Acute effects of light on body temperature and activity in Syrian hamsters: influence of circadian phase

Departments of ¹ Psychology, ² Psychiatry and Pharmacology, Dalhousie University, Halifax, Nova Scotia, Canada B3H 4J1

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Light exposure at night causes an acute increase in human body temperature, which normally falls during the night. This change is largely attributable to the suppression by light of the nocturnal rise in melatonin levels. Little is known, however, about the effects of light on body temperature in nocturnally active mammals in which the nightly peak in melatonin secretion coincides with the circadian phase of elevated, rather than decreased, body temperature. We investigated the effects of a 1-h exposure to light on body temperature and activity of Syrian hamsters, Mesocricetus auratus, at two phases during the night and at two phases during the projected day. Brain or abdominal temperature was recorded continuously using implanted radio transmitters while locomotor activity was monitored simultaneously using a passive infrared movement detector. Responses to light exposure were strongly circadian phase dependent; light during the night caused elevations in both brain and core body temperature, whereas light during the projected day did not. Temperature increases at night could not be attributed solely to activity increases at the onset of light pulses, indicating a contribution from nonbehavioral mechanisms of thermogenesis. These results provide the first evidence for circadian modulation of acute temperature responses to light in a nocturnal mammal.

abdominal temperature; brain temperature; melatonin; Mesocricetus auratus; nocturnal

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

MOST MAMMALIAN circadian rhythms, including those of body temperature (T_b) and activity, are generated by a circadian pacemaker located in the suprachiasmatic nucleus (SCN) of the hypothalamus (45, 56, 59). These internally generated rhythms are synchronized to the 24-h rotation of the earth by environmental cycles acting as zeitgebers (entraining agents), the most important of which is the cycle of light and darkness (48). Light-dark (LD) information is conveyed to the SCN through a set of projections comprising a circadian photic system, which includes a direct retinal projection to the SCN (44, 47).

When animals are kept in constant environmental conditions, without an external synchronizing cycle, their behavioral and physiological rhythms are expressed with near-24-h circadian rhythms, referred to as free-running (6). Brief light pulses presented against a background of constant darkness (DD) can then cause shifts in the phase of these rhythms when presented during the animal's subjective night, but not during the subjective day (41, 43, 64).

Light can also exert physiological and behavioral effects independently of shifting circadian phase. In humans, who normally show a decline in T_b during the night, early evening or nighttime light exposure of appropriate intensity and duration induces an immediate increase in core body temperature (T_co; see Refs. 8, 15, 62). During the daytime, however, when T_b is normally elevated, exposure to bright light does not induce similar increases (8). Light exposure has also been reported to elevate T_b of a diurnally active squirrel monkey (46). Whether these acute effects of light on T_b and its phase-shifting effects are mediated by related or independent mechanisms is not clear (11, 14, 18). Injection of glutamatergic antagonists into the suprachiasmatic region can block light transmission to the SCN that would stimulate brown fat thermogenesis in anesthetized rats (2, 3), suggesting a role for hypothalamic mechanisms in both processes.

The mechanisms underlying thermoregulatory responses to nocturnal light pulses are not well characterized. In humans, the normal nocturnal decline in T_b parallels the nocturnal rise in pineal melatonin secretion (12, 62), which is also regulated by the circadian pacemaker in the SCN (44). Bright light is capable of suppressing melatonin (40, 61), and the effects of light on T_b can be antagonized in a dose-dependent manner by exogenous melatonin (13, 18, 38, 62). Thus it has been proposed that suppression of melatonin levels accounts for much of the temperature-elevating effects of nocturnal light exposure in humans (8, 11).

Melatonin may have different effects on thermoregulatory processes in nocturnal species from those in diurnal species. Melatonin is produced during the dark phase, and light suppresses melatonin synthesis, regardless of whether a species is nocturnal or diurnal in behavior type (21, 26, 34, 35, 40). Thus nocturnal animals exhibit high T_b, whereas diurnal animals exhibit low T_b when melatonin levels are elevated at night. It has been reported recently that daytime melatonin administration elevates the cortical temperature or T_co of several nocturnal rodents, although melatonin-induced hypothermia was also reported in some early studies (4, 5, 20, 32, 33, 55, 60). If melatonin suppresses T_b in nocturnal species, as it appears to do in humans, then the nocturnal rise in temperature normally observed in rats or hamsters must occur despite this influence. Alternatively, melatonin may have opposite thermoregulatory effects in nocturnal and diurnal species. In that case, if light alters T_b via melatonin suppression, nighttime light should decrease, rather than elevate, T_b in nocturnal animals.

Because the effects of light exposure on T_b regulation and their dependence on circadian phase have not been studied in nocturnal rodents, we investigated changes of brain temperature (T_br) and T_co of unrestrained Syrian hamsters, Mesocricetus auratus, in response to light exposure. Light pulses were applied either during the dark phase, when light is known to phase shift circadian rhythms and elevate human T_b, or during the light phase when light does not phase shift rhythms. Because hamsters are relatively small, changes in activity could have significant, rapid impacts on their T_b (52). We therefore monitored motor activity simultaneously with T_br or T_co to quantify possible contributions of changes in activity to T_b changes associated with an experimental treatment.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Animals and experimental procedures. Forty adult male Syrian hamsters, M. auratus (~150 g; LVG:lak; Charles River, St. Constant, Québec), were used. Twenty-two hamsters were implanted surgically with temperature-sensitive FM radio transmitters (~2.5 g with coating; model VM-FH; MiniMitter) in the abdominal cavity to record T_co. The other 18 animals were implanted intracranially to record T_br using temperature-sensitive brain-probe FM transmitters (~1.2 g; model XM-FH; MiniMitter) sealed within the tip of a cannula.

The brain probe transmitters were fixed to the surface of the skull using dental cement and stainless steel screws as anchors. The tip of the cannula bearing the temperature sensor (460 µm diameter) was aimed at the dorsomedial hypothalamic region with the aid of a Kopf stereotaxic device (coordinates: 0.3 mm posterior to bregma, 0.5 mm lateral to midline, and 7.0 mm below the skull surface). This area was targeted to record hypothalamic temperature without risking damage to the region of the circadian pacemaker in the SCN. At the conclusion of the experiment, most animals with brain probes were killed with an anesthetic overdose, and their brains were perfused and prepared for standard cresyl violet histological examination to identify probe locations. The tips of the probes were located in the ventromedial thalamus toward the boundary of the dorsomedial hypothalamic region (Fig. 1). All surgical procedures were conducted under deep pentobarbital sodium anesthesia (Somnotol; ~100 mg/kg ip; MTC Pharmaceuticals), and all procedures were conducted in accordance with Canadian Council on Animal Care recommendations and with the approval of the Dalhousie University Committee on Laboratory Animals.

