INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

PHYSIOLOGICAL PROCESSES such as sleep and activity, core body temperature, and hormonal levels oscillate with predictable rhythmicity to coincide with the solar day. Those oscillations that persist in constant conditions are termed circadian rhythms and are under the primary control of a biological time keeper located in the suprachiasmatic nuclei (SCN) of the anterior hypothalamus. The SCN maintains a self-generated neural output with a period of ~24 h, which is entrained to the external environment by light input at the retina, allowing adjustments in cycle length according to changes in photoperiod.

The timing of output from the SCN, and thus the rhythms under its control, is governed by the cycle of several newly described clock genes (26), and the timing of circadian rhythms can be altered by light stimuli. The major SCN afferents from the retina are the direct monosynaptic retino-hypothalamic tract (19) and an indirect retino-geniculate hypothalamic tract (8) from the retina via the ventral lateral geniculate nucleus/intergeniculate leaflet (IGL) to the SCN. A third prominent projection from the raphe nucleus to the SCN has recently been suggested as an SCN afferent in the rat after studies indicating the existence of a retino-raphe projection (27). These major afferents and their dominant transmitters (retino-hypothalamic tract excitatory amino acids, retino-geniculate hypothalamic tract GABA and neuropeptide Y, and retino-raphe-hypothalamic tract serotonin) are believed to be the major mediators of circadian function in mammals. However, several other neurotransmitters have been identified in the SCN and are possibly important in the control of circadian timing.

ACh has been identified in the SCN, as have both muscarinic and nicotinic cholinergic receptors (30, 32, 33), together with its synthetic enzyme choline acetyltransferase (4). ACh levels were reported to rise threefold in the rat SCN after a light pulse presented at night (21), implicating a role for ACh in light-mediated effects on SCN rhythmicity. Thus the hypothesis that ACh is involved in the control of circadian timing at some level is not new, although the specific mechanisms of action within the system are not completely understood.

A series of studies using a line of rat bred for cholinergic hypersensitivity [Flinders Sensitive Line (FSL)] and reported as an animal model of depression showed that animals with upregulated brain cholinergic systems displayed altered circadian timing under entrained and free-running conditions (28, 29). The line was not developed specifically for analysis of circadian function and, as such, was not maintained in controlled light/dark cycles. It was the aim of the current project to develop an independent line of rat under controlled lighting conditions for the specific purpose of examining the role of cholinergic neurotransmission in circadian timing and light effects on the SCN. This study reports the emergence of altered periodicity in these rats and the response to a light pulse at night.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The breeding program was partly based on that of Overstreet et al. (24), beginning with the offspring of 10 randomly bred Wistar albino rats from the Central Animal House (Univ. of Adelaide). Animals were selected solely on the basis of the thermic response to a muscarinic agonist oxotremorine, as a measure of cholinergic sensitivity. At 28 days of age the offspring were weighed, and a baseline temperature was recorded using a rectal thermistor inserted 5-6 cm. On the following day each animal received a 0.25 µmol/kg injection of oxotremorine sesquifumurate (Research Biochemicals International, Natick, MA), and body temperature was recorded 40 min later, corresponding to the time of the baseline reading. The six males with the greatest degree of hypothermia were mated with the six females showing the largest response to form the basis for the sensitive-to-oxotremorine (S_ox) line. Similarly, the six males having the lowest degree of hypothermia or the most hyperthermia were mated with the corresponding six females to form the resistant-to-oxotremorine (R_ox) line. Brother-to-sister matings were avoided in the pairing of breeders. In subsequent generations (denoted by G, followed by number) all offspring were phenotyped in the same manner at 28 days of age, and S_ox breeders were selected only from S_ox offspring and R_ox breeders from R_ox offspring. Seven breeding pairs were used for each line from the third selection to ensure an adequate number of offspring for experiments and to decrease the possibility of inbreeding. All animals were maintained in 12:12-h light-dark cycle (lights off at 1900).

Timing Studies

Core body temperature rhythmicity in an entrained photoperiod was examined to compare the phase-angle difference in a separate SCN-driven rhythm in males selected from G11. At 35 days of age, five males of each line were implanted with temperature transmitters (model TA10TA-F20, Data Sciences International, St. Paul, MN) under 3% halothane-oxygen anesthesia and placed in individual home cages in a light-controlled environment chamber in a 12:12-h light-dark photoperiod (lights off 1900). Temperature was recorded automatically for 10 s every 10 min continuously under the same light/dark cycle for 5 consecutive days (after a 2-day recovery period) using Dataquest IV LabPro Software (Data Sciences International) (13). The data for 5 days were pooled for each animal, and the average profiles for five animals in each group were combined to produce a single temperature profile for each line. The marker used to measure the phase of the temperature rhythm was the morning decline, defined as the time at which the temperature dropped below the daily average for each animal. The average time of the temperature offset was determined for each animal and differences between groups analyzed by t-test.

