Expression of Clock and Clock-Driven Genes in the Rat Suprachiasmatic Nucleus during Late Fetal and Early Postnatal Development

Transcription

1 Expression of Clock and Clock-Driven Genes in the Rat Suprachiasmatic Nucleus during Late Fetal and Early Postnatal Development Zuzana Kováčiková, 1 Martin Sládek, 1 Zdenka Bendová, Helena Illnerová, and Alena Sumová 2 Institute of Physiology, Academy of Sciences of the Czech Republic, Prague, Czech Republic Abstract The SCN as a site of the circadian clock itself exhibits rhythmicity. A molecular clockwork responsible for the rhythmicity consists of clock genes and their negative and positive transcriptional-translational feedback loops. The authors previous work showed that rhythms in clock gene expression in the rat SCN were not yet detectable at embryonic day (E) 19 but were already present at postnatal day (P) 3. The aim of the present study was to elucidate when during the interval E19-P3 the rhythms start to develop in clock gene expression and in clock-controlled, namely in arginine-vasopressin (AVP), gene expression. Daily profiles of Per1, Per2, Cry1, Bmal1, and Clock mrna and of AVP heteronuclear (hn) RNA as an indicator of AVP gene transcription were assessed in the SCN of fetuses at E20 and of newborn rats at P1 and P2 by the in situ hybridization method. At E20, formation of a rhythm in Per1 expression was indicated, but no rhythms in expression of other clock genes or of the AVP gene were detected. At P1, rhythms in Per1, Bmal1, and AVP and a forming rhythm in Per2 but no rhythm in Cry1 expression were present in the SCN. The Per1 mrna rhythm was, however, only slightly pronounced. The Bmal1 mrna rhythm, although pronounced, exhibited still an atypical shape. Only the AVP hnrna rhythm resembled that of adult rats. At P2, marked rhythms of Per1, Per2, and Bmal1 and a forming rhythm of Cry1, but not of Clock, expression were present. The data suggest that rhythms in clock gene expression for the most part develop postnatally and that other mechanisms besides the core clockwork might be involved in the generation of the rhythmic AVP gene expression in the rat SCN during early ontogenesis. Key words circadian clock, suprachiasmatic nucleus, clock genes, arginine-vasopressin, ontogenesis, rat All mammals exhibit daily rhythms at various levels, ranging from the molecular to the behavioral. These rhythms persist even in a nonperiodic environment with a period close to 24 h and are entrained to the 24-h day mostly by the LD cycle (Pittendrigh, 1981). The circadian rhythms are controlled by a clock located in the SCN of the hypothalamus (Klein et al., 1991). The SCN itself exhibits rhythms in the uptake of 2-deoxyglucose, a marker of metabolic activity (Schwartz, 1991), in electrical activity (Gillette, 1991), 1. These authors contributed equally to the study. 2. To whom all correspondence should be addressed: Alena Sumová, Institute of Physiology, Academy of Sciences of the Czech Republic, Vídeňská 1083, Prague 4, Czech Republic; sumova@biomed.cas.cz. JOURNAL OF BIOLOGICAL RHYTHMS, Vol. 21 No. 2, April DOI: / Sage Publications 140

2 Kováčiková et al. / SCN CLOCK RHYTHMICITY DURING ONTOGENY 141 in spontaneous as well as a light-induced expression of immediate early genes namely c-fos, a marker of neuronal activity (Kornhauser et al., 1993; Schwartz et al., 1995; Sumová et al., 1998), in the production of many peptides for instance, of arginine-vasopressin (AVP) (van den Pol, 1991; Yambe et al., 2002), and other rhythms. This SCN rhythmicity is due to SCN molecular clockwork (for review, see King and Takahashi, 2000; Reppert and Weaver, 2001; Fu and Lee, 2003). A set of mammalian clock genes, namely 3 period genes (Per1, -2, and -3), 2 cryptochrome genes (Cry1 and -2), Clock, Bmal1, casein kinase 1 epsilon (CK1ε), and Rev-erbα, are part of the clockwork. With the exception of Clock and CK1ε, all these genes are expressed in a rhythmic way; the rhythmic expression of Bmal1 is in antiphase to that of Per and Cry genes. Clock genes are thought to be involved in the core clockwork by forming