281

Molecular Brain Research, 15 (1992) 281-290 © 1992 Elsevier Science Publishers B.V. All rights reserved 0169-328x/92/$05.00

BRESM 70482

Circadian and light-induced expression of immediate early gene mRNAs in the rat suprachiasmatic nucleus E l l e n L. S u t i n a n d T h o m a s S. K i l d u f f Sleep Research Center, Department of Psychiatry TD-114, Stanford University School of Medicine, Stanford, CA 94305 (USA) (Accepted 19 May 1992)

Key words: Biological Clock; Suprachiasmatic nucleus; Light; Circadian rhythm; fos; NGFI-A; NGFI-B

Recent work has suggested that immediate early genes are involved in the phase-shifting response of the circadian pacemaker to light. In mammals, the endogenous pacemaker is located in the suprachiasmatic nucleus (SCN). We report here a systematic examination of the expression of the immediate early genes los, NGFI-A, and NGFI-B in the rat SCN, in basal and light-stimulated conditions across the circadian cycle. The photic induction of all three genes examined was found to be gated by the circadian system with maximal induction observed during the mid to late subjective night. The photic-induced expression of N'6FI-B in the SCN was considerably less than that of NGFI-A, despite previous observations suggesting greater induction of NGFI-B by membrane depolarization. The levels of immediate early gene expression induced by light do not directly correspond to the magnitude of previously reported light-induced phase-shifts of the activity rhythm. The results also suggest that los may undergo a circadian rhythm of expression in the rat SCN.

INTRODUCTION Virtually all eukaryotic organisms exhibit daily physiologica! and behavioral rhythms that are synchronized to the environmental light-dark cycle. These circadian rhythms will persist in the absence of external time cues and are entrained by the external light-dark cycle, t n mammals, circadian rhythms are generated by a pacemaker contained within the suprachiasmatic nucleus (SCN) of the anterior hypothalamus22. Since the periodicity of this endogenous clock is not exactly 24 h in most species, the clock must be entrained to the environmental cycle. Entrainment of circadian rhythms occurs by resetting, or phase-shifting, the pacemaker each day. In nocturnal mammals, phase-shifting occurs primarily in response to light stimuli during the dark phase of the cycle in a phase-dependent manner: light pulses delivered early in the active period (the animal's subjective night) will cause phase delays of the pacemaker, whereas pulses later in the active period result in phase advances of rhythms9'28. An important challenge now is to determine the molecular events in-

volved in phase-shifting and entrainment of the circadian clock. Recent work from several laboratories has raised the possibility that immediate early genes are involved in the phase-shifting response to light. Immediate early genes, such as fos, NGFI-A, and NGFI-B, encode transcription factors and have been proposed to act as 'third messengers,' coupling short-term membrane events to long-term changes in gene expression in the central nervous system18'25. Several groups have found that los expression is induced in the SCN of hamsters 15A9'23 and rats 3'11'23 when light pulses are delivered during the animal's subjective night. Kornhauser et al. 15 further demonstrated that increases in fos expression in the SCN occurred only at circadian times at which photic stimuli produce phase-shifts of the activity rhythm, suggesting that los may be a molecular component of the phase-shifting response. Rusak et al. 23 also found that expression of another immediate early gene, NGFI-A (nerve growth factor inducible), was induced in the SCN by photic stimulation at a particular circadian time point in the subjec-

Correspondence: E.L. Sutin, Department of Biological Sciences, Stanford University, Stanford, CA 94305-5020, USA. Fax: (1) (415) 725-5356.

282 tive night. These observations are remarkable because they suggest a clear correlation between a change in gene expression and an observable physiological behavior and, thus, may lead to insights into the molecular machinery underlying the mechanism of circadian rhythm entrainment. To understand the role of immediate early genes in circadian rhythm generation and entrainment, it is necessary to determine basal and light-induced expression of such genes across the circadian cycle. It has been proposed 4 that the different effects of various extracellular stimuli on cell physiology are mediated by activation of distinct subsets of immediate early genes. As applied to the circadian system, it is possible that different combinations of immediate early genes regulate the direction and magnitude of the phase-shift associated with an animal's response to light. This report describes the systematic examination of the expression of three immediate early genes in the rat SCN in basal and light-stimulated conditions across the circadian cycle. The photic induction of all three genes examined, fos, NGFI-A and NGFI-B, was found to be gated by the circadian system with maximal induction observed during the late subjective night. These results also suggest that fos may undergo a circadian rhythm of expression in the rat SCN.

