| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Division of Behavioral Neuroscience, Department of Psychology, University of Colorado, Boulder, Colorado 80309
Address all correspondence and requests for reprints to: Dr. M. Girotti, Department of Psychology, Muenzinger Building Room D244, University of Colorado, UCB 345, Boulder, Colorado 80309. E-mail: girotti{at}colorado.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Based on the above evidence, it was hypothesized that in the AM, CRH neuron stimulatory input to the pituitary is negligible and low plasma ACTH levels reflect constitutive secretion of corticotropes. In the PM, a stimulatory drive from PVN CRH neurons increases ACTH secretion (hypothalamic circadian drive). Therefore, in the AM a tonic-negative feedback effect of low concentration corticosterone is sufficient to maintain ACTH levels in the normal range. In the PM, a higher concentration of corticosterone, likely acting at both pituitary and hypothalamic sites, is required to exert the same effect (11). If the hypothesized circadian drive is present only at one time of day, then it is expected that the basal (and perhaps also stress induced) activity of the CRH neuron changes across the 24-h day. We postulated that immediate early gene (IEG) expression in the PVN would provide a good measure for investigation of the presence of circadian drive upstream of the HPA hormonal output. IEGs are rapidly induced by a change in intracellular state, after cellular stimulation from intercellular signals (12, 13). To determine whether circadian inputs result in selective or global effects on IEG expression, we examined more than one IEG within each anatomical component of the HPA axis. Thus, we measured c-fos, zif268 (also known as NGFI-A or egr-1) and CRH gene expression in the PVN at two times of the day, during the resting phase (2 h after lights on, AM) and just before the initiation of activity (1 h before lights off, PM). Because the circadian rise in corticosterone secretion precedes (anticipates) the onset of the dark period, the choice of these two measurement times ensures that the photoperiod (lights on) and rat’s activity period (inactive) were the same for all measures, whereas basal corticosterone was quite different between the AM and PM measures. To examine whether patterns of IEG expression observed in the PVN are similar in downstream components of the HPA axis, we also examined the basal expression of c-fos, zif268, and POMC genes in the anterior pituitary. In the adrenal cortex, we examined basal and stress-induced expression of c-fos and another widely expressed IEG, NGFI-B (14).
In addition to the circadian rhythm of basal HPA axis hormonal secretion, an acute activation of the axis and release of glucocorticoids is observed in response to stress. It is not known to what extent the circadian corticosterone rhythm directly or indirectly affects basal and stress-induced IEG expression in the axis. The surge in corticosterone that occurs around the onset of the dark photoperiod could directly affect the basal and stress-induced gene expression levels during the ensuing several hours (evening time levels). In this paper we refer to such glucocorticoid effects on HPA axis hormone levels and gene expression changes as acute effects of glucocorticoids. On the other hand, there may also be an influence of this circadian variation of corticosterone secretion such that its prior occurrence affects subsequent cellular activity and gene expression irrespective of recent circulating levels of corticosterone. For example, Jacobson et al. (15) found that giving adrenalectomized (ADX) rats a corticosterone replacement regimen that produced a diurnal variation in corticosterone levels (corticosterone in the drinking water) was sufficient to normalize the duration of an ACTH response to acute stress challenge in the morning. In contrast, replacement with constant levels of corticosterone (sc corticosterone pellets) was ineffective on this measure. An important feature of providing ADX rats with corticosterone in the drinking water is that almost all of the drinking occurs during the first several hours after lights off as the rats engage in feeding (15). Consequently, these ADX rats have a PM surge in circulating corticosterone. However, due to the very short half-life of corticosterone [approximately 20 min in rats; (16)], circulating corticosterone levels are nearly undetectable during the next morning when the rats are inactive. The Jacobson study indicates that a surge of corticosterone during the evening was sufficient to normalize the temporal pattern of the ACTH response to acute stress the following morning. In this study we refer to these relatively long-lasting delayed effects of glucocorticoids as permissive effects.
