Pineal Melatonin Rhythms in Female Turkish Hamsters: Effects of Photoperiod and Hibernation

BIOLOGY OF REPRODUCTION 35, 74-83 (1986) Pineal Melatonin Rhythms in Female Turkish Hamsters: Effects of Photoperiod and Hibernation JANET M. DARROW,2 LAWRENCE TAMARKIN,3 MARILYN J. DUNCAN,4 and BRUCE D. GOLDMAN4 Worcester Foundation for Experimental Biology24 Shrewsbury, Massachusetts 01545 and Clinical Psychobiology Branch3 National Institute of Mental Health Bethesda, Maryland 20205 ABSTRACT Daily rhythms of pineal and serum melatonin content were characteried for adult female Turkish hamsters (Mesocricetus brandti) exposed to long days (16L:8D, 22#{176}C)or after transfer to short days (IOL:14D, 22#{176}C). The nocturnal peak of pineal melatonin content was found to be approximately 3 h greater in duration on short than on long days. Changes in levels of serum melatonin closely paralleled those of pineal melatonin. Thus, an effect of photoperiod on synthesis and secretion of pineal melatonin was demonstrated. In a separate experiment, female hamsters were induced to hibernate by exposure to a short-day, cold environment (1OL: 14D, 6#{176}C).During the 4 to 5-mo hibernation season, Turkish hamsters are known to display 4 to 8-day bouts of torpor (body temperature = 7-9#{176}C) alternating with 1 to 3-day intervals of euthermia (body temperature = 35-37#{176}C). Little evidence of nocturnal synthesis or secretion of pineal melatonin was detected in females sampled during torpor. However, animals sampled during the first day after arousal from a torpor bout displayed melatonin rhythms no different in phase or amplitude from those seen in females held at 22#{176} C. Thus, despite the absence of pineal melatonin output during torpor, the pineal gland of hibernating Turkish hamsters produces an appropriately phased, rhythmic melato nm signal during intervals of euthermia. INTRODUCTION The daily rhythm of pineal melatonin is known to be a component of the photoperiodic mechanism in a number of seasonally breeding mammals (Reiter, 1980; Goldman and Darrow, 1983; Karsch et al., 1984; Stetson and Watson-Whitmyre, 1984; lllnerova and Vanecek, 1985; Tamarkinet al., 1985). In Djungarian hamsters and in sheep, it has been proposed that the duration of the nocturnal pineal melatonin peak is the critical feature of a pineal endocrine signal which conveys daylength information (Bittman et al., 1983; Carter and Goldman, 1983a,b; Goldman et al., 1984). In the Turkish hamster (Mesocricetus brandti), unlike other mammals tested to date, pinealectomy results Accepted November 12, 1985. Received August 12, 1985. This research was supported by NIH Research Grant HD 15912 to BDG. 2 Reprint requests: Janet M. Darrow, Ph.D., Worcester Foundation for Experimental Biology, 222 Maple Avenue, Shrewsbury, MA 01545. in gonadal regression on long days, indicating that the pineal is required for maintenance of gonadal function (Carter et al., 1982). In contrast, removal of the pineal gland in the closely related Syrian hamster (Mesocricetus auratus) results in continuous maintenance of the reproductive system on long days and prevents short-day-induced reproductive regression (Reiter, 1978). Because of the unusual response of Turkish hamsters to pinealectomy, it was of interest to examine the daily rhythm of pineal melatonin content in this species on long vs. short days, for comparison with melatonin patterns already characteried in other species. Furthermore, since it has been suggested that pineal melatonin in mammals is secreted predominantly into the peripheral circulation (Rollag et al., 1978), we utilied a recently developed radioimmunoassay for serum melatonin to measure circulating levels of the hormone. Turkish hamsters hibernate when exposed to a short-day, cold environment in the laboratory (Hall and Goldman, 1980). The hibernation season is 4-5 74

MELATONIN IN TURKISH HAMSTERS 75 mo in duration and is characteried ly alternating bouts of torpor (lasting 4-8 days) and arousal (lasting 1-3 days). During torpor, core body temperature drops to as low as 7-9#{176}C, representing a substantial decline from the 35-37#{176}C range of euthermic body temperatures observed during intervals of arousal. To investigate the consequences of torpor on pineal biosynthetic/secretory activity during hibernation, levels of pineal and serum melatonin were measured in hibernating hamsters during torpor and during intervals of arousal. The daily melatonin profiles observed in hibernating hamsters were compared with those measured in nonhibernating animals housed at Animals 22#{176}C. MATERIALS AND METHODS Young adult female Turkish hamsters (Mesocricetus brandti) were obtained from our breeding colony. The animals were raised from birth on a long-day photoperiod (16L:8D, lights on from 0200 to 1800 h) at 22 ± 2#{176}C.All hamsters were supplied with food (Wayne Lab Blox or Purina Chow No. 5008) and water ad libitum. Experiment I. Pineal Melatonin on Long and Short Days Female hamsters ranging in age from 1.5 