View larger version (158K): [in this window] [in a new window]   Fig. 1.   Coronal section of a hamster brain slice showing a typical location of a brain temperature probe. The tip of the probe (*) was located in the ventromedial thalamus near the border of the dorsomedial hypothalamus. VM, ventromedial thalamic nucleus; VMH, ventromedial hypothalamic nucleus; 3V, third ventricle. Scale bar: 1 mm.

After surgery, animals were housed individually in standard plastic cages (44 × 23 × 14 cm), with wood shavings as bedding materials. The cages were placed inside individual ventilated, light-proof chambers in a temperature-controlled room (25 ± 1°C) under a 14:10 LD photoperiod. A 15-watt incandescent lamp (GE lighting) inside each chamber served as the light source. The light intensity was ~100 lx measured with a photometer (Optikon, Santa Monica, CA) held horizontally 40 cm below the light source. Food (Purina Laboratory Chow) and water were available at all times, and the animals maintained their body weights throughout the experiment.

Animals were divided randomly into four groups that were tested at different times of day. Zeitgeber time (ZT) 0 was defined as the time of lights on, and ZT14 was defined as the time of lights off. Detailed information for each of these groups, including numbers of animals studied, numbers of animals for which T_br or T_co were recorded, lighting conditions, and times of treatments, are provided in Table 1. Animals were allowed to entrain to LD cycles for at least 2 wk before any treatment. Treatments consisted of a 1-h exposure to a light pulse using the same light source as for daily lighting cycles. Animals were exposed to a light pulse at one of four daily phases and were tested again in the same manner at the same daily phase after a 1-wk recovery period.

View this table: [in this window] [in a new window]   Table 1.   Experimental treatments

Animals were exposed to light for 1 h beginning at one of two times during the subjective day (projected light phase; ZT6 or ZT10) or at two phases during the subjective night (ZT16 or ZT22). To allow us to expose animals to a light pulse against a dark background during their subjective days, animals were kept in darkness for 40 h (ZT6 pulses) or 44 h (ZT10 pulses) before these light pulses. The usual LD cycle was reestablished at ZT6 the day after the treatment. Because hamsters have free-running circadian rhythms with periods that are very close to 24 h (49), there would be little spontaneous shift in phase relative to the entraining cycle during the treatment day. The tests during the subjective night were conducted during the dark phase of a LD cycle, and data from these tests were used for analyses. A control experiment was conducted during which animals were maintained in DD for 1-2 days before being tested during the subjective night, and no differences were observed between subjective night responses in these two different situations (see RESULTS). Increases in ambient temperature measured in the middle of a cage were <0.7°C after a 0.5-h light exposure and were <1.1°C after a 1-h light exposure.

Data acquisition. Transmitters used for recording both T_br and T_co were calibrated against a precision-certified mercury thermometer in a water bath to the closest 0.1°C before implantation. Radio frequency signals corresponding to an animal's T_br or T_co from a transmitter were captured and amplified by an antenna system. This system consisted of two-dimensional coils surrounding each animal's cage. The antennas from different measurement channels were multiplexed via multiplexing chips (MS 4051bcn; Fairchild) to a receiver (RLA 3000; Data Sciences International) that had been modified to reduce background noise and to selectively enhance sensitivity to the transmitters' carrying frequency.

Locomotor activity was monitored using passive infrared recorders (PIR; DSC BV300; Bravo) hanging 25 cm above the animals' cages. Movements of a body that has a temperature different from its surroundings trigger a PIR that, in turn, produces an activity count (37). The PIRs detected any movement made by an animal >1 cm in any direction, to a maximum of 5 counts/s.

A programmed microprocessor (MC 68HC 705J1A; Motorola) was used to control switching among the antennas. The microprocessor also read the transistor-transistor logic signals from the transmitters and from the PIRs for each animal and sent the signals to an IBM personal computer via a serial port. The radio transmitter frequency for each of eight animals was measured sequentially for a 1-s interval every 8 s. For each animal, the characteristic radio frequency in a 3-min window was defined as the observed value with the smallest difference from all other frequencies recorded during that interval. This value was calculated and stored every 3 min by computer software written in Visual Basic that also transformed these radio frequency signals into temperature readings according to the regression equations obtained from transmitter calibration. The activity counts from each animal recorded continuously by a PIR were summed and stored by the same software in 3-min intervals, simultaneously with the animal's T_br or T_co data.

Data analysis. The values for T_b and activity counts for each time point recorded during two baseline days before a treatment were averaged. The averaged data were used to construct time series for activity and T_br or T_co, which were plotted for each animal as daily waveforms. Similarly, data for each daily time point from the two replicate treatment days were averaged for each animal. These values were used to calculate other measures: daily mean temperatures and activity counts for each individual were calculated by averaging the data points collected from 1 h before the treatment time to the same time the following day (or the equivalent phase during baseline days). Daily T_br and T_co minimum and maximum values were calculated as the means of the three lowest consecutive values during the subjective day or the three highest consecutive values during the subjective night. Daily temperature amplitude was calculated as the difference between the calculated daily maximum and minimum.

To minimize irrelevant variations across days and to permit calculation of group mean waveforms, T_br and T_co data were normalized for each animal on a daily basis using a Z transformation (68). Each data point was subsequently expressed in units of standard deviation relative to the daily mean. Further calculations and statistical analyses on T_br and T_co were all based on these standardized data. Overall means were calculated, and time series were plotted for groups of animals who were treated with light pulses at a single ZT.