To determine possible glucocorticoid effects due to either the ongoing or previous diurnal phasic surge in corticosterone levels, we compared dependent measures in ADX rats, ADX rats given corticosterone in their drinking water, or sham-operated rats. The choice of a PM measurement time in this study that precedes the onset of the dark period minimizes the extent to which ADX rats given corticosterone in the drinking water have already begun their daily feeding and water intake on the test day. Consequently, these rats, as opposed to sham-ADX rats, lack the concurrent PM surge in corticosterone. On the other hand, these rats, in contrast to noncorticosterone-replaced ADX rats, had a surge in corticosterone approximately 20–22 h earlier as a result of feeding and drinking during the previous night.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Restraint stress procedure
Restraint involved taking rats from the home cage and placing them in adjustable length (15.5 ± 2.5 cm long and 6.3 cm diameter) Plexiglass tubes with air holes in the front, top, and back. This stressor is considered to be primarily psychological in nature because it does not produce pain or direct physical insult (17). Restraint took place in a separate room adjacent to the home cage room.
Adrenalectomy and corticosterone replacement
Adrenalectomies were performed under halothane-induced anesthesia via the dorsolateral approach. Control animals were operated under the same anesthesia, but the adrenals were not removed (sham animals). Adrenalectomized animals were given immediate access to 0.9% saline (ADX animals) or 0.9% saline containing 25 µg/ml corticosterone (Sigma, St. Louis, MO) (ADX + Bw) and were allowed to recover from surgery for 5 d before testing. There was no detectable plasma corticosterone as determined by our corticosterone assay (see below) in ADX rats.
In situ hybridization histochemistry and image analysis
Tissue sections (10 µm thick) were cut on a cryostat (model 1850; Leica Microsystems, Nussloch, Germany) through the extent of the PVN [
1.80 mm posterior to bregma (63)] or through the middle portion of the adrenal gland, thaw mounted onto poly-L-lysine-coated slides and stored at –80 C.
In situ hybridization for nerve growth factor I-B (NGFI-B), c-fos, and zinc finger 268 (zif268) mRNAs and CRH heteronuclear RNA (hnRNA) and image analysis of digitalized images from x-ray films were performed as previously described (18). For the generation of the riboprobes, plasmids containing a fragment of c-fos cDNA (courtesy of Dr. T. Curran, St. Jude Children’s Research Hospital, Memphis, TN) or zif268 cDNA (courtesy of Dr. J. Milbrandt, Washington University School of Medicine, St. Louis, MO) or a portion of the CRH intron (kindly provided by Dr. R. Thompson, University of Michigan, Ann Arbor, MI) were used. The NGFI-B plasmid harboring a 839-bp fragment of the NGFI-B gene spanning nucleotides 188-1027 (GenBank accession no. NM_024388) was generated in house. Briefly, total RNA from rat hippocampus was isolated using the SV Total RNA extraction system (Promega, Fitchburg, WI) and first-strand cDNA was produced using SuperScript II reverse transcriptase reagents (Invitrogen, Carlsbad, CA). After PCR with NGFI-B-specific oligonucleotides, the amplified product was cloned into pCRII-TOPO vector (Invitrogen) according to manufacturer instructions. The identity of the cloned DNA was verified by DNA sequencing (University of Colorado Molecular, Cellular, and Developmental Biology sequencing facility).
ACTH RIA
Blood samples for the ACTH assay were collected into EDTA-coated tubes and stored at –80 C. Plasma concentrations of ACTH were determined by RIA procedures described elsewhere (19) using antiserum (rabbit antibody Rb7, courtesy of Dr. Bill Engeland, University of Minnesota, Minneapolis, MN). The detection limit for this assay was 15 pg/ml for a 50-µl sample; the intraassay coefficient of variability was 6–10% and the interassay coefficient of variability was 9%.
Corticosterone enzyme immunoassay
Measurement of plasma corticosterone was conducted on 20 µl of plasma with an enzyme immunoassay kit (Assay Design, Ann Arbor, MI) according to manufacturer’s instructions. The sensitivity for this assay was 130 ng/100 ml. The intraassay coefficient of variability was 8–10% and the interassay coefficient of variability was 10%.
