Transient entrainment of a circadian pacemaker during development by dopaminergic activation in Syrian hamsters

Transcription

1 Brain Research Bulletin, Vol. 48, No. 2, pp , 1999 Copyright 1999 Elsevier Science Inc. Printed in the USA. All rights reserved /99/$ see front matter PII S (98) Transient entrainment of a circadian pacemaker during development by dopaminergic activation in Syrian hamsters Julia Grosse and Fred C. Davis* Department of Biology, Northeastern University, Boston, MA, USA [Received 14 September 1998; Revised 28 October 1998; Accepted 4 November 1998] ABSTRACT: Maternal cues entrain a circadian pacemaker in fetal Syrian hamsters. These cues may act through dopaminergic activation of the fetal suprachiasmatic nucleus (SCN); injection of the dopamine D1 agonist SKF38393 to pregnant hamsters entrains activity rhythms of their pups and induces expression of c-fos in the fetal SCN. The aim of this study was to examine the ability of SKF38393 to entrain neonatal Syrian hamsters and to determine the age at which this effect is lost. SKF38393 injections given to two groups of pups at opposite times of day on postnatal days (PN) 1 5 entrained the pups activity rhythms to average phases that differed by 9.25 h. SKF38393 failed to establish different average phases when given on PN Injection of SKF38393 on PN 1, but not PN 6, induced expression of Fos. These results demonstrate that dopaminergic activation is a potent entraining stimulus in neonatal hamsters and that its entraining effects, as well as its ability to induce Fos, are lost by PN 6. The phase established by dopaminergic activation was approximately opposite to that previously shown to be established by melatonin injections. Dopaminergic activation and melatonin may mimic separate but complementary maternal entraining signals which represent day and night Elsevier Science Inc. KEY WORDS: Rhythm, Suprachiasmatic nucleus, SKF38393, c-fos, Fos, Melatonin, Dopamine. INTRODUCTION Mammalian circadian rhythms are regulated by a circadian pacemaker within the suprachiasmatic nucleus (SCN) of the hypothalamus. In adults, these rhythms are entrained to the environment primarily by the light-dark cycle. In fetuses, however, the pacemaker is functional before retinal innervation of the SCN [27,32] and is entrained by maternal rhythms [6]. This finding suggests that in humans, as well as in rodents, the fetal circadian pacemaker and the rhythms they ultimately regulate, such as the sleep-wake cycle, may be influenced by maternal circadian rhythms or by drugs that mimic or antagonize the relevant maternal signals. In rats and hamsters, the maternal SCNs are known to be required for normal maternal entrainment [7,28]. The identity of the specific maternal signal(s) is, however, unknown. Three stimuli have been shown to entrain litters born to SCN-lesioned mothers: injection of melatonin [8] or the dopamine agonist SKF38393 (36) to the pregnant mother in Syrian hamsters, and restricted feeding schedules imposed upon the mother in rats [37]. In the case of melatonin, a variety of evidence suggests that it is a physiologically important signal [5], including the observation that the effectiveness of exogenous melatonin is transient during development. In Syrian hamsters, entrainment is possible with melatonin injections to the pregnant mother or directly to pups on postnatal days (PN) 1 5, but not on PN 6 10 [8,16]. In rats, the first two postnatal weeks is also the age when maternal entrainment ends [10,17,26,30] and entrainment by light begins [13]. A response to light, measured as 2-deoxyglucose uptake or induced expression of c-fos within the SCN, can be seen within 48 h of birth [15,22,38], but it is not yet known if entrainment by light occurs at this time. In Syrian and Djungarian hamsters, induction of Fos protein is first seen on PN 5 and 3, respectively [12,20], and changes in the number and distribution of immunoreactive cells continues for 2 weeks (where appropriate, embyonic, and postnatal ages have been adjusted to be consistent with the convention used in this article). Similar to melatonin, entrainment by dopaminergic activation might also be transient during development. Although dopaminergic activation and light both induce c-fos expression within the SCN, a transition occurs during the first postnatal week as to which of these is most effective in doing so [38,39]. In hamsters, prenatal injections SKF38393 cause entrainment, but a preliminary study indicates that injections to adults do not [31,36]. The goals of the present study were to determine whether entrainment by dopaminergic activation extends beyond birth and if so, whether the entrainment ends within the first two postnatal weeks. In addition the study examined whether c-fos induction by SKF38393 is correlated with the drug s ability to cause entrainment. Finally, the study tested the hypothesis that, if SKF38393 causes entrainment, the phase relationship it establishes is different from that established by melatonin. This result would be consistent with the suggestion that dopaminergic activation represents or mimicks a signal for daytime whereas melatonin represents a signal for nighttime [35]. * Address for correspondence: Fred C. Davis, Department of Biology, Northeastern University, 414 Mugar Life Science Building, Boston, MA 02115, USA. Fax: ; f.davis@nunet.neu.edu 185

2 186 GROSSE AND DAVIS FIG. 1. Photomicrograph of a Nissl-stained coronal section through the lesion site of a representative SCNX mother. The actogram of this animal is shown in Fig. 2A. 3V, third ventricle; OCh, optic chiasm. MATERIALS AND METHODS General Methods Animals. Adult male and female Syrian hamsters (8 weeks old, Charles River Laboratories, Kingston, NY, USA) were maintained on a 14L:10D light-dark cycle, with lights off at 1600h EST. Estrous cycles of the females were determined by daily examination of the vaginal discharge, and animals were mated overnight on the night of ovulation. Fertilization was assumed to occur mid-way through the dark portion of the LD cycle, with the day after mating being designated embryonic day 1 (E1). Females were transferred to constant dim light ( 2 lux, dim LL) at the time of lights-off on E1. They were housed individually in cages equipped with a running wheel for the recording of activity rhythms. Revolutions of the wheel, triggering a microswitch mounted on the side of the cage, were collected in 10-min bins and recorded and stored using DataCol III software (Minimitter Co., Sunriver, OR, USA). Birth occurred during the morning (generally before 1000h) of E16, with the 24-h period following the time of birth being designated PN 1. Food and water were available ad libitum throughout the experiment. Surgery. On the seventh day of gestation, pregnant females received electrolytic lesions directed toward the SCN. Animals were anaesthetized with sodium pentobarbital (100 mg/kg body weight) and placed in a stereotaxic frame. Co-ordinates were 8.3 mm below the skull surface and 0.6 mm anterior to bregma (incisor bar 2 mm below ear bar), and a current of 4 ma was passed for 10 s. Animals were returned to dim LL, and activity recording continued. Any female whose activity showed circadian rhythmicity after lesioning was removed from the experiment. Histology. Verification of SCN lesion placement was carried out at the end of the experiment by Nissl staining. Animals were deeply anaesthetized with sodium pentobarbital and perfused intracardially with 0.01 M phosphate-buffered saline (PBS) containing 25,000 U/l heparin, followed by 4% paraformaldehyde in 0.01 M PBS (PFA). Brains were post-fixed for 2 h in PFA and cryoprotected overnight in 20% sucrose in 0.01 M PBS. Coronal sections through the lesion site were cut at 60 m on a freezing microtome and stained with thionin. Immunocytochemistry. Immunocytochemical detection of Fos, the protein product of the immediate early gene c-fos, was carried out as follows: pups were perfused as above and the brains post-fixed for 24 h then cryoprotected in 20% sucrose in PBS for 24 h. Then, 60- sections through the SCN region were cut on a freezing microtome. Free-floating sections were treated for 10 min in 0.5% H 2 O 2, then rinsed in 0.01 M PBS. Sections were then incubated overnight at 4 C in 0.01 M PBS containing 1% bovine serum albumen (BSA) and 0.3% triton-x, anti-fos antibody (DCH1, 1:4000, donated by Dr. D. Hancock, ICRF, London, UK) and 2% normal goat serum (NGS). This antiserum is raised against the N-terminal sequence of 16 amino acids of Fos and is specific for that antigen [14]. Detection of primary antiserum was achieved using the Vectastain Elite avidinbiotin complex method and visualized using diaminobenzidine. Quantification of immunocyto-chemistry (ICC) was achieved by counting the number of Fos immunoreactive (Fos-ir) nuclei in the SCN in three sections per animal, using the total number of Fos-ir cells per animal to generate group means. Differences between group means were assessed by ANOVA with post-hoc analysis by Dunnett s t-test.