to 3 mo were selected for this study. Estrous cyclicity was monitored by daily examination of vaginal discharges following the method of Orsini (1961). At the onset of the experiment, 39% (60/15 3) of the females selected for the study were found to be acyclic on long days. This proportion of anovulatory females is typical for our colony of Turkish hamsters. Thirty-one acyclic and 50 cycling females were transferred to a short-day light schedule (1OL:14D, lights on from 0500-1500 h, 22#{176}C), and the remaining 29 acyclic and 43 cyclic females were left on the longday schedule. All animals were checked daily for a further period of 5-6 wk, at which time they were killed by decapitation at various times of day. At each time point, 2-8 females of each reproductive state were sampled on both long and short days. During the dark portion of the light/dark cycle, hamsters were killed under dim red illumination. Pineals were rapidly excised and froen in separate vials on dry ice, then stored at -20#{176}C until assayed. Uteri were removed from most females and weighed. Radioimmunoassay for pineal melatonin was performed as described previously (Tamarkin et al., 1979), using melatonin antiserum R1055, which was generously supplied by Dr. M. D. Rollag. Experiment II. Pineal and Serum Melatonin on Long and Short Days Seventy-three female hamsters (2-4 mo of age) were selected for this study. Estrous cyclicity was not monitored in these females. Thirty-eight females were transferred to short days (1OL:14D, as above), and the remainder were left on long days (16L:8D). After 4-5 wk, animals were decapitated at various times during the night, under dim red light, for collection of trunk blood and pineals. Animals were sampled during January and February. Pineal melatonin assays were performed at the Worcester Foundation as described previously (Goldman et al., 1984), using 125 1-melatonin analog (Meloy Laboratories, Inc., Springfield, NJ) and R1055 antiserum supplied by M. D. Rollag. Serum melatonin was measured using melatonin antiserum R1056, also supplied by Dr. Rollag. Serum samples (300 p1) were extracted with 2.5 ml dichioromethane according to a procedure described by Brown et al. (1983), and melatonin content was determined by double antibody radioimmunoassay procedures (Darrow and Goldman, 1986). Experiment III.Melatonin Rhythms During Hibernation Ninety females ranging in age from 2-4 mo were selected for study. After a 6-wk prior exposure to short days (1OL:14D, 22#{176}C),groups of females were transferred to a cold room (6 ± 1#{176}C) with an identical short-day photoschedule. Hamsters were housed separately in 48 X 25 X 21 cm polycarbonate cages, containing Absorb-Dri bedding and cotton nesting material. Individuals were checked for torpor each day at 2-4 hours before lights-off. Torpid hamsters were readily identified by their reduced respiration rate and lack of response to a puff of air detected at the animal s back (Hall and Goldman, 1980). This method of detecting torpor has been corroborated by telemetric recording of body temperature (Hall, 1981). Those individuals exhibiting torpor for at least 10 days in a preceding 20-day interval were decapitated at 4-h intervals, either on Day 2 or 3 of a bout of torpor, or during the first day after spontaneous

76 DARROW ET AL. arousal from torpor. Pineals and trunk blood were collected as in previous studies. Some hamsters (16/90) showed less than 10 days of torpor during 60 days of cold exposure. These females were designated nonhibernators, and were sampled at 3 times of day. The cold room portion of the experiment ran from March to July. Data Analysis Statistical tests employed were chi square, linear regression analysis and one-way analysis of variance followed by the Duncan posteriori test for significance. females was found to be anovulatory. Of the cycling hamsters transferred to short days, 62% (31/50) became acyclic during the 6-week study. Thus, despite the rather high incidence of spontaneous acyclicity on long days, it appears that exposure to short days inhibited reproductive activity in a significant number of female Turkish hamsters (p< 0.01). Uterine weights for most of the females whose pineals were sampled are presented in Table 1. On each photoperiod uterine weights of acyclic females were significantly reduced as compared to those of cycling hamsters (p<0.01). When the pineal melatonin profiles were plotted separately for cycling vs. acyclic females, no differences related to reproductive state were apparent in RESULTS Experiment I. Pineal Melatonin on Long vs. Short Days IOL: 14D A significant nocturnal elevation of pineal melatonm content was observed for both short- and long-day female hamsters (Fig. 1). On short days (1OL:14D, upper panel), melatonin levels began to increase within 4-6 h after lights-off and remained high until 12 h after lights-off. Most individuals sampled during the last 2 h of the night showed low melatonin values. On long days (16L:8D, lower panel), hormone levels were