5 nmol oligonucleotide, 50 nmol [35S]dATP (spec. act. 1200 Ci/mmol; New England Nuclear (NEN), Boston, MA), and 150 units terminal deoxynucleotidyl transferase (Bethesda Research Labs. (BRL), Gaithersburg, MD), all in a buffer obtained from B R L After a 15 min reaction at 37°C, the labeled probe was separated from the enzyme and unincorporated radiolabeled dATP residues by phenol/chloroform extraction and ethanol precipitation. 35S-Labeled sense and antisense fos probes were transcribed from a full-length rat cDNA 8 cloned in both orientations in pSP65. Plasmids were linearized with BamHI to produce templates that direct synthesis of approximately 2.1 kb RNAs. Linear DNA templates (1 /zg) were transcribed in a reaction mixture (20/zl) containing 400 ~Ci [35S]UTP (1250 Ci/mmol, Amersham), 2 mM ATP, CTP and GTP (Promega), 2 units RNase block (Stratagene), 10 mM DTT (dithiothreitol), 1 unit of SP6 RNA polymerase, all in a buffer provided by BRL. Following RNA synthesis (1 h, 37°C), the DNA template was degraded by the addition of DNase (10 units, BRL; 20 min 37°C), after which the RNA was purified by phenol/cholorform extraction and ethanol precipitation.

In situ hybridization histochemistry In situ hybridization histochemistry was performed by a modifica-

Male Wistar rats maintained under a light-dark schedule of 12:12 (12 h light: 12 h dark)were transferred into constant darkness for at least 48 h prior to the experiment. At six circadian times (CT, where each day is divided up into 24 circadian h, and CT 0 is the time of lights-on) around the clock, half the animals were exposed to a 30 min pulse of light (approx. 1700 lux) and then sacrificed. Animals were anesthetized with a solution of 7.5% ketamine/0.8% xylazine/0.15% acepromazine and decapitated. Brains were quickly removed, frozen in 2-methylbutane and stored at - 8 0 ° C until sectioning. Control animals at each time point remained in darkness. The dark control animals were sacrificed in the same manner and their brains were removed under dim red illumination ( < 2 lux). Three circadian times were chosen during the animals' subjective day (CT 2, 6, 10) and three during the subjective night (CT 14, 18, 22). At each circadian time, 4 animals were light-stimulated and 4 remained in darkness, except at CT 22 where there were 6 animals in each group.

tion of protocols which have been utilized successfully with other probes 2'24'31. Frozen tissue was warmed to - 18°C in a cryostat and 12 /zm sections were cut through the hypothalamus containing the SCN from its rostro-caudal pole and thaw mounted onto twice-coated gelatin chrome-alum slides. The slides were then placed upon a heat plate at 65°C until dry and stored desiccated at -80°C. Before fixation and hybridization, sections were warmed to room temperature and allowed to dry. Sections were then fixed in 4% formaldehyde in 0.1 M phosphate-buffered saline (PBS, pH 7.4, 4°C) for 10 min. Slides were rinsed twice (1 min each) in PBS, incubated in 0.25% acetic anhydride in 0.1 M triethanolamine/0.9% NaCI (pH 8) for 10 min at room temperature, and then transferred successively through 70% (1 min), 80% (1 min), 95% (2 min), and 100% ethanol (1 min), chloroform (5 min), 100% and 95% ethanol and air dried. 106 cpm probe in 50 /zl buffer containing 0.01% DTT, 50% formamide, 4 x S S C (standard sodium citrate; I × S S C = 0.15 M NaCI, 0.15 M sodium citrate, pH. 7.2), 1 xDenhardt's, 250 /~g/ml yeast tRNA, 500 /zg/ml single-stranded DNA, and 10% dextran sulfate, were applied to each section. Sections were covered with a glass coverslip and incubated overnight at 37°C in a humid chamber. The coverslips were removed and the sections rinsed twice in 1 x SSC with 10 mM Dq"I" at room temp. Slides hybridized with the los riboprobe were then incubated in RNase A (25 ~zg/ml) at 37°C for 30 min. All sections were then washed in four 15-min changes of 2 x S S C / 5 0 % formamide at 40°C, then for two 30 min washes at room temperature in 1 ×SSC, rinsed in water, dehydrated in 70% and then 95% ethanol and air dried. Autoradiographic localization of bound probe was performed by exposure of the sections to XAR-5 film for 4-7 days. For higher resolution analysis, the sections were dipped into Kodak NTB3 nuclear emulsion diluted 1 : 1 with water. Exposures ranged from 2 to 4 weeks after which the slides were developed by using Kodak D-19 developer followed by Kodak fixer and counterstained with 0.1% thionin. Both bright-field and dark-field optics were used to examine the autoradiograms and determine the distribution of labeled cells.