3 TRANSIENT ENTRAINMENT BY DOPAMINE 187 FIG. 2. Double-plotted actograms beginning on PN 21, the day of weaning, of two SCNX female hamsters (A, E) and three of each of their pups (B D, F H). Pups B D received SKF38393 injections at 0800h on PN 1 5, whereas pups F H received SKF38393 at 2000h on PN 1 5. Time of activity onset on the day of weaning is about 12 h out of phase with the time of injection. Arrows show the time at which injections were given, and open circles indicate the time of weaning. Data analysis and statistics. Activity records were analyzed and graphically displayed using Circadia software (Behavioral Cybernetics, Cambridge, MA, USA). The presence or absence of circadian rhythmicity was confirmed, and free-running period calculated, using the 2 periodogram. Onset of pups activity rhythms on the day of weaning was determined by fitting lines by eye to at least 10 consecutive activity onsets and extrapolating back to the day of weaning. Phases were plotted

4 188 GROSSE AND DAVIS FIG. 3. Double-plotted actograms beginning on PN 21, the day of weaning, of two SCNX female hamsters (A, E) and three of each of their pups (B D, F H) that received SKF38393 on PN Pups B D were treated at 0800h, whereas pups F H received injections at 2000h. Time of activity onset differs considerably between pups within the same litter, despite their receiving the same treatment. Symbols are the same as those in Fig. 2. on a circle representing the 24 h of PN 21. Average phase was calculated by vector addition, where the length of the average vector, r, represents the scatter among phases. The Rayleigh test [1] was used to determine if the distribution of phases within litters and treatment groups was significantly different from uniform, and, where there were significant average phases

5 TRANSIENT ENTRAINMENT BY DOPAMINE 189 FIG. 5. Phase on the day of weaning of all pups treated on PN 6 10 plotted on a circle representing the 24 h of PN 21. (A) Animals that received SKF (B) Animals that received vehicle. Symbols are the same as those in Fig. 4. Although average phases are shown for all of the groups, it is only for the 2000 vehicle injection group (B) that average phase is meaniful because in the other groups the distributions were not significantly different from uniform (see text). FIG. 4. Phases on PN 21 of all pups in the PN 1 5 (Experiment 1) and PN 1 (Experiment 2) groups, plotted on a circle representing the 24 h of the day of weaning. (A) Pups that received SKF38393 on PN 1 5. (B) Pups that received vehicle injections on PN 1 5. (C) Pups that received SKF38393 on PN 1 only. Open circles are the phases of individual pups treated at 0800h; closed circles are the phases of pups treated at 2000h. The mean phase of each group is shown by the arrow within the circle (open arrow, 0800 injections; closed arrow, 2000 injections), whereas the length of that arrow and the value of r represent the degree of synchrony within a group. Although an average phase is indicated for the 0800 vehicle group in B, the average phase has little meaning because the distribution was not significantly different from uniform (see text). Arrows outside the circle show the times of the injections. in groups treated at opposite times of days, differences between these phases were determined by the Mardia, Watson, and Wheeler test [1]. Protocols Experiment 1: Injections of SKF38393 over five consecutive days at two times in development (PN 1 5 and PN 6 10). Pups were born to SCN-lesioned (SCNX) mothers and housed in dim LL, so that they received no known external or maternal entraining stimuli prior to or after birth. Litters, containing 4 14 pups, received i.p. injections of SKF38393 (8 mg/kg, RBI, Natick, MA, USA) or vehicle (sterilized, double-distilled water) at 0800h or 2000h for 5 consecutive days, beginning on PN 1, the day of birth (PN 1 5 group, 2 3 litters per treatment), or PN 6 (PN 6 10 group, 2 litters per treatment). All pups within a litter were treated identically. After the last injection, the litters were left undisturbed until weaning on PN 21. On the day of weaning, pups were removed from the mothers and immediately placed in a cage equipped with a running wheel, maintaining the same lighting conditions as before. To assess any effects of the time of weaning on phase of activity rhythms, half the animals in each litter were weaned at 0800h, and the other half at 2000h. At least 7 pups from large litters (those containing 7 animals) and all pups from small litters (those with 7 animals) were tested. Activity rhythms were monitored for 3 weeks. Final group sizes were as follows: PN 1 5 group morning vehicle 25, evening vehicle 21, morning SKF 23, evening SKF 29; PN 6 10 group morning vehi-