elevated by 4 h after lights-off, and had decreased to near baseline in most individuals sampled just before lights-on. A concentration of 200 pg melatonin/pineal was designated as reflecting a substantial elevation of the hormone above daytime levels (i.e., at least 2 standard deviations above baseline levels observed in this and other assays). On this basis, levels of pineal melatonin were elevated for about 6 h on short days (from H 6 to H 12 of the night), and only for about 3 h on long days (from H 4 to H 7 of the night). Although the duration of the peaks was appreciably different on short vs. long days, no difference in the amplitude of the peaks was detected. The population of females represented in Figure 1 varied in reproductive state. Among the cycling females maintained on long days, 30% (13/43) became acyclic during the 6-wk period of examination. This spontaneous loss of estrous cyclicity on long days is commonly observed in Turkish hamsters in our laboratory. At the onset of the experiment, a similar proportion (60/153 or 39%) of the long-day -J 4 w a- w 0. 500 1100 0300 0500 TIME 16L: 8D FIG. 1. Pineal melatonin content in female hamsters sampled on short days (1OL:14D, 22#{176} C, upper panel) and on long days (16L:8D, 22#{176}C,lower panel). Shaded horiontal bars along x-axes indicate the dark portion of the daily light/dark cycle. Verticle lines above the means denote SEM (n = 6-15 females per time point). Samples coincident with the beginning or end of the night were taken several minutes before the light/dark transition.

MELATONIN IN TURKISH HAMSTERS 77 TABLE 1. Uterine weights of cycling and acyclic Turkish hamsters in long and short days. Condition n Uterine wt (mg) ± SE Long days, cycling 26 348.8 ± 19.1 Long days, acyclic 13 238.6 ± 51.2 Short days, cycling 15 296.5 ± 19.0 Short days, acyclic (cycling when moved to short days) 30 168.9 ± 13.3 Short days, acyclic (acyclic when moved to short days) 31 100.6 ± 5.7 either long or short days (data not shown). Therefore, the data for cyclic and anovulatory females were combined for each photoperiod (Fig. 1). Using the 200 pg/pineal value as a reference for determining elevated pineal melatonin, both cyclic and anovulatory hamsters displayed peak durations of 6 h on short days and 3 h on long days. Thus, the duration of the melatonin peak appeared to vary with daylength but not with reproductive state per se. During the time periods when most hamsters displayed elevated pineal melatonin, a sieable proportion of the females showed low levels (<100 pg melatonin/pineal). The incidence of atypically low values was similar (20-30%) in both acyclic and cyclic females, whether on long or short days (Table 2). Thus, no aspect of the pineal melatonin rhythm measured in this study was found to differ as a function of reproductive state on a given photoperiod. Experiment II. Pineal and Serum Melatonin on Long vs. Short Days Profiles of pineal melatonin for females on long and short daylengths in this experiment (Fig. 2) were quite similar in amplitude and duration to those observed in the previous study. Only slight differences were noticed in the phase of the noctural peaks. (For example on 16L: 8D, melatonin levels were still elevated above 200 pg/gland just minutes before lights-on, whereas in Experiment I, levels had declined to baseline by this time.) On short days, pineal melatonin content was elevated above 200 pg/gland for approximately 7 h (from H 5 to H 12 of the night). In contrast, duration of the melatonin peak on long days was only about 4 h (from only H 4 to H 8 of the night). These results indicate an approximate 3-h difference in duration of the pineal melatonin peak between long- and short-day hamsters, similar to the results of Experiment 1. The temporal pattern of changes in serum melatonm (Fig. 2, dark circles) paralleled very closely the rhythm of pineal melatonin content, both in longand in short-day animals. Peak levels in the circulation represented a significant 2- to 3-fold increase (p<o.01) compared to values observed early at night. When individuals were examined for correspondence between pineal and serum values, the correlation coefficients (r) were 0.3 74 for short-day females (p<0.02, n=36) and 0.652 for long-day females (p<0.01, n=34). When pineal melatonin values were highest for the group, atypically low values were not observed as frequently as in Experiment I, i.e., 7% (1/14) on long days, 17% (3/18) on short days. Experiment III. Melatonin Rhythms During Hibernation Most hamsters exposed for 6 wk to a short-day warm environment (1OL: 14D, 22#{176} C) began hibernat- TABLE 2. Proportion of cyclic and acyclic females showing low pineal melatonin content (<100 pg/gland) during the nocturnal pineal melatonin peaks on long vs. short days. 16L:8D 1OL:14D Cyclic females 4/19 (21%) 10/33 (30%) Acyclic females 5/19 (26%) 10/50 (20%) Data includes females sampled from 4 to 6 hours after lights-off on 16L:8D and those sampled from 6 to 12 hours after lights-off on 1OL: 14D.