interacting negative and positive transcriptionaltranslational feedback loops. The mammalian SCN develops gradually (Moore, 1991). In the rat, formation of the SCN begins on embryonic day (E) 14 and continues through E17 from the specialized zone of the ventral diencephalic germinal epithelium as a component of periventricular cell groups. Synaptogenesis in the SCN progresses slowly in the late prenatal and early postnatal periods and then increases noticeably from postnatal day (P) 4 to P10. Rhythms in the SCN may appear as early as in the late embryonic stage. A day-night variation of metabolic activity monitored by a 2-deoxyglucose uptake was detected earlier in the fetal rat SCN from E19 through E21 (Reppert and Schwartz, 1984), of the AVP mrna level at E21 (Reppert and Uhl, 1987), and in the firing rate of the SCN neurons at E22 (Shibata and Moore, 1987). Although all the above-mentioned studies indicate the presence of overt rhythms in the SCN of rat fetuses, our recent work did not reveal detectable rhythms in the expression of clock genes and of their products at E19. However, the rhythms in Per1, Per2, Cry1, and Bmal1, but not in Clock mrna, were expressed at P3 in the SCN (Sládek et al., 2004). The aim of the present study was to map the interval between E19 and P3, that is, to elucidate when during the late embryonic and early postnatal stages the expression of various clock genes in the SCN began to be rhythmic. Daily profiles of Per1, Per2, Cry1, Bmal1, and Clock mrna were determined in the SCN at E20, P1, and P2. Moreover, to compare development of core clock gene expression with that of clock-controlled gene transcription, the SCN daily profile of AVP heteronuclear (hn) RNA was studied both at E20 and P1. The AVP hnrna as an indicator of the AVP gene transcription (Yambe et al., 2002) was chosen as a marker of an overt SCN rhythmicity for its close connection to the core clock mechanism. CLOCK-BMAL1 heterodimers act through an E-box enhancer in the vasopressin gene to activate its transcription (Jin et al., 1999). To detect the AVP transcript levels by in situ hybridization, a probe complementary to part of the intronic AVP sequence was used instead of the exonic sequence. This approach enabled detection of nascent transcripts or hnrna and thus of a pure transcription of the AVP gene not compromised by a potential AVP mrna instability (Yambe et al., 2002). Animals MATERIALS AND METHODS Female Wistar rats (Bio Test s.r.o., Konarovice, Czech Republic) were maintained under an LD cycle with 12 h of light and 12 h of darkness per day (LD12:12) in a temperature of 23 ± 2 C and with free access to food and water for at least 2 months. Light was provided by overhead 40-W fluorescent tubes, and illumination was between 50 and 200 lux, depending on cage position in the animal room. The day the rats were found to be sperm positive was designated embryonic day 0 (E0). Birth occurred on average at E21.5. We used day E20 and not E21 for prenatal studies so as not to kill pregnant rats during delivery. At E20, the morning light was not turned on, and mothers were released into constant darkness. Starting at expected lights-on, a single mother was decapitated every 2 h during the circadian cycle, and 4 fetuses for expression of 3 clock genes and 4 fetuses for expression of another 2 clock genes and an AVP gene per each time point were sampled. Day of delivery was designated the postnatal day 0 (P0). Pups born during the night were sampled during the next night (P1) or 2 nights later (P2); pups born during the day were sampled during the following day (P1) or 2 days later (P2). On the day of sampling, mothers with their pups were released into constant darkness; that is, the morning light was not turned on. At P1, 4 pups for clock gene expression and 4 pups for AVP gene expression from 1 litter and at P2, only 4 pups for clock gene expression from 1 litter per each time point were sampled in darkness every 2 h throughout the whole circadian cycle. The times for expected lights-on and lights-off were designated CT0 and CT12, respectively.

