Animals

Animals were treated in accordance with the guidelines of the University of Tokyo. Newly hatched chicks were purchased from a local supplier, and maintained under various light/dark conditions. The light was provided by fluorescent lamps to give a constant light intensity of ∼300 lux at the level of chicks. The ambient temperature was kept at 28°C±0.5°C irrespective of light/dark conditions. Pineal glands were isolated from decapitated chicks, frozen in liquid nitrogen, and kept at −80°C until subsequent analyses. All of the procedures during the dark-period were performed under dim red light (>640nm).

Differential display-based cloning

Differential display analysis was performed as previously described (Doi et al., 2001). Briefly, total RNA was extracted from each pool of the isolated pineal glands by using guanidine thiocyanate (Takanaka et al., 1998), and it was treated with DNase I (Takara Shuzo, Kyoto, Japan) to digest contaminating genomic DNA. One micro gram of the total RNA was reverse transcribed by using ThermoScript (Invitrogen, CA, USA) with an anchored oligo (dT) primer (H-T₁₁A, H-T₁₁C or H-T₁₁G; RNAimage kit, GenHunter, TN, USA) at 55°C for 60min. One-hundredth of this reaction mixture was subjected to PCR in a reaction mixture (10 μl) composed of 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl₂, 200 μM each of the dNTPs, 1unit AmpliTaq DNA polymerase (Applied Biosystems, Tokyo, Japan), 1 μM anchored oligo (dT) primer (H-T₁₁A, H-T₁₁C or H-T₁₁G), 1 μM H-AP primer (one of eighty kinds of H-AP primers; RNAimage kit, GenHunter). The amplification reaction was comprised of an initial denaturation step of 94°C for 5 min, and 30 cycles of 94°C for 30 sec, 36°C for 2 min and 72°C for 30 sec, with a final extension step of 72°C for 7 min. The PCR products were subjected to 7.5% native polyacrylamide gel electrophoresis (PAGE), stained with SYBR Green I (Molecular Probes, OR, USA) and the band intensities were quantified by FLA-2000 bioimage analyzer (Fuji Film, Tokyo, Japan). The PCR products selected were cut out from the gel and eluted by boiling in 50 μl of distilled water at 95°C for 5 min. An aliquot (2 μl) of the eluted DNA solution was subjected to the secondary PCR under the same conditions as those in the initial analysis. The reamplified DNA fragment was sub-cloned into pCR2.1 TOPO vector (Invitrogen) and sequenced.

Quantitative reverse transcription (RT)-PCR analysis

Quantitative RT-PCR analyses were performed as previously described (Hirota et al., 2001). Primer pairs used were as follows; for cHsp90α (GenBank accession number;  X07265), 5′-GATGA TGAGC AGTAT GCTTG G-3′ (forward) and 5′-TGGTC TTGTT GAGCT CTTCC-3′ (reverse); for chicken TATA-box binding protein (cTbp, GenBank accession number;  D83135), 5′-GTCGA ATATA ATCCC AAGCG-3′ (forward) and 5′-TCTGC TCGAA CTTTA GCACC-3′ (reverse). The optimal cycle numbers for quantitative analyses were 17 and 21 for cHsp90α and cTbp, respectively.