78 DARROW ET AL. -J 4 LU a- 0 I- 4 -j LU 0 0. 800 600 400 200 800 IOL 140 16L 80 I-.1 I I I I I 600 30 400 20.4 remains relatively stable at euthermic levels until the next bout of torpor (Hall, 1981). For females sampled during the first day after arousal from torpor, the melatonin rhythms were 40 quite similar in amplitude and phase to those of females housed at room temperature in the earlier experiments (Fig. 3, upper panel). During the noctur- 30 nal peaks, pineal melatonin was elevated about 10- fold (open circles, p<0.01) and serum melatonin about 3-fold (closed circles, p<0.01) compared to 20 daytime levels. Hamsters which failed to hibernate.o displayed concentrations of melatonin in the pineal and serum similar to those of hibernators sampled iii during an interval of arousal (Fig. 3, upper panel). In contrast to the full-amplitude rhythms of melatonin observed in hamsters sampled during euthermia, those sampled during torpor showed little...,. evidence of nocturnal production or secretion of. pineal melatonin (Fig. 3, lower panel). In one set of females killed during mid-torpor (open triangles), mean values of pineal melatonin were less than 100 C 40 pg/gland at 3 time-points during the night. In another set of females for which both pineal and serum levels were assessed, levels of pineal melatonin were slightly elevated at the 0200-h time-point (256 ± 31 pg/ gland). However, the night-time values of serum melatonin for these females remained at basal levels (<15 pg/ml). 200 0 DISCUSSION Melatonin Rhythms on Long vs. Short Days 500 900 2300 0300 TIME FIG. 2. Concentrations of melatonin in pineals and serum in females sampled on short days (1OL:14D, 22#{176}C,upper panel) and on long days (16L:8D, 22#{176}C,lower panel). Open symbols indicate mean melatonin content of the pineal; closed symbols denote mean concentrations in serum. Vertical bars above symbols denote SEM (n = 4-5 hamsters per time point). Light cycles are depicted as in Figure 1. Dashed line at the top of each panel is an estimate of duration of melatonin peak in the pineal glands, as explained in the text. ing within 1 mo after transfer to a short-day cold room (1OL:14D, 6#{176}C).Bouts of torpor typically lasted for 4-8 days, interspersed by intervals of spontaneous arousal from torpor which lasted for 1-3 days. During arousal from torpor, body temperature increased from near ambient (7-9#{176}C) to euthermic levels (35-37#{176}C), as measured in a separate group of hibernating females. This increase occurs within 2-4 h, after which time body temperature The present study has demonstrated an effect of daylength on the daily rhythm of pineal melatonin in the Turkish hamster (Mesocricetus brandti). In two replicate experiments, the nocturnal peak of pineal melatonin content was approximately 3 h longer in duration on 1OL:l4Dthanon 16L:8D. The magnitude of this durational difference is not as great as the 6-7 h differences observed for Djungarian hamsters (Phodo pus sungorus) on identical photoperiods (Goldman et al., 1982, 1984). However, in the more closely related Syrian hamster (Mesocricetus auratus), a comparable durational difference (i.e., 2-4 h) is suggested in most of the studies in which rhythms of pineal melatonin have been compared on long vs. short days (Rollag et al., 1980; Brainard et al., 1982; Elliott and Tamarkin, 1982; Stetson and Watson- Whitmyre, 1984). The present results in Turkish hamsters also suggest that the rhythm of pineal melatonin content may reflect the daily pattern of

MELATONIN IN TURKISH HAMSTERS 79 V -J 4 w a- 0 I- 4 -J w m -I 0 Cl, m a, 30 120 I0 1000 1400 1800 2200 0200 0600 TIME FIG. 3. Levels of melatonin in pineal glands (open symbols) and serum (closed symbols) of female hamsters exposed to short-day cold (1OL: 14D; Ta = ambient temperature = 6 ± 1#{176}C).In the upper panel, hamsters were euthermic when sampled (i.e., body temperature = 35-37#{176}C). Circles indicate melatonin levels in hibernating females sampled on the first day after spontaneous arousal from a bout of torpor (n = 6 females per time point). Square symbols indicate levels of melatonin in females that had shown no hibernation despite short-day cold exposure (n = 6). In the lower panel, hamsters were sampled on Day 2 or 3 of a torpor bout (body temperature = 7-9#{176}C). Open triangles denote pineal melatonin in one set of females (n = 4-6). Open and closed vertical bars denote pineal and serum melatonin, respectively, in another set (n = 6).