3 142 JOURNAL OF BIOLOGICAL RHYTHMS / April 2006 Fetuses and pups were killed by rapid decapitation. The brains were removed, immediately frozen on dry ice, and stored at 80 C. Pup brains were sectioned into 5 and/or 1 series (at P1) and 5 series (at P2) of 12-µm-thick slices for each brain in alternating order throughout the whole rostrocaudal extent of the SCN and processed for in situ hybridization to determine levels of Per1, Per2, Cry1, Bmal1, and Clock mrnas and AVP hnrna. The fetal brains were cut into 3 series of 12-mm-thick slices for each brain and processed for in situ hybridization as well to determine levels of Per1, Per2, Cry1, Bmal1, and Clock mrnas and AVP hnrna. For all ages, each gene expression at each time point was determined in 4 of the brains. All experiments were conducted under license no. A with the U.S. National Institutes of Health and in accordance with the Animal Protection Law of the Czech Republic (license no. 1020/491/A/00). In Situ Hybridization Histochemistry The cdna fragments of rat rper1 (980 bp; corresponds to nucleotides of the sequence in GenBank accession no. AB002108), rat rper2 (1512 bp; corresponds to nucleotides of the sequence in GenBank accession no. NM ), rat rbmal1 (841 bp; identical to nucleotides of the sequence in GenBank accession no. AB012600), rat rclock (1158 bp; identical to nucleotides of the sequence in GenBank accession no. AB019258), mouse mcry1 (719 bp; corresponds to nucleotides of the sequence in GenBank accession no. NM ), and rat ravp (506 bp; identical to nucleotides of the intronic sequence in GenBank accession no. X01637) were used as templates for in vitro transcription of complementary RNA probes. The rper1, rper2, and mcry1 fragmentcontaining vectors were generously donated by Professor H. Okamura (Kobe University School of Medicine, Japan), and rbmal1, rclock, and ravp were cloned in our laboratory. Briefly, cdna fragments were yielded from the rat hypothalamic total RNA. After reverse transcription, cdna was amplified by standard polymerase chain reaction and ligated into vector pgem-t-easy and pbluescript, respectively. For AVP probe, primers designed to amplify intronic region were used. The cloned inserts were sequenced to verify the amplified products. The probes were labeled by using α-[ 35 S]thio-UTP, and the in situ hybridization was performed as described previously (Shearman et al., 2000; Sládek et al., 2004). The sections were hybridized for 20 h at 60 C (Per1, Per2, Clock, and AVP), 58 C (Bmal1), and 55 C (Cry1). Following a posthybridization wash, the sections were dehydrated in ethanol and air dried. Finally, the slides were exposed to film BIOMAX MR (Kodak) for 8 to 10 days and developed using the developer Fomatol LQN and fixer FOMAFIX (FOMA, Hradec Králové, Czech Republic). As a control, in situ hybridization was performed in parallel with sense probes (apart from Per2 and AVP) on sections containing the SCN. For each age, the whole daily profile of a clock gene expression was measured using the same labeled probe. All sections hybridized with the probe were processed simultaneously under identical conditions. For the determination of AVP gene expression at E20 and P1, the same labeled probe was used for both ages. Therefore, only AVP hnrna profiles at E20 and P1 were compared by 2-way analysis of variance (ANOVA); in all other profiles for each age, only 1-way ANOVA was used to compare absolute values within the profile. Autoradiographs of sections were analyzed using an image analysis system (ImagePro, Olympus, New Hyde Park, NY) to detect the relative optical density (OD) of the specific hybridization signal. In each animal, the mrna or hnrna level was quantified bilaterally, always at the midcaudal SCN section containing the strongest hybridization signal. Each measurement was corrected for a nonspecific background by subtracting the OD values from the same adjacent area in the hypothalamus. This area was expected to be free of the specific signal and thus served as an internal standard. The background signal of the area was consistently low and did not exhibit marked changes with age or the time of day. Finally, slides were counterstained with cresyl violet to check the presence and the midcaudal position of the SCN in each section. In no case did in situ hybridization yield any specific signal using a sense probe. Data were expressed as a mean of OD from 4 animals ± SEM; the OD for each animal was calculated as a mean of the left and right SCN values. Statistical Analysis Data on clock gene mrna profiles were analyzed by 1-way ANOVA for the time difference. Data on AVP hnrna were analyzed by 2-way ANOVA for age and time differences and by 1-way ANOVA for only time differences. Subsequently, the Student- Newman-Keuls multiple range test was used, with p < 0.05 being required for significance. A rhythm

