| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Netherlands Institute for Neuroscience, Amsterdam, The Netherlands;
Anhui Geriatrics Institute, The first Affiliated Hospital of Anhui Medical University, Hefei, P. R. China; and
Department of Neurobiology, School of Life Science, University of Science and Technology of China, Hefei, P.R. China
2Correspondence: Netherlands Institute for Neuroscience, Meibergdreef 47, 1105 BA Amsterdam, The Netherlands. E-mail: d.f.swaab{at}nin.knaw.nl
SPECIFIC AIMS
The rhythmic pineal melatonin production, which is under the control of the biological master clock, the suprachiasmatic nucleus (SCN), is disrupted in Alzheimer disease (AD) patients from the earliest stages onward. This may contribute to their severe circadian rhythm disturbances, such as nightly restlessness. Here we investigate whether the pineal clock gene expression is affected in the process of AD and the underlying mechanisms.
PRINCIPAL FINDINGS
1. Diurnal rhythmic expression of hBmal1, hCry1, hPer1 is lost in the pineal gland of both preclinical (Braak stages I-II) and clinical AD subjects (Braak stages V-VI)
We found a significant diurnal rhythmic profile of clock genes hBmal1, hCry1, hPer1, and hßbeta;1-adrenergic receptor (hßbeta;1-ADR) in control human pineal gland (all P<0.05) (Fig. 1
a, b, c, e). There were no significant daily variations in hClock gene expression (Fig. 1d
, P=0.76). Diurnal variations of hßbeta;1-ADR expression were similar to the hPer1 diurnal expression pattern and both were positively correlated (r=0.50, P=0.02, n=24).
|
Intriguingly, diurnal rhythmic expression of hBmal1, hCry1, hPer1, and hßbeta;1-ADR mRNA was lost, both in preclinical AD (Braak stages I-II) (all P>0.2) and in clinical AD (Braak stages V-VI) (all P>0.1) (Fig. 1)
. No significant correlation was found between hßbeta;1-ADR and hPer1 mRNA, either in Braak I-II (r=0.09, P=0.68, n=22) or Braak V-VI (r=0.25, P=0.26, n=22).
Remarkably, we found that daytime levels of hCry1 mRNA were significantly increased in Braak V-VI compared with those in Braak 0 (P=0.004) and in Braak I-II (P=0.014), whereas no such change was observed between Braak 0 and I-II (P=0.9) (Fig. 1b)
. There was no significant difference of hßbeta;1-ADR, hPer1, hBmal1, hClock mRNA levels among Braak 0, I-II and V-VI, either in 24 h (all P>0.2) or in any time bins (all P>0.05).
2. Rat pineal deprived of SCN control showed similar clock gene alterations as the AD pineal
As the pineal is synchronized to the environment through the SCN, we investigated the effect of the loss of SCN control on the pineal clock gene oscillation. We compared pineal clock gene diurnal expression in the rat before and after disruption of the SCN-pineal functional connection by either superior cervical ganglionectomy (SCG-X) or lesioning the SCN (SCN-X).
In control rat pineal, Cry1, Per1, Bmal1, and ßbeta;1-ADR mRNA showed clear diurnal rhythms (all P<0.02), but not clock (P=0.55). Moreover, Per1 and ßbeta;1-ADR mRNA were highly correlated during 24 h (r=0.61, P<0.001, n=58) or at their trough and peak times (ZT 8 and ZT18) (r=0.76, P<0.001, n=22).
After experimental denervation, the day-night (ZT 8 and ZT 18) differences in Per1, Cry1, ßbeta;1-ADR, and Bmal1 gene expression completely disappeared in both SCN-X and SCG-X rats (see the online full-length article for figures). Moreover, the correlation between Per1 and ßbeta;1-ADR mRNA on ZT8 and ZT18 was lost in both SCN-X (r=0.58, P=0.10; n=9) and SCG-X rats (r=0.46, P=0.17, n=10). Strikingly, daytime (ZT8) Cry1 mRNA levels were higher in SCN-X (P=0.036) and SCG-X rats (P=0.003) compared with control rats (see online article for figures).
Remarkably, these changes in denervated rat pineal mirrored the clock gene alterations in the AD pineal, suggesting that the synchronization of the pineal through output from the SCN is lost in both preclinical and clinical AD.
3. Decreased AVP mRNA levels in the SCN of both preclinical and clinical AD subjects
AVP is a major rhythmic neuropeptide output of the SCN clockwork. The total amount of AVP mRNA in the SCN was decreased by 45% in Braak stages I-II (P=0.038; n=9) and strongly reduced by 70% in Braak V-VI (P=0.004; n=9) compared with Braak 0 (n=9). There was no significant difference in the AVP mRNA amount between Braak I-II and Braak V-VI (P=0.085). The total number of AVP mRNA-expressed cell profiles in the SCN markedly decreased in Braak V-VI to 20% of that in Braak 0 (P=0.001) and to 30% of that in Braak I-II (P=0.003). No difference of the number of AVP mRNA-expressed cell profiles was found between Braak 0 and I-II (P=0.40) (see online article for figures). These data indicate that the activity of the SCN, but not the cell number, is reduced early on in the AD pathogenesis.