Semiquantitative PCR
Relative levels of gene expression from dissected anterior pituitaries were measured by semiquantitative PCR as previously described (20). Briefly, RNA was extracted from anterior pituitary tissue according to the method of Chomczynski and Sacchi (21), using Trizol (Invitrogen). First-strand cDNA was generated with random hexamers using SuperScript II reverse transcriptase (Invitrogen). Hot-start PCR, using Taq-Bead hot start polymerase (Promega) was used to generate fragments of the POMC gene spanning intron 1 and exon 2 and c-fos and zif268 coding regions. Oligonucleotide sequences were as follows: POMC forward, 5'-GAGGCATAAACTGGCACC-3', POMC reverse, 5'-GCTGTTCATCTCCGTTGC-3'; c-fos forward, 5'-CTTCCTTTGTCTTCACCTACC-3', c-fos reverse, 5'-CCTTCTCTGACTGCTCACA-3'; and zif268 forward, 5'-TCACTCACCCACCATGGACAA-3', zif268 reverse, 5'-TCAGCAGCATCATCTCCTCCA-3'. Linear amplifications were obtained at 30 cycles for POMC and c-fos and 25 cycles for zif268. Duplicate reactions for each sample were run on 3–3.5% agarose gels, and digital images of the UV-visualized bands were quantified using the National Institutes of Health Image program. Relative levels of expression are indicated as ratios between the densitometric measure of the test gene and a housekeeping gene (GAPDH).
Experiment 1: diurnal comparison of basal and acute stress activity in adrenal-intact rats
Animals (N = 24) were separated into four groups (n = 6); two groups were subjected to restraint stress for 15 min before rapid decapitation that occurred in the morning, 2 h after lights on [AM = zeitgeber time (ZT) 2], or in the evening, 1 h before lights off (PM = ZT 11). The other two groups were killed at the same times of day, but they were left undisturbed in their cages until death.
Experiment 2: diurnal comparison of basal and acute stress activity in sham and ADX rats
Animals (N = 48) were separated into eight groups (n = 6). Half of the animals were adrenalectomized and half were sham operated. Five days after operation, animals from each group were then subjected to either 15 min restraint stress in the AM or PM and immediately killed or killed at the same times without restraint stress experience.
Experiment 3: diurnal comparison of basal activity in sham, ADX, and ADX + corticosterone drinking water (Bw)-replaced rats
Animals (N = 48) were separated into six groups (n = 8). One third of the animals were ADX, one third were ADX and given corticosterone (25 µg/ml) in the drinking fluid (ADX + Bw), and one third was sham operated. Five days after operation animals from each group were either killed in the AM or PM without further manipulations.
Statistical analysis
ANOVA was used to examine overall main effects and interactions of time of day and stress condition (experiment 1; 2 x 2 factorial design); time of day, stress condition, and glucocorticoid state (experiment 2; 2 x 2 x 2 factorial design); or time of day and glucocorticoid state (experiment 3; 2 x 3 factorial design). Data were log transformed before applying ANOVA because of the greater variability associated with the measures in ADX and stressed rats, compared with basal measures in adrenal-intact rats. Within this framework, our criterion to establish the significance of circadian effects was to apply unpaired Student’s t tests in post hoc pairwise comparisons only in those instances in which the ANOVA yielded a significant main effect for time of day or a significant interaction of time of day with any other factor. The program StatView (SAS Institute Inc., Cary, NC) was used in all the analyses. An alpha level of P < 0.05 was used to determine statistical significance. Data presented represent group means ± SEM.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). There was also a trend for an increase in basal ACTH in the PM (Fig. 1B, inset); however, neither the overall time of day main effect nor time of day by stress condition interaction was statistically significant for this measure. Thus, as previously described (11, 22), the diurnal amplitude was substantially larger for corticosterone than ACTH.
|
) showed an increase in the PM (P < 0.05), suggesting an increase in activation state in PVN neurons as the animal approaches the time of awakening. Unlike c-fos, CRH hnRNA showed a significant decrease in the PM (P < 0.01; Fig. 1D). Finally, zif268 mRNA basal expression in the PVN was low and did not significantly change with time of day (Fig. 1E).
Stress-induced levels
Fifteen minutes of restraint stress produced rapid elevations in plasma hormones levels (corticosterone and ACTH) and PVN gene expression (c-fos mRNA, CRH hnRNA, and zif268 mRNA; Fig. 1, A–E). Stress-induced changes were not significantly different between the two times of day for any of the parameters analyzed, except for CRH hnRNA (Fig. 1D). Stress-induced CRH hnRNA levels were significantly lower in the PM than AM (P < 0.01). In the case of c-fos mRNA, there was a significant time of day by stress condition interaction (P = 0.01), lending support to the notion that the induction of the c-fos gene by acute stress overrides the diurnal difference in basal expression.