Hybridization probes

Data analysis

The hybridization probes used in this study were synthetic DNA oligonucleotides complementary to NGFI-A and NFGI-B messenger RNAs (mRNAs) and a single-stranded cRNA probe complementary to fos mRNA. The NGFI-A probe was a 45- base oligodeoxyribonucleotide complementary to bases 357-399 of the NGFI-A mRNA, corresponding to amino acids 2-1617. A messenger sense strand probe served as a control to measu.re non-specific labeling; no signal above background was detected. The NGFI-B probe was a 45 base oligodeoxyribonucleotide complementary to bases 1683-1728 of the NGFI-B mRNA, corresponding to amino acids 491-506 TM. The oligonucleotide probes were radiolabeled by the addition of 35S-dATP to the 3' termini. Each reaction mixture (50/~1) contained

Specific hybridization was quantified using a computer-assisted image analysis system (MCID; Imaging Research, Inc., St. Catherines, Ont.). For each animal, the autoradiograms from two coronal brain sections were chosen on the basis of thionin stains to contain the mid-region of the SCN. From each autoradiogram, bilateral optical density (O.D.) measurements were made of the ventrolateral SCN, dorsomedial SCN, anterior hypothalamus and optic chiasm. To control for slight variation in section thickness, time of exposure, and specific activity of the probes, relative O.D. (ROD) values were calculated within each section for the ventrolateral SCN, dorsomedial SCN, and optic chiasm using the anterior hypothalamus as denominator. The average ROD values were then calculated for

MATERIALS AND METHODS

Photic stimulation

283 each of these three brain regions for each animal. These ROD values were then used to calculate group mean RODs in order to compare the experimental and control groups by analysis of variance (ANOVA).

RESULTS

The expression of immediate early genes in the SCN following light pulses at circadian times when the endogenous pacemaker can be reset suggests a possible involvement of these genes in the regulation of circa-

dian rhythms. In an effort to describe in quantitative terms the level of immediate early gene expression compared to basal levels, the relative mRNA levels of three of these genes, fos, NGFI-A and NFGI-B, were measured in the SCN of rats maintained either in constant darkness or exposed to a 30 min pulse of light at six circadian times. Representative film autoradiograms are shown in Fig. 1, comparing hybridization of fos and NGFI-A probes to the SCN of two animals sacrificed in the

Fig. 1. Representative film autoradiograms of coronal hypothalamic sections demonstrating regional localization of fos and NGFI-A m R N A expression in the rat ventrolateral SCN after in situ hybridization with 35S-labeled probes. A: rat exposed to 30 rain light at CT 18 and hybridized with a asS-labeled los antisense riboprobe. B: adjacent section from the same rat as in A (12 ~ m apart) but hybridized with a los sense strand RNA probe. C: rat sacrificed in darkness at CT 18 and hybridized with the los antisense riboprobe. D: same rat as in A (section is 60/.tm away) hybridized with a 35S-labeled NGFI-A antisense oligonucleotide probe. E. Adjacent section from the same rat as in A, B, and D but hybridized with a control NGFI-A sense strand probe. F: adjacent section from same rat as in C (96/zm apart) hybridized with a NGFI-A antisense probe. Arrows indicate the location of the SCN in each section. Bar = 2 mm.

