INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

OVER THE COURSE OF THE YEAR, most mammals respond to seasonal changes of day length, i.e., of photoperiod, with altered physiology and behavior. Photoperiod information from the environment is conveyed to the organism by a circadian rhythm of melatonin production in the pineal gland; the production is high during subjective night when nocturnal animals are active and diurnal animals rest and low during subjective day when nocturnal animals rest and diurnal ones are active (15, 16). With longer day lengths, light at an earlier dawn advances the phase of the morning decline of melatonin production, whereas light at a later dusk delays the phase of the evening production rise (14-16). The resulting alteration in the duration of the nocturnal melatonin signal, compressed during long summer days and decompressed during short winter days, appears to serve as an endogenous photoperiodic message (5, 21). After a change from a long to a short photoperiod, decompression of the melatonin signal in rats and hamsters occurs gradually only, mostly into the morning hours (12, 17, 18).

In the rat, rhythmic melatonin production is driven by a circadian rhythm of the activity of pineal N-acetyltransferase, which synthesizes the melatonin precursor N-acetylserotonin (28). The N-acetyltransferase rhythm, as other overt rhythms, is controlled by a circadian pacemaker located in the suprachiasmatic nucleus (SCN) of the hypothalamus (27). The phase and period of the pacemaker are entrained to the natural 24-h day-night cycle by the alternation of environmental light and darkness, mainly by the light part of the day (35). Photic information is conveyed to the SCN by way of a direct retinal projection, the retinohypothalamic tract, and to a lesser extent also by other pathways, namely the geniculohypothalamic tract (10, 32, 33). After a light stimulus, in addition to other processes, immediate early genes, especially c-fos and junB genes, are transcriptionally activated mostly in the retinorecipient, i.e., ventrolateral part of the SCN (29, 38). These genes are believed to function in coupling short-term signals elicited by extracellular events to long-term changes in cellular phenotype by mediating subsequent changes in gene expression (13, 34). After stimulation of a cell and a subsequent transcription of c-fos and junB, corresponding protein products c-Fos and JunB are produced and may, as transcription factors, regulate the transcription of late response genes (29, 37, 38). Presence of c-Fos in the cell may thus serve as a marker of neuronal activation. Importantly, light induces c-fos and junB expression and elevates c-Fos and JunB in the mammalian SCN only during the subjective night, when it also phase shifts circadian rhythmicity (24, 29, 38, 46). The rhythm in the light-elevated c-fos and junB mRNA and c-Fos and JunB protein thus also represents the endogenous rhythm of the SCN sensitivity to light.

The rhythm in c-fos photoinduction in the rat and hamster SCN is photoperiod dependent; duration of the endogenous interval that enables the high c-fos photoinduction is short in long and long in short days (41, 42, 44, 50). The effect of photoperiod is not altered by pinealectomy, indicating that day length affects the functional state of the SCN circadian pacemaking system directly, not via the pineal melatonin signal (40). Similarly as the melatonin signal, the interval enabling the high c-fos photoinduction extends only gradually after a change from a long to a short photoperiod and also mostly into the morning hours (42). All these data indicate that the photoperiod affected pineal melatonin production may mainly reflect the photoperiod modulated state of the SCN circadian pacemaking system (19, 20, 41).

As mentioned earlier, the rhythm in c-fos photoinduction is expressed mostly in the ventrolateral (vl) or ventral part of the SCN, called also a core, which receives direct or indirect photic inputs (10, 30, 32, 33). Besides the vl part, the SCN is composed of a dorsomedial (dm) or a dorsal part called a shell (30, 32). The dm-SCN receives mostly nonphotic input from the cortex, basal forebrain, and hypothalamus (30) and contains many arginine vasopressin (AVP)-immunoreactive cells (22, 47); levels of AVP mRNA and AVP undergo circadian variations rising in the early day (25, 26, 31). Recently, we have found that the dm-SCN exhibits a rhythm in the spontaneous c-Fos immunoreactivity (43). The immunoreactivity is low during the night and starts to increase before the morning light onset. This rhythmicity may represent an elevated dm-neuronal activity during the day.