Experiment 2: diurnal differences in basal and stress-induced HPA axis hormone levels, PVN gene expression of adrenal-intact (sham) and ADX rats, and adrenal cortical gene expression of sham rats
The second experiment was essentially a replication of the first experiment with the added dimension of comparing the responses of adrenal-intact rats 5 d after sham surgery (sham) with rats 5 d after ADX.
Sham rats: basal and stress-induced plasma hormone and PVN gene expression levels
As expected, the diurnal pattern of basal and stress-induced plasma hormone and PVN gene expression levels of sham rats was very similar to that seen in experiment 1 (Fig. 2, A–E). Importantly, this experiment replicated the PVN gene expression patterns of adrenal-intact rats observed in experiment 1. Specifically, basal c-fos mRNA levels were higher in the PM than AM (P < 0.05; Fig. 2C), basal CRH hnRNA levels were lower in the PM than AM (P < 0.001; Fig. 2D), and basal zif268 mRNA levels did not differ between the two times of day (Fig. 2E). Furthermore, stress-induced CRH hnRNA levels were lower in the PM, compared with the AM (P < 0.05; Fig. 2D).
|
, F–H). Under basal conditions, adrenocortical c-fos mRNA showed a trend to increase in the evening (Fig. 2F, inset), although two-way ANOVA indicated no main effect of time of day nor a time of day by stress condition interaction. Acute stress produced a substantial increase in c-fos expression in the adrenal cortex at both times of day, although perhaps to a lesser extent in the PM than AM (Fig. 2, F and H). In the case of NGFI-B gene expression in the adrenal cortex, there was a clear increase in basal mRNA in the evening (P < 0.01), and NGFI-B mRNA was markedly elevated by stress at both times of day (P < 0.001 in AM and P < 0.05 in PM; Fig. 2, G and H).
ADX rats: basal and stress-induced plasma hormone and PVN gene expression levels
ADX dramatically increased basal AM and PM ACTH levels (Fig. 2B), in agreement with previous observations (26). Moreover, in ADX animals, basal ACTH levels were higher in the PM than AM (P < 0.01). In the PVN, ADX increased evening but not morning basal levels of c-fos mRNA (P < 0.001; Fig. 2C), thus producing an increase in the basal diurnal difference amplitude compared with sham animals. In this experiment, ADX did not significantly affect basal levels of CRH hnRNA in the morning (Fig. 2D), but it did increase CRH hnRNA basal evening levels (P < 0.001), bringing them close to the ADX AM basal levels and thereby abolishing a diurnal difference in basal CRH hnRNA levels at these two times of day. There was no significant effect of ADX on basal levels of zif268 mRNA (Fig. 2E).
As in adrenal-intact rats, 15 min of restraint produced a large increase of all measured parameters in ADX rats (Fig. 2, B–E). There was no diurnal difference in the stress-induced levels of these parameters, including CRH hnRNA. Overall levels, however, of stress-induced ACTH, c-fos mRNA (only AM reached significance) and CRH hnRNA (but not zif268 mRNA) were higher in ADX rats than sham rats (P < 0.01).
Experiment 3: effect of ADX ± Bw replacement on diurnal differences in basal HPA axis hormone levels, PVN gene expression, and anterior pituitary gene expression
To discriminate which of the above effects brought about by ADX were due to removal of a permissive corticosterone effect or removal of an acute phasic influence of PM elevations in basal corticosterone, experiment 3 included a group of ADX rats that had corticosterone in their drinking water (ADX + Bw). Permissive corticosterone effects would be manifest in both sham and ADX + Bw groups, whereas phasic influences of corticosterone would affect only the sham group. As shown in Fig. 3A, AM corticosterone levels in ADX + Bw rats were less than sham rats and barely detectable. PM corticosterone levels were slightly higher, averaging around 4 µg/dl, probably as a result of an early start (before lights out) in daily activity and drinking for some of the animals in this group. The average corticosterone levels for this group were, however, well below those measured in sham rats at the same time.
|
, A–E). Thus, there was a large PM increase in basal corticosterone (P < 0.001) and a small statistically nonsignificant PM increase in basal ACTH. For gene expression in the PVN, there was a PM increase in basal c-fos mRNA (P < 0.05), a PM decrease in basal CRH hnRNA (P < 0.001), and no diurnal difference in zif268 mRNA.