Hourly mean T_br, T_co, and activity were calculated as the mean of the 20 data points recorded during each hour. For both T_b and activity, several measures were derived: peak temperature during the light pulse, temperature nadir after the light pulse, and the latency to each of these. The T_b and activity peaks were defined for each animal as the highest values recorded during the hour in which a light pulse was given or in the corresponding hour during baseline days. Similarly, T_b and activity nadirs were defined as the lowest values recorded during the period after the end of the light-pulse hour. For both T_b and activity peak and nadir values, the latency to reach these values was defined as the time elapsed from onset of the light pulse (or equivalent time during a baseline day) until this value was achieved. The length of time that T_b was elevated over baseline values after a treatment was also determined.

Animals were considered to be active whenever any activity was detected within a 3-min interval. Thus the total activity time during a specific period was calculated as the sum of all 3-min intervals during that period in which an animal showed activity. Durations of inactivity were calculated in a corresponding way.

Group means for all variables were obtained by averaging the relevant values from all individuals in the same treatment group. Differences between baseline and treatment days in mean T_b and activity characteristics at each ZT were examined using paired t-tests. T_b and activity characteristics at different ZTs were evaluated using least-squares regression methods, and the contribution of the mean level of motor activity to the elevation in T_b was determined accordingly using the method described by Gordon and Yang (28).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Daily T_br, T_co, and activity patterns. Hamsters showed pronounced daily rhythms in T_br, T_co, and locomotor activity that were entrained to the 24-h LD cycles (Fig. 2). Fluctuations of T_b generally reflected changes in activity levels (Fig. 2; see Table 5 for slopes and correlation coefficients between activity and T_b). Animals usually became active and showed a gradual increase in T_b shortly before the beginning of the dark phase, an anticipation typical of entrained circadian rhythms. Each dark phase was typically characterized by two or three peaks of T_b and activity, with intervening periods of lower activity and T_b. The first peak in the early dark phase was generally the largest and lasted for 2-3 h, whereas the others usually did not exceed 1 h (Fig. 2). During the light phase, animals spent most of their time at rest, and both their T_b and activity were lower than during the dark phase. Nonetheless, they still showed short, irregular bursts of increased activity and T_b at 2- to 4-h intervals, indicating an ultradian rest-activity rhythm characteristic of many small rodents (Fig. 2 and Refs. 7 and 17). These ultradian rhythms may represent feeding pattern-related activity and thermogenesis in the hamster (54).

View larger version (40K): [in this window] [in a new window]   Fig. 2.   Examples of 24-h waveforms showing the baseline patterns of brain temperature (Tbr), core body temperature (Tco), and locomotor activity of Syrian hamsters (Mesocricetus auratus). Tbr (A: hamster 510) or Tco (B: hamster 42) was recorded simultaneously with activity for each individual animal at 3-min intervals. Each point represents the average of values recorded at that time on 2 different days. Black bar represents the daily dark phase of a 14:10-h light-dark (LD) cycle. Abscissa is labeled as Zeitgeber time (ZT) in which ZT0 represents the beginning of the daily light phase. Open boxes indicate the times at which a light pulse could be given on experimental days (see Fig. 3).

The daily patterns of T_br and T_co were similar overall (Fig. 2); however, T_br was generally higher than T_co when the daily mean, maximum, and minimum were compared (Table 2). The daily amplitude of T_br was smaller than that of T_co (Table 2). Rapid, small changes were also detected more frequently for T_br than for T_co (Fig. 2). Nevertheless, once standardized for mean level, these two measures of temperature are quite comparable, and they are analyzed and discussed together below.

View this table: [in this window] [in a new window]   Table 2.   Comparisons of Tbr and Tco characteristics on days with and without light pulses

Responses of T_br, T_co, and activity to light exposure. Light pulses did not significantly alter the overall daily mean, maximum, minimum, or amplitude of either T_br or T_co rhythms (Fig. 3; Table 2), neither did they affect daily mean activity counts (22.6 ± 10.1 for light-pulse days, n = 40; 22.6 ± 10.8 for baseline days, n = 40). Light pulses applied at ZT16 often coincided with the latter part of the first nocturnal activity peak, whereas those at ZT22 often occurred during the last activity peak in the dark phase (Figs. 2 and 3). Light exposures at ZT6 and ZT10 usually did not occur systemically in association with the small-amplitude bursts of activity or T_b elevation that occurred spontaneously during the projected day phase (Fig. 2).

View larger version (38K): [in this window] [in a new window]   Fig. 3.   Examples of 24-h waveforms showing patterns of Tbr, Tco, and locomotor activity of Syrian hamsters (M. auratus) that were exposed to a light pulse lasting for 1 h at ZT16-ZT17 (A) or at ZT22-ZT23 (B). Tbr (A: hamster 510) or Tco (B: hamster 42) was recorded simultaneously with activity at 3-min intervals for the same individual animal as in Fig. 2, and each point represents the average of the values recorded at that time during 2 treatment days. Open box shows the time at which the light pulse was given (see Fig. 2).

Responses to light pulses varied considerably with phase of treatment. At both ZT16 and ZT22 during the dark phase, light exposure caused a brief increase in activity that peaked rapidly (Table 3). This peak was followed by a sustained reduction in activity lasting >1 h and was characterized by lower intensities of activity and/or longer duration of inactivity (Table 3). Interestingly, nighttime light pulses also caused a sustained elevation in T_b lasting for most of the 1-h light exposure (Fig. 3). After a light pulse, T_b was significantly higher than baseline levels for 24.43 ± 18.77 min at ZT16 and for 54.81 ± 54.99 min at ZT22. In contrast, at both ZT6 and ZT10 during the projected light phase, there were few changes in either activity or T_b associated with the light pulses (see below for details). The patterns of responses to light treatments observed at any given ZT were similar for T_br and T_co, although changes in T_br were generally more rapid. The durations of any changes observed in T_br and T_co in response to light exposure were also similar at a given ZT (Fig. 3). Nighttime responses of T_b to light pulses were similar regardless of whether the animals were tested in the dark phase of a LD cycle or after 1-2 days in DD. At either ZT16 or ZT22, the mean T_co during a light pulse was always at least 45% higher than that during the same interval on the baseline day (Fig. 4).