284 subjective night at CT 18. One animal received a 30 min light pulse in the subjective night at CT 18 and one animal was kept in the dark throughout. Discrete localization of bound fos antisense probe to the SCN and to no other hypothalamic region is clearly demonstrated in the light-stimulated animal (Fig. 1A), while the animal kept in the dark shows no hybridization above background (Fig. 1C). The fos sense control probe shows only background hybridization (Fig. 1B). Consistent with previous immunocytochemical and in situ hybridization histochemical findings, los mRNAlabeled neurons were localized to the ventrolateral portion of the SCN, where fibers from retinal gangion cells terminate. The distribution of grain densities in the ventrolateral SCN is illustrated both at the regional level on film autoradiograms (Fig. 1A) and at the cellular level (Fig. 2). Fig. 2 shows the cytoplasmic localization of silver grains in specific cells of the ventrolateral SCN containing fos m R N A after a light pulse at CT 22; light at this time, in particular, produced very strong and specific accumulation of fos message. Representative results obtained with an NGFI-A probe again demonstrate restricted spatial expression within the SCN with little non-specific hybridization above background. Strong hybridization signal is clearly evident in the ventrolateral SCN with a moderate accumulation of N G F I - A m R N A in the dorsomedial SCN of light-stimulated animals (Fig. 1D), and not in animals kept in darkness (Fig. IF). A N G F I - A sense control probe revealed only background hybridization (Fig. 1E). The semi-quantitative analysis of m R N A expression is described below for each gene.

los Light stimulation induced fos to a greater extent in the subjective night relative to the subjective day; in agreement with other reports, light during the subjective night at CT 14, 18 and 22 induced fos expression in all animals studied. The induction was stronger in the late than the early subjective night. Fig. 3 presents levels of fos m R N A i n the ventrolateral SCN, dorsomedial SCN, and optic chiasm across the circadian day. When the effect of photic stimulation was examined across the day, analysis of variance revealed the R O D values from the light-stimulated group to be significantly increased over the dark group at CT 18 and 22 in both the ventrolateral and dorsomedial portions of the SCN. No significant changes were observed in the optic chiasm. The basal expression of fos in the ventrolateral SCN is presented in Fig. 4 by double-plotting the data

Fig. 2. Emulsion-dipped in situ hybridization autoradiograms showing distribution of fosmRNA in individual SCN cells from a rat photically stimulated at CT 22. A: silver grains overlyingcells in the VL-SCN appear white in this dark-field photograph. Black arrows define the lateral border of the dorsal SCN. B. Brightfield photograph (grains appear black) which is an enlargement of the area in A enclosed by the black rectangle. OC, optic chiasma; 3V, third ventricle. Bar = 100 p.m.

for the dark animals from Fig. 3. These data indicate a rhythm of los expression in the ventrolateral SCN in dark control animals with peak levels occuring late in the subjective night at CT 22. At CT 6 through CT 18, none of the animals showed any detectable los expres-

285 sion in the SCN. However, los mRNA was observed in 3 out of 6 animals at CT 22 and in one out of four animals at CT 2. To rule out the possibility that the observed fos induction resulted from the dim red light under which animals are removed from their cages and sacrificed, fos mRNA was measured in an additional 4 animals at CT 22. Two of the animals were kept in the dark until CT 22 at which time they were removed from their cages and handled for a minute under dim red illumination ( < 2 lux) and then returned to their light-tight chambers for 30 min while the other 2 animals were left in the dark up until the time of sacrifice. In this experiment, one animal out of the 4 showed detectable fos expression in the SCN; this was a rat which was left in the dark up until the time of sacrifice•

NGFI-A Light pulses dramatically increased expression of NGFI-A in the ventrolateral SCN during the subjective night at CT 14, 18 and 22 (Figs• 1D -F and 5). Light also caused a moderate accumulation of NGFI-A mRNA in the dorsomedial region of the SCN (Fig. 5B). The apparent maximal effect of light was observed at CT 18 at which time NGFI-A mRNA was increased more than four times over background levels. However, autoradiograms of the tissue from CT 22 were generated in a separate in situ hybridization assay from the animals at CT 2, 6, 10, 14, and 18 and thus preclude quantitative comparisons between CT 22 and other time points due to the variations in hybridization and probe specific activity. The levels of NGFI-A mRNA in the ventrolateral SCN, dorsomedial SCN, and optic

C

A Ventrolateral

SCN:

Foe

Antisense

J

>. I-

Dorsomedial SCN: Foe Antisenee

Light Dark

4

**

U.l ..I

I-

Circadian and light-induced expression of immediate early gene mRNAs in the rat suprachiasmatic nucleus.

Recent work has suggested that immediate early genes are involved in the phase-shifting response of the circadian pacemaker to light. In mammals, the ...
3MB Sizes 0 Downloads 0 Views