ADX ± Bw: basal plasma hormone, thymus weights, and PVN gene expression levels
As observed in experiment 2, ADX produced a large increase in basal ACTH levels, and there was a significant diurnal difference with higher PM basal ACTH levels (P < 0.01; Fig. 3B). AM ACTH in the ADX + Bw group was low and similar to sham levels, indicating a permissive influence of prior nocturnal intake of corticosterone. Interestingly, PM basal ACTH levels of ADX + Bw rats were elevated to nearly ADX levels in the PM. Measure of thymus weights confirmed that corticosterone replacement was appropriate to reduce thymus weight toward adrenal-intact values without producing signs of higher than normal corticosterone exposure (thymus weights: sham, 414 ± 20 g; ADX, 555 ± 22 g; ADX + Bw, 485 ± 20 g; significance of post hoc t test: P < 0.05 for ADX +Bw, compared with either ADX or sham).
The effects of ADX ± Bw on PVN gene expression are summarized in Figs. 3 and 4. As in experiment 2, ADX produced a large increase in PM PVN basal c-fos mRNA (P < 0.01) but had no effect on AM levels. Corticosterone drinking water replacement normalized the PM levels of c-fos expression, thereby eliminating the increased circadian amplitude difference observed in ADX animals (Figs. 3C and 4A). In this experiment, in contrast to the previous experiment, ADX produced a significantly elevated level of basal CRH hnRNA at both times of day, compared with sham rats (P < 0.001; Figs. 3D and 4A). Nevertheless, as in the previous experiment, ADX eliminated the diurnal difference in basal CRH hnRNA. Corticosterone replacement reduced basal CRH hnRNA levels to within range of the sham levels; however, it failed to reinstate the diurnal difference between AM and PM CRH hnRNA levels observed in sham rats. As in experiment 2, there was no corticosterone-dependent effect observed on zif268 mRNA levels (Fig. 3E).
|
, F–H, and 4B). Anterior pituitaries were dissected and mRNA extracts used in semiquantitative reverse transcriptase PCR analysis with primers specific for c-fos mRNA, proopiomelanocortin (POMC) hnRNA, and zif268 mRNA.
The expression pattern of c-fos mRNA in the anterior pituitary showed a marked diurnal difference in ADX + Bw rats that was not evident in sham or ADX rats (P < 0.01; Fig. 3F). There was also a large PM increase in POMC hnRNA levels of ADX + Bw rats (P < 0.001; Figs. 3G and 4B). In addition, sham rats showed a small but significant increase in POMC hnRNA levels in the PM (P < 0.01) and a similar trend, although not significant, was observed in ADX animals. Levels of zif268 mRNA in the anterior pituitary were similar for all groups; however, there was a small but significant decrease in the PM of sham rats (P < 0.05; Fig. 3H).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We found that the diurnal difference in HPA axis hormone secretion was also manifest by a selective elevation of PVN basal c-fos expression in the evening. This was observed in all three experiments (Figs. 1–4). In line with the idea that the increased c-fos gene expression in the PM was associated with increased CRH neuron activity and hence CRH secretion, a parallel increase in corticotrope POMC hnRNA (Figs. 3 and 4) was also observed in the evening. The POMC gene acts as a corticotrope-specific activity marker when examining levels of the primary transcript (POMC hnRNA) (32). Thus, PVN basal c-fos expression may reflect a circadian central drive that marks the activity of the CRH neuron during the 24 h. Interestingly, despite evidence for increased PVN neuronal activity, other genes in the PVN (CRH and zif268) did not display the same trend of increased evening expression observed with c-fos (discussed later).
In adrenal-intact rats, there was only a small increase in basal ACTH levels at the PM time point sampled. However, in ADX rats, there was a strong diurnal increase in basal ACTH in the PM. The fact that this diurnal difference was still present (and even enhanced) in ADX rats given corticosterone in the drinking water suggests that the normal diurnal surge of endogenous corticosterone in the evening largely limits the concurrent secretion of ACTH. Perhaps consistent with this conclusion is the observation from other studies that replacement of ADX rats with moderately low constant levels of corticosterone (sc corticosterone pellets) is sufficient to normalize basal ACTH in the AM but not PM (22, 33). Interestingly, the large PM elevation of ACTH in ADX rats with corticosterone pellets is substantially delayed after the onset of the dark period.