View this table: [in this window] [in a new window]   Table 3.   Statistical description and analysis of results on activity

View larger version (41K): [in this window] [in a new window]   Fig. 4.   Twenty-four-hour group mean (n = 6 animals) waveforms showing patterns of Tco and locomotor activity of Syrian hamsters (M. auratus) that were kept in constant darkness (DD). The animals were exposed to a light pulse lasting for 1 h at ZT16-ZT17 after 50 h in DD (A and B) or at ZT22-ZT23 after 56 h in DD (C and D). Tco was recorded simultaneously with activity at 3-min intervals. A and C show Tco and activity baselines; B and D show data collected on experimental days. Dashed open boxes indicate the times at which a light pulse could be given on experimental days, whereas the solid open boxes show the time at which the light pulse was actually given. A: ZT16, no light, n = 6; B: ZT16, with light, n = 6; C: ZT22, no light, n = 6; D: ZT22, with light, n = 6.

To examine further the changes in T_b and activity caused by light treatment at various ZTs, Figs. 5 and 6 show individual mean standardized T_b and activity waveforms during a 4-h period beginning 1 h before and continuing until 3 h after the initiation of a light pulse. At ZT16 (Fig. 5), light exposure evoked an immediate, brief increase in activity, followed rapidly by a strong suppression of activity. As a result, the latency to reach peak activity during the hour was significantly shorter during light-pulse days than during baseline days. Mean activity level and duration during the light pulse, and activity duration during the following hour, were nevertheless significantly reduced during treatment days compared with the equivalent phases on baseline days (Fig. 5; Table 3).

View larger version (24K): [in this window] [in a new window]   Fig. 5.   Changes in Tb (A and B, top) and locomotor activity (A and B, bottom) in hamsters during the time from 1 h before to 3 h after a light pulse was applied during the dark phase at ZT16 (A) and ZT22 (B). Solid lines represent data collected on days with a light pulse (see Fig. 2), and dotted lines represent data from baseline days without a light pulse (see Fig. 2). The onset of the light pulse is labeled as elapsed time 0. Tb and activity data were recorded every 3 min and are presented as group means (n = 16 at ZT16; n = 8 at ZT22). Tb data were normalized and are expressed in units of SD from daily mean.

View larger version (23K): [in this window] [in a new window]   Fig. 6.   Changes in Tb (A and B, top) and locomotor activity (A and B, bottom) in hamsters during the time from 1 h before to 3 h after a light pulse was applied during the subjective day phase at ZT6 (A) and ZT10 (B). Solid lines represent data collected on days with a light pulse (see Fig. 3), and dotted lines represent data from baseline days without a light pulse (see Fig. 2). Group mean data are presented with n = 8 at both ZTs. See Fig. 4 for more information.

In contrast to the effects on activity, light pulses at ZT16 increased both peak T_b and mean T_b during the light-pulse hour compared with equivalent phases on baseline days (Table 4). The light-induced elevation in T_b persisted so that the hourly mean T_b was also higher during the hour after the light pulse by comparison with the equivalent phase on baseline days (Table 4).

View this table: [in this window] [in a new window]   Table 4.   Statistical description and analysis of results on temperature

Light pulses at ZT22 had effects similar to those at ZT16 in causing a significantly elevated T_b, as reflected in higher peak T_b levels and higher hourly mean T_b during light exposure (Fig. 5 and Table 4). There were also differences, however. The increase in T_b during the light pulse hour at ZT22 appeared to be more robust than at ZT16 (Fig. 5). Furthermore, this increase in T_b was accompanied by a marked increase in activity and a shortened latency to peak activity during the light pulse (Table 3). There were, however, no significant differences between light-pulse and baseline conditions in amount of activity (Fig. 5 and Table 3). Another contrast to the response at ZT16 was that at ZT22 the elevated T_b started to decrease more rapidly before the end of the light pulse and continued to drop during the subsequent hours. Both the T_b nadir and the mean T_b during the hour after the light pulse were therefore lower than the values during baseline (Fig. 5 and Table 4). These differences may be due to a phase advance of the circadian activity and T_b rhythms induced by the late-night light pulse. A phase advance would accelerate the onset of the subjective day phase, characterized by lower activity (as seen in the second half of the light pulse) and T_b levels.

In contrast to the highly significant effects of light exposure during the dark phase on T_b and activity, animals showed few responses to light pulses during the projected light phase at either ZT6 or ZT10 (Fig. 6). Despite the occasional brief increase in activity in response to light onset, there were no statistically significant differences relative to baseline in any measure of either activity or T_b during the light pulse nor during the hour after the light pulse (Fig. 6; Tables 3 and 4).

Relations between responses of T_b and activity to light exposure. In a small animal like a hamster, increased activity levels often lead to intense metabolic heat production and increased T_b, although T_b does not necessarily reflect immediate, ongoing activity but rather reflects the amount of activity integrated over a period of time and after a variable delay. The initial rise in T_b observed immediately after light onset during the dark may have been related to an initial burst of activity at that time. This relation was not observed consistently, however, and temperature could be elevated independently of activity level (Figs. 5 and 6; Tables 3 and 4). For example, activity changes cannot account for the sustained increases in T_b throughout the light pulse and into the following hour at ZT16, because activity was suppressed during most of this period. In addition, at both ZT16 and ZT22, there was no significant increase in total activity during the light pulse relative to baseline, but T_b was still elevated significantly.

Analysis of the linear correlations between activity and T_b, which can quantify the influence of activity changes on T_b changes (28), was applied to extract the effects of light on T_b from any possible masking effects of activity over various periods of time. Comparisons were made between light-pulse and baseline conditions of the intercepts, slopes, correlation coefficients, and contributions of the mean level of activity to the rise in T_b (Table 5). After a light pulse, the observed contributions of activity to T_b did not increase, except in relation to one measure: the length of time that T_b exceeded baseline levels after light at ZT22. The slopes of the linear functions relating T_b to activity did not increase, but the intercept was almost always significantly higher after a light pulse (Table 5). The results suggested that nighttime light induced an elevation in T_b, and this elevation did not depend solely on concurrent increases in activity. The analysis thus confirmed that, while light pulses may stimulate activity transiently, resulting in transient increases in T_b in some cases, most of the T_b changes triggered by light were not reflected in changes in activity.