4 Kováčiková et al. / SCN CLOCK RHYTHMICITY DURING ONTOGENY 143 was considered when the 1-way ANOVA revealed a significant effect of time and at the same time the maximum and minimum values clustered into 2 separate, roughly out-of-phase time intervals. The intervals included at least 2 successive maximum and minimum values, respectively. RESULTS Daily Profiles of Clock Gene mrnas Figure 1 shows representative in situ hybridization studies of Per1, Per2, Cry1, and Bmal1 mrna in the SCN of 20-day-old fetuses and 1- and 2-day-old rats as well as Clock mrna in the SCN of 20-day-old fetuses and 2-day-old rats at CT8 (i.e., during the subjective day) and CT16 (i.e., during the subjective night). From these and other similar autoradiographs, relative OD levels, that is, relative mrna amounts of the above mentioned clock genes, were estimated. In 20-day-old fetuses, the 1-way ANOVA revealed a significant effect of time for Per1, Per2, and Bmal1 (p < 0.01) (Fig. 2A, 2D, 2J) and Cry1 (p < 0.05) (Fig. 2G) but not for Clock (Fig. 2M). Per1 mrna at CT2 was significantly higher than all other values except those at CT0 and CT4 (Fig. 2A). Elevated levels might thus begin to cluster in the CT0-4 interval. Per2 mrna levels at CT4, 8, and 12 were significantly elevated as compared with the value at CT10 (Fig. 2D). Neither Figure 1. Representative autoradiographs of coronal brain sections at the level of the SCN; Per1, Per2, Cry1, and Bmal1 mrnas in 20-day-old fetuses (E20) and in 1-day-old (P1) and 2-day-old (P2) rats, and Clock mrna at E20 and P2 were examined at CT8 or CT16 by in situ hybridization. the maximum nor the minimum levels thus clustered into separate intervals. The same holds true for the Cry1 (Fig. 2G) and Bmal1 (Fig. 2J) mrna profiles. The Cry1 mrna level at CT10 was significantly higher than those at CT14, 20, and 24. The Bmal1 mrna level at CT22 was significantly elevated when compared with those at CT2, 6, and 20 and the level at CT10 was significantly higher than those at CT6 and 20. Hence, at E20 only, forming of an initial rhythm in Per1 mrna expression was indicated. In 1-day-old rats, the 1-way ANOVA also revealed a significant effect of time for Per1, Per2, and Bmal1 (p < 0.01) (Fig. 2B, E, K) and Cry1 (p < 0.05) (Fig. 2H). Per1 mrna levels at CT7, 8, and 9 were significantly elevated as compared with those at CT16, 22, and 23 (Fig. 2B). Hence, elevated and also low values clustered into 2 separate, roughly out-of-phase time intervals. Per2 mrna levels at CT7, 8, 9, and 13 were significantly higher than that at CT22 (Fig. 2E). The maximum level at CT9 was significantly higher than levels at CT15, 19, and 22 but not those at CT16, 17, or 21. Although elevated values fell into 1 interval, this was not the case for low levels. For the Cry1 mrna profile, the post hoc analysis did not reveal significant differences among various time points (Fig. 2H). Bmal1 mrna levels at CT16, 17, 19, 21, 22, and 23 were significantly elevated when compared with those at CT7, 8, 11, and 13, but they did not differ from that at CT9 (Fig. 2K). The elevated and also low values thus clustered into 2 separate time intervals, with the exception of the CT9 value. It appears that at P1, a slight rhythm in Per1 expression, a pronounced rhythm in Bmal1 expression, and a forming rhythm in Per2 expression were already present. In 2-day-old rats, the 1-way ANOVA revealed a significant effect of time for Per1, Per2, Cry1, and Bmal1 mrna profiles (p < 0.01) (Fig. 2C, 2F, 2I, 2L) but not for the Clock mrna profile (Fig. 2N). Per1 mrna levels at CT2, 4, 6, and 8 were significantly higher than those at CT16, 18, 20, and 22 (Fig. 2C). Hence, the elevated values as well as low ones fell into 2 separate time intervals of almost polar opposites. Per2 mrna levels at CT6, 8, 10, and 12 were significantly elevated as compared with those at CT18, 20, 22, and 24 (Fig. 2F). The maximum and minimum values clustered into 2 separate, roughly out-of-phase time intervals, as was the case with Per1 mrna. Cry1 mrna levels at CT10 and CT14 were significantly higher than the minimum value at CT22 (Fig. 2I). Elevated values might thus start to cluster into 1 time interval. Bmal1 mrna levels at CT16 and CT18 were significantly elevated when compared

5 144 JOURNAL OF BIOLOGICAL RHYTHMS / April 2006 Figure 2. Daily profiles of Per1 (A,B,C), Per2 (D,E,F), Cry1 (G,H,I), Bmal1 (J,K,L), and Clock (M,N) mrna levels in the SCN of 20-dayold fetuses (E20; A,D,G,J,M) and 1-day-old (P1; B,E,H,K) and 2-day-old (P2; C,F,I,L,N) rats maintained in LD12:12 and released into darkness at the time of the expected DL transition (CT0). The brain sections were assayed for mrna by in situ hybridization. Data are expressed as mean ± SEM from 4 animals. Per1 mrna: on E20, the level at CT2 was significantly higher than all other levels (p < 0.05) except those at CT0 and CT4; on P1, levels at CT7, 8, and 9 were higher than those at CT16, 22, and 23 (p < 0.05); on P2, levels at CT2, 4, 6, and 8 were higher than those at CT16, 18, 20, and 22 (p < 0.05). Per2 mrna: on E20, levels at CT4, 8, and 12 were significantly higher than that at CT10 (p < 0.05); on P1, levels at CT7, 8, 9, and 13 were higher than that at CT22 (p < 0.05); and the maximum level at CT9 was higher than levels at CT15, 19, and 22 (p < 0.01). Cry1 mrna: on E20, the level at CT10 was significantly higher than those at CT14, 20, and 24 (p < 0.01); on P2, levels at CT10 and CT14 were higher than that at CT22 (p < 0.05). Bmal1 mrna: on E20, the level at CT22 was higher than those at CT6 and CT20 (p < 0.01); on P1, levels at CT16, 17, 19, 21, and 23 were higher than those at CT7, 8, 11, and 13 (p < 0.01); on P2, levels at CT16 and CT18 were higher than those at CT8 and CT10 (p < 0.05).