80 DARROW ET AL. pineal melatonin secretion, since changes in serum melatonin titers paralleled very closely changes in pineal melatonin concentration. A close correlation between levels of serum and pineal melatonin has also been observed in the Djungarian hamster, both for individual animals and for population profiles (Darrow and Goldman, 1986). Furthermore, in Djungarian hamsters, the rhythm of pineal melatonin content parallels the activity of N-acetyltransferase (NAT), an important enyme in the melatonin biosynthetic pathway (Ilinerova et al., 1983). Thus, measurements of pineal NAT, pineal melatonin and serum melatonin all yield similar estimates of the daily pattern of production and secretion of pineal melatonin. A close correlation among the same three measures has also been reported for the rat (Wilkinson et al., 1977), suggesting that this correspondence may hold for rodents in general. The pineal melatonin rhythms described here for Turkish hamsters are comparable to those reported for a number of other mammalian species in which long- vs. short-day melatonin patterns have been studied, e.g., sheep (Arendt, 1979; Bittman et al., 1983), rats (illnerova and Vanecek, 1980), and whitefooted mice (Petterborg et al., 1981; Lynch et al., 1982), as well as Syrian and Djungarian hamsters. All show a strictly nocturnal elevation in pineal or serum melatonin, and most show a direct proportionality between duration of the melatonin peak and length of the night. Despite the similarity in melatonin rhythms, however, these species show a variety of reproductive responses following pinealectomy. The Turkish hamster is the only mammal tested to date in which removal of the pineal gland leads to gonadal regression, even when the animals have been maintained exclusively on a stimulatory daylength (Carter et al., 1982). In contrast, the reproductive system of the closely related Syrian hamster remains activated indefinitely after pinealectomy, regardless of photoperiodic manipulations (Reiter, 1978). It has been suggested that some minimal output of pineal melatonin may be needed to maintain gonadal function in the Turkish hamster, since treatments that abolish the melatonin rhythm, such as pinealectomy or constant light, cause testicular regression (Carter et al., 1982). A progonadal role for the pineal has also been reported for Djungarian hamsters transferred from short to long days (Brackmann and Hoffmann, 1977; Hoffmann, 1977; Carter and Goldman, 1983b). Any comprehensive explanation of the role of the pineal melatonin signal in photoperiodic time measurement will need to account for such species differences in response to pinealectomy. However, with regard to events that are more likely to be encountered by hamsters in the field, such as the transition from long to short days, there may be more similarity among the species. Daily administration of melatonin in the late afternoon to pineal-intact Turkish hamsters causes testicular regression on long days, possibly by adding to, and thus lengthening, the peak of endogenous melatonin (Carter et al., 1982). Similar results have been obtained with timed daily injections or infusions of melatonin in pineal-intact Syrian and Djungarian hamsters (Tamarkin et al., 1976; Goldman et al., 1984). The results with Turkish hamsters were interpreted to suggest that changes in duration of the melatonin peak regulated by photoperiod may be causally involved in the mediation of daylength effects on reproduction in this species, similarly to what has been proposed for Djungarian hamsters and sheep (Bittman et al., 1983; Carter and Goldman, 1983a and b; Goldman et al., 1984). However, it should be noted that some of the results obtained with daily injections of melatonin in Syrian hamsters cannot readily be explained by a hypothesis based on duration of melatonin peak and have been interpreted to indicate a circadian rhythm of sensitivity to melatonin (Stetson and Tay, 1983). In this context, it would be of interest to know whether the duration of daily infusions of melatonin could drive the photoperiodic reponse of pinealectomied Turkish and Syrian hamsters. We have observed that a substantial proportion (3 0-40%) of female Turkish hamsters held on long days become spontaneously acyclic. This loss of estrous cyclicity is not easily explained in terms of an aberrant rhythm in pineal melatonin, since no difference was detected between melatonin profiles of acyclic vs. cyclic females on long days. Similarly, about 40% of the females transferred to short days continued to exhibit estrous cyclicity, yet showed a melatonin profile appropriate to the short day photoperiod. This intra-individual variation in reproductive state may be attributable to an altered responsiveness of target tissue(s) to the pineal melatonin signal. Alternatively, the mixed gonadal responses of Turkish hamsters could derive from individual differences unrelated to photoperiod and/or pineal function.