CONCLUSIONS AND SIGNIFICANCE
This is the first study to show a rhythmic expression of a number of clock genes (hBmal1, hCry1, and hPer1) in the human pineal of controls, i.e., subjects without neurological disease or AD neuropathological changes. Furthermore, we found that the rhythmic expression of pineal clock genes was lost in both preclinical and clinical AD patient groups, which may be due to a decreased output of the master clock, the SCN.
We observed a diurnal rhythmic expression of clock genes in the human pineal, which is in agreement with previous studies in the rodent pineal. In the rodent, the pineal clock gene Per1 is under the noradrenergic control of the SCN through a ßbeta;-adrenergic cAMP signaling pathway. ßbeta;1-ADR acts as a key factor in this cAMP signaling pathway and is directly linked to the multisynaptic noradrenergic pathway derived from the SCN. The highly positive correlation between Per1 and ßbeta;1-ADR seen in the intact rat pineal, which actually disappeared in the rat pineal deprived of SCN control, may thus reflect the SCN regulation of Per1. Intriguingly, in control human pineal, hPer1 and hßbeta;1-ADR were also positively correlated and showed a similar daily expression pattern. Moreover, functionally active cAMP-responsive elements (CREs) have been found in the human hPer1 promoter. It therefore seems likely that hPer1 in the human pineal is also controlled by the SCN via the ßbeta;-adrenergic cAMP signaling pathway. Thus, by activating hPer1, the sympathetic input could potentially trigger the clockwork oscillation in the human pineal.
The diurnal rhythmic expression of hPer1, hCry1, and hBmal1 was lost in both preclinical and clinical AD, which suggests that pineal clock gene oscillation is disrupted early on in the AD process. Moreover, the positive correlation between hPer1 and hßbeta;1-ADR mRNA, which probably indicates SCN control of hPer1, is disturbed in both preclinical and clinical AD. The pineal itself does not suffer from AD neuropathology. Therefore, we propose that the changes in pineal clock gene expression in the AD process are due to a disrupted SCN control. This possibility is strongly supported by our animal experimental data. The rat pineal deprived of SCN control showed alterations of clock gene expression remarkably similar to those observed in the AD pineal. This holds for the loss of rhythmic clock gene expression, the loss of correlation between Per1 and ßbeta;1-ADR mRNA, and in particular for the increased Cry1 mRNA. Although the mechanism underlying the increased hCry1 in AD pineal is not yet clear, our studies suggest it may be due to a complete lack of sympathetic input.
Remarkably, we found that the AVP mRNA levels in the SCN decreased from the earliest AD neuropathological stages onward. AVP is a major rhythmic neuropeptide output of the SCN clockwork and regulates the rhythm of activity within the SCN and in other brain regions. AVP expression in the SCN is driven by the SCN molecular clock, as its rhythm had disappeared and its mRNA levels were dramatically decreased in clock mutated mice. Therefore, our data suggest that the SCN has a diminished output and a disrupted clock function from the earliest AD stages onward. Moreover, it supports the possibility that the SCN control of the pineal is disturbed early on in the AD process.
The finding that human pineal clock gene changes in the AD process are mimicked in the rat pineal by SCN lesion or SCG ectomy strongly suggests that the functional connection between the SCN and the pineal is affected in preclinical and clinical AD stages. Other functional disturbances of brain network connections could underlie the cognitive deficits in Alzheimer’s disease. Indeed, a loss of functional connectivity between prefrontal cortex and hippocampus has been reported recently in living Alzheimer patients, and a loss of synapses has been proposed to be one of the earliest changes in the disease process. Synchronization of the pineal clock gene oscillation to environmental cues has provided us with a rare opportunity to study functional connectivity in the postmortem brain.
In conclusion, the decreased activity of the SCN—already present at the moment of the occurrence of the very first tangles in the transentorhinal cortex (Braak stage I)—most likely affects pineal clock gene synchronization. We propose that the circulating melatonin levels and their daily rhythmicity may thus provide information about the very first AD stages that cannot be monitored in any other way at this time.
|
FOOTNOTES
1 Present address: BioFocus, Galapagos Genomics, Leiden, The Netherlands. 
To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.05-4446fje
This article has been cited by other articles:
|
|
 |
|
  C. Bertolucci, N. Cavallari, I. Colognesi, J. Aguzzi, Z. Chen, P. Caruso, A. Foa, G. Tosini, F. Bernardi, and M. Pinotti Evidence for an Overlapping Role of CLOCK and NPAS2 Transcription Factors in Liver Circadian Oscillators Mol. Cell. Biol., May 1, 2008; 28(9): 3070 - 3075. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
: Braak 0), preclinical AD (
: Braak I-II) and clinical AD (
: Braak V-VI). The significant daily variation of hBmal1, hCry1, hPer1, hßbeta;1-ADR mRNA only appears in Braak 0 (aged controls). *Among Braak 0, values significantly higher than 1000–1600 (a–c) or significantly higher than 1000–1600 and 2200–0400 (e). #hCry1 mRNA levels are higher in Braak V-VI than Braak 0 and I-II (b). ## hCry1 mRNA levels are higher in Braak V-VI and Braak I-II than Braak 0 (b).
= a decrease of levels. 
= daily rhythm; 
= disrupted daily rhythm.