Immunoblot analysis

Each pool of the isolated pineal glands (n=5) was homogenized with 400 μl of ice-cold 20 mM Tris-HCl buffer (pH 7.4) containing 2.5 mM EDTA, 100 mM NaCl, 1 μg/ml leupeptin, 1 μg/ml aprotinin and 0.1 mM phenylmethylsulfonyl fluoride, and the homogenate was centrifuged at 700 g for 10 min. These manipulations for the pineal glands isolated during the dark period were performed under dim red light (>640 nm). The supernatant contained almost all of the detectable cHSP90α protein expressed in the pineal gland (data not shown). An aliquot of the supernatant was subjected to sodium dodecyl sulfate (SDS)-PAGE (7.5% acrylamide) after boiling for 8 min in SDS-PAGE sample buffer (10 mM Tris-HCl, 6% [v/v] glycerol, 2% [w/v] SDS, 50 mM DTT, 2 mM EDTA, 0.002% [w/v] Coomassie Brilliant Blue, pH 6.8). Proteins in the gel were electro-transferred to a polyvinylidene difluoride membrane (Immobilon transfer membrane; Millipore Corp., MA, USA), which was then incubated for 1hr at 37°C in blocking solution I (10 mM sodium phosphate, 150 mM NaCl, 4% bovine serum albumin, pH 7.4), followed by incubation at 4°C overnight with HSP90α-specific antibody (final concentration, 400 ng/ml; StressGen Biotechnologies, Victoria, Canada) diluted in a reaction buffer (10 mM sodium phosphate, 150 mM NaCl, 0.1% bovine serum albumin, pH 7.4). After washing 3 times for 5 min each with the reaction buffer, the blot was incubated for 1 hr at 37°C with alkaline phosphatase-conjugated anti-rabbit IgG antibody (New England Biolabs, MA, USA) diluted in the reaction buffer (1:2,500 dilution). The immunoreactivities were visualized by using CDP-star detection system (New England Biolabs).

Preparation of tissue sections and immunofluorescence analysis

The isolated pineal glands were fixed with 4% (w/v) paraformaldehyde in 0.1 M sodium phosphate buffer (pH 7.4) for 4hr at 4°C. Tissues isolated during the dark period were fixed in complete darkness. The fixed tissues were incubated for cryoprotection in dehydrating solution I (10 mM sodium phosphate, 10% sucrose [w/v], 140mM NaCl, pH 7.2) at 4°C for 4 hr, and subsequently incubated in dehydrating solution II (10 mM sodium phosphate, 30% sucrose [w/v], 140 mM NaCl, pH 7.2) at 4°C overnight. Then, the tissues were embedded in a 2:1 (v/v) mixture of dehydrating solution II and OCT mounting medium (Sakura, Tokyo, Japan), frozen by using liquid nitrogen, and stored at −80°C until use. Finally, 10-μm-thick sections were cut out from the frozen tissues, mounted on gelatin-coated glass slides, and air-dried. The sections on glass slides were pretreated with blocking solution II (10 mM sodium phosphate, 1% bovine serum albumin, 3% horse normal serum, 3% goat normal serum, 0.05% Tween20, 140 mM NaCl, 1 mM MgCl₂, pH 7.4) for 30 min at room temperature, and then incubated with a mixture of rabbit anti-HSP90α antibody (final concentration, 1 μg/ml; Stress-Gen Biotechnologies) and mouse anti-pinopsin antibody P9 (1:400 dilution; Okano et al., 1997) diluted in blocking solution II at 4°C overnight. After rinsing with phosphate-buffered saline buffer (10 mM sodium phosphate, 140 mM NaCl, 1 mM MgCl₂, pH 7.4), the sections were incubated with a mixture of fluorescein isothiocyanate-conjugated anti-rabbit IgG antibody (final concentration, 1.2 μg/ml; Vector Laboratories, CA, USA) and Texas Red-conjugated anti-mouse IgG antibody (final concentration, 6 μg/ml; Vector Laboratories) diluted in blocking solution II at 4°C overnight. The sections were coverslipped with Vectashield mounting medium (Vector Laboratories) containing 4′, 6-diamidino-2-phenylindole (DAPI) for nuclear staining.

In vivo measurement of the local temperature adjacent to the pineal gland

The chick body temperature at a local region adjacent to the pineal gland was measured by using a small sensor tip of thermistor (globular shaped, ø 1.0 mm; 503ET-1, Ishizuka Electronics, Japan), which was connected to a digital monitor (DM1001C, Ishizuka Electronics, Japan) with a pair of electric wires (ø 0.8 mm, 1.0 m long) in a single silicon tube (1 mm in inside diameter, 2 mm in outside diameter; AS ONE, Tokyo, Japan). Fourteen-days-old chicks were subjected to the in vivo temperature measurement. The sensor tip was surgically placed in a brain position close to the pineal gland (1.0 mm anterior and 1.0 mm lateral to the protuberantia occipitalis interna, 1.0 mm ventral to the cortical surface), and the connecting tube was attached to the skull with the aid of dental cement (GC Corp., Tokyo, Japan) so that the sensor was fixed at the position. Activities of the chick were restricted by placing it in a small columnar arena (80 mm in diameter, 80 mm high) with a cover of a circular cloth loosely bound to the neck of the bird like a collar. Food and water were freely available, and the ambient temperature was monitored and kept at 28°C±0.5°C.