MELATONIN IN TURKISH HAMSTERS 81 Melatonin Rhythms During Hibernation In hibernating Turkish hamsters, the daily rhythm of pineal melatoni#{241} production and secretion appears to be abolished during torpor, as judged by the low levels of pineal and serum melatonin observed in torpid hamsters sampled at various times during the night. This finding is in agreement with the absence of a pineal melatonin rhythm on the second day of a torpor bout in hibernating Syrian hamsters (Vanecek et al., 1984) and with the absence of a plasma melatonin rhythm in marmots serially sampled during torpor (Florant et al., 1984). However, Vanecek et al. (1984) reported a low amplitude rhythm of pineal melatonin content in Syrian hamsters during the first night of a torpor bout. In some of our torpid hamsters killed during the night, slightly elevated levels of pineal, but not serum melatonin, were observed. This elevation may be due to a retention of hormone which was synthesied before or during entry into torpor and was then only partially secreted by the gland. In support of this possibility, preliminary studies in our laboratory indicate that Turkish hamsters usually enter torpor during the latter half of the night (J. A. Elliott and B. D. Goldman, unpublished results), when melatonin production and secretion have been observed during intervals of arousal. Whether the slight elevation of pineal melatonin we observed in some torpid individuals reflects hormone actually synthesied during torpor, or a residual amount rcmaining from an earlier interval of euthermia, it is clear from the serum measurements that little melatonin is secreted into the peripheral circulation during torpor. ln contrast to the lack of pineal melatonin output during torpor, full-amplitude nocturnal peaks of pineal and serum melatonin were evident within the first day after spontaneous arousal from a bout of torpor. A similar phenomenon has been reported for marmots after arousal from torpor (Florant et al., 1984). Although the 4-h sampling interval utilied in the present study precludes precise characteriation of the phase of the melatonin peaks, the rhythm did not appear to differ substantially from that observed in hamsters housed on 1OL:14D at 22#{176}C. Thus, the phase of the reappearing melatonin rhythm does not seem to have been drastically altered by the prior 4-8-day period of reduced body temperature. These findings indicate that during the hibernation season, the pineal gland is capable of producing an appropriately phased, rhythmic melatonin signal during intervals of arousal, despite the virtual shutdown of the gland during bouts of torpor. It has been proposed that circadian oscillations are relatively unaffected by the low core body temperatures characteriing torpor in mammals, as indicated by the persistence of a low-amplitude circadian rhythm of body temperature in torpid bats (Menaker, 1959) and retention of the rhythm of nocturnal burrow emergence in hibernating doormice (St. Girons, 1965), both under conditions of constant darkness. However, recent data from hibernating Syrian hamsters have been interpreted to indicate that the circadian pacemaker governing the rhythm of pineal melatonin is arrested during torpor (Vanecek et al., 1985). Hamsters that were artificially aroused from torpor, transferred to a warm room, and maintained thereafter in constant darkness showed a peak of pineal melatonin about 12 h later, whether aroused during the light or the dark portion of their previous short-day photoperiod. Interestingly, a second peak of melatonin occurred approximately 24 h after the first peak in each group, suggesting a resetting of the circadian rhythm rather than merely a nonspecific increase of melatonin in response to artificial arousal. It is unlikely that our results in Turkish hamsters were due to a resetting of the pineal rhythm at the time of arousal, since separate studies have indicated that spontaneous arousal from torpor occurs randomly throughout the day and night in Turkish hamsters held on 1OL:14D at 6#{176}CU. A. Elliott and B. D. Goldman, unpublished results). However, our results do not rule out the possibility that circadian oscillators might be arrested during torpor, since our animals were maintained on a