6 Kováčiková et al. / SCN CLOCK RHYTHMICITY DURING ONTOGENY 145 Figure 3. AVP heteronuclear (hn) RNA levels in 20-day-old fetuses (E20) and in 1-day-old rats (P1) maintained in LD12:12 and sampled in darkness either at CT2 (i.e., during the subjective day) or at CT16 (i.e., during the subjective night). Representative coronal brain sections at the level of the SCN examined by in situ hybridization are depicted. Note that in contrast to the SCN, the signal in supraoptic nuclei does not vary with time. with those at CT8 and CT10 (Fig. 2L). Elevated levels and the low ones thus clustered into 2 separate time intervals. The Clock mrna profile did not show any rhythm whatsoever. It appears that at P2, rhythms in Per1, Per2, and Bmal1 expression were fully present, while the rhythm in Cry1 expression might only start to form. Daily Profiles of AVP hnrna Figure 3 shows representative in situ hybridization studies of AVP hnrna in the SCN of 20-day-old fetuses and 1-day-old rats at CT2 during the subjective day and at CT16 during the subjective night. From these and other similar autoradiographs, relative OD levels, that is, relative AVP hnrna amounts, were estimated. The daily profile of AVP hnrna of 20-day-old fetuses (Fig. 4A) was compared with that of 1-dayold rats (Fig. 4B). The 2-way ANOVA revealed a significant effect of age (F = 172.1, p < 0.01), effect of time (F = 6.4, p < 0.01), as well as a significant interaction effect (F = 4.1, p < 0.01). The AVP hnrna profile of 20-day-old fetuses thus differed significantly from that of 1-day-old rats. In 20-day-old fetuses, the 1-way ANOVA did not reveal a significant effect of time, and hence no circadian rhythm in AVP hnrna expression was detected. In 1-day-old rats, the effect of time was by now highly significant (p < 0.01). The AVP hnrna levels at CT2 and again at CT22 were significantly higher than those at CT10, 12, 14, 16, 18, and 20, while levels at CT0, 4, 6, 8, and 24 did not differ significantly from that at CT22. Elevated AVP Figure 4. Daily profiles of AVP heteronuclear (hn) RNA in the SCN of 20-day-old fetuses (A) and in 1-day-old (P1) rats (B). Rats were maintained in LD12:12 and released into darkness at CT0. The brain sections were assayed for hnrna by in situ hybridization. Data are expressed as mean ± SEM from 4 animals. For P1, AVP hnrna levels at CT2 and CT22 were significantly higher than those at CT10, 12, 14, 16, 18, and 20 (p < 0.05). hnrna levels might thus fall in the CT0-8 and again in the CT22-24 interval, whereas low levels might fall in the CT10-20 interval. Apparently, in 1-day-old rats, a circadian rhythm in AVP hnrna expression was present. DISCUSSION In agreement with our previous study (Sládek et al., 2004), the rhythmic expression of canonical core clock genes in the rat SCN could be clearly detected only after birth. In 20-day-old fetuses, Per1 expression was still as low as that in 19-day-old fetuses, but a rhythm of Per1 mrna might start to form. In 1-day-old rats, the Per1 mrna rhythm was clearly discernible, with higher levels clustered in the CT7-9 interval. However, only in 2-day-old rats was a more robust rhythm in Per1 expression, similar in waveform and phase to those in 3- and 10-day-old rats (Sládek et al., 2004) already present. At P2, analogously as at P3 and P10, maximum values fell in the daytime CT2-8 interval. Similarly, in the adult mouse

7 146 JOURNAL OF BIOLOGICAL RHYTHMS / April 2006 SCN, elevated mrna levels clustered in the CT3-12 interval (Reppert and Weaver, 2001). No rhythmic Per2 expression could be detected in 20-day-old fetuses. In 1-day-old rats, formation of a rhythm was already indicated, with elevated Per2 mrna levels falling in the CT7-13 interval. In 2-day-old rats, a more robust Per2 mrna rhythm was present, similar to those in 3- and 10-day-old rats (Sládek et al., 2004). At P2 and P3, high levels fell in the CT6-12 interval, and at P10, they fell in the CT6-14 interval. Analogously, in adult mice, high levels were clustered in the CT5-12 interval (Reppert and Weaver, 2001). No rhythms in Cry1 expression could be detected in 20-day-old fetuses or 1-day-old rats. In 2-dayold rats, formation of a rhythm was indicated, with elevated levels clustered in the CT10-14 interval. In 3- and 10-day-old rats (Sládek et al., 2004) and in adult mice (Reppert and Weaver, 2001), high