Identification of chick pineal Hsp90α as a light responsive gene

To explore the molecular mechanism underlying the phase-dependent phase-shifting effect of light on the circa-dian clock, we employed a differential display-based screening of the chick pineal genes that are induced by light in a phase-dependent manner. Newly hatched chicks were raised in LD cycles for 7days and transferred to DD on day 8. Then animals were exposed to a 90-min light stimulus from various CT points (CT0, CT14 or CT20) on day 8, and the pineal glands were collected from animals (30 chicks each) at two distinct time periods after the light onset (20–50 min and 60–90 min after the light onset). The control animals were maintained in the dark, and their pineal glands were collected at CT1.5, CT15.5 and CT21.5. This sampling schedule was chosen to identify various types of light-responsive genes such as those showing a CT-dependent light-induction and those induced rapidly or slowly after the light onset. We screened 5,699 bands of the PCR products and found 99 bands as light-induced products. As shown in Table 1, these products were classified into four types on the basis of the time point(s) when they were light-induced: (i) induced only at CT14 (early night), (ii) induced only at CT20 (the late night), (iii) induced both at CT14 and CT20 (night), but not at CT0 (morning), and (iv) induced at every time point tested. All the light-induced products showed higher band intensities at 60min than those at 20min after the light onset. We cloned three products, two from type (iv) (Fig. 1) and one from type (i) (Doi et al., 2001) because these three showed remarkable light induction. Sequencing analyses revealed that the latter one was derived from a message of cE4bp4 gene (Doi et al., 2001). The sequences of the former two products (indicated by arrows in Fig. 1A and 1B) matched completely the cHsp90α cDNA sequence (GenBank accession no.  X07265) within the 339 bases (+2513 to +2851; +1 indicates transcription initiation site) and within the 646 bases (+1650 to +2295), respectively. These results strongly suggest that the mRNA levels of the pineal cHsp90α increase profoundly in response to the light stimulus given at CT0, 14 and 20.

Table 1

Classification of PCR products in differential display analyses

Fig. 1

Identification of chicken Hsp90α gene by differential display analyses. Chicks were exposed to light for 90 min from indicated CT points of a day, and their pineal glands were isolated at 20–50 min (L20–50) or 60-90min (L60–90) after the light onset. Control animals were kept in the dark (D). Amplification reactions were performed with a pair of H-T11C primer (5′-AAGCT TTTTT TTTTT C-3′) and H-AP1 primer (5′-AAGCT TGATT GCC-3′) in (A), and with a pair of H-T11G primer (5′-AAGCT TTTTT TTTTT G-3′) and H-AP14 primer (5′-AAGCT TGGAG CTT-3′) in (B). Arrows indicate the positions of the amplified DNA fragments derived from cHsp90α transcript.

Light-dependent and time-of-day-dependent expression of cHsp90α mRNA

A time course of the light-dependent cHsp90α induction was examined by quantitative RT-PCR analyses, in which the mRNA levels of the chick pineal cHsp90α were monitored during and after 60-min light exposure from CT20 (Fig. 2A). The cHsp90α mRNA level began to increase within 5min after the light onset, and at 90 min it reached a peak level ∼4 fold higher than the basal level. Then, the cHsp90α mRNA level decreased gradually and reached to a level close to that of the control (dark-kept) animals within 3 hr after the light offset. To examine whether the pineal cHsp90α expression shows a CT-dependent light response, chicks were exposed to a 60-min light stimulus at various CT points of a day (CT2, CT6, CT10, CT14, CT18 and CT22), and the level of the pineal cHsp90α mRNA was evaluated at 60 min after the light onset (Fig. 2B). The light stimuli given at CT10, CT14, CT18 and CT22 produced significant increases in the level of the pineal cHsp90α mRNA, but no significant effect was observed by light given at early morning to midday (CT2 and CT6) under the experimental conditions. In addition to the acute response to light, a daily change in the pineal cHsp90α mRNA level was examined. In a LD cycle (Fig. 3; on day 8), the pineal cHsp90α mRNA exhibited a profound daily fluctuation with a peak in the early morning, and a similar fluctuation with lower amplitude was observed in animals maintained in DD (Fig. 3; on day 9). These results indicate that the expression of cHsp90α is regulated not only by the external light but also by the circa-dian clock.