light/dark cycle throughout the study. Thus, the extent to which the reappearing melatonin rhythm might have been phased by exposure to the light cycle during torpor and/or following arousal is unknown. In addition to the different species and different lighting conditions used in the present study (compared to those used by Vanacek et al., 1985), it may be of significance that the animals in our study were examined after spontaneous arousal while still in a cold environment, and the Syrian hamsters in the earlier study were aroused by manual stimulation, transferred to a warm environment, and then examined. Thus, it is difficult to know which of these several differences in experimental protocol might have been responsible for the apparent differences in results between the two studies. Further work

82 DARROW ET AL. is need to determine if the circadian pacemaker system of Turkish hamsters is affected by the low body temperature experienced during hibernation. Of interest in this context are results from a separate study in which hibernating female Turkish hamsters were allowed to self-select their exposure to environmental light, i.e., by emerging from a totally darkened simulated burrow into an outer cage exposed to natural changes in daylength (Darrow and Goldman, in preparation). During intervals of spontaneous arousal from bouts of torpor, the animals emerged from the burrow frequently during daylight hours as well as at night. Hence, information about external lighting conditions and time of day was apparently obtained by these animals during their hibernation season. Whether or not the circadian clock governing the pineal gland might stop during torpor, sufficient exposure to environmental light may be available to bring the rhythm of pineal melatonin (and perhaps other physiological rhythms as well) into synchrony with the external 24-h day/night cycle. ACKNOWLEDGMENT We wish to express appreciation to Dr. Mark D. Rollag for supplying the antisera for the melatonin radioimmunoassays. REFERENCES Arendt J, 1979. Radioimmunoassayable melatonin: circulating patterns in man and sheep. Prog Brain Res 52:249-58 Bittman EL, Dempsey RJ, Karsch FJ, 1983. Pineal melatonin secretion drives the reproductive response to daylength in the ewe. Endocrinology 1132276-83 Brackmann M, Hoffmann K, 1977. Pinealectomy and photoperiod influence testicular development in the Djungarian hamster. Naturwissenschaften 64:341-42 Brainard GC, Petterborg U, Richardson BA, Reiter RJ, 1982. Pineal melatonin in Syrian hamsters: circadian and seasonal rhythms in animals maintained under laboratory and natural conditions. Neuroendocrinology 35:342-48 Brown GM, Grota U, Pulido 0, Burns TG, Niles LP, Snieckus V, 1983. Application of immunologic techniques to the study of pineal indolealkylamines. Pineal Res Rev 1:207-46 Carter DS, Goldman BD, 1983a. Antigonadal effects of timed melatonm infusion in pinealectomied male Djungarian hamsters (Phodopus sungor-us sungorus): duration is the critical parameter. Endocrinology 113:1261-67 Carter DS, Goldman BD, 1983b. Progonadal role of the pineal in the Djungarian hamster (Phodopus sungorus sungorus): mediation by melatonin. Endocrinology 113:1268-73 Carter DS, Hall VD, Tamarkin L, Goldman BD, 1982. Pineal is required for testicular maintenance in the Turkish hamster (Mesocricetus brandti). Endocrinology 111:863-71 Darrow JM, Goldman BD, 1986. Circadian regulation of pineal melatonm rhythms and reproduction in the Djungarian hamster. J Biol Rhythms 1:39-54 Elliott JA, Tamarkin L, 1982. Phase relationship of two circadian oscillators regulates pineal melatonin rhythm in Syrian hamsters. Endocrinology 1 1O:A3 26, Abst. 988 Florant CL, Rivera ML, Lawrence AK, Tarnarkin L, 1984. Plasma melatonin concentrations in hibernating marmots: absence of a plasma melatonin rhythm. Am J Physiol 247:R1062-R1066 Goldman BD, Carter DS, Hall VD, Roychoudhury P, Yellon SM, 1982. Physiology of pmneal melatonin in three hamster species. In: Klein DC (ed.), Melatonin Rhythm Generating System, Basel: S. Karger, pp. 210-31 Goldman BD, Darrow JM, 1983. The pineal gland and mammalian photoperiodism. Neuroendocrinology 37:386-96 Goldman BD, Darrow JM, Yogev L, 1984. Effects of timed melatonin infusions on reproductive