View this table: [in this window] [in a new window]   Table 5.   Effect of nighttime light exposure on correlations between activity and Tb and the rise in Tb attributable to changes in mean level of locomotor activity

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Phase dependence of acute effects of light. The photically induced hyperthermia observed in this study depended strongly on circadian phase, with significant responses observed only during the night. A diurnal species, the squirrel monkey, Saimiri sciureus, has also been reported to show marked elevation in T_b when exposed to light at night but only slight increases in T_b during the subjective day, presumably resulting from increases in activity (46). Human subjects who remain inactive during bright light exposures also show hyperthermia only during the nighttime (8).

The thermogenic effects of nocturnal light in humans have been attributed to suppression of melatonin, which is secreted spontaneously at night. Exogenous melatonin will lower human T_b during the day and occlude the effects of light at night (12). This explanation is, however, not applicable to hamsters or other nocturnal species in which melatonin and T_b normally rise in parallel during the night. Thus the thermogenic effects of light in hamsters are unlikely to be related to melatonin suppression, unless melatonin has very different effects on thermoregulation in hamsters and humans (see below).

The rhythmic sensitivity to the hyperthermic effects of light may be related in both diurnal and nocturnal species to a role for retinal input to the SCN in mediating these effects (2). It is well known that many aspects of SCN neuronal function are rhythmic, with high sensitivity during the subjective night and relative, or complete, unresponsiveness during the subjective day. This rhythmic sensitivity characterizes circadian rhythm phase shifting (43, 64), photic regulation of immediate-early gene expression in the SCN (57, 58), photic regulation of SCN neuronal firing rates (42), and SCN sensitivity to other stimuli (22).

The circadian phases at which light induced hyperthermia in this study are also phases at which it can induce delay or advance shifts of circadian rhythms and SCN gene expression, whereas those at which it was ineffective are phases at which it does not cause substantial phase shifts or gene expression in rodents. In humans, light pulses capable of shifting the circadian system and elevating T_b can also have immediate effects on other aspects of physiology and behavior (8, 16).

Although it is possible that the SCN are involved in both phase-shift and acute responses to light, it is not clear that these share a common mechanism. Based on studies of human subjects, it has been proposed that acute changes in T_b may be primary events mediating circadian phase-shift responses (18, 19). There is not, however, a great deal of experimental support for this hypothesis, and there is some contrary evidence. Prior administration of melatonin to humans can completely reverse the acute T_b elevation induced by nighttime bright light; in contrast, it may only modify or not alter light-induced phase shifts of human (13, 38) or immediate-early gene expression in the rodent SCN (39). Thus the mechanisms by which light alters physiology acutely and affects circadian phase appear to diverge at some level.

In addition, our results show that, despite the fact that light at both ZT16 and ZT22 increased T_b, there were differences in the patterns of change in activity and T_b at these phases (Fig. 5; Tables 3 and 4). Specifically, hyperthermia lasted longer and activity was more depressed during light exposure at ZT16, whereas activity and temperature in the hour after the light pulse were lower than baseline at ZT22 but not at ZT16. The results at ZT22 might be attributed to the light pulse causing an immediate phase advance of the T_b rhythm. If a rapid advance were induced late in the subjective night, it might accelerate the decline in T_b and activity that normally occurs at this time. An immediate phase shift by light could not, however, account for the central observation of T_b elevation at both nocturnal phases.

The differences between ZT16 and ZT22 imply a divergence of the effects of light on behavioral and physiological regulatory systems at these phases. Differences have also been described in the effects of light on SCN cell immediate-early gene expression early and late in the subjective night (29, 51) and in the ability of neurochemical manipulations to alter photic phase-shifting effects at these phases (50). Thus there does not appear to be a unitary nocturnal response to light, but a variety of responses are evoked differentially at different circadian phases during the night.

Light-induced T_b elevation. Nocturnal light pulses affected both T_b and activity of Syrian hamsters. The onset of nocturnal light pulses triggered a brief activity increase, which could result in an increase in heat production. The idea that increased activity leads directly to elevated T_b is supported by the observation of fluctuations of T_b that typically follow fluctuations in spontaneous activity (Figs. 2-4). After the initial brief increase, however, hamsters usually showed a prolonged reduction in activity during most of the light pulse and beyond, whereas their T_b showed a sustained increase. In addition, a detailed quantitative analysis of the effect of light on the linear regression of T_b on activity indicates that alterations in activity cannot account for most of the light-induced T_b elevation (Table 5). In fact, light pulses did not alter the slopes of T_b vs. activity, although it often significantly increased the intercepts for the linear regression of T_b on activity, suggesting that T_b would be higher during a light pulse even if there were no activity. Thus the sustained hyperthermic response of T_b to light cannot be attributed directly to changes in activity. Because sleep was not recorded in this study, we cannot address the possibility that changes in sleep/waking status might contribute to observed changes in T_b or activity.

The hyperthermic response in humans appears to depend on photic suppression of the normal nocturnal increase in melatonin levels (61). The effect of melatonin on human T_b is likely to be the result of effects on both thermoregulatory centers and on peripheral thermoregulatory processes involved in heat production and/or heat loss (13). High levels of melatonin may inhibit thyroid hormone secretion and corticosterone metabolism (thereby reducing heat production), increase serotonin levels, and enhance peripheral blood flow (thereby increasing heat loss; see Refs. 38, 53, 66, 67). Light-induced blockade of the melatonin rise will have the opposite effect, resulting in an increase in human T_b. One inference that has been drawn from these observations is that a substantial proportion of the normal nocturnal decline in human T_b is related to increasing levels of melatonin at that time (11).

As a nocturnal species, however, hamsters show spontaneous increases in activity and T_b during their subjective night, whereas their melatonin secretion (like that of all species studied) increases at the same time. Despite the fact that both T_b and melatonin normally rise together at night in hamsters, we found that light pulses still further elevated T_b. Whether this response relates to photic suppression of melatonin secretion as in humans (11) is not known. Melatonin has been reported to have complex effects on thermoregulation in nocturnal rodents, decreasing T_b of mice and rats and increasing T_b of Siberian hamsters, Phodopus sungorus sungorus (5, 30, 55, 60). The finding that melatonin decreases T_b in rats does not seem to be consistent with evidence that melatonin causes constriction of the isolated tail artery of rats because the tail is a major thermoregulatory organ for rats (65, 67). The diverse effects of melatonin on T_b of rodents may reflect complexity in its target mechanisms: melatonin activates two different receptor subtypes in vascular smooth muscle that mediate either vascular relaxation or constriction (23).