Fig. 2

Light-dependent changes in cHsp90α mRNA levels. One-day-old chicks were maintained in LD cycles for 8 days and then transferred to DD on day 9, when animals were exposed to light for 60min from each indicated time point [Light (+)] or they were kept in the dark for control [Light (–)]. (A) Time course of cHsp90α induction during and after light exposure at CT20-21. The pineal glands were isolated at indicated time points, and the relative mRNA levels of cHsp90α and cTbp in the isolated pineal glands were measured by RT-PCR. The band intensities of the amplified products of cHsp90α were normalized to those of cTbp, and the value at CT20 (0 min) was set to 1. **p<0.01, compared with the value at CT20 (0 min) by Student's t-test. (B) Light-induction of cHsp90α at various CT points of a day. The pineal glands were isolated from light-exposed animals (open bars) or from dark-kept animals (solid bars) 60 min after indicated time point. Relative mRNA levels of cHsp90α and cTbp were measured by RT-PCR, and the value of dark-kept animals at CT14 was set to 1. All of the values are the mean±SEM from three independent experiments. **p<0.01, *p<0.02, compared with the value of dark-kept animals by Student's t-test.

Fig. 3

Temporal changes in mRNA levels of cHsp90α in the chicken pineal gland. One-day-old chicks were maintained in LD cycles for 8 days and then transferred to DD. The pineal glands were isolated at indicated time points of day 7–8 in LD and day 9 in DD. Relative mRNA levels of cHsp90α and cTbp in the isolated pineal glands were evaluated by RT-PCR, and each value was expressed as described in Fig. 2 except that the minimal value of cHsp90α mRNA was set to 1. All of the values are the mean±SEM from three independent experiments.

Analyses of the chick pineal HSP90α protein

A possible change in cHSP90α protein level was examined by evaluating the total amount of HSP90α protein expressed in the chick pineal gland. In immunoblot analysis (Fig. 4), the expression of cHSP90α protein was detected in the pineal gland of chicks that were maintained in LD cycles at a constant room temperature, and the levels showed no significant change throughout the day in contrast to the profound fluctuation of its mRNA level.

Fig. 4

Immunoblot analysis of the temporal change in the chick pineal HSP90α protein level. One-day-old chicks were maintained in LD cycles for 8 days and their pineal glands were isolated at indicated time points of day 8 (from 5 chicks for each time point). Equal amount of proteins (16 μg) of the pineal homogenate were immuno-blotted with anti-HSP90α antibody (Upper). The band densities were quantified by densitometry and the value of ZT1 was set to 1 (Lower). All of the values are the mean±SEM from three independent experiments.

Although the total amount of the cHSP90α protein accumulated in the whole pineal tissue was apparently constant (Fig. 4), the level of the cHSP90α protein may diurnally fluctuate within a subset of the pineal cells. To test this, localization of cHSP90α protein in the chick pineal gland was investigated by immunohistochemical analyses (Figs. 5 and 6). Examination of a possible colocalization of cHSP90α with pinopsin, a pineal specific photoreceptive molecule (Okano et al., 1994) revealed cHSP90α immunoreactivity in almost all the pineal cells including the follicular pinealocytes (indicated by solid arrowheads in Fig. 5) whose rudimentary outer segments showed pinopsin immunoreactivity (indicated by open arrowheads in Fig. 5). This suggests cHSP90α protein expression in the photosensitive clock cells in the pineal gland (Takahashi et al., 1989; Okano et al., 1994; Bernard et al., 1997). Pineal cHSP90α proteins were localized mainly in the cytosol and diffusely in the nucleus in the follicular and parafollicular pinealocytes (indicated by arrows in Fig. 5). The immunohistochemical analysis of the temporal change in the localization demonstrated that cHSP90α immunoreactivities were kept at a nearly constant level in immunopositive pineal cells of the pineal glands isolated at ZT1, ZT7, ZT13 and ZT19 (Fig. 6). At every ZT point tested, the subcellular localization of cHSP90a protein was mainly cytoplasmic as observed at ZT6 (Fig. 5A). These observations indicate that not only the total amount of the pineal cHSP90α protein but also its sub-cellular localization in each pineal cell is kept almost constant all day long in LD cycles.