6 190 GROSSE AND DAVIS Experiment 3: Fos expression in the SCN following SKF38393 injections. Pups were gestated and born under the same conditions as the above experiments. On PN 1 or PN 6, pups received a single injection of SKF38393 or vehicle at 0800h or 2000h and were returned to their mothers. Pups from at least two litters contributed to each group. One hour after injection, pups were perfused as above (see Surgery), and the brains were processed for immunocytochemical detection of Fos as above (see Immunocytochemistry). RESULTS General All mothers that showed a loss of circadian rhythmicity also had complete lesions of the SCN, as determined by post-mortem histology. Figure 1 shows the lesion site of a representative mother, whose actogram is also shown (Fig. 2A). Actograms from three other lesioned mothers are also shown in Figs. 2 and 3. All of the pups showed robust circadian rhythms of activity, with the mean ( SEM) free-running period being h (n 195). FIG. 6. Representative photomicrographs of coronal sections of pup brains processed for immunocytochemical detection of Fos after various treatments on PN 1. (A) Fos-ir in the SCN of a pup that received SKF38393 at 0800h. (B) Fos-ir in the SCN of a pup that received SKF38393 at 2000h. (C) Fos-ir in the SCN of a pup that received vehicle at 0800h. OCh, optic chiasm; 3V, third ventricle. cle 11, evening vehicle 14, morning SKF 17, evening SKF 19. Experiment 2: Single injections of SKF Pups were gestated, born, and raised as in Experiment 1. On the day of birth, PN 1, pups received a single injection of SKF38393 at 0800h or 2000h (n 2 litters per group). As above, all litter-mates were treated identically. Litters were left undisturbed until the day of weaning, and activity rhythms were monitored as above. Final group sizes were: morning SKF 16, evening SKF 20. Experiment 1 PN 1 5 injections. The mean phase at weaning of the pups that received SKF38393 at 0800h was 17:45h (r 0.80), whereas that of the group that received SKF38393 at 2000h was 08:30h (r 0.83). Clustering of phases within both of these groups was highly significant ( p 0.001, Rayleigh test), and the distribution of phases between the two groups was also significantly different ( p 0.001, Mardia, Watson, and Wheeler test). The mean phases of the two groups differed by 9.25 h (Fig. 4A). Actograms from two mothers and three of each of their pups (SKF38393 at 0800 for one litter and 2000 for the other) are shown in Fig. 2. Pups that received vehicle injections were less synchronized within a treatment group than were those that received SKF38393, and there was more overlap between the morning and evening groups (Fig. 4B). The morning group was not significantly clustered ( p 0.05, Rayleigh test), and clustering of the evening group barely achieved significance ( p 0.05, r 0.44). In none of the four PN 1 5 experimental groups were the pups phases clustered around the time that they were weaned. When phases were plotted relative to the time of weaning, none of the distributions was significantly different from uniform (SKF38393: 0800, r 0.18; 2000, r 0.08; vehicle: 0800, r 0.04; 2000, r 0.026, all groups, p 0.10). PN 6 10 injections. Actograms from two mothers and three of each of their pups (SKF38393 at 0800 for one litter and 2000 for the other) are shown in Figure 3. Pups injected with SKF38393 at either 0800 or 2000 failed to show significantly clustered phases (0800, r 0.24; 2000, r 0.35; p 0.05 in both cases, Rayleigh test) (Fig. 5A). Pups injected with vehicle in the morning on PN 6 10 also did not show significantly clustered phases (r 0.38, p 0.05, Rayleigh test), whereas pups injected with vehicle in the evening did (r 0.50, p 0.05, Rayleigh test) (Fig. 5B). Although significant clustering among pups that received vehicle injections in the evening was seen with both the PN 1 5 and PN 6 10 injections, the average phases in the two groups were quite different (Figs. 4B and 5B). In none of the four PN 6 10 experimental groups were the pups phases clustered around the time that they were weaned. When phases were plotted relative to the time of weaning, none of the distributions was significantly different from uniform (SKF38393: 0800, r 0.02; 2000, r 0.07; vehicle: 0800, r 0.17; 2000, r 0.18, all groups, p 0.10).