The aim of the present study was to determine whether the photoperiod affects also the dm-SCN rhythmicity, namely the rhythm in the spontaneous c-Fos immunoreactivity, and if so, how the dm-SCN rhythm adjusts to a change from a long to a short photoperiod. In addition, to characterize further the rhythm of c-Fos immunoreactivity, a question of whether the spontaneous c-Fos production is localized in vasopressin-producing cells in the dm-SCN or in some other cells is tackled as well.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals

Experimental Paradigms

Experiment 1. Rats maintained under LD 12:12 were perfused through the ascending aorta either in the early night or in the early morning; in the latter case, the morning light was not turned on. Alternate brain sections were processed for immunohistochemistry with antibodies against c-Fos, against AVP, and against c-Fos and AVP, respectively, to find out whether c-Fos and AVP are colocalized in the same cells.

Experiment 2. Rats were maintained in LD 16:8 or LD 8:16 for at least 4 wk before the experiment. On the day of the experiment, the morning light was not turned on and animals maintained in darkness for at least the next 24 h were killed every 2 h for c-Fos immunoreactivity determination, always two rats per one time point. For each photoperiod, the killings were done in 1 day and c-Fos immunoreactivity was determined in one assay. For the LD 8:16 photoperiod, the experiment was repeated, but just the afternoon c-Fos decline and the morning rise were followed, and data from both experiments were pooled.

Experiment 3. Rats maintained under a long photoperiod (LD 16:8) were either killed in the morning hours for determination of the morning rise in the spontaneous c-Fos immunoreactivity or were transferred to a short photoperiod (LD 8:16) and killed in the morning hours 3, 6, 13, and 35 days, respectively, after the transfer. On the day when the animals were killed, the morning light was not turned on. Two rats were always killed per one time point. Studies on c-Fos rise in LD 16:8 and 35 days after the transfer to LD 8:16 were repeated three times, studies on c-Fos rise 3, 6, and 13 days after the transfer were repeated two times, and the data were pooled.

Immunocytochemistry

Dual-Labeling Immunocytochemistry

Data Analysis

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Experiment 1. Localization of c-Fos immunoreactivity

View larger version (110K): [in this window] [in a new window]   Fig. 1.   c-Fos (a-f) and arginine vasopressin (AVP; c-f) immunopositive cells in the suprachiasmatic nucleus (SCN). Representative photomicrographs of coronal sections of SCN of rats entrained to 12:12-h light-dark cycle (LD 12:12) and perfused in darkness in the early morning reveal c-Fos cells (brown immunoreaction; a and c) as well as AVP cells (violet immunoreaction; c) in the dorsomedial part of the SCN; b and d are enlarged photomicrographs of a and c, respectively, showing just unilateral SCN with the third ventricle to the left of the image. e And f represent enlarged c-Fos and AVP immunopositive cells, either single (arrows) or double (feathered arrows) labeled in the SCN of rats perfused in darkness in the early morning (e) or in the early night (f).

Experiment 2. Rhythm of c-Fos Immunoreactivity Under a Long and a Short Photoperiod

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Circadian rhythms in the number of c-Fos immunoreactive (ir) cells in the whole SCN (A), in the dorsomedial SCN (B), and in the ventrolateral SCN (C) of rats maintained originally in LD 16:8 () or in LD 8:16 (), released into darkness at the time of the expected dark-light transition and assayed for c-Fos immunoreactivity during the first cycle in darkness. For the whole LD 16:8 profile and for the nighttime minimum under LD 8:16 from 2200 to 0400, each point represents mean ± SD from 2 animals. For the afternoon c-Fos decline from 1200 to 2000 and for the morning c-Fos rise from 0600 to 0800 under LD 8:16, each point represents mean ± SD from 4 animals. Solid bars, original dark periods in LD 16:8 or in LD 8:16.

Under the original LD 16:8 photoperiod, c-Fos immunoreactivity continuously declined from a high daytime value at 1400 to a nighttime level at 2200 and was at its lowest level at 2200, 2400, and 0200 vs. 1400 (P < 0.05); minimum of c-Fos immunoreactivity thus lasted for about 4 h (Fig. 2). The morning rise was indicated already at 0400, i.e., at the time of the usual morning light onset, but was significant only at 0800 vs. 2400 (P < 0.05). Under the original LD 8:16 photoperiod, c-Fos immunoreactivity decreased at about the same time as under LD 16:8, maybe just slightly earlier; c-Fos immunoreactivity declined from a high daytime level at 1400 to a low nighttime level at 2000 and was at its lowest level at 2000, 2200, 2400, 0200, 0400, and 0600 vs. 1400 (P < 0.05); minimum of c-Fos immunoreactivity thus lasted for about 10 h. In the morning, a significant rise occurred at 0800 vs. 0600 (P < 0.05), i.e., at the time of the usual morning light onset, and hence about 4 h later than under LD 16:8 (Fig. 2). At 0200 and at 0600 in the morning, the Student's t-test revealed a significantly lower c-Fos immunoreactivity under LD 8:16 than under LD 16:8 (P < 0.05). Thus the main difference between the rhythm of c-Fos immunoreactivity under a long photoperiod and that under a short photoperiod was in the time of the morning c-Fos rise and, consequently, in duration of the interval when c-Fos immunoreactivity was below a certain level.