An alternative interpretation is that photically induced changes in melatonin levels are not related directly to changes in hamster T_b during a nocturnal light pulse. It is not known what other thermogenic mechanisms may have been activated by the light pulse. One possibility is that nocturnal light exposure is stressful to hamsters and leads to sympathetic arousal and thermogenesis associated with sympathetic mechanisms. Nonshivering thermogenesis in brown adipose tissue of anesthetized rats can be induced by electrical stimulation of the retinohypothalamic tract (the direct retinal projection to the SCN), which activates glutamatergic receptors in the SCN, or by chemical stimulation of the posterior hypothalamus (an area involved in regulation of heat production; see Refs. 1 and 3). These observations suggest that photic input to SCN neurons, which project to several hypothalamic nuclei involved in thermoregulation (36), can regulate thermogenesis in rat brown adipose tissue (2). The sympathetic system outflow from the SCN and other nuclei to brown adipose tissue has recently been described in a rodent, P. sungorus sungorus, supporting this interpretation (9). Further experiments are needed to investigate whether the hypothalamic nuclei that control thermogenesis and those that monitor T_b levels are involved in light-induced T_b elevation in Syrian hamsters.

Comparisons between T_br and T_co. Daily patterns of T_br and T_co in hamsters, as well as their responses to photic stimuli, were generally very similar, but T_br was maintained in a narrower range across the day. Similar findings have been reported for T_co and cortical temperature in rats (63). In some circumstances, however, discrepancies between T_co and T_br temperatures have been reported. Thus selective hypothalamic cooling in hamsters has been induced by exercise or by antidepressant drugs (24, 27), and temperature changes have been reported in male rat hypothalamic nuclei during copulations that were only marginally detected in T_co (10).

There were some differences between T_br and T_co of hamsters. T_br was generally higher, as reported in most other species in which hypothalamic temperature was recorded (24, 25, 31, 61); however, depending on position and type of thermosensor, T_br may sometimes be lower than T_co (27). In addition, T_br of our hamsters sometimes showed more rapid, bursting changes in response to light, suggesting a greater responsiveness of T_br than T_co. These differences might, however, also be attributable to differences in the thermal properties of the sensors, because the sensor at the tip of the brain probe was not surrounded by a plastic capsule and paraffin coating, as was the abdominal sensor/transmitter.

In conclusion, light exposure has acute effects on T_br, T_co, and activity of hamsters at only some circadian phases. Hyperthermic responses to light exposure appeared only during the subjective night, coincident with times at which light can also phase shift circadian rhythms and induce gene expression in the SCN pacemaker. Further studies will be needed to assess how photic effects on temperature and circadian rhythms are related to each other and whether melatonin plays a mediating role in these effects.

Perspectives

There has been a long history of interest in the effects of environmental lighting on mammalian physiology and behavior. There are effects of light on humans that do not appear to involve either of the well-characterized mechanisms by which environmental lighting changes alter mammalian physiology: photoperiodic time measurement and circadian rhythm entrainment. These energizing and mood-altering effects have been employed in the treatment of clinical conditions such as seasonal affective disorder. Studies showing that nocturnal light exposure elevates human T_b by depressing melatonin levels provide a potential mechanism for these effects. Our results demonstrate, however, that in nocturnal rodents this same temperature-elevating effect of light is found during the subjective night. For reasons reviewed above, these effects are not likely to be attributable to suppression of melatonin in nocturnal species, although this hypothesis requires an explicit test. It is more likely that light acts to elevate T_co in nocturnal rodents by affecting autonomically regulated heat loss and gain mechanisms. It remains to be established whether similar mechanisms operate in diurnal species and whether the pineal/melatonin system plays different roles in temperature regulation among nocturnal and diurnal mammals.

	ACKNOWLEDGEMENTS

We thank William Lonc, Chris Wright, and Gerhard Körtner for help with the design of the data acquisition system and Donna Goguen, Debbie Fice, and Tanya Myers for assistance with surgical and histological procedures.

	FOOTNOTES

This research was supported by a postdoctoral fellowship to X. Song and a research grant (A0305) to B. Rusak from the Natural Sciences and Engineering Research Council of Canada.

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. §1734 solely to indicate this fact.

Address for reprint requests and other correspondence: B. Rusak, Dept. of Psychology, Dalhousie Univ., Halifax, NS B3H 4J1, Canada (E-mail: rusak{at}is.dal.ca).

Received 17 May 1999; accepted in final form 10 December 1999.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

1.   Amir, S. Retinohypothalamic tract stimulation activates thermogenesis in brown adipose tissue in the rat. Brain Res 503: 163-166, 1989[ISI][Medline].

2.   Amir, S. Blocking NMDA receptors or nitric oxide production disrupts light transmission to the suprachiasmatic nucleus. Brain Res 586: 336-339, 1992[ISI][Medline].

3.   Amir, S, Shizgal P, and Rompré P-P. Glutamate injection into the suprachiasmatic nucleus stimulates brown fat thermogenesis in the rat. Brain Res 498: 140-144, 1989[ISI][Medline].

4.   Amstrong, SM. Melatonin and circadian control in mammals. Experimentia 45: 932-938, 1989[ISI][Medline].

5.   Aruntyunyan, GS, Mashkovskii MD, and Roshchina LF. Pharmacological properties of melatonin. Fed Proc 23: 1330-1332, 1964.

6.   Aschoff, J. Exogenous and endogenous components in circadian rhythms. Cold Spring Harbor Symp Quant Biol 25: 11-28, 1960[ISI][Medline].

7.   Aschoff, J, and Gerkema M. On diversity and uniformity of ultradian rhythms. Exp Brain Res Suppl 12: 321-334, 1985.

8.   Badia, P, Myers B, Boecker M, and Culpepper J. Bright light effects on body temperature, alertness, EEG and behavior. Physiol Behav 50: 583-588, 1991[Medline].

9.   Bamshad, M, Song CK, and Bartness TJ. CSN origins of the sympathetic nervous system outflow to brown adipose tissue. Am J Physiol Regulatory Integrative Comp Physiol 276: R1569-R1578, 1999[Abstract/Free Full Text].

10.   Blumberg, MS, Mennella JA, and Moltz H. Hypothalamic temperature and deep body temperature during copulation in the male rat. Physiol Behav 39: 367-370, 1987[Medline].