7 TRANSIENT ENTRAINMENT BY DOPAMINE 191 FIG. 7. Histogram showing the mean ( SEM) number of Fos positive (Fos ve) nuclei in the SCN of pups injected with SKF38393 or vehicle in the morning (am, 0800h) or evening (pm, 2000h) of PN 1 or PN 6. Experiment 2 Single injections of SKF38393 were able to induce significant clustering of phases in both groups of pups that received single injections of SKF38393 on PN 1 (0800, r 0.52, phase 02:00 h, p 0.05; 2000, r 0.62, phase 10:10 h, p 0.01) (Fig. 4C). The two distributions were significantly different from each other ( p 0.05), suggesting that even a single injection of SKF38393 was able to set the pups rhythms to different phases. In neither group were the pups phases clustered around the time that they were weaned. When phases were plotted relative to the time of weaning, neither of the distributions was significantly different from uniform (SKF38393: 0800, r 0.11; 2000, r 0.31, p 0.10). Experiment 3 Injections of SKF38393 at both 0800 and 2000 were able to induce the expression of c-fos in the SCN when administered on PN 1 to pups born in constant conditions and born to SCNX mothers. However, no Fos-ir was seen in the SCN of pups that received vehicle injections at either time (Figs. 6 and 7). In SKF-injected animals, Fos-ir nuclei were most dense in the dorsal part of the SCN, and extended dorsally toward the paraventricular nucleus of the hypothalamus. Injection of SKF38393 at 0800 or 2000 on PN 6 resulted in very little expression of Fos in the SCN, and there was no difference in the mean number of Fos-ir nuclei between drug- and vehicle-treated animals (Fig. 7). DISCUSSION The dopamine D1 agonist SKF38393 was able to entrain the free-running activity rhythms of weanling hamsters when injected on PN 1 5. Injection of the drug at opposite times of day to two groups of pups resulted in a 9.25-h difference in the mean phases of activity onset of the two groups, and both of the groups were significantly clustered. Pups that were identically treated except for receiving vehicle rather than SKF38393 differed in two important ways. The groups were not significantly clustered or were less clustered, and the overlap between phases in the groups was greater. Even a single injection of SKF38393 had an effect on phase greater than that of the procedure and vehicle alone; single injections given at opposite times of the day on PN 1 to separate groups of pups were able to induce significant clustering in both groups and the groups were significantly different. The p.m. group was somewhat more clustered than the a.m. group and showed an average phase more similar to its respective group in the multiple

8 192 GROSSE AND DAVIS FIG. 8. Composite of results from the current study and from Grosse et al. [16] showing the phases on PN 21 of pups injected with SKF38393 or melatonin in the morning (0800h) or evening (2000h) of PN 1 5. Phases are plotted on a circle representing the 24 h of the day of weaning. (A) Phases of animals treated at 0800h. (B) Phases of animals treated at 2000h. Note that injection of the drugs at the same time of day results in almost opposite mean phases. Open symbols, pups injected with SKF38393; closed symbols, pups injected with melatonin. The large filled arrow outside of the circle indicates the time of the injections, whether melatonin or SKF injection experiment than did the a.m. group (Fig. 2A,C). This could have been because the single p.m. injection was given at an earlier age than the single a.m. injection when sensitivity to SKF38393 could be higher. Alternatively, even though the mothers had SCN lesions, and this is known to disrupt synchrony among pups [7], there might have been some non-randomness among phases (e.g. related to developmental synchrony) that places phases closer to the average phase of the p.m. group. This could also contribute to weak, but significant, clustering in the multiple-injection vehicle p.m. groups. In contrast to the effect of SKF38393 on PN 1 5, administration of the drug on PN 6 10 was unable to entrain the activity rhythms of pups. Neither the morning nor the evening treatment groups showed significantly clustered phases. These results demonstrate that the entraining effect of SKF38393 is lost by at least day 10 and probably by day 6. The ability of a treatment to cause entrainment was correlated with its ability to induce Fos. Administration of SKF38393 to pups born to SCNX mothers at 0800 or 2000 on PN 1, but not on PN 6, induced the expression of Fos in the SCN. In