Constant condition. It is not practical to conduct long-term melatonin studies using urinary 6-SMT as a marker due to daily feeding requirements. Thus the rhythm of core body temperature was studied under constant dark conditions. Five S_ox and five R_ox males from G12 were implanted with transmitters and after 3 days (in light/dark conditions) were monitored for 21 days in constant dark conditions. Feeding, watering, and cage maintenance were carried out in complete darkness every 4-5 days using infrared vision equipment. Temperature data from the two lines were analyzed using the TAU computer package (Mini-Mitter, Sunriver, OR). The differences in period length (as determined for each animal using the TAU package) were analyzed using t-tests ( = 0.05).

Light Pulse Studies

The two experiments were 1) five S_ox and five R_ox males from G9 were exposed to a 1-min, 2-lx light pulse at circadian time (CT) 16 (4 h after subjective lights off) on the treatment night and 2) five S_ox and five R_ox males from G10 were exposed to a 1-min, 2-lx light pulse at CT18 (6 h after subjective lights off) on the treatment night.

c-Fos. Two experiments were conducted to determine the induction of the immediate early gene c-fos. Animals were housed 4 or 5 to a cage in 12:12-h light-dark lighting conditions before the experiment. 1) Twenty-three S_ox and twenty-three R_ox males from G8 and G9 and five normal males were exposed to a 1-min, 2-lx light pulse 4 h after lights off [zeitgeber time (ZT) 16] and 10 S_ox and R_ox and 5 normal animals were left untreated. 2) Twelve S_ox and twelve R_ox from G10 and G11 and five normal males were exposed to a 1-min, 2-lx light pulse 6 h after lights off (ZT18), and five of each group were left untreated.

Two hours after treatment, animals were decapitated and brains were rapidly removed and fixed by immersion in 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4). Four 70-µm coronal sections encompassing the SCN were cut from each brain on a Vibroslice microtome and processed for c-Fos protein using a protocol reported previously (20). Results are expressed as the number of immunopositive cells per animal (calculated by averaging the number of positive cells in the left and right SCN). The rectangular counting frame was 480 × 330 µm with the long side tangential to the ventral indenture of the SCN into the optic chiasm and the shorter side of the rectangle in line with the third ventricle. Data were analyzed by Kruskal-Wallis nonparametric ANOVA and Mann-Whitney U tests post hoc.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The thermic responses to oxotremorine administration in the founder population were normally distributed, ranging from a hypothermic response of 2.6°C to a hyperthermic response of 1.1°C, with an average drop of 1.0°C. In subsequent generations, S_ox animals exhibited an increased hypothermic response, whereas the response of the R_ox population remained relatively stable. The difference in thermic response between S_ox and R_ox breeders (i.e., the most extreme responders) was significant at the first selection and remained so up to G11 (Fig. 1A). The thermic response of all offspring of a single generation was significantly different between S_ox and R_ox at G4 and remained so to G11 (Fig. 1B).

View larger version (23K): [in this window] [in a new window]   Fig. 1.   Temperature change from baseline in 28-day-old animals subjected to cholinergic challenge. A: data from animals selected as breeders (n = 6 or 7). B: data from all offspring for that generation (G; n = 45-85). Generation number (G) is on x-axis, and change in core body temperature from a baseline recording is on y-axis. Open symbols represent animals sensitive to oxotremorine (Sox); closed symbols represent animals resistant to oxotremorine (Rox). Data points represent means ± SD for each group.

Timing Studies

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Onset time of melatonin production for Sox (open symbols) and Rox (closed symbols) males determined by 6-sulphatoxymelatonin (6-SMT) excretion rate. Abscissa shows generation number of the animals used. Horizontal line indicates onset time for normal outbred animals in our laboratory [zeitgeber time (ZT) 16.1]. Onset time was defined as time at which 6-SMT excretion rate rose to >20 pmol/h. Light cycle was 12:12-h light to dark with lights off at 1900 (ZT12). All points are means ± SE for 5-10 animals.