11.   Cagnacci, A. Influences of melatonin on human circadian rhythms. Chronobiol Int 14: 205-220, 1997[ISI][Medline].

12.   Cagnacci, A, Elliott JA, and Yen SSC Melatonin: a major regulator of the circadian rhythm of core temperature in humans. J Clin Endocrinol Metab 75: 447-452, 1992[Abstract].

13.   Cagnacci, A, Soldani R, and Yen SSC Contemporaneous melatonin administration modifies the circadian response to nocturnal bright light stimuli. Am J Physiol Regulatory Integrative Comp Physiol 272: R482-R486, 1997[Abstract/Free Full Text].

14.   Cajochen, C, Kräuchi K, Danilenko KV, and Wirz-Justice A. Evening administration of melatonin and bright light: interactions on the EEG during sleep and wakefulness. J Sleep Res 7: 145-157, 1998[ISI][Medline].

15.   Campbell, SS, Terman M, Lewy AJ, Dijk D-J, Eastman CI, and Boulos Z. Light treatment for sleep disorders: consensus report. V. Age-related disturbances. J Biol Rhythms 10: 151-154, 1995[Abstract/Free Full Text].

16.   Czeisler, CA, Kronauer RE, Allan JS, Duffy JF, Jewett ME, Brown EN, and Ronda JM. Bright light induction of strong (type 0) resetting of the human circadian pacemaker. Science 244: 1328-1332, 1989[Abstract/Free Full Text].

17.   Daan, S, and Aschoff J. Short-term rhythms in activity. In: Handbook of Behavioral Neurobiology, edited by Aschoff J.. New York: Plenum, 1981, p. 491-498.

18.   Deacon, S, and Arendt J. Melatonin-induced temperature suppression and its acute phase-shifting effects correlated in a dose-dependent manner in humans. Brain Res 688: 77-85, 1995[ISI][Medline].

19.   Deacon, S, English J, and Arendt J. Acute phase-shifting effects of melatonin associated with suppression of core temperature in humans. Neurosci Lett 178: 32-34, 1994[ISI][Medline].

20.   Deboer, T, and Tobler I. Chronic administration of melatonin reduces REM sleep in the Djungarian hamster (Phodopus sungorus). Neurosci Lett 231: 118-122, 1997[ISI][Medline].

21.   Deprés-Brummer, P, Lévi F, Metzger G, and Touitou Y. Light-induced suppression of the rat circadian system. Am J Physiol Regulatory Integrative Comp Physiol 268: R1111-R1116, 1995[Abstract/Free Full Text].

22.   Derambure, PS, and Boulant JA. Circadian thermosensitive characteristics of suprachiasmatic neurons in vitro. Am J Physiol Regulatory Integrative Comp Physiol 266: R1876-R1884, 1994[Abstract/Free Full Text].

23.   Doolen, S, Krause DN, Dubocovich ML, and Duckles SP. Melatonin mediates two responses in vascular smooth muscle. Eur J Pharmacol 345: 67-69, 1998[ISI][Medline].

24.   Duncan, WC, Johnson KA, and Wehr TA. Antidepressant drug-induced hypothalamic cooling in Syrian hamsters. Neuropsychopharmacology 12: 17-37, 1995[ISI][Medline].

25.   Fuller, CA. Circadian brain and body temperature rhythms in the squirrel monkey. Am J Physiol Regulatory Integrative Comp Physiol 246: R242-R246, 1984.

26.   Goldman, BD, Hall V, Hollister C, Yellon SM, and Tamarkin L. Diurnal changes in pineal melatonin content in four rodent species: relationship to photoperiodism. Biol Reprod 24: 778-783, 1981[Abstract].

27.   Gordon, CJ, Rezvani AH, Fruin ME, Trautwein S, and Heath JE. Rapid brain cooling in the free-running hamster Mesocricetus auratus. J Appl Physiol 51: 1349-1354, 1981[Abstract/Free Full Text].

28.   Gordon, CJ, and Yang Y. Contribution of spontaneous motor activity to the 24 hour control of body temperature in male and female rats. J Therm Biol 22: 59-68, 1997.

29.   Guido, ME, Rusak B, and Robertson HA. Spontaneous circadian and light-induced expression of junB mRNA in the hamster suprachiasmatic nucleus. Brain Res 732: 215-222, 1996[ISI][Medline].

30.   Haim, A, and Fourie FR. Effects of melatonin on heat production and enzymatic activity in diurnal and in nocturnal rodents. Comp Biochem. Physiol. 71A: 473-475, 1982.

31.   Hayward, JN, and Baker AN. A comparative study of the role of the cerebral arteriole blood in the regulation of brain temperature in five mammals. Brain Res 16: 417-440, 1969[ISI][Medline].

32.   Heldmaier, G, and Lynch R. Pineal involvement in thermoregulation and acclimatization. Pineal Res Rev 4: 97-139, 1986.

33.   Huber, R, Deboer T, Schwierin B, and Tobler I. Effect of melatonin on sleep and brain temperature in the Djungarian hamster and the rat. Physiol Behav 65: 77-82, 1998[Medline].

34.   Illnerová, H. Entrainment of the circadian rhythms in melatonin production by light. Pineal Res Rev 6: 173-177, 1988.

35.   Illnerová, H, and Vanécek J. Entrainment of the circadian rhythm in the rat pineal N-acetyltransferase activity by prolonged periods of light. J Comp Physiol [A] 161: 495-510, 1987[Medline].

36.   Kita, M, Shibata S, Oomura Y, and Ohki K. Excitatory effects of the suprachiasmatic nucleus on the ventromedial nucleus in the rat hypothalamic slice. Brain Res 235: 137-141, 1982[ISI][Medline].

37.   Körtner, G, and Geiser F. Body temperature rhythm and activity in reproductive Antechinus (Marsupialia). Physiol Behav 58: 31-36, 1995[Medline].

38.   Kräuchi, K, Cajochen C, Danilenko KV, and Wirz-Justice A. The hypothermic effect of late evening melatonin does not block the phase delay induced by concurrent bright light in human subjects. Neurosci Lett 232: 57-61, 1997[ISI][Medline].

