained cold may be due to either reductions in stroke volume (see below) or systemic vascular resistance.

A possible explanation of reduction in MAP during sustained cold is pressure-mediated natriuresis and diuresis, leading to a reduction in circulating blood volume. Cold exposure is known to produce diuresis (11). The possibility of reduced blood volume is indirectly supported by the delayed increase in fluid consumption that occurred 3 days after initiation of cold exposure. However, SD rats also showed a delayed increase in fluid intake, although MAP remained significantly elevated compared with the precold baseline. The postcold hypotension that we observed within hours of beginning the rewarming in T_a may also indirectly support a role for reduced blood volume during cold exposure. Two recent reports have indicated that blood volume is actually increased by 1 wk of cold exposure (28, 33). However, in one of these reports, it is clear that animals were rewarmed during surgical implantation of catheters and measurement of blood volume (28). We did not measure the blood volume response to cold exposure.

The greater decrease in MAP shown by the SHR strain during cold exposure might be related to that strain's exaggerated sympathetic reactivity (16, 20). The SHRs responded more strongly in the MAP measure than did SD rats when cold exposure began, and SHRs have been shown to have a greater plasma catecholamine response to acute cold stress (25). Thus it could be argued that the fall in MAP during the week of cold exposure simply reflects a gradual dissipation of the exaggerated sympathetic response to cold. However, this explanation is tempered by the fact that the HR response peaked on the third day of cold exposure in SHRs, whereas MAP began declining toward baseline after the first 24 h in cold. It is also possible that chronic sympathoexcitation during cold resulted in a reduced vasoconstrictor responsiveness, causing a decrease in vascular resistance and MAP. If this were the case, then the mechanisms involved would have to operate within just a few days to explain both the decrease of MAP in SHRs during cold and the recovery of MAP during the postcold period. The possibility should also be noted that changes in parasympathetic activity unrelated to sympathetic activity could mediate some of the observed strain differences in the MAP response to cold.

T_a has sometimes been reported to have different effects on metabolic rate (O₂) in SHRs and normotensive Wistar-Kyoto controls. Strain differences were reported in adult rats when T_a was in the range 25-32°C (38) and in infant rats when the T_a range was 17-23°C (18). However, others have reported no differences in O₂ between these strains in cold (5°C) or warm (35°C) conditions (37). Furthermore, an acute (90-min) exposure to cold (6°C) was found to result in no strain difference in either HR or lumbar sympathetic nerve activity (34). Thus whether differential sympathoexcitation between SHRs and SD rats during cold exposure can explain the MAP response in SHRs appears to be an unresolved issue. Our findings in the SD strain, similar to previous results by Fregly and Schechtman (12), show that MAP in the SD strain did not return to baseline levels.

Although it seems clear that the activation of the sympathetic nervous system is a major component of the cardiovascular responses during 1 wk of cold exposure (7, 23, 40), other mechanisms contribute to the cardiovascular outcomes of this homeostatic challenge. For example, intravenous administration of a neuropeptide Y receptor antagonist transiently lowers MAP in the cold-exposed borderline-hypertensive rat (14). Thus peripheral neuropeptide Y release from sympathetic nerve terminals may contribute to cold-induced increases in blood pressure. An additional ramification of chronic elevation of efferent sympathetic nerve activity may be activation of the renin-angiotensin-aldosterone system. One week of cold exposure produced a severalfold increase in plasma ANG II levels (7). Interestingly, the increases in plasma ANG II and norepinephrine caused by cold exposure were prevented by pair feeding. We speculate that pair-fed animals were placed in severe negative energy balance and that the normal cold-induced elevation of sympathetic activity was prevented by the activation of homeostatic mechanisms due to diminished energy reserves.

In animals allowed to increase food intake during cold exposure, angiotensin-converting enzyme inhibition, angiotensin receptor blockade, and aldosterone receptor blockade have been shown to reduce MAP (see Ref. 31 for review). Thus there is substantial evidence for a role of the peripheral renin-angiotensin-aldosterone system in cold-induced hypertension. Furthermore, central angiotensin mechanisms also may be involved in cold-induced hypertension. Central administration of antisense oligonucleotides to both angiotensinogen mRNA and the AT₁ receptor mRNA were shown to reduce MAP in animals with chronic (5 wk) cold-induced hypertension (24). Additional work is needed to elucidate how these mechanisms interact to alter MAP during acute and chronic exposure to cold.

There is a heterogeneous pattern of sympathoexcitation during cold exposure (9, 10, 17, 19, 22, 39). The evidence indicates that sympathetic activity is increased to both BAT and the heart but not the kidneys or the liver. Effector mechanisms for temperature regulation, including sympathetic outflow to BAT, are under control of the preoptic hypothalamus (6, 30). In addition to the preoptic area, multiple hypothalamic nuclei, including the paraventricular and ventromedial nuclei, appear to comprise an organized central network innervating BAT (2). Coronal transection caudal to the preoptic area strongly activates nonshivering thermogenesis (8), and thus pathways that emerge from the preoptic hypothalamus must chronically inhibit sympathetic activity to BAT. At least one target for these projections may be the rostral raphe pallidus. Disinhibition of GABAergic activity in this region by microinjection of bicuculline strongly increases sympathetic activity to BAT (22). Furthermore, acute exposure to cold (4°C for 4 h) increases fos expression in rostral raphe pallidus (22). It is tempting to speculate that this pathway could provide a mechanism for the concurrent increases in HR and nonshivering thermogenesis observed during cold exposure and rewarming.

In summary, the overall pattern of our results indicates that variations of T_a produce rapid changes in a suite of cardiovascular and behavioral responses that have many similarities in hypertensive and normotensive strains of rats. The findings are consistent with the general concept that the cardiovascular responses to cold exposure in rats are closely related to, and perhaps a secondary consequence of, the mechanisms responsible for increasing heat production.