View larger version (29K): [in this window] [in a new window]   Fig. 3.   A: 6-SMT excretion profiles in 12:12-h light-to-dark cycle for Sox (n = 20, open symbols) and Rox (n = 20, closed symbols) males selected from G8 and G9 and normal outbred males (n = 20, cross symbols). Collections ran from 1800 to 0900 with lights off at 1900 (ZT12). Data points represent means ± SE. Where SE bars are not visible, they are obscured by symbol. B: average 24-h core body temperature profiles for Sox (n = 5, thin line) and Rox (n = 5, bold line) males from G11 and normal (n = 10, dashed line) male rats. Recordings were taken over 5 consecutive days, and an average profile was produced for each animal. Group average was then calculated for each time point. Profiles represented are 3-point, moving average 24-h profiles for each group. Light cycle as in A.

Constant dark conditions. S_ox animals from G12 displayed a significantly shorter free-running period of the core body temperature rhythm than the R_ox animals (23.9 ± 0.04 and 24.3 ± 0.1 h, respectively). Figure 4 shows representative actograms of the core body temperature of S_ox and R_ox animals.

View larger version (55K): [in this window] [in a new window]   Fig. 4.   Representative actograms of temperature rhythms of Sox (top) and Rox (bottom) animals selected from G12. Animals were maintained in constant darkness throughout 21-day period. X-axis represents a double-plotted 48-h period. Y-axis represents consecutive days.

Light Pulse Studies

View larger version (30K): [in this window] [in a new window]   Fig. 5.   6-SMT excretion profiles for G10 Sox and Rox animals treated with a 1-min, 2-lx light pulse. Four consecutive nights are shown. In each panel, data from night 1 (solid symbols) overlays data from nights 2, 3 and 4 (open symbols). A: contains data from Sox animals treated with a pulse at circadian time (CT) 16; B: data from Rox animals treated at CT16; C: Sox animals treated with a pulse at CT18; and D: Rox animals treated at CT18. Data sets represent means ± SE for 5 animals.

The light pulse at CT18 caused a similar degree of suppression of melatonin production on the night of treatment in both groups (Figs. 5, C and D). A significant phase delay in the timing of onset of 6-SMT excretion rate on nights 3 and 4 was also seen in both groups compared with night 1 (Table 1). There was no difference in response times between the groups and the large phase delay recorded on night 3 in the R_ox group is a result of the reduced excretion from some animals. The altered excretion pattern of 6-SMT as a response to a light pulse on the first posttreatment night is unique to the R_ox line, and the excretion profile returned to normal on night 4, giving a more accurate measure of the phase delay.

c-Fos. A light pulse at ZT16 resulted in the appearance of c-Fos-positive cells in the SCN region of both S_ox and R_ox animals (Fig. 6A); however, there were significantly less cells recorded in the S_ox group. Treatment at ZT18 resulted in a similar number of cells being labeled in S_ox, R_ox, and in normal animals (Fig. 6B).

View larger version (19K): [in this window] [in a new window]   Fig. 6.   Number of immunopositive cells in suprachiasmatic nuclei of rats after exposure to a light pulse (1 min, 2 lx) at either ZT16 (A) or ZT18 (B). Treatment groups on x-axis are normal outbred (NL) animals exposed to light. ND, normal animals left untreated; SL, Sox animals exposed to light; SD, Sox animals left untreated; RL, Rox animals exposed to light; RD, Rox animals left untreated. Data sets represent means ± SE for 5-23 animals.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We have developed a line of rat with genetically inherited cholinergic supersensitivity. S_ox rats displayed an early onset of the melatonin production rhythm compared with the codeveloped R_ox line from the third generation, in addition to an earlier offset of the temperature rhythm in G11 animals under a light-dark cycle. Furthermore, S_ox animals maintained under constant dark conditions showed a shorter free-running period of the temperature rhythm compared with both R_ox and normal outbred animals from our laboratory (13, 25). The response to a light pulse applied 4 h after subjective dark onset was also reduced in S_ox animals as measured by melatonin production and c-Fos expression in the SCN. Thus the decreased phase-angle difference of the melatonin and temperature rhythms, shorter free-running period of the temperature rhythm, and altered sensitivity to light in S_ox animals may be related to the inherited cholinergic supersensitivity affecting SCN rhythmicity. The S_ox line may provide a unique model for further analysis of the neural control of circadian timing.

