Introduction

Rhythmic, daily fluctuations of blood melatonin levels have been demonstrated in a number of avian species in both light:dark (LD) cycles and in constant conditions (Norris, 1981; Cassone, 1983; Underwood et al., 1984; Foa and Menaker, 1988). Significantly, a stable phase relationship between peak melatonin blood levels and peak locomotor activity exists in birds in both LD and in free-running conditions (Norris, 1981; Oshima et al., 1987). Removal of endogenous sources of rhythmic melatonin such as the retinae and/or the pineal gland abolishes locomotor rhythmicity in a number of avian species (Gaston and Menaker, 1968; Gaston, 1971; Ebihara et al., 1984). The implantation of melatonin-filled capsules abolishes locomotor rhythmicity in house sparrows (Turek et al., 1976) and suppresses locomotor rhythms in pigeons (Oshima et al., 1989a). In addition, daily melatonin injections can entrain the perch hopping rhythm in starlings (Gwinner and Benzinger, 1978). Collectively, these data suggest that endogenous melatonin is involved in the regulation of circadian locomotor rhythms in birds (Cassone and Menaker, 1984). On the other hand, there is as yet no direct demonstration that physiological levels of melatonin presented in naturally occurring temporal patterns are sufficient to regulate circadian locomotor rhythmicity.

Until such experiments have been performed it would be premature to conclude that rhythms of endogenous melatonin regulate behavioral rhythmicity in any species. In the present work, pinealectomized (P-X) homing pigeons were used to test the hypothesis that rhythmic infusions of physiological amounts of melatonin can entrain behavioral rhythms in pigeons. Since P-X pigeons still have a significant rhythm of circulating melatonin (Foa and Menaker, 1988) our experimental design depends on overriding this rhythm and shifting its' phase using exogenously infused melatonin.

Materials and Methods

Experimental Birds and Housing Conditions

The homing pigeons (Columba livia; males and females, 400 - 600 grams) used in this experiment were maintained indoors under a light-dark (LD) cycle of 12 hours light: 12 hours dark prior to experimentation. The birds were obtained and housed during experiments as described in Effects of Pinealectomy, Melatonin Implants, and Exposure to Constant Light on Feeding and Locomotor Rhythms in the Pigeon, Columba livia. Behavioral data were collected as described in Effects of Pinealectomy, Melatonin Implants, and Exposure to Constant Light on Feeding and Locomotor Rhythms in the Pigeon, Columba livia.

Surgeries:

Pigeons were P-X as previously described (Ebihara et al,1984). Intact or P-X pigeons held in LD cycles were anesthetized and fitted with either 1) a sub-cutaneous cannula (for delivery of melatonin in vehicle or vehicle alone; [singly cannulated]) or 2) both an intravascular cannula for blood removal and a subcutaneous cannula (doubly cannulated). Occasionally, doubly cannulated pigeons had to be re-cannulated due to a loss of patency in the intravascular cannula. Complete descriptions of the procedures are included in Appendix A.

Infusions:

Infusions of either vehicle (phosphate buffered saline; PBS) or melatonindissolved in vehicle into P-X pigeons was accomplished in the following manner: Doubly or singly cannulated P-X pigeons, entrained to an LD cycle, were then subjected to DD conditions. On the first day in DD, and each day thereafter, the birds received a 10 hour continuous infusion of vehicle or melatonin/vehicle solution. The infusion cycle began either 6 hours before or 6 hours after the onset of behavioral activity (subjective dawn). The timing of the infusions was controlled by a timer adjusted to have a period shorter than, equal to, or longer than 24 hours (T = 23.67 - 24.77). The melatonin dosages infused (0.46, 0.7, 0.93, 1.86 and 3.72 uglhr) were controlled by adjusting either the concentration of the melatonin solution (between 1.5 - 3.0 X 10-2 gIl) or by adjusting the pump rate between 22 and 124 ul/hr. Seventeen pigeons were infused with melatonin; while thirteen were infused with more than one timing pattern or dose, no pigeon was infused with more than 3 different regimes. Infusion of vehicle alone was employed as a control procedure: in 6 of 22 trials vehicle alone was infused prior to melatonin infusion; in the remaining 16 trials vehicle alone was infused after melatonin infusion. Control pigeons receiving vehicle only were infused at the same rates as those receiving melatonin. Since none of the pigeons in either of the two control groups were affected by infusion of vehicle alone, the data were pooled. Overall, twelve birds received vehicle only infusions; while half of these pigeons were infused with more than one timing pattern or dose, only one pigeon was infused with as many as four.