Cry1 mrna levels fell in CT8-12, CT8-16, and CT8-14 intervals, respectively. Similarly as at E19 (Sládek et al., 2004), expression of Bmal1 was high at E20, but no rhythm could be detected. In 1- and 2-day-old rats, however, a clear Bmal1 mrna rhythm was present, with maximum values falling at night to CT16-23 and CT16-18 intervals, respectively. In 3- and 10-day-old rats (Sládek et al., 2004) and in adult mice (Reppert and Weaver, 2001), maximum values were clustered in CT14-20, CT14-22, and CT12-21 intervals, respectively (i.e., always during the night hours). Expression of Clock did not reveal any rhythm in 20-day-old fetuses or 2-day-old rats, as was the case in 3- and 10-day-old and adult rats. At all ages, Clock appeared to be expressed constitutively rather than in a cyclical manner. Nevertheless, when adult rats are maintained under a short photoperiod, Clock expression may also become rhythmic (Sumová et al., 2003). The data indicate that since the 1st appearance of rhythms in clock gene expression in the rat SCN, the phase of the rhythms is roughly in agreement with that of 10-day-old rats (Sládek et al., 2004) and of adult mice (Reppert and Weaver, 2001). The phase might rather be set by the mother than by an LD cycle. Although a newborn rat can perceive light and respond to a photic stimulus by induction of the c-fos gene in the SCN, it does not yet exhibit any rhythm in the SCN photosensitivity (Bendová et al., 2004). The data also confirm our previous finding that rhythms in canonical clock gene expression in the rat SCN emerge mostly post- and not prenatally (Sládek et al., 2004). Maturation of the rhythms may proceed by increasing the rhythms amplitude, either by increasing the rhythm maximum, decreasing the minimum, or both. Also, in the fetal SCN of Syrian hamsters, molecular oscillations equivalent to those observed in adults were not detected (Li and Davis, 2005). However, Ohta and colleagues (2002, 2003) reported on rhythms in Per1 and Per2 expression in the fetal rat SCN. The above-mentioned authors sampled 20-day-old fetuses at 4-h intervals and found peak levels of Per1 mrna at ZT8 and Per2 mrna at ZT12 and ZT16. In our more densely sampled 20-day-old fetuses, only a forming rhythm in Per1 expression with an indicated peak at CT2 (but no rhythm in Per2 expression) was present. Since Ohta and colleagues (2002, 2003) presented only relative values pertaining to the maximum expression, a detailed comparison of their data with the present study is not possible. The discrepancy between both data might possibly be explained by how pregnant rats were maintained. In our study, rats were released into constant darkness before sampling, whereas in the Ohta and colleagues studies, rats were killed during an LD cycle. Although the fetal clock is believed to be entrained by a maternal cue and probably does not use an LD cycle for synchronization (Davis and Mannion, 1988; Reppert and Weaver, 1991; Weaver and Reppert, 1995), the direct effect of light on the developing fetal circadian clock cannot be excluded. Biologically relevant light wavelengths may reach the fetus in the uterus and influence its development (Jacques et al., 1987). No rhythm of the clock-controlled expression of the AVP gene could be detected in 20-day-old fetuses. However, a significant rhythm of AVP hnrna with pronounced amplitude appeared in the SCN of 1-day-old rats. Elevated levels fell in the CT0-8 interval and again in the CT22-24 interval and low levels in the CT10-20 interval. In adult rats, AVP hnrna levels start to rise after CT21, are elevated until CT9, and become undetectable at CT13 and CT17; the AVP mrna rhythm is phase delayed by about 4 h, as compared with the AVP hnrna rhythm (Yambe et al., 2002). It seems that since the 1st appearance of clock-controlled AVP gene expression, the phase of the rhythm has been roughly in phase with that in adult rats, similarly as has been the case with rhythmic clock gene expression. Our data indicate that the rhythm of AVP gene expression in the rat SCN develops only after E20. Reppert and Uhl (1987) found a significant difference between a daytime AVP mrna level at CT5 and a nighttime level at CT17 already in the SCN of 21-day-old fetuses. In our study, we used in situ hybridization with an intronic