addition, Fos expression within the SCN was low in all of the groups that received vehicle only. These results suggest that there is a developmental loss in the sensitivity of SCN cells to dopaminergic activation resulting in both the loss of Fos expression and the loss of entrainment of the circadian pacemaker. In addition, dopamine has been shown to increase phosphorylation of the calcium/cyclic AMP response-element-binding-protein (CREB) in neonatal Syrian hamster SCN [23]. Together these results provide correlative evidence for the involvement of c-fos expression in entrainment, as has been previously obtained for entrainment by light in adults [21]. It is likely that entrainment by some non-photic stimuli, such as arousal [24] and melatonin [35], do not require an increase in c-fos expression. Whether any stimulus that does induce c-fos in the SCN is also able to cause entrainment is unknown. Prenatal exposure to nicotine in rats or to cocaine in rats and mice induces c-fos expression in the fetal SCN [2,4,39], but it is not yet known whether either of these drugs causes entrainment. Previous studies demonstrated entrainment of weanling hamster activity rhythms following injection of SKF38393 to their mothers late in gestation [35,36]. The results of the current study suggest that the entrainment observed in those earlier experiments was the result of a direct effect of the drug on the fetus rather than on the mother and provide further evidence for the hypothesis that entrainment of the fetal and neonatal pacemaker is mediated by maternal activation of dopaminergic systems within the fetus/ neonate [39]. It is likely that SKF38393 is acting through D1 receptors, because the SCNs of fetal mice with a targeted deletion of the D1 receptor gene no longer show c-fos induction following prenatal injection of the drug to the mother [2]. The pups used in the current experiments were born to SCNX mothers and raised in constant conditions. It is likely, therefore, that, before exposure to SKF38393, the pups circadian pacemakers were at widely different phases [7] and that the drug subsequently produced clustering among these previously scattered phases. This suggests that SKF38393 was able to cause large phase shifts and was able to do so during at least several circadian phases. Large phase shifts are also suggested by Experiment 2 in which only a single injection of the drug on PN 1 had an effect. In a previous study, single prenatal injections of either SKF38393 or melatonin were also able to produce significant clustering among pups born to SCNX mothers [35]. In that study, simulations suggested that significant clustering could be produced from randomly distributed phases if large phase shifts were induced at phases covering about half of the circadian cycle [35]. The simulations also indicated that the drugs could be without effect at some phases, in particular, phases (encountered by the drug) when pups happen to already be at the final entrained phase at the time of the injection. In Experiment 3 of the current study, the induction of Fos was observed in the SCN of all pups injected with SKF38393 on PN 1. This suggests that, if Fos induction is always correlated with phase shifting, then the phases at which SKF38393 does not induce at least some phase shifts are restricted to a relatively small fraction of the circadian cycle. The reason for the loss of response to SKF38393 by PN 6 is unclear. D1 dopamine receptors are present in fetal and neonatal SCN [2,34,36], but are also present in the SCN of adult rats and primates, including humans [29,40], suggesting that the loss of response to dopaminergic activation is not due to loss of receptors during development. Whether there is a loss of receptors in Syrian hamsters is not known; it is also unknown whether there is a change in receptor number or affinity. It appears that, by whatever mechanism, there is a general loss of response to stimuli that may be involved in maternal entrainment. In rats, maternal entrainment ends during the first 2 weeks [17,26,30], whereas in hamsters, both melatonin [16] and SKF38396 lose their ability to entrain (current results). At the same time, rats begin to show entrainment by light