39.   Kumar, V, Goguen DM, Guido ME, and Rusak B. Melatonin does not influence the expression of c-fos in the suprachiasmatic nucleus of rats and hamsters. Mol Brain Res 52: 242-248, 1997[Medline].

40.   Lewy, AJ, Wehr TA, and Goodwin FK. Light suppresses melatonin secretion in humans. Science 210: 1267-1269, 1980[Abstract/Free Full Text].

41.   Meijer, JH, and De Vries MJ. Light-induced phase shifts in onset and offset of running-wheel activity in the Syrian hamsters. J Biol Rhythms 10: 4-16, 1995[Abstract/Free Full Text].

42.   Meijer, JH, Watanabe K, Détàri L, and Schaap J. Circadian rhythms in light response in suprachiasmatic nucleus neurons of freely moving rats. Brain Res 741: 352-355, 1996[ISI][Medline].

43.   Minors, DS, Waterhouse JM, and Wirz-Justice A. A human phase-response curve to light. Neurosci Lett 133: 36-40, 1991[ISI][Medline].

44.   Moore, RY, and Card JP. Visual pathways and the entrainment of circadian rhythms. Ann NY Acad Sci 453: 123-133, 1985[ISI][Medline].

45.   Moore, RY, and Klein DC. Visual pathways and the central neural control of a circadian rhythm in pineal serotonin N-acetyltransferase activity. Brain Res 71: 17-33, 1974[ISI][Medline].

46.   Moore-Ede, M, Sulzman FM, and Fuller CA. Circadian timing of physiological systems. In: The Clocks That Time Us, edited by Moore-Ede MC, Sulzman FM, and Fuller C.. Cambridge, MA: Harvard Univ. Press, 1982, p. 201-294.

47.   Morin, LP. The circadian visual system. Brain Res Rev 19: 102-127, 1994[Medline].

48.   Pittendrigh, CS. Circadian surfaces and the diversity of possible roles of circadian organization in photoperiodic induction. Proc Natl Acad Sci USA 69: 2734-2737, 1972[Abstract/Free Full Text].

49.   Pittendrigh, CS, and Daan S. A functional analysis of circadian pacemakers in nocturnal rodents. I. The stability and liability of spontaneous frequency. J Comp Physiol [A] 106: 223-252, 1976.

50.   Ralph, MR, and Menaker M. Bicuculline blocks circadian phase delays but not advances. Brain Res 325: 362-365, 1985[ISI][Medline].

51.   Rea, MA. Different populations of cells in the suprachiasmatic nuclei express c-fos in association with light-induced phase delays and advances of the free-running activity rhythm in hamsters. Brain Res 579: 107-112, 1992[ISI][Medline].

52.   Refinetti, R, and Menaker M. The circadian rhythms of body temperature. Physiol Behav 51: 613-637, 1992[Medline].

53.   Reiter, RJ. Pineal melatonin: cell biology of its synthesis and of its physiological interactions. Endocr Rev 12: 151-180, 1991[ISI][Medline].

54.   Rowland, N. Metabolic fuel homeostasis in Syrian hamsters: nycthermal and exercise variables. Physiol Behav 33: 243-252, 1984[Medline].

55.   Ruby, NF, Kang T, and Heller HC. Melatonin attenuates photic disruption of circadian rhythms in Siberian hamsters. Am J Physiol Regulatory Integrative Comp Physiol 273: R1540-R1549, 1997[Abstract/Free Full Text].

56.   Rusak, B. The role of the suprachiasmatic nuclei in the generation of circadian rhythms in the golden hamster, Mesocricetus auratus. J Comp Physiol [A] 118: 145-164, 1977.

57.   Rusak, B, McNaughton L, Robertson HA, and Hunt SP. Circadian variation in photic regulation of immediate-early gene mRNA in rat suprachiasmatic nucleus cells. Mol Brain Res 14: 124-130, 1992[Medline].

58.   Rusak, B, Robertson HA, Wisden W, and Hunt SP. Light pulses that shift rhythms induce gene expression in the suprachiasmatic nucleus. Science 248: 1237-1240, 1990[Abstract/Free Full Text].

59.   Rusak, B, and Zucker I. Neural regulation of circadian rhythms. Physiol Rev 59: 449-526, 1979[Free Full Text].

60.   Saarela, S, and Reiter R. Function of melatonin in thermoregulatory processes. Life Sci 54: 295-311, 1994[ISI][Medline].

61.   Shanahan, TL, and Czeisler CA. Light exposure induces equivalent phase shifts of the endogenous circadian rhythms of circulating plasma melatonin and core body temperature in men. J Clin Endocrinol Metab 73: 227-235, 1991[Abstract].

62.   Strassman, RJ, Qualls CR, Lisansky EJ, and Peake GT. Elevated rectal temperature produced by all-night bright light is reversed by melatonin infusion in men. J Appl Physiol 71: 2178-2182, 1991[Abstract/Free Full Text].

63.   Sundgren-Andersson, AK, Östlund P, and Bartfai T. Simultaneous measurement of brain and core temperature in the rat during fever, hyperthermia, hypothermia and sleep. Neuroimmunomodulation 5: 241-247, 1998[ISI][Medline].

64.   Takahashi, JS, DeCoursey PJ, Bauman L, and Menaker M. Spectral sensitivity of a novel photoreceptive system mediating entrainment of mammalian circadian rhythms. Nature 308: 186-188, 1984[Medline].

65.   Ting, KN, Davies DJ, Sugden D, Delagrange F, Guardiola-Lemaître B, and Scalbert E. Studies on the vasoconstrictor action of melatonin and putative melatonin receptor ligands in the tail artery of juvenile Wistar rats. Br J Pharmacol 122: 1299-1306, 1997[ISI][Medline].

66.   Viswanathan, M, Hissa R, and George JC. Suppression of sympathetic nervous system by short photoperiod and melatonin in the Syrian hamster. Life Sci 38: 73-79, 1986[ISI][Medline].

67.   Viswanathan, M, Laitinen JT, and Saavedra JM. Expression of melatonin receptors in arteries involved in thermoregulation. Proc Natl Acad Sci USA 87: 6200-6203, 1990[Abstract/Free Full Text].

68.   Zwillinger, D. CRC Standard Mathematical Tables and Formulae (30^th ed), edited by Zwillinger D.. London: CRC, 1996.