8 Kováčiková et al. / SCN CLOCK RHYTHMICITY DURING ONTOGENY 147 probe and thus detected hnrna, that is, nascent AVP transcript and hence AVP gene transcription (Sherman et al., 1986; Yambe et al., 2002). The diurnal changes of AVP mrna levels in the SCN may be mostly regulated by transcriptional activities (Carter and Murphy, 1992) but also by mrna stability associated with changes in the length of the poly-a-tail (Robinson et al., 1988; Carter and Murphy, 1989). Therefore, it cannot be excluded that mrna degradation might contribute to the finding of diurnal AVP mrna changes in 21-day-old fetuses. It is also possible that a circadian variation in AVP expression might start as early as during the very late prenatal period. From a comparison of the profiles of clock gene expression with the profile of AVP gene expression at P1, it seems that the rhythms of AVP hnrna and Bmal1 mrna were the most pronounced in 1-day-old rats. It is tempting to speculate that the rhythm in Bmal1 expression, which peaked during the subjective night, might be the driving force for the rhythm in AVP gene expression. However, it also cannot be excluded that some maternal cues might trigger the rhythm of AVP expression in the very late prenatal stage. The rhythms of AVP gene expression and AVP production (Jáč et al., 2000; Sumová et al., 2000) in the adult rat SCN are present, similar to the rhythm of endogenous c-fos production (Sumová et al., 1998), in the dorsomedial (dm) but not in the ventrolateral (vl) subdivision of the SCN. And during postnatal development, it is just the endogenous rhythm in c-fos production, a marker of the dm-scn rhythmicity, which develops earlier postnatally than the vl-scn rhythm in c-fos photoinduction (Bendová et al., 2004). In conclusion, our data indicate that the detectable rhythmic expression of clock genes in the rat SCN starts mostly after birth. We cannot, however, exclude the possibility of oscillations in clock gene expression in a relatively small number of cells in the fetal SCN. The appearance of a pronounced rhythm in the AVP hnrna already in the SCN of 1-day-old rats suggests that other mechanisms besides rhythmic clock gene expression might contribute to generation of the rhythm in AVP gene transcription during the early developmental stage. ACKNOWLEDGMENTS The authors thank Lucie Heppnerová and Eva Suchanová for their excellent technical assistance, Mr. John Novotney for his careful reading of the article, and Professor Hitoshi Okamura (Kobe University School of Medicine, Japan) for his generous gift of the plasmid templates used for the synthesis of rper1, rper2, and mcry1 riboprobes. Our work is supported by the Grant Agency of the Czech Republic, Grant Nos and , Research Project Nos. LC554 and AV0Z , and by the EU 6th Framework Project EUCLOCK No REFERENCES Bendová Z, Sumová A, and Illnerová H (2004) Development of circadian rhythmicity and photoperiodic response in subdivisions of the rat suprachiasmatic nucleus. Brain Res 148: Carter DA and Murphy D (1989) Independent regulation of neuropeptide mrna level and poly(a) tail length. J Biol Chem 264: Carter DA and Murphy D (1992) Nuclear mechanisms mediate rhythmic changes in vasopressin mrna expression in the rat suprachiasmatic nucleus. Mol Brain Res 12: Davis FC and Mannion J (1988) Entrainment of hamster pup circadian rhythms by prenatal melatonin injections to the mother. Am J Physiol 255:R439-R448. Fu L and Lee CC (2003) The circadian clock: pacemaker and tumour suppressor. Nat Rev Cancer 3: Gillette MU (1991) SCN electrophysiology in vitro: rhythmic activity and endogenous clock properties. In Suprachiasmatic Nucleus: The Mind s Clock, Klein DC, Moore RY, and Reppert SM, eds, pp , New York, Oxford University Press. Jáč M, Kiss A, Sumová A, Illnerová H, and Ježová D (2000) Daily profiles of arginine vasopressin mrna in the suprachiasmatic, supraoptic and paraventricular nuclei of the rat hypothalamus under various photoperiods. Brain Res 887: Jacques SL, Weaver DR, and Reppert SM (1987) Penetration of light into the uterus of pregnant animals. Photochem Photobiol 45: Jin X, Shearman LP, Weaver DR, Zylka MJ, de Vries GJ, and Reppert SM (1999) A molecular mechanism regulating rhythmic output from the suprachiasmatic circadian clock. Cell 96: King DP and Takahashi JS (2000) Molecular genetics of circadian rhythms in mammals. Annu Rev Neurosci 23: Klein DC, Moore RY, and Reppert SM, eds (1991) Suprachiasmatic Nucleus: The Mind s Clock, New York, Oxford University Press. Kornhauser JM, Mayo KE, and Takahashi JS (1993) Immediate-early gene expression in a mammalian circadian pacemaker: the suprachiasmatic nucleus. In Molecular Genetics of Biochemical Rhythms, Young MW, ed, pp , New York, Dekker. Li X and Davis FC (2005) Developmental expression of clock genes in the Syrian hamster. Dev Brain Res 158: Moore RY (1991) Development of the suprachiasmatic nucleus. In Suprachiasmatic Nucleus: the Mind s Clock,