9 TRANSIENT ENTRAINMENT BY DOPAMINE 193 [13] and the SCN begins to show c-fos induction in response to light [12,20,22,38]. The average phase established by SKF38393 injections in the present study occurred at times that were approximately opposite to that seen following postnatal injections of melatonin [16]. This difference is illustrated in Fig. 8, which is a composite of results from the current study and from the previous melatonin study. The different phases established by dopaminergic activation and melatonin suggest that they have opposite effects on the circadian pacemaker. If they acted on the pacemaker at opposite times of day, their effects would be complementary. In the intact hamster, melatonin release occurs during the dark phase of a light-dark cycle and during subjective night (active time) of a hamster in constant conditions. Thus, the presence of melatonin may serve as a cue for the onset of activity by signifying nighttime. Dopaminergic activation, on the other hand, may be associated with day. SKF38393, like light pulses, induces c-fos expression in the SCN, and it is the subjective day phase of the pups rhythms that becomes coincident with the SKF38393 injection. There is precedence for a similar relationship between melatonin and dopamine in the retina. In the retina, dopamine synthesis and release are stimulated by light [3,18], and dopamine promotes light-adaptive retinal responses [9,25,33]. Melatonin, on the other hand, stimulates dark-adapted retinal responses and modulates dopaminergic neurotransmission [11,33]. An antagonistic interaction between melatonin and dopamine receptors on the same cell in the chick retina has been hypothesized [19]. In addition, melatonin was recently shown to antagonize the phosphorylation of CREB, which is otherwise induced by dopamine in primary cultures of neonatal Syrian hamster hypothalamus [23]. Thus, maternal melatonin and dopaminergic activation may mediate maternal entrainment, possibly by having opposing effects upon cellular processes that are coupled to the fetal and neonatal pacemaker. The maternal signal that might lead to dopaminergic activation in the fetal SCN is not known. ACKNOWLEDGEMENTS This work was supported by NIH grant HD18686 to F.C.D. The authors thank Dr. D. C. Hancock for the generous gift of DCH1 antibody. REFERENCES 1. Batschelet, E. Recent statistical methods for orientation data. In: Galler, S. R., ed. Animal orientation and navigation. Washington, DC: U.S. Government Printing Office; 1972: Bender, M.; Drago, J.; Rivkees, S. A. D1 receptors mediate dopamine action in the fetal suprachiasmatic nuclei: Studies of mice with targeted deletion of the D1 dopamine receptor gene. Mol. Brain Res. 49: ; Boatright, J. H.; Hoel, M. J.; Iuvone, P. M. Stimulation of endogenous dopamine release and metabolism in amphibian retina by light and K -evoked depolarization. Brain Res. 482: ; Clegg, D. A.; O Hara, B. F.; Heller, H. C.; Kilduff, T. S. Nicotine administration differentially affects gene expression in the maternal and fetal circadian clock. Develop. Brain Res. 84:46 54; Davis, F. C. Melatonin: Role in development. J. Biol. Rhythms 12: ; Davis, F. C.; Frank, M. G.; Heller, H. C. Ontogeny of sleep and circadian rhythms. In: Turek, F. W.; Zee, P. C., eds. Neurobiology of sleep and circadian rhythms. New York: Marcel Dekker; in press. 7. Davis, F. C.; Gorski, R. A. Development of hamster circadian rhythms: Role of the maternal suprachiasmatic nucleus. J. Comp. Physiol. A 162: ; Davis, F. C.; Mannion, J. Entrainment of hamster pup circadian rhythms by prenatal melatonin injections to the mother. Am. J. Physiol. 255:R439 R448; Dearry, A.; Burnside, B. Dopaminergic regulation of cone retinomotor movement in isolated teleost retinas I: Induction of cone contraction is mediated by D2 receptors. J. Neurochem. 46: ; Deguchi, T. Ontogenesis of a biological clock for serotonin: Acetyl coenzyme A N-acetyltransferase in pineal gland of rat. Proc. Natl. Acad. Sci. USA 72: ; Dubocovich, M. L. Melatonin is a potent modulator of dopamine release in the retina. Nature 306: ; Duffield, G. E.; Dickerson, J. M.; Alexander, I. H. M.; Ebling, F. J. P. Ontogeny of a photic response in the suprachiasmatic nucleus in the Siberian hamster (Phodopus sungorus). Eur. J. Neurosci. 7: ; Duncan, M. J.; Banister, M. J.; Reppert, S. M. Developmental appearance of light-dark entrainment in the rat. Brain Res. 369: ; Ebling, F. J. P.; Maywood, E. S.; Staley, K.; Humby, T.; Hancock, D. C.; Waters, C. M.; Evan, G. I.; Hastings, M. H. The role of N-methyl-D-aspartate-type glutamatergic neurotransmission in the suprachiasmatic nuclei of the syrian hamster. J. Neuroendocrinol. 