The current project was developed on the basis of results from circadian studies in the cholinergically supersensitive FSL (24). Throughout the development of the FSL, the animals were maintained in a constant-light photoperiod (24). Under such conditions, the circadian rhythmicity of melatonin production in the rat is abolished (9), and other SCN generated cycles can also become arrhythmic (3, 22). It is possible, therefore, that the altered timing of circadian cycles in the FSL animals was not solely or directly related to the increased cholinergic sensitivity. Consequently, the S_ox and R_ox animals were maintained in a 12:12-h light-dark photoperiod from the outset of the program to ensure environmental lighting conditions were conducive to proper SCN functioning and identical to those of the randomly bred colony. In addition to the current unavailability of the FSL animals in Australia, another important factor in the establishment of an independent line was our unique ability to study the evolution of any rhythm differences noninvasively during the selection period using the rhythm of melatonin production.

Melatonin onset was used as a marker of SCN function in the S_ox and R_ox lines for several reasons. The rhythm of melatonin production is not masked by activity, temperature, or stress in rats. It is a quantifiable measure, highly reproducible within and between animals, and our unique urine collection system provides reliable data using very few animals. The use of the melatonin rhythm in a cyclic photoperiod allowed the analysis of the circadian timing of breeders from each generation (i.e., the extreme responders from each line) with minimal stress to the animals and in much less time than temperature or running activity studies would have allowed. The temperature profiles of G11 animals in 5 days of a light/dark cycle confirmed an altered phase-angle difference in the S_ox animals. It is interesting to note that the melatonin profile of R_ox animals selected from G8 and G9 was also different from normal animals and that the larger phase-angle difference in the R_ox occurred concurrently with a longer free-running period of the temperature rhythm. The R_ox animals may be an equally useful tool in this area of research.

The FSL provided intriguing evidence for a correlation between cholinergic supersensitivity and altered timing of circadian rhythms in both entrained and constant conditions (28, 29), and the S_ox animals have displayed similar altered circadian timing. Aschoff (2) first associated the phenomenon of a short (or negative) phase-angle difference occurring concurrently with a shorter free-running period of activity in the same animal, suggesting altered timing of SCN rhythms under both cyclic and constant conditions. Our laboratory has previously reported free-running locomotor activity rhythms in Wistar rats from our colony having a period of 24.24 h (25) and temperature rhythms of 24.2-24.3 h (13), significantly longer than the period recorded in the S_ox animals in the current study. At present, the sole measure of the free-running period of the circadian timing system in the S_ox and R_ox animals is the core body temperature rhythm. Long-term melatonin studies are difficult to conduct using our system due to the risk of daily cage maintenance and feeding affecting the rhythms. Nevertheless, we have no reason to suspect that the period of the melatonin rhythm would not also be different between the lines. It is possible that the altered cholinergic state of the S_ox animals affected the temperature regulation system alone rather than the circadian timing system. This appears unlikely however, because our data show an altered period of the temperature rhythm of the S_ox rats and not merely a change in amplitude or waveform of the rhythm. Furthermore, a similar phase-angle difference of both temperature and melatonin rhythms in light/dark conditions suggests that the timing of both rhythms was altered and that the SCN, which controls the timing of both rhythms, has been affected.

The circadian timing system of S_ox rats also responded differently to a light pulse 4 h after (subjective) dark as determined by melatonin rhythmicity and c-Fos expression. The results indicate that both S_ox and R_ox rats incurred phase changes similar to normal outbred animals after treatment at CT18, however after a CT16 pulse, the S_ox rats showed both a reduced number of c-Fos-positive cells in the SCN and reduced suppression of melatonin production acutely, as well as a smaller phase shift in the melatonin production rhythm on the posttreatment nights. We are confident that the changes in melatonin onset on the two nights posttreatment represent true phase shifts in the rhythm. It has previously been argued that transient shifts in laboratory animals generally are in the same direction as the resultant steady-state phase shifts (5, 6). Furthermore, Honma et al. (10) showed that activity onset on the first posttreatment cycle was similar to that obtained after several cycles, especially in the delay portion of the phase response curve (i.e., the first half of the night). Similarly, Illnerova and Vanecek (11) did not observe transients in the N-acetyltransferase rhythms of rats after a light pulse administered in the first one-half of the night. A separate study further validated the use of immediate responses by showing that phase delays were 0.3 h less when analyzed on the first posttreatment night compared with the "steady state," indicating that immediate shifts (especially delays) possibly underestimated the magnitude of the response but not the direction (18). In addition to previous data published using this protocol in our laboratory (7, 15, 25), these reports support the use of two posttreatment nights, and we therefore believe the data gained from the current light-pulse studies to be physiologically significant.

Daan and Pittendrigh (6) first showed that the SCN of animals responds differently to stimuli according to the phase of the circadian cycle it is applied. With only two time points studied thus far it is difficult to explain the different responses to light at night between the lines. One possibility is that the S_ox animals may exhibit a different phase-response curve to light compared with R_ox and normal animals. The largest light-induced phase shifts in melatonin rhythmicity in rats were reported at 4, 6, or 8 h after lights off, whereas a pulse at 2 or 10 h after lights off failed to alter the rhythm (14). However, on the basis of the shorter period of melatonin and temperature rhythms in S_ox rats, one might hypothesize an increased sensitivity to light early in the evening, in contrast to the current results. Thus this altered sensitivity to light at CT16 remains unexplained, yet very intriguing. Examination of a full phase-response curve of S_ox and R_ox rats to pulses will shed further "light" on this subject.

There may be differences between the S_ox and R_ox lines that are not solely related to altered cholinergic sensitivity. The breeding program selected for an upregulated brain cholinergic system rather than a mutation in a single gene that is related to cholinergic function. Interestingly, the FSL animals were reported to be more sensitive to serotonergic stimulation, which was hypothesized to be the "consequence of a primary change in the cholinergic system during the breeding program" (23). This may also be true for the S_ox animals as research in our laboratory suggests that the serotonergic projection to the SCN via the raphe nucleus is important in light-mediated phase changes in the first part of the night (13), and the S_ox animals appear to be less sensitive to light at that time. Further manipulation of the circadian timing system of these animals with both cholinergic and serotonergic agents will be important in addressing these questions.

The cholinergic supersensitivity of S_ox animals is most likely a result of global differences in cholinergic mechanisms throughout the brain. Discreet brain nuclei outside the SCN have a role in the modulation of light effects on SCN timing (IGL, raphe nucleus); however, the intrinsic period of the clock is primarily governed by the length of the cycle of translation/transcription of genetic material within the SCN (26). Regions such as the brain stem and basal forebrain extend cholinergic projections to the SCN (16, 31), which could account for an altered level of cholinergic stimulus at the SCN. We hypothesize that in both S_ox and R_ox animals the normal level of cholinergic stimulus impinging on the SCN is altered, possibly resulting in altered endogenous period in the SCN.

In conclusion, we have developed a line of rat that displays a heightened sensitivity to the cholinergic agonist oxotremorine. The current data confirm the early reports from the Flinders studies with regard to thermic sensitivity (24) and expand the observations of altered SCN rhythmicity in the FSL rats (28, 29). The S_ox animals display a decreased phase-angle difference of both the melatonin and temperature rhythms in an entrained photoperiod, whereas the R_ox animals have a larger phase-angle difference. The S_ox animals also exhibit lower sensitivity of the circadian timing system to light at CT16. Most importantly, the S_ox animals have a shorter free-running period of the temperature rhythm under constant dark conditions. Thus the extremes of cholinergic sensitivity are correlated with altered timing of SCN-controlled circadian rhythms, suggesting an important role for ACh in the circadian timing system.

Perspectives

ACh is not considered a major transmitter of the circadian timing system; however, this study demonstrates that the cholinergic system is important in biological time keeping at some level. Detailed analyses of the neural mechanisms involved in the timing of biological rhythms and the mediation of rhythmicity by light will ultimately allow pharmacological manipulation of the SCN. Shift workers, transcontinental travellers, and, most importantly, sufferers of sleep disorders would benefit greatly from treatments that address the underlying circadian state, rather than symptoms only. The S_ox and R_ox lines will be useful in future study of the role of both the cholinergic and serotonergic neurotransmitter systems in SCN function. Furthermore, the S_ox line provides a unique opportunity to examine the possibility of a genetic predisposition to disorders of sleep timing. A circadian timing system that does not respond appropriately to light stimuli (including the entraining effects of morning light) may predispose an individual to certain insomnias. There is also a large literature regarding the high prevalence of circadian rhythm disorders in psychiatric illnesses such as major depression, schizophrenia, and Alzheimers disease. As the cholinergic and serotonergic systems are reported to be involved in these illnesses, the S_ox animals also provide a unique model for research in this area.