Experiment 3. Adjustment of the Rhythm of c-Fos Immunoreactivity to a Change From a Long to a Short Photoperiod

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Adjustment of the spontaneous morning c-Fos rise in the dorsomedial SCN to a change from a long to a short photoperiod. Rats maintained in LD 16:8 were transferred to LD 8:16, and the morning rise in the number of c-Fos immunopositive cells was followed in LD 16:8 () and 3 (), 6 (), 13 (), and 35 () days, respectively, following the transfer. For LD 16:8 and for 35 days after the transfer, each point represents mean ± SE from 6 animals. For 3, 6, and 13 days following the transfer, each point represents mean ± SE from 4 animals. Solid bars, later parts of dark periods in LD 16:8 or in LD 8:16.

Whereas in LD 16:8, a significant morning rise occurred at 0400 vs. 0300 (P < 0.01), after 3, 13, and 35 days in LD 8:16, a significant increase occurred 2 h later, i.e., at 0600 vs. 0300 (P < 0.01, 0.05, and 0.05, respectively), and after 6 days in LD 8:16, a significant rise occurred only at 0700 vs. 0300 (P < 0.01). However, after 3, 6, and 13 days in LD 8:16, but not after 35 days, a c-Fos increase was already indicated at 0500. After 6 days in LD 8:16, a further c-Fos rise occurred at 0700 vs. 0600 (P < 0.01), and after 13 days in LD 8:16, such a rise occurred at 0800 vs. 0700 (P < 0.01); at 0700, the difference between the 6- and 13-day groups was significant (P < 0.01). Also, after 35 days in LD 8:16, a further c-Fos increase occurred at 0800 vs. 0700 (P < 0.01); the 0800 value was, however, lower than that after 13 days in LD 8:16 (P < 0.01).

Altogether, the data indicate that whereas the rough adjustment of the spontaneous morning c-Fos rise to a change from a long to a short photoperiod was completed within 3-6 days after the change, a fine adjustment might continue even after 13 days in LD 8:16. It appears that not just the phase, but also the slope of the morning c-Fos rise may change during the adjustment.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The rhythm of spontaneous c-Fos immunoreactivity detected in darkness in the SCN of rats maintained originally in LD 12:12 (43) was expressed also in rats maintained originally in long or short days. The rhythm is probably due to changes in c-Fos synthesis inasmuch as recently a difference in the rat SCN c-fos mRNA between the subjective day and night was reported (6). From immediate early genes in the rat and hamster SCN, just c-fos and junB, but not fos B and NGFI-A, are expressed rhythmically in the absence of photic input (6-9, 43).

c-fos is spontaneously expressed and c-Fos produced mostly in the dm-SCN (Refs. 6 and 43 and current study). Our data show that spontaneously c-Fos immunoreactive cells were anatomically located in the same SCN region as AVP-producing neurons. However, most of the c-Fos immunoreactive cells were devoid of AVP immunoreactivity and vice versa, just few of vasopressinergic cells contained c-Fos. This held true for the morning cells labeling in darkness as well as for the early night labeling. The results might indicate that the degree of c-Fos and AVP colocalization is not time dependent and suggest that AVP-producing cells in the rat SCN are mostly not identical with those producing c-Fos. In contrast, labeling of vasopressinergic neurons with antibody against a putative mammalian clock protein mPER 1 in the mouse SCN appears to be time dependent; vasopressinergic neurons are devoid of mPER 1 immunoreactivity at the very beginning of the subjective day, when the immunoreactivity is at its minimum; however, at the time of the mPER 1 peak at the beginning of the subjective night, most of the vasopressinergic neurons exhibit nuclear mPER 1 immunoreactivity (11).