3: ; Fuchs, J. L.; Moore, R. Y. Development of circadian rhythmicity and light responsiveness in the rat suprachiasmatic nucleus: A study using the 2-deoxy[1-14C]glucose method. Proc. Natl. Acad. Sci. USA 77: ; Grosse, J.; Velickovic, A.; Davis, F. C. Entrainment of Syrian hamster circadian activity rhythms by neonatal melatonin injections. Am. J. Physiol. 270:R533 R540; Honma, S.; Honma, K.; Hiroshige, T. Restricted daily feeding during nursing period resets circadian locomotor rhythm of infant rats. Am. J. Physiol. 252:R262 R268; Iuvone, P. M.; Galli, C. L.; Garrison-Gund, C. K.; Neff, N. H. Light stimulates tyrosine hydroxylase activity and dopamine synthesis in retinal amacrine neurons. Science 202: ; Iuvone, P. M.; Gan, J. Functional interaction of melatonin receptors and D1 dopamine receptors in cultured chick retinal neurons. J. Neurosci. 15: ; Kaufman, C. M.; Menaker, M. Ontogeny of light-induced Fos-like immunoreactivity in the hamster suprachiasmatic nucleus. Brain Res. 633: ; Kornhauser, J. M.; Mayo, K. E.; Takahashi, J. S. Light, immediateearly genes, and circadian rhythms. Behav. Genet. 26: ; Leard, L. E.; Macdonald, E. S.; Heller, H. C.; Kilduff, T. S. Ontogeny of photic-induced c-fos mrna expression in rat suprachiasmatic nuclei. Neuroreport 5: ; McNulty, S.; Schurov, I. L.; Sloper, P. J.; Hastings, M. H. Stimuli which entrain the circadian clock of the neonatal Syrian hamster in vivo regulate the phosphorylation of the transcription factor CREB in the suprachiasmatic nucleus in vitro. Euro. J. Neurosci. 10: ; Mead, S. M.; Ebling, F. J. P.; Maywood, E. S.; Humby, T.; Herbert, J.; Hastings, M. H. A non-photic stimulus causes instantaneous phaseadvances of the light-entrainable circadian oscillator of the Syrian hamster, but does not induce the expression of c-fos in the suprachiasmatic nuclei. J. Neurosci. 12: ; Pierce, M. E.; Besharse, J. C. Circadian regulation of retinomotor movements I: Interaction of melatonin and dopamine in control of cone length. J. Gen. Physiol. 86: ; Reppert, S. M.; Coleman, R. J.; Heath, H. W.; Swedlow, J. R. Pineal N-acetyltransferase activity in 10-day-old rats: A paradigm for studying the developing circadian system. Endocrinology 115: ; Reppert, S. M.; Schwartz, W. J. The suprachiasmatic nuclei of the fetal rat: Characterization of a functional circadian clock using 14C-labeled deoxyglucose. J. Neurosci. 4: ; Reppert, S. M.; Schwartz, W. J. Maternal suprachiasmatic nuclei are necessary for maternal coordination of the developing circadian system. J. Neurosci. 6: ; Rivkees, S. A.; Lachowicz, J. E. Functional D1 and D5 dopamine receptors are expressed in the suprachiasmatic, supraoptic, and paraventricular nuclei of primates. Synapse 26:1 10; Sasaki, Y.; Murakami, N.; Takahashi, K. Critical period for the en-

10 194 GROSSE AND DAVIS trainment of the circadian rhythm in blinded pups by dams. Physiol. Behav. 33: ; Snodgrass, P.; Davis, F. C. Dopaminergic activation is a transient mechanism for entrainment in Syrian hamsters [Abstract]. Physiologist 38:A 23; Speh, J. C.; Moore, R. Y. Retinohypothalamic tract development in the hamster and rat. Develop. Brain Res. 76: ; Stenkamp, D. L.; Iuvone, P. M.; Adler, R. Photomechanical movements of cultured embryonic photoreceptors: Regulation by exogenous neuromodulators and by a regulable source of endogenous dopamine. J. Neurosci. 14: ; Strother, W. N.; Zimmer, K. A.; Norman, A. B.; Duncan, H. J.; Lehman, M. N. Autoradiographic localization of D1 dopamine receptors in the fetal and neonatal hamster suprachiasmatic nucleus [Abstract]. Soc. Neurosci. Abstr. 19:1816; Viswanathan, N.; Davis, F. C. Single prenatal injections of melatonin or the D1-dopamine receptor agonist SKF to pregnant hamsters sets the offsprings circadian rhythms to phases 180 apart. J. Comp. Physiol. [A] 180: ; Viswanathan, N.; Weaver, D. R.; Reppert, S. M.; Davis, F. C. Entrainment of the fetal hamster circadian pacemaker by prenatal injections of the dopamine agonist, SKF J. Neurosci. 14: ; Weaver, D. R.; Reppert, S. M. Periodic feeding of SCN-lesioned pregnant rats entrains the fetal biological clock. Develop. Brain Res. 46: ; Weaver, D. R.; Reppert, S. M. Definition of the developmental transition from dopaminergic to photic regulation of c-fos gene expression in the rat suprachiasmatic nucleus. Mol. Brain Res. 33: ; Weaver, D. R.; Rivkees, S. A.; Reppert, S. M. D1-dopamine receptors activate c-fos expression in the fetal suprachiasmatic nuclei. Proc. Natl. Acad. Sci. USA 89: ; Weiner, D. M.; Levey, A. I.; Sunahara, R. K.; Niznik, H. B.; O Dowd, B. F.; Seeman, P.; Brann, M. R. D1 and D2 dopamine receptor mrna in rat brain. Proc. Natl. Acad. Sci. USA 88: ; 1991.