Differential gene expression response to circadian drive, stress, and glucocorticoid effects in the PVN
Each of the three genes examined in the PVN had a unique response profile to the diurnal, acute stress and glucocorticoid conditions of this study. Thus, each gene reported different aspects of PVN activation and regulation. Here we summarize the features of each gene expression pattern.
c-fos Gene expression patterns.
c-fos Gene expression patterns. As described As described above, c-fos mRNA expression in the PVN was elevated in the evening under basal conditions, in phase with diurnal HPA axis hormone secretion and pituitary POMC hnRNA levels. Stress caused an even greater elevation of c-fos levels at both times of day, with no diurnal difference in this gene’s peak stress-induced expression. Basal PVN c-fos mRNA levels in ADX animals were sensitive to removal of glucocorticoids but only in the evening. Consequently, basal c-fos mRNA levels in the PVN of ADX rats were indistinguishable from those of sham rats in the AM but considerably elevated in the PM. This suggests that, whereas a constant low level of c-fos expression in the AM is not modified by changes in corticosterone status, endogenous corticosterone normally constrains the extent to which c-fos gene expression is increased in the PM in response to increased circadian drive. Administration of corticosterone in the drinking water to ADX rats normalized PM levels of c-fos mRNA. Thus, corticosterone has a permissive effect on PM basal c-fos expression that does not require the concomitant presence of high levels of circulating corticosterone but rather a surge in circulating corticosterone the previous evening. Glucocorticoid receptor is highly expressed within CRH neurons of the PVN (34); it remains to be determined, however, whether this circadian-dependent effect of glucocorticoids on c-fos expression is a result of a direct action of glucocorticoids in the PVN.
We also found that ADX increased stress-induced c-fos in the PVN, compared with adrenal-intact animals (experiment 2, Fig. 2). This result confirmed previous studies conducted in the AM (30, 35, 36, 37). The results have extended previous analysis by demonstrating a similar enhancement of stress-induced c-fos in the PM of ADX rats and an absence of a diurnal difference in the extent of c-fos induction by stress. Taken together these results suggest that: 1) the relatively recent (within the past 24 h), but not concurrent, presence of glucocorticoid is sufficient to exert an inhibitory influence on either the magnitude of a net neural stimulatory circadian drive or the intrinsic PVN response to that drive, and 2) activation of stress neurocircuitry overrides the diurnal difference in PVN basal c-fos expression, irrespective of adrenal status.
zif268 Gene expression patterns.
A somewhat unexpected finding of this study was that zif268 gene expression displayed no significant diurnal change in the PVN (Figs. 1–3). However, such lack of response was not due to a generalized insensitivity of this gene to neural stimulation because zif268 gene expression in the PVN reacted promptly and prominently to acute stress in a manner similar to c-fos (Figs. 1 and 2). The fact that these two IEGs display such different patterns of expression in response to circadian input is interesting, given the fact that generally c-fos and zif268 have been shown to share very similar activation patterns and to be induced by similar sets of stimuli (38). It is possible that zif268 and c-fos are expressed in different subpopulations of parvocellular PVN neurons, and further colocalization studies are required to address this point. If, on the other hand, the two genes are expressed in the same cells, these observations indicate that whereas stress activates a signaling cascade that converges on some common regulator, circadian drive can selectively activate signaling to c-fos but not to zif268 (or CRH gene expression; see below). The neurochemical connections between SCN outputs and PVN are not yet completely mapped out and appear to include both direct and indirect connections, with perhaps a predominance of the latter on parvocellular CRH neurons (6, 7, 9, 39, 40). Within this framework, it is likely that neurochemical signals that relay circadian information to the PVN vary between AM and PM, triggering differential signaling cascades and selective responses in gene expression.
Finally, zif268 response in the PVN differed from c-fos in that it was insensitive to glucocorticoid status, under both basal and stress conditions. This is in agreement with previous studies showing that stress-induced c-fos gene expression in the PVN can be suppressed by long-term elevation of glucocorticoids, whereas stress-induced zif268 expression is relatively insensitive to glucocorticoid elevation (27).
CRH gene expression patterns.
Whereas c-fos expression rose in the PM under basal conditions, CRH primary transcript showed a decline at this time of day. This pattern agrees with previous observations that the rhythm of CRH basal gene expression (primary and mature transcript accumulation) is in opposite phase to corticosterone and ACTH diurnal secretion patterns (22, 41). These observations have been taken to indicate that basal CRH gene transcription and peptide secretion are under separate control and are relatively uncoupled (22, 42).
It is striking that not only basal CRH hnRNA levels but also stress-induced levels were lower in the PM than AM. This is in contrast to the lack of a diurnal difference in stress-induced plasma hormone, c-fos mRNA, and zif268 mRNA levels. The fact that stress produced equal degrees of c-fos and zif268 activation at both times of day suggests that a similar amount of restraint-induced neural stimulation was imparted to the CRH neuron in the morning and evening. Thus, one or more factors restrict basal and stress-induced CRH gene expression in the evening. One possible candidate for this effect is corticosterone, the levels of which increase in the PM. This prospect is supported by the fact that the long-term absence (5 d) of glucocorticoids not only increased basal CRH hnRNA levels, compared with sham levels (particularly evident in Fig. 3), but also completely abolished the AM vs. PM difference in basal (Figs. 2–4) and stress-induced (Fig. 2) CRH hnRNA levels. Placing corticosterone in the drinking water of ADX rats normalized the overall basal CRH hnRNA levels but did not reinstate the AM/PM difference observed in sham animals (Fig. 3). This suggests that the acute elevation of basal corticosterone levels in the evening probably suppressed both basal and stress-induced CRH hnRNA levels at this time of day. This conclusion is also supported by the observation that within the corticosterone replaced ADX group some animals had slightly higher corticosterone levels than others, maybe as a result of some early diurnal drinking before the lights out transition. Interestingly, in this group corticosterone levels and CRH expression showed a strong inverse correlation (r = –0.81, P = 0.01). Thus, an acute elevation of corticosterone levels can inhibit CRH gene expression under basal conditions and can limit its stress-dependent induction. There is support for glucocorticoids to have a direct effect on CRH gene expression. Glucocorticoids have been shown to inhibit transcriptional activation of the human CRH promoter through a negative glucocorticoid response element (43, 44). Glucocorticoids may also regulate CRH mRNA stability in the rat (45).
Contrary to our observations, other studies have reported a lack of dependence of basal CRH gene expression circadian fluctuations on glucocorticoids (22, 46). Whereas the reason for this discrepancy with the results of our study remains to be established, it is possible that because our experiment examined only two time points, we missed a potentially shifted peak and trough of CRH expression in the ADX animals. If this is the case, then glucocorticoids may interact with circadian stimuli in setting the phase of circadian CRH gene expression changes. In this respect it is interesting to note that Watts et al. (22), using a regimen of constant corticosterone replacement via sc pellets in ADX rats, still observed a circadian difference in CRH expression but with a marked phase shift from sham animals. Thus, it is also possible that different outcomes may be expected, depending on the type of corticosterone replacement used (in drinking water, as in our study, vs. constant release pellets, as in Refs. 22 and 46).
Finally, although not examined in this study, it is worth noting for comparison purposes that arginine vasopressin hnRNA in the parvocellular PVN has been shown to display a circadian rhythm of expression in ADX rats and ADX rats supplemented with constant release corticosterone pellets (22) but not adrenal-intact animals. This result raises the intriguing possibility that the normal diurnal rhythm in corticosterone secretion masks an endogenous rhythm in arginine vasopressin gene expression in the PVN.
Anterior pituitary gene expression
In the anterior pituitary, relative POMC gene expression displayed a similar profile to HPA axis hormone secretion (Fig. 3), suggesting that transcriptional regulation of ACTH production and ACTH secretion may be relatively well coupled. Levels of POMC hnRNA exhibited a clear diurnal pattern under basal conditions, with increased levels in the evening, in line with presumed increased CRH peptide secretion at this time of day. POMC transcription can be induced by a CRH receptor-dependent signaling cascade that involves protein kinase A and calcium (47, 48, 49), whereas glucocorticoids have the potential to prevent transcriptional activation of the POMC gene by interfering with positive regulators on its promoter (50, 51). In our experiment, ADX induced elevated basal POMC hnRNA expression at both times of day, suggesting a tonic suppressive effect of glucocorticoids on this gene. Very likely, the tonic inhibition that glucocorticoids exert on ACTH plasma levels, observed in the present (Figs. 2 and 3) and previous studies (11), is, at least in part, due to an effect at the level of POMC gene expression. Interestingly, whereas placement of corticosterone in the drinking water of ADX rats normalized the morning levels of POMC hnRNA, it did not correct the large increase in basal POMC hnRNA levels observed in the PM of ADX animals (Fig. 3G). A very similar pattern of ACTH secretion was evident in the same rats (Fig. 3B). This suggests that the acute presence of high concentrations of corticosterone in the PM suppresses the levels of POMC hnRNA and ACTH secretion at this time of day, when presumably POMC transcription and ACTH secretion are otherwise enhanced by a sustained increase in CRH secretion. In a previous report (46), PM levels of POMC mRNA were still elevated, compared with sham rats, in ADX rats replaced with pellets at low doses of corticosterone (1.5 µg/dl) but not with higher doses of corticosterone (2.7 µg/dl). It is noteworthy that there was also a significant diurnal difference in anterior pituitary c-fos mRNA of ADX rats given corticosterone replacement (Fig. 3F), even though there was no diurnal difference in c-fos mRNA in the anterior pituitary of sham and ADX rats. It is possible that the presence of several c-fos expressing cell types in the anterior pituitary, most of which do not respond to CRH stimulation, masked small diurnal CRH-dependent stimulatory effects on c-fos expression in corticotropes of sham and ADX rats.