Blood Samples:

Singly cannulated pigeons. Blood samples were collected from individually housed intact, P-X and P.X/melatonin infused pigeons by brachial vein venipuncture as previously described (Foa and Menaker, 1988). Sampling was performed with the aid of an infra-red viewer after birds had been in DD for at least 24 hours. A series of 4 blood samples was drawn in a 24 hour period (one every six hours) and at least one week elapsed between each series of samples. Blood was also sampled every two hours in four intact pigeons. Since these data were highly variable they were not included in the data presented in this Chapter (see Appendix B). Blood samples were spun in a centrifuge at 4 C for 15 minutes at 2900 g. 100 ul serum samples were then aliquotted into borosilicate glass tubes and stored at -70 C until extraction and radioimmunoassay for melatonin.

Doubly cannulated pigeons - Blood samples were collected from individually housed intact, P-X and P-X/melatonin infused pigeons with the use of an indwelling cannula. These samples were taken after the pigeons had been exposed to at least 24 hours of DD. The standardized procedure employed is described in Appendix A A series of 12 blood samples was drawn in a 24 hour period (one sample every two hours). 400 ul blood samples were placed into borosilicate glass tubes and spun, aliquotted and frozen as above. Periodically, the whole blood fractions were re-suspended in heparinized PBS and reintroduced into the pigeons.

Extraction, Radioimmunoassay and Validation:

The melatonin in the serum and plasma samples were extracted with chloroform within two weeks of sampling and a radioimmunoassay for melatonin was performed on the extracted samples as previously described (Foa and Menaker, 1988). Known quantities of melatonin in PBS were also extracted and/or assayed and these values were used to correct for extraction efficiency and as inter-assay controls. The average extraction efficiency (10 extractions) was 82%. Intra-assay coefficient of variation (CV) for melatonin standards (40 pg/100 ul) in PBS was 6%. (10 assays). The inter-assay CV for the melatonin standards was 21% and the limit of detection was 37 pg/ml (for 100 ul samples). The values for 20%, 50%, and 80% binding on standard curves were 1679 + 101, 341 + 21, and 70 + 5 pg/ml respectively. This assay (R1055) was previously validated for specificity of pigeon serum melatonin using high performance liquid chromatography and for parallelism of inhibition curves of pigeon serum (Foa and Menaker, 1988).

Data Analysis

Behavioral records were used in the determination of entrainment only if the pigeons were allowed to free-run after the infusion cycle. By this criterion, the behavioral results from several (n=8) pigeons receiving melatonin infusions were excluded since, although their behaviors were synchronized to the infusion cycle, entrainment was not demonstrated (ie. - no free-run). The periods of free-running rhythmicity after entrainment were calculated by drawing an best eye-fit line through the offsets of feeding behavior and calculating the slope of the line (feeding behavior was always used since the feeding offsets were clearer than locomotor offsets). The phase angles of entrainment to melatonin infusion T cycles were determined from feeding records of pigeons that entrained clearly for at least 10 days. To determine phase angles, a best eye-fit line was drawn through the offsets of the last 10 days of entrainment. The time differential between this behavioral offset and the onset of melatonin infusion was recorded. Blood melatonin levels of P-X birds were included in the analysis only if the behavior was rhythmic enough to allow measurement of behavioral offset (22/24 trials; n = 17). Three out of 32 infused pigeons had blood melatonin levels that indicated they were not receiving exogenous melatonin. These were not used in the analysis. The significance of differences between means was determined using Student's t test (p < 0.05).

Results

I. Behavioral Responses to Melatonin Infusion Cycles

Representative feeding records from P-X pigeons entrained to LD cycles and then subjected to melatonin infusion T cycles in DD are presented in Figure la-e. Circadian rhythmicity, as measured by feeding activity, synchronized to daily infusions of melatonin delivered cyclically in the circadian range (23.67 < T < 24.77). When the melatonin infusions were terminated the feeding activities free-ran with periods different than that of the previous melatonin infusion cycle even though vehicle infusion was continued with the same period and phase (Figure 1 a,b,c,e). Entrainment was seen in all 29 trials in which pigeons were given melatonin infusions followed by vehicle infusion or no infusion (Table 1).

In Figure 2a-d are shown feeding and locomotor records from P-X pigeons entrained to LD cycles and then exposed to melatonin infusion cycles in DD. Both feeding and locomotor activity synchronized to daily infusions of melatonin in these and all other pigeons that received melatonin infusions. These behaviors remained synchronized as long as the melatonin infusion cycle was continued. [In one case, the feeding activity (locomotor activity not measured) became desynchronized from the melatonin infusion cycle (T = 24.65) for 20 days; the feeding activity of this bird later re-entrained to the infusion cycle.] When the melatonin cycle was terminated both behavioral activities free-ran with virtually identical periods (feeding; range - 23.33 - 24.56, x = 23.98 + 0.09; locomotor; range - 23.30 - 24.65, x = 23.98 + 0.09). Although we did not conduct controlled experiments to test for after-effects, the free-running periods of the behavioral rhythms did not appear to be consistently related to the period (T cycle) of the previous melatonin infusion. Several melatonin doses (0.46 ug - 3.6 ugfhr) and T cycles were all able to entrain or synchronize feeding and locomotor activities. The average phase angle of entrainment (length of time between feeding offset and melatonin infusion onset) in different melatonin T cycles was -1.63 + 0.22 hrs (T < 24 hrs, n=3), -0.14 + 0.43 (T = 24, n=7), and 2.53 + 0.44 hrs (T > 24 hrs, n = 11). Although locomotor rhythmicity was often less clear than feeding rhythmicity during melatonin infusion, the effects of melatonin infusion on these two rhythms were indistinguishable.

Typical feeding records of pigeons receiving vehicle infusions on the first day in DD are presented in Figure 3a-c. Neither feeding nor locomotor (data not shown) activities synchronized to vehicle infusions whether the infusions were delivered to the pigeon on the first day in DD (Figure 3a-c - top box) or several weeks later (Figure 3a-c - bottom box) after entrainment to melatonin infusions (Figure 3a-c - middle box). Saline vehicle infusion was not an effective entraining agent at any period (Table 1).

II. Blood melatonin levels in Intact, P-X and P-X/Melatonin

infused pigeons:

Six hour samples
Representative melatonin profiles of blood samples taken via venipuncture from an individual pigeon - intact, following P-X, and with subsequent melatonin infusion are presented in Figure 4 (top). All of the profiles exhibited rhythmic changes in absolute melatonin levels over the course of a 24 hour period. In the case shown, P-X reduced the amplitude of the melatonin rhythm and melatonin infusion restored it to its' normal amplitude and shape (Figure 4, top). While the amplitudes of both intact and P-X/melatonin infused profiles were nearly identical, the amplitude of the P-X profile was suppressed. The peaks (maximal levels) of blood melatonin were always approximately 180 degrees out of phase with the peaks (major bouts) of behavioral activity in P-X/melatonin infused birds (Figures la-d, 2a-d, 3a,b). The peaks of blood melatonin and behavioral activity in P-X and intact birds were always antiphasic as well (data not shown). The average values of the blood melatonin content of samples from pineal intact, P-X and P-X/melatonin infused birds are presented in Figure 4 (bottom). The same trends that were apparent in individual pigeons (Figure 4, top) were also apparent among the population of birds sampled. The differences between the peak and trough values were significant for intact, P-x, and P X/melatonin infused pigeons (p < .001). P-X peak levels were on average lower than intact levels although the differences were not significant. The peak levels of P-X/melatonin infused individuals for this dose (0.93 uglhr) were slightly higher than the peak intact levels (p < .05). The peak blood melatonin levels of P-X/melatonin infused for lower doses (0.70, 0.46 uglhr) were not significantly different than peak intact levels. The peak levels of circulating melatonin in both intact and P-X pigeons are in close agreement with those previously published by Foa and Menaker (1988).

Two Hour Samples

Typical blood melatonin profiles from samples taken at two hour intervals from intact birds, the same birds after P-X and during P-X/melatonin infusion (0.70 uglhr) are presented in Figure 5. These results (obtained by cannulation) closely parallel those obtained at six hour intervals via venipuncture (Figure 4). Significant differences were found between the peaks and troughs of intact (p < .01) P-X (p < .001) and P-X/melatonin infused (p > .001) pigeons. Peak P-X levels were on average lower than peak intact levels although the differences were not significant. The differences between intact and P-X/melatonin infused peak levels were not significant.

The average peak blood melatonin levels of P-X pigeons receiving measured melatonin infusions are presented in Figure 6. Although two different methods and frequencies of blood sampling were used (Figure 6 top versus bottom), the measured blood levels are similar for the two methods for the range of doses used (0.46 - 3.8 ug/hr).

Discussion

Our results indicate that feeding and locomotor rhythms of the pigeon can be entrained with cyclic melatonin infusions in the circadian range. In this study, cyclical melatonin infusions were effective in entraining the behavioral rhythms in all 29 trials in which pigeons were subjected to melatonin infusions and allowed to free-run to confirm entrainment. In all but one trial, the behavioral rhythms of these pigeons maintained a stable relationship to the melatonin infusion cycle until it was terminated. Vehicle infusions alone were not sufficient to entrain behavioral activity in any case demonstrating that melatonin was the important timing signal in these infusions. Gwinner and Benzinger (1978) have previously demonstrated locomotor activity entrainment of starlings to daily melatonin injections. Since blood melatonin levels after injection were not determined in this study, the physiological significance of these melatonin injections was not demonstrated. While melatonin has long been hypothesized to be centrally important to the circadian system of birds, our results are the first direct demonstration in any avian species that physiologically significant amounts of exogenous melatonin delivered in normal temporal patterns are sufficient to entrain behavioral rhythms.

Our results also demonstrate that feeding and locomotor rhythms of this species are affected by melatonin infusions in a similar manner. Melatonin infusions generally entrain both feeding and locomotor rhythms (Figure 1a-d, Table 1). These findings parallel results from Effects of Pinealectomy, Melatonin Implants, and Exposure to Constant Light on Feeding and Locomotor Rhythms in the Pigeon, Columba livia in which these two behavioral activities were affected in a similar manner by P-x, exposure to various intensities of LL, and melatonin capsule implantation.

Our results confirm Foa and Menaker's (1988) results that a large amplitude blood melatonin rhythm persists in P-X pigeons. Underwood et aI., (1988) report that the eyes of Japanese quail contain circadian oscillators and Foa and Menaker (1988) have demonstrated that the eyes are the source of rhythmic blood melatonin in P-X pigeons. We did not observe dual, or especially broad, peaks of melatonin in P-X/melatonin infused pigeons - as might be expected if the exogenous (infused) and the endogenous peak (from the retinae) remained completely independent of one another. This suggests that the rhythm of retinal melatonin was entrained by the melatonin infusion cycle. Because blood was drawn from P-X/melatonin infused pigeons a few weeks after the initiation of the melatonin infusion regimen it is possible that there were dual or especially broad melatonin peaks for some portion of the period between the start of melatonin infusion and the time of our measurements. Indeed it is likely that this would occur prior to the steady state entrainment of the retinal melatonin rhythm. Furthermore, the sloppiness of the behavioral rhythms occasionally observed during this time (Figure 1b,2a,b) may have been due to this phenomenon.

Our data show that feeding and locomotor rhythms in pigeons entrain to melatonin infusion cycles. There is also a trend indicating differences in the measured phase angles of entrainment to different T cycles of melatonin infusion. These results suggest that the system or structure upon which melatonin acts in pigeons is an oscillator (Pittendrigh and Daan, 1976a). A synthesis of the results from a number of experiments on avian species has led Cassone and Menaker (1984) to suggest that the avian suprachiasmatic nuclei are damped oscillators that are dependant on rhythmic input, perhaps from melatonin, to maintain behavioral rhythmicity. In the pigeon, there is evidence for an extra-pineal, extra-retinal oscillator since P-XlE-X pigeons entrain to an LD cycle and locomotor rhythmicity persists for several cycles in DD (Ebihara et al., 1984). However, results from hypothalamic lesion studies (Ebihara et al., 1987) have not enabled the identification of the site(s) of this oscillator.

The study reported here greatly strengthens the argument that melatonin is an integral part of the pigeon circadian system that regulates behavioral rhythms. The direct demonstration that the circadian behavior of pigeons can be entrained by physiological amounts of infused melatonin in temporally normal patterns strongly supports the hypothesis that endogenous melatonin rhythms regulate the circadian rhythms of feeding and locomotion in intact pigeons. While there is indirect evidence suggesting that melatonin plays a similar role in the circadian systems of other avian species, direct confirmation awaits future studies.