development in the Djungarian hamster (Phodopus sungorus). Endocrinology 114: 2074-83 Hall V, 1981. Seasonal cycles in gonadal function and hibernation in the Turkish hamster (Mesocricetus brandti). PhD Thesis, Univ. Conn. Hall V, Goldman 1980. Effects of gonadal steroid hormones on hibernation in the Turkish hamster (Mesocricetus brandci). J Comp Physiol 135:107-114 Hoffman K, 1977. Photoperiodic mechanism in hamsters: the participation of the pineal gland. In: Assenmacher I, Farner DS (eds.), Environmental Endocrinology. Berlin: Springer, pp. 94-102 lllnerova H, Vanecek J, 1980. Pineal rhythm in N-acetyltransferase activity in rats under different artificial photoperiods and in natural daylight in the course of a year. Neuroendocrinology 31: 321-26 Illnerova H, Vanecek J, 1985. Complex control of the circadian rhythm in pineal melatonin production. In: Mess B, Ruada C, Tima U, Pevet P (eds.), The Pineal Gland, Current State of Pineal Research. Budapest: Akademiai Kiado, pp. 137-53 lllnerova H, Vanecek J, Hoffmann K, 1983. Regulation of the pineal melatonin concentration in the rat (Rattus norvegicus) and in the Djungarian hamster (Phodopus sungorus). Comp Biochem Physiol 74A:155-59 Karsch FJ, Bittman EL, Foster DL, Goodman RL, Legan SJ, Robinson JE, 1984. Neuroendocrine Basis of Seasonal Reproduction. Rec Prog Horm Res 40:185-232 Lynch GR, Sullivan JK, Heath HW, Tamarkin U, 1982. Daily melatonin rhythms in photoperiod sensitive and insensitive white-footed mice (Peromyscus leucopus). In: Rejter RJ (ed.), The Pineal and Its Hormones. New York: A. R. Liss, Inc., pp. 67-73 Menaker M, 1959. Endogenous rhythms of body temperature in hibernating bats. Nature 184:1251-52 Orsini MW, 1961. The external vaginal phenomena characteriing the stages of the estrous cycle, pregnancy and pseudopregnancy, lactation and the anestrus hamster, Mesocricetus auratus Waterhouse. Proc Anim Care Panel 11:193-206 Petterborg U, Richardson BA, Reiter RJ, 1981. Effect of long or short photoperiod on pineal melatonin content in the white-footed mouse, Peromyscus leucopus. Life Sci 29:1623-27 Reiter RJ, 1978. Interaction of photoperiod, pineal and seasonal reproduction as exemplified by findings in the hamster. In: Reiter RJ (ed.), The Pineal and Reproduction. Karger, Basel, pp. 169-190 Reiter RJ, 1980. The pineal and its hormones in the control of reproduction in mammals. Endocr Rev 1:109-31 Rollag MD, Morgan RJ, Niswender GD, 1978. Route of melatonin secretion in sheep. Endocrinology 102:1-8 Rollag MD, Panke ES, Reiter R, 1980. Pineal melatonin content in male hamsters throughout the seasonal reproductive cycle. Proc Soc Exp Biol Med 165:330-34 Saint Girons MC, 1965. On the persistence of circadian rhythms in hibernating mammals. In: Aschoff J (ed.), Circadian Clocks. Amsterdam: North-Holland Publishing Co., pp. 321-23 Stetson MH, Tay DE, 1983. Time course of sensitivity of golden hamsters to melatonin injections throughout the day. Biol Reprod 29:432-38 Stetson MH, Watson-Whitrnyre M, 1984. Physiology of the pineal and its hormone melatonin in annual reproduction in rodents. In: Reiter RJ (ed.), The Pineal Gland, New York: Raven Press, pp. 109-153 Tamarkin L, Baird CJ, Almeida OFX, 1985. Melatonin: a coordinating signal for mainmalin reproduction? Science 227:714-20 Tamarkin L, Reppert SM, Klein DC, 1979. Regulation of pineal melatonin in the Syrian hamster. Endocrinology 104:385-89 Tamarkin L, Westrom WK, Hamill Al, Goldman BD, 1976. Effect of melatonin on the reproductive systems of male and female Syrian hamsters: a diurnal rhythm of sensitivity to melatonin. Endocrin-

MELATONIN IN TURKISH HAMSTERS 83 ology 99:1534-41 Vanecek J, Jansky U, Illnerova H, Hoffmann K, 1984. Pineal melatonin in hibernating and aroused golden hamsters (Mesocricetus auratus). Comp Biochem Physiol 77A:759-62 Vanecek J, Jansky U, Illnerova H, Hoffmann K, 1985. Arrest of the circadian pacemaker driving the pineal melatonin rhythm in hibernating golden hamsters, Mesocricetus auratus. Comp Biochem Physiol 80A:21-23 Wilkinson M, Arendt J, Bradtke J, deziegler D, 1977. Determination of dark-induced increase of pineal N-acetyltransferase activity and simultaneous radioimmunoassay of melatonin in pineal, serum and pituitary tissue of the male rat. J Endocr 72:243-45