In contrast to the other pituitary-related measures, there was little evidence for a diurnal difference in zif268 mRNA expression in the anterior pituitary. A previous study has shown that increases in zif268 mRNA levels can be detected in the anterior pituitary after acute stress (32).
Adrenal gland gene expression
Besides gene expression changes in the PVN and anterior pituitary, we examined IEG expression within the adrenocortex. We chose to analyze c-fos and NGFI-B, a transcription factor that plays a role in the ACTH-dependent regulation of some adrenocortical steroidogenic enzymes (52, 53). NGFI-B expression in cortical areas of the adrenal gland (zona fasciculata and reticularis) has been previously shown to increase after stress (24); however, no direct measure of diurnal expression levels of this gene has been reported. Overall, c-fos mRNA pattern of expression had a similar trend to plasma ACTH, with small basal differences, and therefore seemed to reflect the acute stimulation of the adrenocortex by ACTH (Fig. 2). NGFI-B gene expression in the adrenal cortex showed a much more pronounced diurnal change under basal conditions, and it was markedly elevated by stress. Thus, NGFI-B gene expression was reminiscent of the profile of corticosterone secretion. These data suggest the possibility that the expression of different IEGs in the adrenocortex may reflect different aspects of adrenal function. Changes in adrenal gland sensitivity with time of day have long been proposed (54, 55); however, the mechanism remains obscure. Recent evidence for a direct role of SCN in controlling adrenal sensitivity has been produced (4, 56) and corroborated by anatomical evidence of SCN-derived autonomic inputs to the gland (57, 58, 59, 60, 61). In addition, it appears that light stimulation is capable of inducing steroidogenic enzymes and the NGFI-B family of transcription factors, as well as corticosterone secretion in the absence of ACTH secretion, via an SCN-dependent mechanism that requires intact autonomic innervations (62).
Conclusion
In brief, we have shown that a circadian drive that regulates HPA axis basal hormone secretion is also manifest on basal c-fos gene expression in the PVN. We show that different IEGs within the HPA axis anatomical components display different patterns of diurnal rhythmicity; in part, these differential patterns of gene expression are the result of gene-selective responses to diurnal-dependent intercellular signals and acute and/or permissive glucocorticoid actions.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure Statement: The authors have nothing to disclose.
First Published Online February 15, 2007
Abbreviations: ADX, Adrenalectomy/adrenalectomized; AM, morning; Bw, corticosterone drinking water; HPA, hypothalamus-pituitary-adrenal; IEG, immediate early genes; hnRNA, heteronuclear RNA; NGFI-B, nerve growth factor I-B; PM, evening; POMC, proopiomelanocortin; PVN, paraventricular nucleus of the hypothalamus; SCN, suprachiasmatic nucleus; zif268, zinc finger 268; ZT, Zeitgeber time.
Received September 22, 2006.
Accepted for publication February 2, 2007.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
4-steroid hydrogenase activity on adrenal cortical function. Endocrinology 68:747–758[Medline]This article has been cited by other articles:
 |
|
 |
|
  A. Kapoor and S. G. Matthews Prenatal Stress Modifies Behavior and Hypothalamic-Pituitary-Adrenal Function in Female Guinea Pig Offspring: Effects of Timing of Prenatal Stress and Stage of Reproductive Cycle Endocrinology, December 1, 2008; 149(12): 6406 - 6415. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||