Fig. 5

Localization of the chick pineal HSP90α protein. One-day-old chicks were maintained in LD cycles for 8 days and tissue sections derived from the pineal gland isolated at ZT6 of day 8 were double-labeled with anti-HSP90α antibody (A, green) and anti-pinopsin antibody P9 (B, red). The nucleus was detected by staining with DAPI (D, blue). Panel C is a merged image of A and B, demonstrating the distinctive distribution of HSP90α and pinopsin in the follicular pinealocytes. Solid arrowheads indicate the region of inner segments of the follicular pinealocytes, which were immunopositive to anti-HSP90α. Open arrowheads indicate rudimentary outer segments of the follicular pinealocytes, which were immunopositive to anti-pinopsin antibody. Arrows indicate the parafollicular pinealocytes, which were immunopositive to anti-HSP90α but immunonegative to anti-pinopsin. L, follicular lumen; F, follicular zone; Pf, parafollicular zone. Scale bar=50 μm

Fig. 6

Immunohistochemical analysis of the temporal changes in the chick pineal HSP90α protein level. One-day-old chicks were maintained in LD cycles for 8 days and their pineal glands were isolated at ZT1 (A), ZT7 (B), ZT13 (C) and ZT19 (D) on day 8. The pineal sections were stained with anti-HSP90α antibody. L, follicular lumen; F, follicular zone; Pf, parafollicular zone. Scale bar=100 μm

Light-dependent and time-of-day-dependent changes in the in vivo temperature of the pineal gland

Considering that the expression of heat shock proteins is generally responsive to a temperature change, we assumed that the changes in the mRNA level of the pineal cHsp90α (Figs. 2 and 3) might be ascribed to the changes in the in vivo temperature of the pineal gland (designated T_P). To test this, T_P was measured continuously with the aid of a thermometer fixed at a brain position close to the pineal gland of the chick (see Materials and Methods). When chicks were maintained in LD cycles, T_P exhibited an overt diurnal fluctuation. As shown in Fig. 7A, T_P rapidly decreased and reached trough levels of 39.2–39.4°C within 1 hr after the light offset (on day 14 and 15). Then T_P gradually increased during the late dark period. After the light onset (on day 15), T_P was rapidly elevated and kept at 40.5-40.9°C during the light period. T_P also exhibited a diurnal fluctuation in constant dark condition (on day 16) albeit with a lower peak/trough ratio (Fig. 7A). In response to light exposure during the nighttime (on day 16), T_P rapidly increased by 1.1°C from 39.9°C to 41.0°C (Fig. 7A). Notably, T_P began to increase within 2min after the light onset, and at 20 min it reached a plateau of 40.9–41.0°C (Fig. 7B). The light-dependent and time-of-day-dependent changes of the pineal temperature (Figs. 7A and 7B) indicate a possible association of the expression profile of the pineal cHsp90α mRNA with the temperature change of the pineal gland.

Fig. 7

Light-dependent and time-of-day-dependent changes in the temperature of the chick brain close to the pineal gland. One-day-old chicks were maintained in LD cycles for 15 days and then kept in constant dark. The animals were exposed to light late at night of day 16 (time of day=20). A sensor tip of thermistor was placed in a brain position adjacent to the pineal gland at the beginning of day 14 (see Materials and Methods), and the temperature was measured continuously from the light offset on day 14. (A) A temporal profile of the pineal temperature from day 14 to day 16. (B) A time course of the pineal temperature change before and after the light exposure on day 16. All of the values are the mean±SEM from the temperature data of three chicks.