The spontaneous morning rise of c-Fos immunoreactivity in the rat SCN, which may indicate an increase in neuronal activity, precedes apparently the daytime rise of mPER 1 mRNA reported for the mouse SCN (11, 45, 54), as well as the morning rise of AVP mRNA and AVP release in the rat and mouse SCN (25, 26, 31, 47). Whereas the cycle of mPER 1 expression is involved in rhythmic transcription of the AVP gene (25), the rhythmic expression of c-Fos and AVP may represent two independent clock-driven rhythms because they are mostly not colocalized in the same cells. Alternatively, the spontaneous morning c-Fos increase might affect the AVP production indirectly, via the transcriptional machinery of the core clockwork.

The rhythm of c-Fos immunoreactivity in the dm-SCN and in the whole SCN in darkness was affected by the previous photoperiod. Although the time of the afternoon and evening c-Fos decline under a long photoperiod differed just slightly from that under a short photoperiod, the morning c-Fos rise occurred later under a short than under a long photoperiod. The dm-SCN rhythm in c-Fos immunoreactivity was thus locked to the original morning light onset, similarly as is the vl-SCN rhythm in c-Fos photoinduction (42, 44) or the pineal rhythm in melatonin production (14-16). Also, the morning c-Fos rise in the dm-SCN under an actual long or short photoperiod occurs spontaneously just before the light onset (23). Due mostly to the difference in the morning c-Fos rise, c-Fos immunoreactivity was spontaneously elevated for a longer time under a long than under a short photoperiod or, in other words, the interval when c-Fos immunoreactivity was low was shorter under a long than under a short photoperiod. The latter interval resembles the interval enabling high c-Fos photoinduction in the vl-SCN (42, 44) as well as the interval when the nocturnal pineal melatonin production is elevated (14-16); duration of all three above-mentioned intervals is by 4-6 h longer under a short than under a long photoperiod. The intervals may represent three different expressions of the subjective night duration, which is conventionally expressed mostly as the period of locomotor activity in nocturnal and of inactivity in diurnal animals. According to these expressions, the subjective night is longer under a short rather than under a long photoperiod.

In the rat, it is the vl-SCN that receives photic inputs, mostly by retinal but also by thalamic terminals and serotonergic afferents from the midbrain raphe (10, 30, 32, 33). Photoperiod affects, however, not just the intrinsic vl-SCN rhythmicity (42, 44), but also as our present results show, the dm-SCN rhythmicity. Recent observations show that the ventral part of the SCN (core) projects densely to the dorsal part (shell) although there is little reciprocal innervation (30). Hence the vl- to dm-SCN projections might transduce information about photoperiod on the dm-SCN neurons and consequently both the SCN subdivisions may respond to the photoperiod.

After a transfer of rats from a long to a short photoperiod, a rough adjustment of the spontaneous morning c-Fos rise in the dm-SCN to the photoperiod change was accomplished within 3-6 days, whereas a fine adjustment might continue even after 13 days following the transfer. Similarly, a rough adjustment of the morning decline in the pineal melatonin production to a change from a long to a short photoperiod is roughly completed within 6 days (16, 18), whereas adjustment of the morning decline in the vl-SCN c-Fos photoinduction may take 13 days (42). The adjustment of the pineal rhythm to short days may be species specific as it lasts considerably more days in hamsters than in rats (12, 17, 18). As changes of the pineal rhythms mostly just reflect changes of the intrinsic SCN rhythmicity (19, 20), adjustment of the latter rhythmicity to short days may be species specific as well.

Our previous findings of the gradual extension of the interval enabling high c-Fos photoinduction following a change from long to short days suggest that memory on long days is stored in the vl-SCN (42); this suggestion is supported by an observation that in vitro stimulation of hypothalamus induces c-Fos in the vl-SCN only during the subjective night (2). The present data on the gradual adjustment of the rhythm in the spontaneous c-Fos immunoreactivity to short days suggest that memory on long days also may be partly stored in the dm-SCN. Alternatively, only the vl-SCN may store the memory on long days and convey the information to the dm-SCN.

In conclusion, the present data show that the intrinsic dm-SCN rhythmicity is affected by the photoperiod as is the vl-SCN rhythmicity. Because the whole SCN rhythmicity is thus photoperiod dependent, all overt rhythms driven by the circadian SCN pacemaking system, and not just the melatonin (14-16, 21) or locomotor activity (4) rhythms, might be photoperiod dependent as well.

Perspectives