9 148 JOURNAL OF BIOLOGICAL RHYTHMS / April 2006 Klein DC, Moore RY, and Reppert SM, eds, pp , New York, Oxford University Press. Ohta H, Honma S, Abe H, and Honma K (2002) Effects of nursing mothers on rper1 and rper2 circadian expressions in the neonatal rat suprachiasmatic nuclei vary with developmental stage. Eur J Neurosci 15: Ohta H, Honma S, Abe H, and Honma K (2003) Periodic absence of nursing mothers phase-shifts circadian rhythms of clock genes in the suprachiasmatic nucleus of rat pups. Eur J Neurosci 17: Pittendrigh CS (1981) Circadian systems: entrainment. In Handbook of Behavioural Neurobiology, vol. 4: Biological Rhythms, Aschoff J, ed, pp , New York, Plenum. Reppert SM and Schwartz WJ (1984) The suprachiasmatic nuclei of the fetal rat: characterization of a functional circadian clock using 14C-labeled deoxyglucose. J Neurosci 4: Reppert SM and Uhl GR (1987) Vasopressin messenger ribonucleic acid in supraoptic and suprachiasmatic nuclei: appearance and circadian regulation during development. Endocrinology 120: Reppert SM and Weaver DR (1991) A biological clock is oscillating in the fetal suprachiasmatic nuclei. In Suprachiasmatic Nucleus: The Mind s Clock, Klein DC, Moore RY, and Reppert SM, eds, pp , New York, Oxford University Press. Reppert SM and Weaver DR (2001) Molecular analysis of mammalian circadian rhythms. Annu Rev Physiol 63: Robinson BG, Frim DM, Schwartz WJ, and Majzoub JA (1988) Vasopressin mrna in the suprachiasmatic nuclei: daily regulation of polyadenylate tail length. Science 241: Schwartz WJ (1991) SCN metabolic activity in vitro. In Suprachiasmatic Nucleus: the Mind s Clock, Klein DC, Moore RY, and Reppert SM, eds, pp , New York, Oxford University Press. Schwartz WJ, Aronin N, Takeuchi J, Bennet MR, and Peters RJ (1995) Towards a molecular biology of the suprachiasmatic nucleus: photic and temporal regulation of c-fos expression. Semin Neurol 7: Shearman LP, Sriram S, Weaver DR, Maywood ES, Chaves I, Zheng B, Kume K, Lee CC, van der Horst GT, Hastings MH, et al. (2000) Interacting molecular loops in the mammalian circadian clock. Science 288: Sherman TG, McKelvy JF, and Watson SJ (1986) Vasopressin mrna regulation in individual hypothalamic nuclei: a northern and in situ hybridization analysis. J Neurosci 6: Shibata S and Moore RY (1987) Development of neuronal activity in the rat suprachiasmatic nucleus. Brain Res 431: Sládek M, Sumová A, Kováčiková Z, Bendová Z, Laurinová K, and Illnerová H (2004) Insight into molecular core clock mechanism of embryonic and early postnatal rat suprachiasmatic nucleus. Proc Natl Acad Sci U S A 101: Sumová A, Jáč M, Sládek M, Šauman I, and Illnerová H (2003) Clock gene daily profiles and their phase relationship in the rat suprachiasmatic nucleus are affected by photoperiod. J Biol Rhythms 18: Sumová A, Trávníčková Z, and Illnerová H (2000) Spontaneous c-fos rhythm in the rat suprachiasmatic nucleus: location and effect of photoperiod. Am J Physiol Regul Integr Comp Physiol 279:R2262-R2269. Sumová A, Trávníčková Z, Mikkelsen JD, and Illnerová H (1998) Spontaneous rhythm in c-fos immunoreactivity in the dorsomedial part of the rat suprachiasmatic nucleus. Brain Res 801: van den Pol AN (1991) The suprachiasmatic nucleus: morphological and cytochemical substrates for cellular interaction. In Suprachiasmatic Nucleus: The Mind s Clock, Klein DC, Moore RY, and Reppert SM, eds, pp 17-50, New York, Oxford University Press. Weaver DR and Reppert SM (1995) Definition of the developmental transition from dopaminergic to photic regulation of c-fos gene expression in the rat suprachiasmatic nucleus. Mol Brain Res 33: Yambe Y, Arima H, Kakiya S, Murase T, and Oiso Y (2002) Diurnal changes in arginine vasopressin gene transcription in the rat suprachiasmatic nucleus. Mol Brain Res 104: