Identification of genes that restrict astrocyte differentiation of midgestational neural precursor cells.

During development of the mammalian CNS, neurons and glial cells (astrocytes and oligodendrocytes) are generated from common neural precursor cells (NPCs). However, neurogenesis precedes gliogenesis, which normally commences at later stages of fetal telencephalic development. Astrocyte differentiation of mouse NPCs at embryonic day (E) 14.5 (relatively late gestation) is induced by activation of the transcription factor signal transducer and activator of transcription (STAT) 3, whereas at E11.5 (mid-gestation) NPCs do not differentiate into astrocytes even when stimulated by STAT3-activating cytokines such as leukemia inhibitory factor (LIF). This can be explained in part by the fact that astrocyte-specific gene promoters are highly methylated in NPCs at E11.5, but other mechanisms are also likely to play a role. We therefore sought to identify genes involved in the inhibition of astrocyte differentiation of NPCs at midgestation. We first examined gene expression profiles in E11.5 and E14.5 NPCs, using Affymetrix GeneChip analysis, applying the Percellome method to normalize gene expression level. We then conducted in situ hybridization analysis for selected genes found to be highly expressed in NPCs at midgestation. Among these genes, we found that N-myc and high mobility group AT-hook 2 (Hmga2) were highly expressed in the E11.5 but not the E14.5 ventricular zone of mouse brain, where NPCs reside. Transduction of N-myc and Hmga2 by retroviruses into E14.5 NPCs, which normally differentiate into astrocytes in response to LIF, resulted in suppression of astrocyte differentiation. However, sustained expression of N-myc and Hmga2 in E11.5 NPCs failed to maintain the hypermethylated status of an astrocyte-specific gene promoter. Taken together, our data suggest that astrocyte differentiation of NPCs is regulated not only by DNA methylation but also by genes whose expression is controlled spatio-temporally during brain development.

DNA methylation is a critical cell-intrinsic determinant of astrocyte differentiation in the fetal brain.

Astrocyte differentiation, which occurs late in brain development, is largely dependent on the activation of a transcription factor, STAT3. We show that astrocytes, as judged by glial fibrillary acidic protein (GFAP) expression, never emerge from neuroepithelial cells on embryonic day (E) 11.5 even when STAT3 is activated, in contrast to E14.5 neuroepithelial cells. A CpG dinucleotide within a STAT3 binding element in the GFAP promoter is highly methylated in E11.5 neuroepithelial cells, but is demethylated in cells responsive to the STAT3 activation signal to express GFAP. This CpG methylation leads to inaccessibility of STAT3 to the binding element. We suggest that methylation of a cell type-specific gene promoter is a pivotal event in regulating lineage specification in the developing brain.

Behavioural rhythm splitting in the CS mouse is related to clock gene expression outside the suprachiasmatic nucleus.

CS mice exhibit a spontaneous splitting in the circadian rhythm of locomotor activity under constant darkness, suggesting that they contain two weakly coupled oscillators in the circadian clock system regulating locomotor activity rhythm. In order to clarify whether the two oscillators are located in the suprachiasmatic nucleus (SCN), a site of the master circadian pacemaker in mammals, circadian rhythms in mRNA of mouse Period genes (mPer1, mPer2 and mPer3) in the SCN and cerebral cortex were examined during rhythm splitting by in situ hybridization. In the SCN, mPer1 and mPer2 showed a circadian rhythm with a single peak in both split and unsplit mice. The rhythms of mPer1 and mPer2 were slightly phase delayed during rhythm splitting in reference to the activity onset, but the phase relationship between the two rhythms was not changed. In the cerebral cortex, the expression of mPer1 and mPer2 underwent the bimodal fluctuation with peaks temporally corresponding to split activity components. The unsplit mice showed the circadian rhythms with a single peak. There was no difference in the mPer3 rhythms in either the SCN or the cerebral cortex between the split and unsplit mice. These results indicate that the circadian oscillations of mPer1, mPer2 and mPer3 in the SCN are not related to the rhythm splitting of CS mice. The split rhythms of the CS mice are suggested to be caused by uncoupling of oscillators located outside the SCN from the SCN circadian pacemaker.

Circadian pattern, light responsiveness and localization of rPer1 and rPer2 gene expression in the rat retina.

Circadian expression, light-responsiveness and localization of clock genes, rPer1 and rPer2, were examined in the rat retina under constant darkness. A significant circadian variation was detected in rPer2 transcript levels with a peak at ZT14, but not in the rPer1. A light pulse given after constant darkness of 3 days increased both rPer1 and rPer2 expression phase-dependently, while rPer1 was induced at more times than rPer2. A major site of these gene expression within the retina was the inner nuclear layer. These findings indicate that rPer1 and rPer2 genes play different roles in the generation and regulation of circadian rhythms in the retina from those in the suprachiasmatic nucleus.

Clock gene expressions in the suprachiasmatic nucleus and other areas of the brain during rhythm splitting in CS mice.

The CS mouse is a mutant strain which displays spontaneous splitting in the circadian locomotor rhythm under continuous darkness. To clarify whether the rhythm splitting occurs in the suprachiasmatic nucleus (SCN) where the mammalian circadian clock is located, the circadian rhythmicities of mammalian clock genes, mPer1, mBMAL1 and mClock, were examined in the SCN and cerebral cortex during rhythm splitting. The circadian profiles of the clock genes during rhythm splitting were essentially the same as those observed under unsplit conditions. However, the mPer1 gene expression throughout the day was bimodal in the piriform and cingulate cortices, peaking in correspondence with two split components of behavioral rhythm. These results indicate that the circadian profiles of three clock gene expressions in the SCN are not consistent with the overt circadian locomotor rhythm, suggesting that the site of rhythm splitting is somewhere outside the SCN, or alternatively different subregions or other clock genes in the SCN are involved in rhythm splitting.

Clock genes outside the suprachiasmatic nucleus involved in manifestation of locomotor activity rhythm in rats.

Chronic treatment of methamphetamine (MAP) in rats desynchronized the locomotor activity rhythm from the light-dark cycle. When the activity rhythm was completely phase-reversed with respect to a light dark-cycle, 24 h profiles were examined for the clock gene (rPer1, rPer2, rBMAL1, rClock) expressions in several brain structures by in situ hybridization, and for the pineal as well as plasma melatonin levels. In the MAP-treated rats, the rPer1 expression in the suprachiasmatic nucleus (SCN) showed a robust circadian rhythm which was essentially identical to that in the control rats. Circadian rhythms in pineal as well as plasma melatonin were not changed significantly in the MAP-treated rats. However, robust circadian rhythms in the rPer1, rPer2 and rBMAL1 expressions detected in the caudate-putamen (CPU) and parietal cortex were completely phase-reversed in the MAP-treated rats, compared with those in the control rats, indicating desynchronization from the SCN rhythm. Such desynchronization was not observed in the circadian rhythms of clock gene expression in the nucleus accumbens and cingulate cortex. The circadian rClock expression rhythm in the MAP-treated rats was not phase-reversed in the CPU and parietal cortex. These findings indicate that the locomotor activity rhythm in rats is directly driven by the pacemaker outside the SCN, in which rPer1, rPer2 and rBMAL1 in the CPU and parietal cortex are involved.

Circadian rhythms and light responsiveness of mammalian clock gene, Clock and BMAL1, transcripts in the rat retina.

Circadian expression and light-responsiveness of the mammalian clock genes, Clock and BMAL1, in the rat retina were examined by in situ hydbribization under constant darkness. A small but significant daily variation was detected in the Clock transcript level, but not in BMAL1. Light increased the Clock and BMAL1 expressions significantly when examined 60 min after exposure. The light-induced gene expression was phase-dependent for Clock and peaked at ZT2, while rather constant throughout the day for BMAL1. These findings suggest that Clock and BMAL1 play different roles in the generation of circadian rhytm in the retina from those in the suprachiasmatic nucleus. Different roles are also suggested between the two genes in the photic signal transduction in the retina.

Daily variation and light responsiveness of mammalian clock gene, Clock and BMAL1, transcripts in the pineal body and different areas of brain in rats.

Expression patterns of mammalian clock genes, Clock and BMAL1, were examined by in situ hybridization in the pineal body, olfactory bulb, hippocampus and cerebellum in rats under constant darkness. In the pineal, the level of Clock transcript was significantly higher at ZT18 (subjective night) than at ZT6 (subjective day), while the level of BMAL1 transcript was significantly higher at ZT6 than at ZT18. A 30 min light pulse did not affect the transcript levels of Clock nor of BMAL1. The Clock expression in the cerebellum was significantly increased at ZT6 than at ZT18, while no difference was detected in the olfactory bulb and hippocampus at these two phases. The BMAL1 expressions in these areas were similar to the case in the pineal. These findings indicate that the mammalian clock gene, Clock and BMAL1, are expressed differently in the different areas of the brain and the pineal.

Phase-dependent induction by light of rat Clock gene expression in the suprachiasmatic nucleus.

To clarify the role of Clock in the photic signal transduction of rat circadian clock, we cloned and sequenced rat Clock and examined the effect of a single light pulse on the Clock mRNA expression in the suprachiasmatic nucleus (SCN) by in situ hybridization. Rats were exposed to a 30 min light pulse ( approximately 300 lx) at one of six circadian phases in constant darkness (DD), and sacrificed 60 min after the light on. In the rats without light exposure, the mRNA level in the SCN was high at ZT (Zeitgeber time) 6 and low at ZT 18 and 22. Light exposure increased Clock mRNA level in the SCN in phase dependent manner. The mRNA level was significantly increased during the subjective night (ZT10-22). The light had no effect on the mRNA level during the subjective day (ZT2 and 6). The Clock mRNA was also detected in the piriform cortex (PC), and increased by light at ZT14. These results suggest that Clock transcription in the SCN is involved in the photic signal transduction of circadian clock in rats.

Circadian oscillation of BMAL1, a partner of a mammalian clock gene Clock, in rat suprachiasmatic nucleus.

A superfamily gene which encodes a bHLH (basic helix-loop-helix)/PAS transcription factor, BMAL1, was cloned and sequenced from rat cDNA. A robust circadian rhythm of rat BMAL1 expression was detected by in situ hybridization in the suprachiasmatic nucleus (SCN), the site of the circadian clock, with the highest level at the subjective night. Less prominent and completely reversed circadian rhythms of rBMAL1 mRNA were observed in the piriform and parietal cortices. The hybridization signals of rBMAL1 mRNA were also detected in the olfactory bulb, hippocampus, and cerebellum. Since the product of rBMAL1 was recently demonstrated to dimerize with the protein of a mammalian clock gene, Clock, and the protein complex was shown to bind the E Box in the promoter region of mPer1 (a mouse homologue to Drosophila clock gene, Per), rBMAL1 possibly plays a critical role in the clock mechanism generating the circadian oscillation in rats.

Circadian periods of single suprachiasmatic neurons in rats.

Neuronal activity of a single neuron was monitored continuously for more than 5 days by means of a multi-electrode dish in dispersed cell culture of the rat suprachiasmatic nucleus (SCN). Sixty-seven out of 88 neurons showed a robust circadian rhythm in firing rate. The mean circadian period was 24.2 h, which was almost identical to that of the locomotor activity rhythm in 114 weanling rats blinded on the day of birth. However, the circadian period in individual SCN neurons was scattered from 20.0 to 28.3 h (SD, 1.4 h), while the period of activity rhythm clustered from 24.0 to 24.8 h (SD, 0.2 h). It is concluded that a large number of SCN neurons contain the circadian oscillator, the period of which is more variable than the circadian period of the SCN as a whole. It is suggested that the circadian rhythms in individual SCN neurons are capable of synchronizing to each other and are integrated to constitute a multiple oscillator system(s) within the SCN.

Circadian rhythm and light responsiveness of BMAL1 expression, a partner of mammalian clock gene Clock, in the suprachiasmatic nucleus of rats.

To clarify whether BMAL1 is involved in the photic signal transduction in the mammalian circadian clock, we examined the effects of a single light pulse on the level of BMAL1 mRNA in the suprachiasmatic nucleus (SCN) of rats by in situ hybridization. Rats were exposed to 30 min light of ca. 300 lux at six different phases in constant darkness and decapitated 60 min later. BMAL1 transcripts in the SCN of the control animals showed a robust circadian oscillation with the highest expression at ZT (Zeitgeber time) 18 and the lowest at ZT2. The light pulse slightly increased the level of BMAL1 transcripts in the SCN. However, the increment did not depend on the phase of light pulse. There was no significant change in the BMAL1 mRNA level up to 120 min after a light pulse at ZT14 and ZT22. These results indicate that BMAL1 transcription is not involved in the photic signal transduction responsible for non-parametric entrainment of the circadian clock in rats.